9 The Internet platforms’ impact on the Diversity of Cultural Expressions: to the Long Tail, and beyond!

Heritiana Ranaivoson[1]

Introduction

While there is a common agreement on the necessity to promote and protect the diversity of cultural expressions, this requires knowing more about how to achieve such a result, in particular in the digital era. First results have been provided regarding the impact of policies (CEIM 2015; Thiec 2014). However, the impact of non-state actors remains understudied, in particular of online platforms.

This is all the more important that digitization is deeply modifying the way cultural sectors are functioning, with an ambiguous impact on the diversity of cultural expressions. These sectors are among the first that have been heavily hit by the digital shift, i.e. the digitisation of information, the generalisation of the Internet protocol, and the rapid take-up of these technologies (Simon & Bogdanowicz 2012), all this in a context of creative destruction and disruptive innovation.

Besides, the digital shift has brought about a reconfiguration of the cultural industries’ value networks (Ballon et al. 2012), in turn leading to uncertainty, conflicts and strategic shifts. In particular, digital technologies are likely to threaten traditional players (creators, intermediaries) to the benefit of Internet giants and specialized platforms (Zhu & Seamans 2010). It is claimed that these players’ strategies are likely to lead to more homogeneity in content supply and consumption (Guèvremont et al. 2013).

This paper proposes to analyse how Internet platforms contribute to, or hinder, the online diversity of cultural expressions: do they propose a more diverse offer? Do they induce a more diverse consumption of cultural content? What are the mechanisms at work? It does so via a literature review focused on the main theory elaborated to assess the impact of digitization on the diversity of cultural expressions, namely the Theory of the Long Tail. The paper opposes it to the Theories of Superstars, since they provide opposite predictions on the impact of digitization on the diversity of cultural expressions. Finally, it proposes ways to go beyond the results and approaches of these theories.

In the remainder, section 2 describes the methodology used, in particular the Stirling Model. Section 3 analyses the Theories of Superstars and the Theory of the Long Tail, and provides an overview of recent studies on the Long Tail. Section 4 discusses the main differences between these theories and ways to go beyond their limitations.

I – Methodology

Based on desk research, the paper consists in an overview of papers analysing the impact of platforms on the diversity of cultural expressions. The papers were found using Google Scholar, and in the references of each paper. The paper notably updates and expands Brynjolfsson et al (2010).

In particular, it proposes to reframe the Long Tail’s theoretical framework and point at its limitations by using the Stirling Model. The Stirling Model considers diversity as a mix of variety, balance and disparity (Stirling 2007). In this approach, the diversity of a system (e.g. Netflix’s feature films catalogue) can only be assessed when its elements (e.g. the feature films) have been grouped into categories (e.g. these films’ nationality). Once this categorization has been done, variety corresponds to the number of categories; balance to the way the elements are spread among categories (e.g. the share of films released per nationality); disparity to the level of difference between the categories (e.g. between every pair of them or between the two most distinct). This Model is increasingly used in the cultural field (see e.g. Peltier & Moreau 2012; Ranaivoson 2007).

As it is somehow too generic to analyse the diversity of cultural expressions, the paper also distinguishes between supplied diversity and consumed diversity (Van Cuilenburg & Van der Wurff 2001). Supplied diversity corresponds to the diversity of what is made available. Consumed diversity refers to diversity as it is actually consumed, thus depending on both consumer tastes and supplied diversity.

II – Theoretical background

To assess the impact of platforms on the online diversity of cultural expressions, the paper proposes two theories that have contrary analyses of the implications of digital technology for content diversity: the Theories of Superstars and the Long Tail Theory.

A – The Theories of Superstars

The Theories of Superstars aim at explaining why consumption is focused on a restricted number of products or creators, the so-called Superstars. There are two theories with different, though compatible, approaches: Rosen (1981) and Adler (1985).

According to Rosen (1981), some creators (or products) are Superstars because they are more talented and benefit from technology that allow them to reach a great number of consumers at a low cost. First, there is a limited substitutability for consumers between two creators with different talents (Rosen 1981). Therefore, a slight superiority in talent leads to a much larger revenue. In other words, revenue is a convex function of talent. However, the assumption of a common agreement among all consumers regarding the distribution of talents is problematic (Benhamou 2012 & Moureau 2006) notably as it lacks realism. More interestingly, Rosen (1981) argues that it is crucial that distribution and consumption technologies rely on low marginal costs. Actually this allows better products or more talented creators to benefit from economies of scale. Therefore, technology plays a crucial role in Stars’ success (Moureau 2006), by allowing to reduce congestion costs in consumption (Schulze 2003).

The Theory of Superstars in Adler’s (1985) approach puts information at the core of the choice by consumers and hence of the resulting (lack of) consumed diversity. For Adler (1985), what is important is “the need on the part of consumers to consume the same art that others do” (Adler 2006: 3). This results from the assumption that consumers get increasingly satisfied the more they know about what they are consuming (Adler 1985; Stigler & Becker 1977). Therefore, information is at the core of Adler’s (1985) theory. For this reason, in his theory, talent is not a crucial assumption.

