WHAT HAS THE ECONOMIC LITERATURE PRODUCED SO FAR, AND WHAT DATA ISSUES REMAIN AS OBSTACLES?

The empirical literature on copyright is admittedly relatively nascent. Economists have thus far largely focused their empirical work on the economics of the patent system.

Indeed, even independent of the impacts of new technologies and models as described in section 11.2, the basic effects of copyright on sector outcomes remain largely untested.

As outlined in Towse (2011) and Handke (2011) as well as in Chapter 7 of this volume, all excellent reviews of the empirical work on copyright, the literature on copyright as an economic incentive and hence as a motivator for content creators is not particularly rich.¹⁶ Little empirical evidence is available concerning the effect on economic incentives of the current or proposed duration of rights, with some quarters believing it is too long and others thinking it too short (Akerlof et al., 2002). While economists criticize the ‘one-size-fits-all’ aspect of copyright that does not reflect the differences between crea­tive sectors, no concrete recommendations on more appropriate copyright length(s) or sector-specific variations of copyright terms have been proposed. A few economists have made some headway in estimating the earnings of artists from copyright, which is signifi­cant for answering questions about the importance of the incentives it offers to creators (Kretschmer, 2002; Watt, 2000, 2004). Yet, more work on the creators’ remuneration is needed. Likewise, few attempts have been made to estimate the costs of copyright to users or final consumers in terms of higher prices or reduced accessibility. Studies on the transaction costs of searching and clearing the rights to protected materials and related entry barriers are scarce, as discussed in more detail below.

How then has the economic literature reacted to the role of copyright in an era of digi­tization and electronic networks? In a somewhat unfortunate turn of events, the majority of empirical economic studies since the advent of the Internet have focused on the effects of unauthorized downloading of creative works on the sales of the creative industries, with a particular focus on music and more recently films (OECD, 2005; Png, 2006; Liebowitz, 2007; Handke, 2010). In this strand of research the focus is almost exclusively on creative industry revenues and not on the individual creator or his or her ability to make a living, or on the actual production of creative works, the nurturing of new talent, or innovation in the process of creativity.

This new ‘peer-to-peer download and copyright’ economics literature has produced some tentative but ambiguous results and two camps of economists: those who assert that unauthorized file-sharing has substantially decreased music industry revenues and those who argue the opposite (for reviews see OECD, 2005; Fink et al., 2010; Handke, 2011; Towse, 2011). In terms of quantity, the largest body of work confirms the nega­tive effect of unauthorized downloads or streaming on music sales (i.e., displacement is taking place). At the same time, the literature also shows that the effect is not a one-to- one replacement of unauthorized downloads for purchased songs; some downloads do not harm music industry revenues, and some indeed have a sampling and tasting effect. In other words, free access to music via unauthorized downloading creates follow-on music purchases or offsetting increases in demand for complementary income streams such as concerts (Oberholzer-Gee and Strumpf, 2009).

A significant number of methodological issues are also at stake, such as the difficulty in causally linking declining industry revenues to increased peer-to-peer network activity and determining whether lower music industry revenues due to falling CD sales are or are not made up for by increased concert revenues.

Transposing what happens in the field of music to other creative sectors is also a stretch as the lessons learnt could be sector specific.¹⁷ With technology changing swiftly the ques­tion is whether throwing darts at this moving target, that is, conducting research on it, is a worthwhile pursuit. More fundamentally, and stepping outside of the peer-to-peer literature, the effects of the new digital set-up on creative supply - and thus the long-term sustainability of this digital ecosystem - have hardly been assessed from a solid empirical point of view (Wunsch-Vincent, 2010a, 2010b). However, this knowledge gap needs to be closed to better inform policy-making.

Based on these observations, a large research task lies ahead of economists when it comes to assessing copyright law in the context of digitization and the Internet. What are the obstacles that need to be overcome to further this research agenda on copyright in the digital age? This chapter concludes by pointing out five main data challenges on our way to assessing the motivational/incentive function of copyright and the main costs and benefits relating to copyright in the digital age.¹⁸

First, reliable statistics are missing on the number of copyrighted works and hence the supply of creative works that is meant to be positively stimulated by copyright law remains unknown. As suggested earlier, no formalities are involved when obtaining copyright; the recognition of copyright ownership is automatic.¹⁹ It is true that some countries can oblige or give procedural advantages to national creators who register their work,²⁰ yet, at the moment, the data available in a few jurisdictions are generally not used for economic analysis, since they are often not accessible in the right format or the data themselves might not be representative.

An alternative route to estimating the number of existing original works could be to obtain unit counts of the number of creative works produced in a country or in a given subsector. In reality, data on units of creative works produced (whether sold or not) are also hard to obtain. For instance, internationally, it is nearly impossible to identify the number of books or the number of songs written in a given country in a valid and inter­nationally comparable fashion.²¹

This problem is compounded in the Internet age, where no central entity exists that could monitor and report the production of creative works. Indeed, statisticians and economists have been struggling with trying to assess the true extent of user-created content online (OECD, 2007). Using simple count data (e.g., the number of online videos uploaded on popular video platforms) is misleading. Amateur videos or uploads to online video platforms are not always original creative works; some videos are unau­thorized copies of television series or other protected works. Likewise, amateurs create or remix content over various decentralized and sometimes overlapping online platforms, causing problems of double-counting uploads. Also, new genres of content emerge that are unaccounted for, and thus difficult to categorize and count (Bruegge, 2011; INSEAD and WIPO, 2012). Finally, data on the uptake and intensity of the use of streaming serv­ices, the number of works accessed and revenues generated are unavailable.

