Abstract: An ongoing debate among firms, rightsholders, particularly in the music industry, and policymakers in the United States and the European Union concerns potential changes to the regulation of user-generated content (UGC) video streaming platforms (e.g., YouTube). Currently, safe harbor provisions protect platforms from liability for copyright-infringing content uploaded by users, and requirements for compensating rightsholders for UGC are weak, resulting in comparatively low payouts. At the same time, it is unclear how a change in these regulations would affect consumer demand for this content on other platforms with higher payouts (e.g., Spotify), that is, whether UGC platforms stimulate or displace demand on other platforms. We study a quasi-experiment that occurred when numerous songs became available as UGC on YouTube after an agreement between YouTube and the German royalty collecting society. Our analysis of an unprecedented data set covering 600,000 songs by 38,000 artists reveals an intriguing finding: Although UGC availability stimulates demand in other streaming channels for most songs, cannibalization occurs for recent releases and hit releases, turning the overall revenue effect negative. We discuss how policymakers can use these findings to understand the implications of changes in regulation, and how labels and artists can decide which content to block or allow on UGC platforms.

Abstract: An ongoing debate among firms, rightsholders, particularly in the music industry, and policymakers in the United States and the European Union concerns potential changes to the regulation of user-generated content (UGC) video streaming platforms (e.g., YouTube). Currently, safe harbor provisions protect platforms from liability for copyright-infringing content uploaded by users, and requirements for compensating rightsholders for UGC are weak, resulting in comparatively low payouts. At the same time, it is unclear how a change in these regulations would affect consumer demand for this content on other platforms with higher payouts (e.g., Spotify), that is, whether UGC platforms stimulate or displace demand on other platforms. We study a quasi-experiment that occurred when numerous songs became available as UGC on YouTube after an agreement between YouTube and the German royalty collecting society. Our analysis of an unprecedented data set covering 600,000 songs by 38,000 artists reveals an intriguing finding: Although UGC availability stimulates demand in other streaming channels for most songs, cannibalization occurs for recent releases and hit releases, turning the overall revenue effect negative. We discuss how policymakers can use these findings to understand the implications of changes in regulation, and how labels and artists can decide which content to block or allow on UGC platforms.

• Commentary by Rebecca Tushnet (2024):  Comment on “Frontiers: The Interplay of User-Generated Content, Content Industry Revenues, and Platform Regulation: Quasi-Experimental Evidence from YouTube” 

• Rejoinder (2024): Rejoinder on “Frontiers: The Interplay of User-Generated Content, Content Industry Revenues, and Platform Regulation: Quasi-Experimental Evidence from YouTube” 

Abstract: Broadband Internet has fundamentally changed business models in many industries. In the music industry, for instance, old business models were challenged by illegal competitors, and broadband Internet has enabled value creation through new business models. The changes that established business models experienced in the wake of broadband Internet, however, differed vastly across national markets, and these differences are not well understood. We build a conceptual framework and study the extent to which differences in economic and cultural factors are associated with different market outcomes in the wake of the proliferation of broadband Internet. Thus, we compile two unique data sets from the music industry, comprising (1) revenue data for 36 countries and 22 years and (2) piracy data for 47 countries and >2 years. We use a Bayesian multilevel model to explore between-country heterogeneity in the associations between these variables and broadband Internet adoption and business model innovations. Our results show that the negative association between broadband Internet penetration and music revenue is weaker in high-income countries, where income restrictions are less likely to drive demand towards illegitimate piracy services. In terms of cultural factors, we find that a market's response to the introduction of broadband Internet is less negative in countries scoring high on Hofstede's individualism and uncertainty avoidance dimensions. Furthermore, we find that overall revenues only recover after the latest generation of streaming services (e.g., Spotify) has been introduced, and the adoption of these services is associated with lower levels of online music piracy.

Abstract: On-demand streaming services that rely on subscription fees or advertising as a revenue source (e.g., Spotify) are a topic of ongoing controversial debate in the music industry because their addition to the distribution mix entails the risk of cannibalization of other distribution channels (e.g., purchases of downloads or CDs) and might reduce overall revenues. To date, no research has assessed the effect of streaming services on revenue, and whether cannibalization indeed takes place. Our research fills this void and assesses the impact of free and paid streaming services on music expenditures and on total music industry revenue. To this end, we constructed a research design in which we observed a panel of more than 2500 music consumers repeatedly over more than one year. This approach allows us to eliminate individual-specific unobserved effects that may otherwise confound the identification of a cannibalization effect. Our results show that the adoption of a free streaming service as well as the adoption of a paid streaming service cannibalizes consumers' music expenditures. The net effect of paid streaming services on revenue, however, is clearly net positive. In contrast, the net effect of free streaming services on revenue is only positive for consumers who were relatively inactive before the adoption. On the industry level, our findings suggest that the negative effect of free streaming on industry revenue is offset by the positive effect of paid streaming in the context that we analyze. Hence, in the market that we study and under the assumptions that we make, we estimate that the overall effect of streaming on industry revenue is positive.

