Hello, this is Penny Fractions, a weekly newsletter on the wacky, wonderful world of music streaming. If you’re a new reader, then check out the expanded outro with a couple more details about myself. I’d like to recommend Board Games, a book by my friend Eric Thurm on...well, the title sort of says it right there (pre-order here). This is shameless promotion but Eric’s one of my oldest New York City friends, so please forgive me! Now that promo is out of the way, let's talk about what might be the music industry’s new favorite phrase: user-centric streaming.
In August, Deezer announced plans to introduce user-centric streaming on its platform in early 2020. A month earlier, Emmanuel de Buretel, founder of the record label Because Music, championed the model as a way to weed out fakes and more evenly distribute music. (Apple has reportedly tested the model.) The train of user-centric streaming propels forward. I’ve offered critiques of the model before, but with the increased likelihood of it entering the mainstream, I wanted to reexamine my thoughts and offer a next-step solution. Now, before we get to the user-centric paradigm, I’d like to establish how the current streaming model operates.
One Market to Fit All
The current model of music streaming is truthfully rather simple: all of the money generated by the streaming service is divided by the total streaming plays, and the money is proportionally distributed to creators. The “pro-rata model” is what the current system is called, but how exactly did we get to it? For that let’s go back to 1997.
Earlier this year, when I wrote about the RIAA’s decade-long campaign of bringing lawsuits against music fans over piracy, I mentioned an often overlooked fact about Sony Music. A couple of issues of Billboard from the summer of 1997 mentioned that the record label was already experimenting with the idea of a pay-per-play music “jukebox” that predated Napster by over a year and Spotify by nearly a decade.
While early online experiments gave away MP3s—or in the case of David Bowie, even digitally sold an entire album—legalized music streaming didn’t take real shape until MusicNet and Pressplay launched in 2001 in response to the peer-to-peer file-sharing of Napster. The services were ultimately unsuccessful, as it only took a couple of months post-launch for artists to realize an issue stood in front of their collective class. A singular subscription fee for music, while a steady check for labels, rather quickly depreciates the value of recorded music for artists and puts all the power in the hands of labels. The New York Times in February 2002 reported just how unhappy artists were with this new arrangement. Major artists quickly phoned their lawyers before the first paycheck was sent out by either platform. Ultimately both platforms fell apart and an early digital confrontation was averted, but that didn’t stop major labels’ continued drive towards this goal.
When YouTube arrived in the mid-2000s, it was eventually corralled into negotiating contracts with major labels in order to host the most popular content on its platform, music. This oligopolistic strong-arming by major labels occurred with the emergence of each new streaming service, ensuring the royalty setup would be pro-label, not musicians. The pro-rata model would continue to populate subsequent platforms from the 00s until the present day. Artists have known the deck was stacked against them for well over a century, and while streaming—and, broadly speaking, the internet—was allegedly going to even the playing field, it in fact has only heightened inequality in music. Thus, I cannot advise enough against viewing user-centric streaming as a magic solution to fundamental power imbalances within the industry.
Hellooooo, User-Centric Streaming
The concept of user-centric streaming isn’t a new one. Examples of it could be found when the early music download store OD2 offered per-stream payments. Even if a niche area, a handful of academic papers on the subject have been produced and offer compelling insights into the model’s potential effectiveness. The paper by Arnt Maasø (“User-centric settlement for music streaming”) presented at SXSW in March 2014highlighted the boost user-centric streaming provided local musicians in Norway via their more actively engaged fanbases. The top-level takeaway from the paper, however, is that there was little impact on major artists’ overall market share (76%, pro-rata vs. 75%, user-centric) and that the top 5000 artists still received the same amount of revenue. The study showed a small bump in earnings for the middle and top tier performing artists, most of which was siphoned away from smaller acts outside of the top 5000. So, while there was some redistribution, the overall shifts in payouts were fairly negligible.
Only a couple of months later a similar study was conducted in Denmark by Rasmus Rex Pedersen that focused specifically on the effects of a user-centric model on the top 5000 artists (Pedersen calls it “per user”, but it implies the same method). Pedersen and Maasø arrive at similar conclusions that see increases for top performers, a decrease for the least popular musicians, and a modest rise in the middle. Both papers show the conflicting nature of the user-centric model: while they praise it for connecting fans to artists by allowing them to “vote with their wallets”, they also show the limits of such a consumerist way of perceiving social impact when the top acts are the ones that’ll end up with even more money.
A few years later, a number of Finnish music groups, including Finland’s musicians’ union, released a paper titled “Pro Rata and User-Centric Distribution Models: A Comparative Study”, which involved a similar study that looked at Spotify users. At the top level, the paper diverged a bit from the others and found that top artists in the user-centric streaming model earned a smaller proportion of the overall payouts (9.9%, pro-rata vs. 5.6%, user-centric). This would suggest that smaller artists should see an increase, but the paper repeatedly noted that such increases are very much contingent on the musician’s fanbase. This introduces an element of volatility into the equation that doesn’t exist within the current pro-rata model making it harder to conclude how user-centric streaming would affect individual artists in practice.