Since consumers try to know as much as possible about what they consume, the most famous creators or products are advantaged, and this is a self-reinforcing feature (Adler 1985). Adler’s (1985) model also allows to understand why distributors concentrate their marketing means on a few creators, instead of spreading them equally among all creators. Actually the aim is that the creator reaches the threshold that will trigger increasing returns. That is why those artists who may seem to need it the least (due to their notoriety) will benefit from the maximal coverage.

A related consequence of Adler’s (1985) model is that Superstars allow to erect barriers to prevent market entry. More precisely the most important players will try to acquire Superstars as they attract most of the attention – and most revenues. Such strategy does not prevent to resort to proliferation of novelties though. Having both Superstars and a proliferation of novelties allows to saturate attention while possibly benefitting from the surprise success of one of those novelties (Benghozi 2006).

While neither Rosen (1981) nor Adler (1985) could have envisaged the impact of the Internet on Superstars, it is possible to extrapolate their arguments to show that the Internet can reinforce Superstars (Brynjolfsson et al. 2010). With digitization, marginal costs for distribution and consumption are even more decreasing. As for information over Superstars, it becomes even more ubiquitous due to the Internet. Whereas the Theories of Superstars have been developed before the Internet era, the theory of the Long Tail derives directly from this technological revolution.

B – The Long Tail and its consequences

Anderson coined the Long Tail to predict that digital technology will allow consumption to become much more diverse (Anderson 2006). The Long Tail consists in two trends: (i) the decreasing importance of Superstars in relative or even absolute terms (e.g. respectively the decrease of their market shares or of their sales volumes, see also 4.5); (ii) the increase of the Tail, i.e. the increase of niche products. The latter idea is also argued for by Brynjolfsson et al. (2003) when they discuss the importance of obscure works in online sales.

There are various reasons for such trends to take place. The Long Tail can emerge first thanks to a democratization of production means, for different types of content (Anderson 2006). Personal Computer and more recently mobile devices have been instrumental in such a trend. Second, there is also a reduction in costs to access content, notably thanks to the Internet (Anderson 2006). Actually, in the offline world, space (or time e.g. for broadcasting) must be reserved to best-sellers rather than left to works that take as much space but sell less (Anderson 2006; Brynjolfsson et al. 2003). In other words, supplied content diversity is broader online than offline as physical space restrictions and logistics are reduced (Le Lec et al. 2015). Third, digitization allows to group enough consumers to create market niches of a sufficient size (Anderson 2006). Finally, relevant filters exist that help consumers find what is likely to please them in spite of the abundant supply.

According to Anderson (2006), the development of the Long Tail benefits consumers and platforms. Consumers benefit from the Long Tail through this much larger choice. Brynjolfsson et al. (2003) thus assess that in the US book market, consumers have benefited more from the increase in supplied variety than in price reduction. The Long Tail also leads to the constant emergence of new services relying on innovative business models (Masnick & Ho 2012). Such services or new activities are at all steps in the value chain from creation to distribution.

The greatest beneficiaries, however, are those companies that give consumers access to a great variety of goods or services (Anderson 2006; Brynjolfsson et al. 2003). Such companies are generally platforms that act as intermediaries between different types of users, e.g. for Amazon between consumers and third-party retailers who sell on its platform. Furthermore, they benefit from economies of scope, i.e. the marginal costs of adding content to their catalogue is very low. This also incites them to increase the diversity of their offer. Another advantage of such a strategy is that a diversified catalogue allows companies to reduce risks, alike with financial assets (Markowitz 1952). Finally, it is also a way to reduce competition since saturation of the market allows erecting barriers to entry (Lancaster 1979 & Schmalensee 1978). Therefore, the impact on traditional intermediaries (bookstores, record producers, etc.) is ambiguous, as is discussed in III.F. Finally, while Anderson (2006) predicts a positive impact on creators, it should be on average of limited scale. Marcone (2010) does not predict major changes for independent creators since the Tail is not developing fast enough.

C – Is there a Long Tail?

The following table provides an overview of studies that have tested the existence of a Long Tail in the cultural sector, mostly in content industries, although there are a few exceptions (e.g. Brynjolfsson et al. 2011). Also, only empirical papers were kept, therefore papers relying on experiments were excluded (e.g. Le Lec et al. 2015). Most papers are based on US data, several others on French data. All papers focus on Variety (V) and Balance (B), and only Bourreau et al. (2011) take disparity (D) into account. Finally, there is no unique trend towards either reinforcement of Superstars or of the Long Tail. One reason might be the type of index used to assess diversity. We come back to both issues in III.E.

Table: Overview of empirical studies on the Long Tail

Article

Sectors

Country

Offered / Consu-med

Stirling Defini-tion

Long Tail Effe-cts?