Abstract: In this paper, we compare the standard, single-response choice-based conjoint (CBC) approach with three extended CBC procedures in terms of their external predictive validity and their ability to realistically capture consumers’ willingness to pay: (1) an incentive-aligned CBC mechanism (IA-CBC), (2) a dual-response CBC procedure (DR-CBC), and (3) an incentive-aligned dual-response CBC approach (IA-DR-CBC). Our empirical study features a unique sample of 2,679 music consumers who participated in a conjoint choice experiment prior to the market entry of a new music streaming service. To judge the predictive accuracy, we contacted the same respondents again 5 months after the launch and compared the predictions with the actual adoption decisions. The results demonstrate that IA-CBC and DR-CBC both increase the predictive accuracy. This result is promising because IA-CBC is not applicable to every research context so that DR-CBC provides a viable alternative. While we do not find an additional external validity improvement through the combination of both extensions, the IA-DR-CBC approach yields the most realistic willingness-to-pay estimates and should therefore be preferred when incentive alignment is feasible.

Abstract: The market for digital content (e.g., music or movies) has been affected by large numbers of Internet users downloading content for free from illegitimate sources. The music industry has been exposed most severely to these developments and has reacted with several different online business models but with only limited success thus far. These business models include attempts to attract consumers by offering free downloads while relying on advertising as a revenue source. Using a latent-class choice-based conjoint analysis, we analyze the attractiveness of these business models from the consumer’s perspective. Our findings indicate that advertising-based models have the potential to attract consumers who would otherwise refrain from commercial downloading, that they cannot threaten the dominance of download models like iTunes, and that current market prices for subscription services are unattractive to most consumers.

Abstract: This paper investigates how the consumption of an artist's creative work is impacted when there's a movement to "cancel" the artist on social media due to their misconduct. Unlike product brands, human brands are particularly vulnerable to reputation risks, yet how misconduct affects their consumption remains poorly understood. Using R. Kelly's case, we examine the demand for his music following interrelated publicity and platform sanction shocks-specifically, the removal of his songs from major playlists on the largest global streaming platform. A cursory examination of music consumption after these scandals would lead to the erroneous conclusion that consumers are intentionally boycotting the disgraced artist. We propose an identification strategy to disentangle platform curation and intentional listening effects, leveraging variation in song removal status and geographic demand. Our findings show that the decrease in music consumption is primarily driven by supply-side factors due to playlist removals rather than changes in intentional listening. Media coverage and calls for boycott have promotional effects, suggesting that social media boycotts can inadvertently increase music demand. The analysis of other cancellation cases involving Morgan Wallen and Rammstein shows no long-term decline in music demand, reinforcing the potential promotional effects of scandals in the absence of supply-side sanctions.

Abstract: The digital age has significantly changed how music is consumed, promoted, and monetized. Social media platforms like TikTok are playing a pivotal role in this transformation. This shift has sparked a debate within the music industry: While some stakeholders see social media platforms like TikTok as opportunities to boost songs to viral status, others raise concerns about potential cannibalization effects, fearing that such exposure might reduce revenue from streaming services like Spotify. In this paper, we evaluate the effect of a song's presence - or absence - on social media on its demand on music streaming services using a quasi-natural experiment: Universal Music Group's (UMG) - one of "The Big 3" record labels - decision to remove its entire content library from TikTok in February 2024. We use representative samples covering close to 50% of the US and 94% of the German streaming markets, employing a difference-in-differences approach to compare the streaming consumption of songs that were removed from TikTok with those that were not. We find that UMG's removal of music from TikTok led to a 2-3% increase in streams on audio platforms for affected songs, indicating substitution effects. However, this average treatment effect masks significant heterogeneity: older songs and songs with less promotional support elsewhere saw a decrease in streaming consumption, suggesting that TikTok helps consumers discover or rediscover content that is not top of mind for consumers.

Abstract: Music streaming platforms such as Spotify provide access to millions of songs. Playlists, as curated collections, play a key role in driving both music consumption and discovery, but the success factors behind their impact on song demand remain insufficiently understood. Using difference-in-differences analyses on 204,036 quasi-experiments, where songs are listed on playlists and later delisted, we find a significant average listing effect of 12% more streams and a carry-over effect of 3% after delisting. These effects show considerable variability, with some songs seeing increases of up to 30%, while others, particularly those on less popular playlists or with extensive prior playlist exposure, experience only marginal gains. Playlists curated by context—focusing on activities or situations rather than genres or artists—lead to stronger uplifts during listing and carry-over effects due to their focus on diverse listening contexts. While playlists with more predictable or homogeneous song selections tend to generate more streams during the listing, as they align well with listener expectations, these same characteristics reduce the carry-over effect as they limit the potential for new music discovery. These findings advance our understanding of playlist curation strategies and offer valuable implications for artists, labels, and curators aiming to lift long-term song performance.

Abstract: Stricter privacy regulations are prompting firms to reevaluate targeted advertising strategies, offering consumers the choice to opt in or out of personalization. This study investigates the long-term impact of discontinuing targeted promotions on shopping behavior, employing two large-scale field experiments with multi-vendor loyalty program (MVLP) participants. We explore spillover effects within the MVLP and use additional, observational data to analyze opt-in/opt-out decisions to assess the economic viability of enhancing data transparency and control through privacy tiers. Our findings provide insights into the relations between privacy regulations, targeted advertising cessation, and consumer behavior in the context of MVLPs.