The reason for my unfiltered skepticism can be found right in the name: user-centric. The model’s neoliberal assumption is that music streaming’s big issue is that consumers are uninformed and don’t understand that their subscription fee isn’t being allocated correctly towards artists. This consumerist solution removes the responsibility to fairly compensate artists from record labels/streaming services and reassigns it to individual fans. If the issue was simply that fans weren’t putting their dollars into the correct funnel, then every artist should just start a Patreon as that’d be a much more sustainable/personal way of pushing towards such hyper-individualism.
The OC Streaming Model
Despite my critiques, if user-centric streaming were adopted by a mainstream streaming service I’d welcome the change. The shift would represent such a major pivot in how music streaming works and would allow for a much healthier conversation around the entire business model. Still, I’d like to offer my own take on user-centric streaming that I’ll call the OC Streaming Model.
The basic idea would be to introduce a streaming cap that would function similarly to a marginal tax rate (if you’re wondering how the name was derived). Until a certain threshold, earnings from an individual song’s streams would be distributed to the songwriters and recording artists involved, while remaining earnings would be distributed to the artist’s respective label and/or publisher. (I’ll admit that I’m flexible when it comes to how the latter pool of revenue might be allocated, but I’ll stick with my initial suggestion going forward.)
For example, if Post Malone received 592 million streams in a single month, a cap of 250 million streams would mean the money made from the remaining 342 million streams would be up for redistribution. The excess streaming revenue would then be distributed to other acts signed with the same label/publisher, with a majority going towards smaller acts and less going towards more successful ones. One concern discussed in the book Rockonomics was that music success is often only for the 0.1% of artists. User-centric streaming, as currently suggested, could make that worse, so this is an idea for how to truly help smaller artists.
One, Or Maybe, Two Last Things
A few months ago, I wrote about the Telecommunications Act of 1996 and the disastrous effects of government deregulation on the entire music industry. A lesson to learn is that the government can play a pretty big role in either helping or hurting the industry for music workers. Thus, while there are calls for user-centric streaming, I’ve suggested and still support federal regulation of these platforms. Even light government auditing of streaming services could go a long way towards combating the issues of fraud that user-centric streaming is touted to alleviate, such that there won’t be any more allegations of places like Tidal fudging numbers, Spotify bot accounts racking up views, or labels buying YouTube ad impressions for views.
A discussion could also be centered around the replacement of play count as the metric by which these services calculate payouts. One such example is IDAGIO, a classical music-focused streaming platform that pays out according to listening time rather than plays. I’m curious to see if listening time differs greatly from plays, cause I remain a bit skeptical about it making a noticeable difference. That’s why I’d recommend the OC Streaming Model as it takes what is so exciting about the user-centric model to another level and also could help foster growth within music, so there isn’t a sense that artists cannot play this system perfectly will forever be held back.
In last week’s newsletter when writing broadly about copyright, I ended up conflating it with issues of right of publicity and trademark. Honestly, I do hope to write about the topic again, so hopefully next time I’ll devote more time to my research and not make such an unforced error. Also, Robin Kaiser-Schatzlein, a newsletter reader and writer, recommended this paper by Dean Baker called “Is Intellectual Property the Root of All Evil? Patents, Copyrights, and Inequality.” I highly recommend it and would love to write more on this topic in the future.
6 Links 2 Read
Neither, and New: Lessons from Uber and Vision Fund - Stratechery
This week’s links I’ll admit are a bit “not-music” themed. Still, I couldn’t help but include this piece by Ben Thompson in which he admits to smelling his own farts too hard when he wrote about WeWork last month in an effort to make a bull case for a company that’s falling apart in front of our eyes.
Facial recognition is racist and bad, so good on these artists and festivals for opposing it.
Honestly, I read this story multiple times and I’m not exactly sure how this would work in practice...but you know what, godspeed.
This piece does a great job to acknowledge there are much larger forces at play with regard to climate change that can’t simply be fixed by banning plastic straws on one’s tour, though it’s a nice gesture!
Tim Ingham last week arrived at a conclusion I wrote about a couple of months ago: that the rise of streaming consolidation will eventually result in catalog splintering. I know some people have disagreed with me on this topic but I’m not sure I see a world in which this doesn’t happen in the 2020s.
Facebook Blocks Users from Sharing Pirate Bay Links - Torrent Freaks
The back-and-forth between the Pirate Bay and Facebook over links isn’t too shocking, but I was a little surprised to learn this feud is over a decade old. It does, however, indicate Facebook’s custom of pleasing major corporations isn’t a newly turned leaf.