Results

Anderson (2006)

music, video, book

US

S&C

V, B

+

LT has an increasing share of sales (40% on Rhapsody, 21% on Netflix, 25% on Amazon)

Bear Sterns (2007)

TV

US

C

B

+

Decrease of Big 3 in Total Day TV Viewing

Benghozi (2008)

DVD, CD

France

C

V, B

+

DVD sales are concentrated online. LT effect stronger for CDs

Benghozi & Benhamou (2008)

CD, DVD

France

S&C

V, B

+

LT effect although less strong during periods when sales are the highest

Bourreau et al. (2011)

Recorded music

France

S&C

V, B, D

+

Increased consumed variety. Weight of top 100 and 100 is lower online, as well as Hirschman Herfindahl Index

Brynjolfsson et al. (2011)

feminine clothes

US

C

B

+

More balanced sales for Internet purchases compared to catalog purchases

Elberse (2008)

music

US

S&C

V, B

+/-

Longer Tail but online transactions are even more concentrated, e.g. 1% of titles available on Rhapsody = 32% of titles listened to

Elberse & Oberholzer-Gee (2006)

video

US

S&C

V, B

+/-

Longer Tail. More videos that do not sell at all. Growing importance of best sellers

Given & McCutcheon (2014)

DVD, books

Australia

S&C

V, B

+/-

LT if absolute values. But more concentration if relative values

Goel et al. (2010)

movies, music, Web search & browsing

n/a

C

B

+

Nearly everyone is at least a bit eccentric (the paper deals more with consumer satisfaction)

Hinz et al. (2011)

VOD (transactional)

Germany

S&C

V, B

+/-

A growing assortment leads to greater demand per customer, although on a diminishing scale. A growing assortment size does not necessarily lead to the end of the “blockbuster era”. Strong influence of search technologies on the demand distribution. Niche demand is mainly generated by heavy users

Kumar et al. (2011)

motion picture

US

C

B

+

Shift towards niche titles

Leskovec et al. (2007)

book, DVD

US

C

B

+

Top 100 = 11,4%. Top 1000 = 27%. 67% of all products have a single purchase. They account for 30% of recommendations. The tail is a bit longer (the paper deals with recommendations, not with sales)

Marcone (2010)

music (Billboard data)

US

C

B

Online sales have become more concentrated, and the hits matter more each year. Overall album sales have fallen over 30% since 2004, and popular album sales have faired even worse than overall album sales

Moreau & Peltier (2011)

books

France

S&C

V, B

+

Sales have decreased for top 500 and all more restrictive tops. They have increased for all other categories. Increase in the number of titles

Mulligan (2014)

Music (on- and offline)

US

S&C

V, B

In 2013 the top 1% of repertoire accounted for 77% of all artist recorded music income. Bias of digital platforms towards the top 1% (streaming, download, but also radio). Weakest concentration for physical music. This does not correspond to a dominance of majors over independents

Page & Garland (2009)

Music

UK

C

B

Volume (legal sales or ‘pirate’ swaps) is concentrated amongst a small proportion of the available tracks. The gap between hits and niches is widening

Peltier & Moreau (2012)

book

France

C

B

+

Bestsellers got smaller market shares online than offline, contrary to medium- and low-sellers. Both online and offline sales shift from the head of the distribution to the tail with increasing magnitude. The LT appears to be more than just a short-lived phenomenon caused by the specific preferences of early adopters of e-commerce

Smyrnaios et al. (2010)

online news

FR (French-speaking)

S

V, B

French-speaking news websites have quite similar characteristics to those of traditional media. News appears to be both varied and very unevenly distributed

Tan et al. (2015)

movie rental

US

S&C

V, B

Product variety is likely to increase demand concentration. Increasing product variety diversifies the demand away from each movie title, but less significantly for hits than for niche products

Walls (2010)

DVD

North America

C

B

+

The DVD market is less heavy-tailed and exhibits less of a winner-take-all property than the theatrical market for motion-pictures

III – Long Tail vs Superstars: beyond the dichotomy

The Theory of the Long Tail and the Theories of Superstars provide quite opposite predictions regarding the impact of digital technology on the diversity of cultural expressions. Before analysing the results of recent empirical research, this section reviews the common points in these theories. This allows us to emphasize the point made by Brynjolfsson et al. (2010) that these theories can and should be analysed as part of an integrated research agenda on the impact of digital technology on online product diversity.

A – An increase of supplied diversity

The Theory of the Long Tail and Theories of Superstars all take as a starting point the supplied diversity of products. This is an important aspect of the Theories of Superstars that they explain the discrepancy between supplied diversity on one side and consumption concentration on the other side.

This increase of supplied diversity results from the observations already made by the Theory of the Long Tail (see 3.2). Following Stirling’s (2007) definition, it is easy to see that Variety has greatly increased. Digital technologies and in particular the Internet have vastly expanded the variety of products that can be profitably made available (Brynjolfsson et al. 2010), leading to a dramatic increase in assortment sizes (Hinz et al. 2011). This may include content that was not supposed to be made available in a permanent way, as used to be the case for most TV content.

