Chapter 1 Empirical Measurement Challenges
It is impossible to understand music streaming earnings outside the context of the music market or the music business ecosystem. Music streaming royalty rates and distribution methods largely follow long-established practices and regulations concerning valuing intellectual property, accounting for royalties, and regulating the administration of copyrights and neighbouring rights. Music streaming competes with other uses, such as public performances in television and radio broadcasting, or digital downloads. Music streaming cannot be analysed in isolation from the volume and price trends of public performance, mechanical licensing, and home copying.
So, what is the correct market to analyse? Who are the listeners (users) of music, and who are the sellers of records and subscriptions who supply them with music? In competition law, the market under consideration is the relevant market.
The concept of the relevant market is in practice the most litigated aspect of competition law. Sometimes it is suggested that the application of the competition law is determined at the moment when the market is defined: should the geographical or (substitute) product element be too narrowly defined, and almost all suppliers will fail the SSNIP test. If the definition is too wide, almost no market activity will be illegal. According to the European Commission Notice, which has been the most important document in the creation of the UK Market definition guidelines, “the definition of the relevant market in both its product and its geographic dimensions often has a decisive influence on the assessment of a competition case.” (EUR-Lex 1997; Office of Fair Trading 2004). It is also a widely held view in the U.S. (Pitofsky 1990; Baker 2007)
The practice of defining the relevant market is often connected to some assumption on market power, and eventually the dispute of the relevant market may become circuitous: the best definition that yields market power the best. Some academics have gone as far as suggesting the abolition of the relevant market concept from competition law (Kaplow 2010).
Our paper does not aim to add arguments to this debate. We take for granted that the UK, EU, US, and major competition jurisdiction apply the concept of relevant markets in a similar way. However, we take on some arguments from the critical literature to highlight some difficulties when we want to make sense of music streaming markets.
For the purposes at hand, we examine late 20th-century studies on the role of economic modelling. The critique of the relevant market doctrine–the SSNIP test applied by the UK and EU competition authorities–is in fact a critique of the application of market model, not a critique of the market itself. When economists talk about markets, they do not necessarily talk about a piece of land in Camden Town where people can buy used records from stalls after a bit of haggling. They are using this “marketplace” as a metaphor. The “relevant market” is a metaphoric place that works as if sellers and buyers could meet and agree on transactions. It assumes people are trading securities from their home computers worldwide and a server clears the transaction which is physically located in Northern Sweden to keep the processors cool but under the trading license of a firm based in the City of London.
In the currently applied definition of market, the possibility of substitution on both sides is an important consideration. If my Spotify app disappears from my phone, can I still listen to music on the radio? If YouTube blocks my account where I uploaded my song, can I still make it available on Apple Music? These questions are particularly difficult to answer in the case of music because it is a digital unit that is reproduced at zero cost, and which gains its value through intellectual property (copyright) protection. It is sold in many configurations, often in opaquely-priced bundles, with incompatible price quotations.
Understanding what the actual market for music is, as well as the role of music streaming in this market, is not only an important question in the application of competition law. For the operator of a streaming platform, or for a record label, economic decisions should be based on prices, demand elasticity and substitution. These prices must be observed and the recording of the empirical observations must be relevant for the purpose of any market study. Even when we are sceptical about the usefulness of applying a market power test on a relevant market, we still need a working definition of the market to observe its prices, quantities, and the behaviour of buyers and sellers.
In the past years, there has been an increased interest in analysing online platforms where consumers do not pay a price. Watching a music video on YouTube, reading amusing content on Facebook, or finding accommodation on AirBnB does not cost oney for the buyer. The buyer is however committed to viewing advertisements, give up her data. In a monetary sense, a commission from a two-sided product (the supply side of a rented room) offers compensation for the platform. As many analysts remarked, such practices probably originated from the use of music and audiovisual content on radio and television (Competition & Markets Authority 2020, pp 44–45; Hausemer et al. 2021). Most recorded music use takes place without the user paying for fit.
The new platforms that make music available for literally billions of people are global: YouTube can be accessed almost anywhere on Earth, and Spotify is selling subscriptions in more than 200 countries and territories (Spotify 2022). From an economic or competition law point of view, it is particularly challenging to contend with zero prices and geographically unrestricted access to a service.