Second, the quality and ‘value’ of copyrighted works are hard to objectively assess, in particular when one moves beyond economic value for the content industry alone, for instance if one is required to assess the economic value for the creator or the artistic value for society at large. A Hollywood movie that cost several million US dollars to produce, a Czech movie shot on a shoestring budget, or a home-made video put on an online platform might generate similar copyright entitlements.

Third, data are missing on the revenues generated on the basis of copyright and the respective distribution of these revenues between creators, the creative industries, and other intermediaries. Admittedly, private sector associations have made tremendous progress in accounting for sales revenues generated by particular content sectors. The International Federation of the Phonographic Industry (IFPI), for instance, has been making available detailed data on sales - and more recently online sales and subscription revenues.²²

International organizations are measuring the economic contribution of the copyright­based industries, including their value-added and their contribution to international trade. WIPO’s Creative Industries Division undertakes activities to better conceptual­ize and measure the creative industries, and measures their contribution to economic growth, trade and development based on the methodology outlined in the 2003 WIPO Guide on Surveying the Economic Contribution of the Copyright-Based Industries.²³

Obtaining a clear picture from these figures on who earns what (i.e., what the creator or performer, the creative industries, other intermediaries such as online music stores and online video platforms earn from a given copyrighted work), and hence deducing some­thing on the incentive effect of copyright, is hard. Indeed, determining the revenue of artists is challenging, as some is based on contractual income negotiated with the content industry, some earnings are based on collectively negotiated contracts from CMOs, and some income is based on the exploitation of copyright (such as income from concerts, readings, and, to use legal terms, performing, mechanical and other rights).

Similarly, in the online context revenues are not generated from the sale of songs via music platforms alone. Watching a video or listening to a song on a streaming service, like YouTube or Spotify (be it based on a subscription payment or supported by advertis­ing) will technically generate a stream of revenue that is hard to elucidate in the current data context (Lanteri, 2013). Online videos often do not generate direct revenues for their creator, and if they do, the money generated does not show in traditional content industry revenue statistics, but rather in data such as revenues reported by CMOs. Certain music subscription services pay the rights holders in advance (advance payment) before any music is streamed and consumed, part of which - in principle - reaches the artist.

Conducting research on the artist and industry split has been complex in the past, and determining the revenue splits between creators, various industries and distributors requires true detective work.²⁴ Each artist’s or creator’s contract is different, and indi­vidual cases gleaned from private contracts - be it for exploitation of the work online or offline or on all media - cannot be generalized. In the case of films or online multiplayer games, rights are owned across the board and payments are not associated with rewards for owning the copyright.

Importantly, more detailed information about the cost structure of different content industries and creators is required to assess the motivational and incentive effects of copyright on the revenue generated (Merrill and Raduchel, 2013). As outlined earlier, the costs of the creative sector can be high at times. Artists rely on the financial and organiza­tional capacity of the content industry to leverage the risk of the sector. The high cost of content production and delivery is particularly timely with respect to the online distribu­tion of content. While earlier assumptions were of a costless, immediate and free global distribution in the digital age, reality has shown that many new costs have emerged (e.g., costs related to digital rights management, online payments, revenue shares paid to online platforms, and so forth).

Interestingly, artists and the creative sector as such also have non-monetary incentives that have to be factored in. While one could see non-monetary motives as a competing incentive to copyright, they are indeed often linked: for example, when a creator foregoes remuneration for the viewing of their works or their particular creative contribution, they might rely on attribution, and thus moral rights to build their reputation, generate other revenue streams or achieve non-monetary rewards.

Fourth, and related to the point on costs, few credible data are available on the admin­istrative and transaction costs related to copyright, and the ways in which the system facilitates or creates barriers. Arguably the transaction costs have to be set against the potential incentive effect of copyright across the economy, which is difficult. One would want to assess the relative costs and benefits of various arrangements for managing transaction costs (Merrill and Raduchel, 2013). Moreover, one should assess the existing and potential effect of new technologies on current licensing procedures and collective rights management institutions.

At the moment the data situation does not allow this easily. A few studies and policy reports have stressed the difficulty of securing rights across the board, and one study in the European context has even put a figure on it, namely that services available in several countries and that offer more than one million titles can face transaction costs of several hundred thousand euros, requiring significant human resources and time (KEA and Vrije Universiteit Brussel, 2012). Any assessment of the administrative effect also has to factor in the economic and other implications of the exceptions and limitations that copyright law foresees. Technically, the latter decreases the costs of copyright as an access barrier. It is precisely the function of exceptions and limitations to make works accessible for various purposes such as parody, education, and other functions.

Finally, data on the pricing and the consumption of creative works would be needed to assess the access constraints potentially imposed by copyrights. These data do not exist beyond broad industry revenue aggregates or have not yet been used in solid economic studies.²⁵

11.5