Supplied disparity is also arguably increasing with consumers being given access to content from all over the world, traditional or resolutely modern. Smyrnaios et al. (2010) give the argument in the case of newspapers that the high distribution costs in offline markets are particularly detrimental to marginal newspapers and magazines that have low sales are disadvantaged in such a system. Therefore, the lower costs of distribution in online markets allow for such newspapers and magazines to be present, with their more differentiated content. However, it is difficult to assess whether online supply is more balanced.

While there is an agreement on the increase of supplied diversity, there are opposite predictions regarding the impact on consumed diversity.

B – The question of whether consumers as a whole like diversity

The two theories are first opposed in terms of whether consumers as a whole value diversity. This itself corresponds to the fact that they have diverse preferences or that every consumer values diversity (Ranaivoson 2012). In theories of Superstars, consumers do not value diversity. On the contrary, according to Anderson (2006), the Internet allows consumers to realize that as a whole they like diversity more than they expect.

Theories of Superstars do not consider that consumers as a whole value diversity. In the case of Rosen’s (1981) theory, this is already noticed by Schulze who regrets the absence of “heterogeneous tastes or a love of variety [that are] an important limitation to star power” (Schulze 2003: 432). In fact, Rosen (1981) evokes the assumption of a taste for diversity but thinks it would not change much to the model’s results. On the other hand, Adler (1985) evokes the existence of niches, and therefore recognizes that consumers may have different tastes – although not necessarily that each consumer may like diversity. This vision of a lack of taste for diversity is crudely expressed by Mulligan (2014) when he compares consumers to sheep who need herding and to be led by the hand, be it offline or online.

Anderson (2006) agrees there are reasons for the Internet to further promote Superstars, in particular through word-of-mouth that induces positive retroactions. However, he thinks such retroactions will take place at the level of market niches, e.g. music genres. Tan et al. (2015) propose that a larger product variety may satisfy heterogeneous consumers’ increasingly varying tastes. From an economic point of view, the Long Tail assumes an underexploited spectrum of customer tastes that has not been addressed sufficiently or cost-effectively by pre-Internet retailers (Hinz et al. 2011).

Actually, Anderson (2006) believes individuals want more than only Superstars. Le Lec et al. (2015) indeed confirm that, as individuals are offered a larger choice set containing a variety of products, their aggregated consumption mechanically evolves towards a less concentrated distribution. Goel et al. (2010) show the diversity of individual tastes, calling into question the conventional view that niche products appeal only to a minority of consumers. For example, consumers may be interested in consuming products that are not recent (Poirier 2010) and more generally not available in physical stores (Bourreau et al. 2011). Furthermore, this may be the case even more for heavy users, as Hinz et al. (2011) find in particular a shift in their demand from blockbusters to niches. This relates to assumptions developed in sociology around the idea that “omnivorous” consumers tend to be “voracious” too (Sullivan & Katz-Gerro 2006; for an overview, see Ranaivoson 2012).

C – The impact of technology as it drives costs down

There is an agreement between Anderson (2006) and Rosen (1981) regarding the importance of technology in driving marginal costs of distribution and consumption down. The Internet enables retailers and manufacturers to increase their assortment sizes (Hinz et al. 2011). Actually, it reduces distribution costs, in particular because there is no longer a risk for retailers to end up with cumbersome unsold articles (Bourreau & Labarthe-Piol 2003). In the same way, storage costs have decreased, also on consumer side (Tepper et al. 2007). In the case of news provision, in online markets, distribution costs of informational goods are very low compared to offline markets (Smyrnaios et al. 2010).

However, the consequences of such technological changes in terms of consumed diversity are quite opposite. For Rosen (1981), reduction in marginal costs of distribution makes it easier for Superstars to get an access to a larger audience. By creating nationally and globally interconnected markets, technology may create incentives for retailers and distributors to disproportionately promote Superstars (Brynjolfsson et al. 2010). It is also possible in some instances that diversity remains low because the costs of producing original content remains high, as Smyrnaios et al. (2010) explain for online news. On the contrary, for Anderson (2006), such reduction benefits above all the works in the Tail whose storage and distribution are made easier. Without digitization, storing and distributing these works would not be profitable (Brynjolfsson et al. 2010). Customer demand increases for the products that belong in these new, larger assortments (Hinz et al. 2011).

D – Accessing to, and computing, information

Anderson (2006) as well as Adler (1985) puts the acquisition of information at the core of their theory, again with opposite consequences. Indeed, digital technology provides access to a virtually unlimited amount of information. An increase in the diversity of available options makes it harder to cope with the related increase in the load of information as research in psychology shows (Ranaivoson 2012). A strategy for users can therefore be to limit the riskiness of their choice and opt for what they already know: Superstars. All the more so that there is even more information available online on these Superstars. A study on peer-to-peer networks of early 2000’s thus shows that consumption there focuses on Superstars because getting information has become more costly (Bourreau & Labarthe-Piol 2003). More recently, Mulligan (2014) has argued that size of the online music services’ catalogue are an inconvenience for their users. This is, he argues, notably due to the small amounts of visual display space digital services have compared to stores’ feet of window space and of front-of-store display space. Therefore, it is possible that the Internet reinforces the position of Superstars as reassuring landmarks for the users, thus transposing the advantages they already have in the offline world.