Our paper is organized around this topic.
We review the methodology of empirically observing or estimating music use volumes, prices, revenues, and concentration.
Second, in the 3 chapter, we review how music is sold in different formats. Our value chain analysis follows the traditional three income stream model which is widely applied both in the U.S. and in Europe. Then we review the contractual (licensing) norm in these channels, and how prices are set. In 3.2 Fair Value we argue that from an economic point of view, the IFRS and WIPO standards on the fair valuation of intangible assets, including copyrights and neighbouring rights, are practically more important for a market analysis, than the issue of equitable remuneration—and we believe that in the policy debate often these terms are used in confusion. Then we introduce the problems of value transfer to unlicensed platforms and the differences of the UUC model, and we make the case that music streaming and UUC should not be treated as one market segment.
Chapter (simulation-results) contains simulation results that we prepared for the Music Creator Earnings’ Project, however, they are even more useful for competition analysis. We argue that for numerous reasons, the best geographical scope for analysis, and the best definition of the relevant geographical data is the national market; in our case, the UK music market. We demonstrate the practical difficulties of the correct separation of the national, domestic, and export revenues. In our view, most publicly available music market information is too aggregated to establish UK volumes, prices, and concentrations. The subchapter 2.2 introduces the problems of estimating the most basic economic variable: the quantity or volume of the music used in the market. Most publicly available information on music neither contains volumes nor prices. We will investigate the possibility of observing or estimating prices in subchapter 2.3. We show that because of the highly globalized nature of the music industry, careful attention must be paid to understanding the functional currency used by entities in the market to report revenues (See 2.3.1). Subchapter 2.4 then overviews the potential empirical data sources and empirical estimation strategies for obtaining quantity and price information for the relevant product/service market, from a relevant geographical segment, for an evidence-based market analysis.
Our analysis shows that in many streaming markets, rightsholders, expressed in British pounds, experienced flat earnings, which was by a large extent due to the benevolent USD/GBP, EUR/GBP, and CHF/GBP exchange rates. It is likely that the majority of individual rightsholders experienced flat or diminishing revenues, and later releases brought less and less revenue per thousand streams. A less benevolent exchange rate environment could have brought very large losses of revenue (See 4.5 for more details.) - Our computer simulations show that UK rightsholders are experiencing a rather varied picture of music streaming. Because of the generally falling value of streams, short-lived hits are more impacted by value decline than perennial hits (See subchapters 4.2 and 4.3.) Rightsholders with a diversified international audience are partly shielded by these effects (See 4.4.) Younger stars and new genres generally experience a windfall from the pro-rata distribution system and older artists with more conventional genres such as rock or blues are experiencing headwinds. The publicly available market totals are hiding completely different market experiences.
Our report highlights some important lessons. First, we show that in the era of global music sales platforms it is impossible to understand the economics of music streaming without international data harmonization and advanced surveying and sampling. Paradoxically, without careful adjustments for accruals, market shares in jurisdictions, and disaggregation of price and volume changes, the UK industry cannot analyse its own economics because of its high level of integration to the global music economy. Furthermore, the replacement of former public performances, mechanical licensing, and private copying remunerations (which has been available for British rightsholders in their European markets for decades) with less valuable streaming licenses has left many rightsholders poorer. Making adjustments on the distribution system without modifying the definition of equitable remuneration rights or the pro-rata distribution scheme of streaming platforms opens up many conflicts while solving not enough fundamental problems. Therefore, we suggest participation in international data harmonization and policy coordination to help regain the historical value of music.
This paper grew out of the analysis we have provided for the Music Creator Earnings’ Project, part of the Intellectual Property Office’s Music Creator Earnings’ in the Digital Era report (Antal 2021b). We were tasked with providing longitudinal analysis of earnings development and relating our findings to equitable remuneration. The starting point of our work was centred around a very broadly defined problem: how much money music creators (rightsholders) earn from streaming, how these earnings are distributed, and how the earnings and their distribution have developed during the last decade. We are reproducing some of our findings in this working paper, because they are related to the very same price and quantity movements that the CMA is investigating.