Music Theory Professor Eric Drott’s research spans several subjects: contemporary music cultures, streaming music platforms, music and protest, genre theory, digital music, and the political economy of music. His most recent book, Streaming Music, Streaming Capital (Duke University Press, 2023), examines the economics of music streaming platforms. We recently spoke with him about the research for his new publication.
Tell us about the genesis of your new book, Streaming Music, Streaming Capital.
It came about through my previous work on musical genre. One thing I noticed over the years is that keyword searches for genre in scholarly databases were increasingly dominated by research in music information retrieval, or MIR. In many cases these were computer scientists, many of whom would go on to work for streaming platforms, trying to develop models for how to automatically classify music as a way of recommending it to listeners. The question they were asking was: How can we use an algorithm to take unlabeled audio files and classify them in a particular way and label them to say this belongs to this category or that category?
These attempts to automate the identification of musical genres became very interesting to me. After that, one thing led to another as I started to look at how music is classified on streaming platforms. This was right around the time that Spotify was starting to really take off in the United States and trying to attract a much bigger mass audience. They started to emphasize playlists that were associated with particular kinds of moods or activities. That raised a lot of questions about what was motivating this change in how music is being categorized.
As someone who is very much a materialist, I think that economic incentives are important. It made me ask questions about the economic logic that’s driving this way of organizing music for listeners. What do streaming platforms get out of it? What kind of data are they collecting on users? What kind of things do they use that data for beyond just giving us recommendations for what we might want to listen to next? How do those recommendations reshape how we listen to music? How is a company like Spotify able to keep going all these years, even though they rarely turn a profit? These questions introduced me to literature about venture capital and private equity markets and how they invest money into these apps.
What are some key concepts in your book that you hope come through if someone is reading it?
Well, there are a couple of things. I realized that trying to do a comprehensive, final statement on streaming and the streaming economy was a futile endeavor because it changes so fast. I needed to take a step back and not necessarily focus just on streaming, but rather use streaming as a way of thinking about the ways in which music relates to markets, to the economy and to capitalism.
Music occupies this really ambivalent position in the market, which is manifest in a lot of ways and has some implications that may be unexpected. It has one foot in the market system and one foot outside of it. Streaming platforms take advantage of this ambiguous status of music because they effectively treat it like what economists would call a public good, as opposed to private goods that sit more comfortably in a market economy.
Can you talk about the difference between a public good and private good and how it relates to music in streaming context?
A lot of the distinction hinges upon whether or not the good in question is something you can exclude other people from using. So, if I have a music CD in my possession, I can keep other people from having it or using it. And related to this is the question of rivalry: If I use a good, does that use it up? For instance, if I have an apple and I eat it, that precludes other people from doing so. Public goods are non-rival and non-excludable. A classic example is air. It’s hard to exclude people from accessing it, and there is so much of it that we don’t need to worry about using it all up. So, it is really hard to charge people money for it.
Historically, public goods are much more difficult to commodify and turn into a marketable good than private goods, which are rivalrous and excludable. And music has a lot of the characteristics of a public good. It’s very easy to copy music. It’s very hard to exclude people from using music. And this has especially been a problem for the music industry since the rise of the internet and file sharing back in the early 2000s. It has made copyright harder to enforce.
And you know what’s interesting about streaming services? Their strategy is to just say, “OK, the proverbial horse is out of the barn. There’s no going back to distributing music by means of excludable, private goods like vinyl records or CDs. Instead, let’s treat it as a public good. But let’s privatize that public good. Let’s put a fence around this massive library of music that we’ve created in collaboration with record labels, and let’s charge people $10 or $11 a month for access to it.”
That changes our relationship to music in a pretty profound way because when I log on to my Spotify account, I’m no longer in the same economic relationship to music that I was back in the day when the main way I accessed music was to go to a record store and plunk down $10 for a record.
On streaming platforms, by contrast, there’s no money that ever changes hands to pay for music. True, I pay for the service. But the music is all free for me to listen to from that point forward. To use another economics term, there is no marginal cost attached to streaming one more song. In short, music no longer appears to listeners as commodities.
But if these changes depend on platforms’ treating music as a public good or utility, that then raises questions about why we need a private company to own and operate these utilities. After all, lots of public goods have historically been managed by public bodies.
Can you talk about some of the challenges this dynamic has created for artists?
Historically, it’s always been hard for artists to make money from selling records. There are a couple of things that make it even more difficult nowadays. First, there is an argument to be made that the way platforms work reinforces a winner-take-all dynamic. So, the global superstars become even more global and even bigger superstars: the Taylor Swifts, Beyoncés, Drakes, Bad Bunnies — they’re doing fine, right? They might be doing better if they were still selling CDs, but I’m not losing sleep for them, since a lot of the revenue that these platforms generate goes to this very small sliver of artists.
The other thing that’s easy to overlook is that the structure of payments is radically different on streaming platforms, compared to record sales. Before streaming, whenever you bought an album, the payment for the good was up front. Assuming you were an artist who didn’t owe an advance to your label (which often wasn’t the case, I should add), whenever somebody bought a CD or tape, you got a certain percentage of that sale.
But streaming payments are indexed to use, not purchase. So, every time somebody streams a song, that gets counted, and until very recently, whatever percentage of the total amount of streaming this accounts for on the platform in a given month determines the percentage of the platform’s overall revenue that the rights owner will receive.
If we look at the long term, maybe those two amounts come out as even. Maybe over time people stream your music consistently over 10, 15, 20, 25 years, and you earn as much as you would make from album sales. But money spread out over time isn’t worth the same as money that’s deposited in your bank account at the beginning of the month. The fact that these payments are being spread out so thinly over a longer period of time means that it becomes very hard for artists because — to paraphrase John Maynard Keynes — in the long run, we’re dead. So, for a lot of people, getting that money in your hands up front is worth more than having it stretch out into the future.
Now, for some artists at the end of their career who sold most of their albums like 40 or 50 years ago, streaming is just the icing on the cake, because you have this new revenue stream. But the people who really benefit are the people who sit on large quantities of intellectual property. Each individual song or album may only amount to a small trickle of revenue here and there, but if you add those up, it becomes this gushing torrent of money.
This is one reason why major record labels, major publishers, investment funds and asset management companies are buying up music rights. A notable example is when Bob Dylan sold his catalog to Universal Music for more than $300 million in 2020. One thing streaming has made possible is an economic model where music is regarded as an asset. And it is an asset that’s particularly attractive because it’s not linked to the ups and downs of the economic cycle. People listen to music in good times and bad times: It’s consistent. It’s predictable. It’s steady. It’s growing. It’s everything investors love.