Anderson (2006) rather argues that the lower costs in acquiring information concern the products and creators for which it is more difficult to get information offline, i.e. those that belong to the Tail. Furthermore, they benefit from decentralised prescription and promotion, contrarily to more centralized traditional media.

The question is also about how digital technology allows us to be given access to information. In other words, this is the question of the filters set up and used, and how they can lead consumers to either the Superstars or the Tail. While the edition process may remain important to guide users (Poirier 2010), such new filters are required to enable online retailers to serve diverse tastes profitably in a context of increasingly big offers (Hinz et al. 2011). Their aim is to reduce customer search costs. According to Hinz et al. (2011), there is a lack of research on the effects of search technologies on individual demand or whether they might influence shares of purchased products. The existing literature suggests that lower search costs enable shoppers to find more products that better fit their preferences (Hinz et al. 2011) and may therefore lead to higher demand for niche products (Tan et al. 2015).

More generally Hinz et al. (2011) find that small changes in search technologies may have significant effects on the distribution of demand. According to them, search functionalities, such as additional filters, can lead to a shift in demand from blockbusters to niches; while systems based on recommendations may shift demand from niches to blockbusters (Hinz et al. 2011). The same way Tan et al. (2015) explain that selection-biased recommendation systems can reduce sales diversity because these systems tend to recommend products with sufficient historical data (and only hits have enough historical data).

E – The measurement of the Long Tail and its limitations

An explanation for the opposite results of the studies on the Long Tail may lie in the fact that different measures of the Long Tail can lead to seemingly contradictory outcomes (Tan et al. 2015).

Brynjolfsson et al. (2010) distinguish three ways to define and measure the Long Tail. First, the Absolute Long Tail measures changes in the absolute number of products sold, e.g. the measure of sales above an absolute cut-off of 100,000 titles (Brynjolfsson et al. 2003). This is also the approach most commonly followed by Anderson (2006). Second, the Relative Long Tail focuses on the relative share of sales above or below a certain rank, e.g. using the Gini coefficient.[2] According to Hinz et al. (2011), existing Long Tail research often relies on the Gini coefficient to assess whether such Long Tail exists. Elberse’s (2008) work belongs to this second category. Third, it is possible to compare the relative importance of the head versus the tail by looking at the value of the slope of the relationship between ordinal rank and cardinal sales[3] (Brynjolfsson et al. 2010). These approaches are not interchangeable, in particular leading to different results. Brynjolfsson et al. (2010) thus notice that the first approach usually leads to conclude that the Long Tail is important, whereas both other approaches lead to conclude that the Long Tail is not so important.

None of these approaches however takes disparity into account. Interestingly, this absence is pointed out by various scholars, although without any reference to the Stirling model. Smyrnaios et al. (2010) oppose the quantitative growth of online information circulation (i.e. variety) to the spectrum of social, political and economic issues covered (i.e. disparity). Brynjolfsson et al. (2010) suggest to apply distance metrics in product space to assess “true product variety” (i.e. disparity). In fact this simply reminds of the spatial model (Hotelling 1929).[4]

F – Considering the role of platforms

Finally, there is a need to go beyond the Long Tail by considering not only the impact of new online platforms on the diversity of cultural expressions but their industrial role, and how they reconfigure cultural sectors. Cultural sectors are getting increasingly organised as two-sided markets (Rochet & Tirole 2002), where new online companies play the role of platforms mediating between different categories of users (e.g. advertisers and readers). As such, they behave in a different way from traditional cultural industries since they have to take into account the interactions between their various categories of users (e.g. consumers, advertisers, app developers, etc.).

According to the proponents of the Theory of the Long Tail, platforms are incited to provide a huge diversity of products as it gives them a competitive advantage towards their competitors (Brynjolfsson et al. 2010). Indeed online platforms typically offer a large selection of niche products and provide the relevant filters to discover niche products, while their traditional competitors only aim at the Head (Brynjolfsson et al. 2010).

However, while platforms may provide a huge diversity of cultural content, they tend to become more than actors in value chains, with an impact difficult to assess on consumption choices (Kulesz 2015). Goel et al. (2010) believe that there is a risk that the online platforms’ increasing control over access to cultural works may threaten the visibility and promotion of marginal cultural works even compared to the current situation.

It should also be investigated how offline and online markets interact, in particular for traditional cultural players who have now entered into online markets by providing their own platforms. To our knowledge, only the research conducted by Doyle and Champion (2014) has provided a serious, evidence-based attempt to compare media companies’ strategies and their impact on content diversity across platforms.

Conclusion

The paper has aimed at showing that there is already an interesting stream of literature that can be used to address the impact of online platforms on the diversity of cultural expressions. This literature has built around the Theory of the Long Tail and more or less implicitly around its opposite Theories of Superstars. The existence of the Long Tail in cultural sectors, or of a trend towards it, is somehow dependent on the indexes used to assess it. Nevertheless, this literature has important findings and raises interesting questions.

First, there is a common agreement on the increase of the supplied diversity of cultural expressions. Digital technologies make it cheaper to produce and distribute content, thus also democratizing tools for more and more citizens all over the world.

The impact on the consumed diversity of cultural expressions is at the core of the debate between proponents of, and opponents to, the Theory of the Long Tail. It depends on whether consumers as a whole value diversity. Technology also plays a role. First because it reduces costs. But more ambiguously because the online production and distribution of cultural content requires the set-up of filters, some of which did not exist in the pre-Internet era (recommendation systems, search engines, social networks, etc.). And each filter may have a different way of orientating consumers towards blockbusters or more obscure works.

The Theory of the Long Tail and the Theories of Superstars, however, do not take the issue in its entirety. First, they do not take disparity into account, for example this component of diversity is not assessed in most studies on the Long Tail. Second, they miss the industrial view of cultural sectors. Actually, the surge of online platforms not only change the way citizens have access to cultural works, but they dramatically modify the relationships between players in the cultural sectors.

A renewed approach is therefore needed that could use the Theory of the Long Tail as a starting point but needs to consider the width of choices available to citizens and how it impacts their consumption decisions (disparity); and the relationships between the diversity of cultural expressions and industrial reconfigurations now taking place. This notably to see whether the development of the Long Tail could be a sustainable process. It needs then to be seen how this could feed into the political process, in particular in relation to the implementation of the UNESCO 2005 Convention.

References

Adler, M. (1985) ‘Stardom and Talent’, Am. Econ. Rev, 75: 208–212.

Adler, M. (2006) Chapter 25 ‘Stardom and Talent’, in V. A. Ginsburgh and D. Throsby (eds.) Handbook of the Economics of Art and Culture, Elsevier: 895-906.

Anderson, C. (2006) The Long Tail. How endless choice is creating unlimited demand, London: Random House Business Books.

Ballon, P., Bleyen, V.-A., Donders, K. & Lindmark, S. (2012) ‘The double “layered” platform structure of the audiovisual media industry: A case study of Flanders’, 23rd European Regional Conference of the International Telecommunication Society, Vienna, Austria.

Bear Stearns (2007) ‘Entertainment Industry. A Longer Look at the Long Tail’, Bear, Stears & Co. Inc.

Benghozi, P.-J. (2006) ‘Mutations et articulations contemporaines des industries culturelles’, in Création et Diversité Au Miroir Des Industries Culturelles, Paris: La Documentation Française.

Benghozi, P.-J. (2008) Effet Long Tail ou effet podium : une analyse empirique des ventes de produits culturels en France, Pôle de Recherche en Economie et gestion, Ecole Polytechnique, CNRS.

Benghozi, P.-J. & Benhamou, F. (2008) ‘Longue traîne: levier numérique de la diversité culturelle?‘, Cult. Prospect, 1–11.

Benhamou, F. (2012) L’Economie du star-system, Paris: Odile Jacob, .

Bourreau, M. & Labarthe-Piol, B. (2003) Un point sur musique et Internet, mimeo, ENST Paris et Université de Paris-Dauphine.

Bourreau, M., Moreau, F. & Senellart, P. (2011) ‘La diversité culturelle dans l’industrie de la musique enregistrée en France (2003-2008)‘, Cult. Études, 5: 1–16.

Brynjolfsson, E., Hu, Y. (Jeffrey) & Simester, D. (2011) ‘Goodbye Pareto Principle, Hello Long Tail: The Effect of Search Costs on the Concentration of Product Sales’, Manag. Sci, 57(8): 1373-1386.

Brynjolfsson, E., Hu, Y. & Smith, M.D. (2003) ‘Consumer surplus in the digital economy: Estimating the value of increased product variety at online booksellers’, Manag. Sci., 49: 1580–1596.

Brynjolfsson, E., Hu, Y. (Jeffrey) & Smith, M.D. (2010) ‘Research Commentary— Long Tails vs. Superstars: The Effect of Information Technology on Product Variety and Sales Concentration Patterns’, Inf. Syst. Res., 21: 736–747.

CEIM (2015) Pour une culture en réseaux diversifiée. Appliquer la Convention sur la protection et la promotion de la diversité des expressions culturelles (CDEC) à l’ère du numérique, Centre d’études sur l’intégration et la mondialisation (Université du Québec à Montréal). <http://www.ieim.uqam.ca/spip.php?page=article-ceim&id_article=9809&lang=fr> (accessed 06 October 2016).

Doyle, G. & Champion, K. (2014) ‘Content diversity in the age of multi-platform delivery’, presenté au Seminario internacional El sector audiovisual en la era digital: Políticas y estrategias para la diversidad, Universidad Carlos III de Madrid (UC3M), 5-7 November.

Elberse, A. & Oberholzer-Gee, F. (2007) ‘Superstars and Underdogs: An Examination of the Long Tail Phenomenon in Video Sales’, MSI Reports: Working Paper Series, 4:49-72.

Elberse, A. (2008) ‘Should You Invest in the Long Tail?’, Harv. Bus. Rev, July-August. <https://hbr.org/2008/07/should-you-invest-in-the-long-tail> (accessed 06 October 2016).

Given, J. & McCutcheon, M. (2014) ‘Heads or Tails?’ [WWW Document], Inside Story, 7 May. <http://insidestory.org.au/heads-or-tails/> (accessed 06 October 2016).

Goel, S., Broder, A., Gabrilovich, E. & Pang, B. (2010) ‘Anatomy of the Long Tail: ordinary people with extraordinary tastes’, in Proceedings of the Third ACM International Conference on Web Search and Data Mining, ACM, New York, USA, February 04-06, 2010: 201-210.

Guèvremont, V., Bernier, I., Burri, M., Cornu, M., Richieri Hanania, L. & Ruiz Fabri, H. (2013) La mise en oeuvre de la Convention sur la Protection et la Promotion de la Diversité des Expressions Culturelles à l’Ere Numérique : Enjeux, Actions Prioritaires et Recommandation, Rapport présenté au Comité intergouvernemental de la Convention sur la protection et la promotion de la diversité des expressions culturelles Septième session ordinaire, Paris, 10-13 décembre 2013. <https://goo.gl/lpd1FM> (accessed 06 October 2016).

Hinz, O., Eckert, J. & Skiera, B. (2011) ‘Drivers of the Long Tail phenomenon: an empirical analysis’, J. Manag. Inf. Syst., 27: 43–70.

Hotelling, H. (1929) ‘Stability in Competition’, The Economic Journal, 39: 41–57.

Kretschmer, M. (2005) ‘Artists’ earnings and copyright: A review of British and German music industry data in the context of digital technologies’, First Monday 10 (1). <http://firstmonday.org/article/view/1200/1120> (accessed 06 October 2016).

Kulesz, O. (2014) ‘UNESCO Convention on the Protection and Promotion of the Diversity of Cultural Expressions: Transversal analysis of Parties’ periodic reports on digital issues and trends’, CE/14/8.IGC/INF.5, Paris, 18 novembre 2014. <http://fr.unesco.org/creativity/sites/creativity/files/8%20IGC-INF%205_Digital_FR.pdf> (accessed 06 October 2016).

Kumar, A., Smith, M.D. & Telang, R. (2011) ‘Information discovery and the Long Tail of motion picture content’, SSRN, June 23. <http://dx.doi.org/10.2139/ssrn.1871090> (accessed 06 October 2016).

Lancaster, K. (1979) Variety, Equity and Efficiency: Product Variety in an Industrial Society, New York: Columbia University Press.

Le Lec, F., Lumeau, M. & Tarroux, B. (2015) ‘Information, choice and consumption: an experimental study of the Long Tail effect’, Ecole d’été numérique (3EN), 6 et 7 juillet 2015, Rennes.

Leskovec, J., McGlohon, M., Faloutsos, C., Glance, N.S. & Hurst, M. (2007) ‘Patterns of Cascading behavior in large blog graphs’, Proceedings of the 2007 SIAM International Conference on Data Mining: 551-556.

Marcone, S. (2010) ‘Is the Long Tail really wagging?’, MEIEA J., 10: 99–105.

Markowitz, H.M. (1952) ‘Portfolio selection’, J. Finance, VII: 77–91.

Masnick, M. & Ho, M. (2012) ‘The Sky Is Rising. A Detailed Look at the State of the Entertainment Industry’, Floor 64. <https://www.techdirt.com/skyisrising/> (accessed 06 October 2016).

Moreau, F. & Peltier, S. (2011) ‘La diversité culturelle dans l’industrie du livre en France (2003-2007)’, Cult. Études, 1–16.

Moureau, N. (2006) Société de l’information et modèles de star system, Hermès, 44: 183–189.

Mulligan, M. (2014) The Death of the Long Tail. The Superstar Music Economy, MIDiA Consulting, March.

Page, W. & Garland, E. (2009) The Long Tail of P2P. (No. Issue 14), Economic Insight.

Peltier, S. & Moreau, F. (2012) ‘Internet and the “Long Tail versus superstar effect” debate: evidence from the French book market’, Appl. Econ. Lett., 19: 711–715.

Poirier, C. (2010) Panel 3 – L’économie de la culture: modèles qui perdurent ou changement de paradigme?, Observatoire de la culture et des communications du Québec, 25 October 2010.

Ranaivoson, H. (2007) Measuring cultural diversity: a review of existing definitions, UNESCO, Paris. <http://www.uis.unesco.org/culture/Documents/Ranaivoson.pdf> (accessed 06 October 2016).

Ranaivoson, H. (2012) ‘Does the consumer value diversity? How the economists’ standard hypothesis is being challenged’, in The Consumption of Culture, The Culture of Consumption. A Collection of Contributions on Cultural Consumption and Cultural Markets, Lambert Academic Publishing: 70-95.

Rochet, J.-C. & Tirole, J. (2002) ‘Cooperation among competitors: Some economics of payment card associations’, RAND J. Econ., 33: 549–570.

Rosen, S. (1981) ‘The Economics of Superstars’, Am. Econ. Rev., 71: 845–858.

Schmalensee, R. (1978) ‘Entry deterrence in the ready-to-eat breakfast cereal industry’, Bell J. Econ., 9: 305–327.

Schulze, G.G. (2003) ‘Superstars’, in: R. Towse (ed.) A Handbook of Cultural Economics. Cheltenham : Edward Elgar Publishing Limited: 431-436.

Simon, J.P. & Bogdanowicz, M. (2012) ‘The Digital Shift in the Media and Content Industries’, JRC Scientific and Policy Reports.

Smyrnaios, N., Marty, E. & Rebillard, F. (2010) ‘Does the Long Tail apply to online news? A quantitative study of French-speaking news websites’, New Media Soc., 12 : 1244–1261.

Stigler, G.J. & Becker, G.S. (1977) ‘De Gustibus Non Est Disputandum’, Am. Econ. Rev., 67: 76–90.

Stirling, A. (2007) ‘A general framework for analysing diversity in science, technology and society’, J. R. Soc. Interface, 4: 707–719.

Sullivan, O. & Katz-Gerro, T. (2006) ‘The Omnivore Thesis Revisited: Voracious Cultural Consumers’, Eur. Sociol. Rev., 23: 123–137.

Tan, T., Netessine, S. & Hitt, L.M. (2015) Plenty is no plague, or is it? an empirical study of the impact of product variety on demand concentration, Faculty and Research Working Paper, Alliance Center for Global Research and Education, Insead & Wharton. <https://sites.insead.edu/facultyresearch/research/doc.cfm?did=51405> (accessed 06 October 2016).

Tepper, S., Hargittai, E. & Touve, D. (2007) ‘Music, Mavens and Technology’, in: S.J Tepper, B. Ivey (eds) Engaging Art: The Next Great Transformation of America’s Cultural Life, New York: Routledge: 171-193.

Thiec, Y. (2014) ‘From a moratorium in the conflict between trade and culture to the creation of a digital deal for cultural diversity’, in L. Richieri Hanania (ed.) Cultural Diversity in International Law: The Effectiveness of the UNESCO Convention on the Protection and Promotion of the Diversity of Cultural Expressions, London: Routledge: 95-106.

Van Cuilenburg, J. & Van der Wurff, R. (2001) ‘Impact of moderate and ruinous competition on diversity: The Dutch television market’, J. Media Econ., 14: 213–229.

Walls, W.D. (2010) ‘Superstars and heavy tails in recorded entertainment: empirical analysis of the market for DVDs’, J. Cult. Econ., 34: 261–279.

Zhu, F. & Seamans, R. (2010) ‘Technology Shocks in Multi-Sided Markets: The Impact of Craigslist on Local Newspapers’, EconPapers, Net Institute, Working Paper #10-11. <http://econpapers.repec.org/paper/netwpaper/1011.htm> (accessed 06 October 2016).


  1. Heritiana Ranaivoson is Senior Researcher and Project Leader at iMinds-SMIT since 2010. He has been a researcher for more than 10 years, and has led several projects at international, national and local levels, funded by public (e.g. European Commission, UNESCO) or private (e.g. Google) organizations. His main research focus are cultural diversity, media innovation and the economic impact of digital technology on cultural industries. Before joining iMinds, he was associate researcher at Cerna, the Centre for Industrial Economics at Mines ParisTech (2008-2010). He holds a MSc in Economics and Management from the Ecole Normale Supérieure de Cachan (ENS Cachan) and a PhD in Industrial Economics from University Paris 1, Panthéon-Sorbonne.
  2. The Gini coefficient measures the inequality among values of a frequency distribution, e.g. the inequality in income distribution. It varies between 0 (perfect equality) and 1 (maximal inequality). In this context, it may be used to assess sales concentration.
  3. To illustrate this, if one wants to compare inequalities in sales for two retailers, one will get two curves representing how sales are spread for each retailer. The third way consists in comparing the slopes of both curves. This assumes that the relationship follows a Power Law distribution.
  4. Hotelling’s model assumes that a line can be used to represent one characteristic of a product. For example, he applies it to cider: the left side would correspond to sweetness and the right side to sourness. Going to the left would mean providing a sweeter cider, to the right a sourer cider. It is then also possible to position consumers based on their tastes (here for more or less sweet cider).


Leave a comment