“Sick of my money funding crap.”
A Fan

“These companies are taking power away from listeners, because listeners don’t have any say where their money goes,” Keating said. “If you only listen to me, I should get all the percentage of the money you spend on music.”
Zoë Keating

“User-centric” is rapidly becoming all the rage in the streaming royalty debate.  To its credit, Deezer is the only streaming service that has announced it is working on trying to implement a user-centric model.  (Others may be, but no public announcements.)

It’s not surprising that a small casual poll by Artist Rights Watch showed that when asked what would help artists the most, the overwhelming choice (83.9%) was buying a CD directly from the artist compared to streaming on Spotify (16.1%) or on YouTube (0%).  More rigorous data is required to reach a conclusion, but the lopsided nature of  the responses suggests that music fans very well may get it.  Not only do they get it, they may be woke enough to try something new.

Streaming is not a way to help independent artists.  And it’s definitely not a way to help songwriters who are still waking up from a century of sleeping on their rights.  Fans are getting the message that the economics of streaming are not sustainable for independent artists (and probably not other artists who are so unrecouped they may not care).  When Daniel Ek makes over $600 million in stock accretion in a single day during a pandemic that has reduced touring revenue to zero, Spotify’s income inequality just seems grotesquely unfair and unethical.  (Ek made more than Saudi Arabia paid for its stake in Live Nation).  This imbalance seems straight out of San Francisco’s Big Four railroad robber barons of the Gilded Age. Both artists and fans feel more than a little dirty contributing to it.

I wrote the first “Ethical Pool” post in 2018, and given the rumors about difficulties Deezer is having with implementation of a user-centric model, I thought it was a good idea to recap the Ethical Pool approach.  I’m not going into the same level of detail as I did in Ethical Pool I, and I’m going to assume that you know certain elements that I won’t get into again here.  If you need more detail, I recommend reading the first piece first.  In this post, we will briefly recap the royalty pool allocation compared to the Ethical Pool, and then focus more on implementation suggestions.

The Perception of Unfairness

Why do we care?  There are a few reasons why streaming royalties are paltry and therefore controversial.  The income inequality is brought into sharp relief with Spotify due to the distribution of shares in Spotify’s public stock sale.  (The problem being Spotify’s typically short-sighted business decision, i.e., failure, to share their good fortune with all creators on the platform, not with the fact that only the majors and big indies got stock.  You can’t take what was never offered.)  However, the problem is not with one particular platform, it’s with all the platforms because they all use some version of the revenue share model of calculating royalties.

At a high level, there are two approaches to paying royalties—up front and back end.  Up front payments are often fixed price per project or fixed price per unit.  Examples of an up-front fixed price royalty would be the statutory mechanical royalty for songwriters or the program license fee for television.  The up front model says, the price for my work is $X, I don’t really care if you make a penny, that’s your problem.  If it doesn’t profit you to pay my up front fee, get another one.  If the up front royalty is small and contemplates numerous copies being sold, there’s often a minimum guarantee or “advance” which does not get paid back, but which is recoupable from the sale of those copies.  A television program may have just a lump sum license fee with some distant profit participation.

The bet is that the user will make money and has the means to finance (or deficit finance) the project to get to the cash flow positive stage.

A backend deal usually has a royalty rate attached instead of a fixed penny rate which is a form of revenue sharing.  The minimum guarantee appears again, and may be recoupable or nonrecoupable or some combination of the two.  Still, the typical backend deal depends on how the licensed work commercially performs in isolation.  In other words, if I license a song to a record company or a television program to a network, I don’t really care how the rest of the songwriters or television producers are compensated (although I will sometimes protect my upside by asking for “Most Favored Nation” treatment on some financial terms which can have its own legal problems).

Unintended Consequences of the Big Pool:  A hyper efficient market share distribution

Streaming backend deals have an additional twist, and it is this twist that creates this hyper efficient marketshare distribution.  Unlike the typical backend deal, streaming backend deals depend on a revenue share.  You could ask why is that different, it’s all a revenue share of some kind.

True.  But this is a very particular kind of revenue share.  Streaming deals establish a pool of service revenue on a monthly basis and that pool is allocated among sound recordings based on plays by the service’s users.

Streaming revenue share is essentially a popularity contest for who gets the most money from the pool at the end of the month.  So rather than having a fixed price that varies based on how you do, how you do depends entirely on how you do relative to everyone else in competition for a fixed pool.

Naturally, the more leverage you have, the less Malthusian your deal will be, and you will have lots of downside protection bells and whistles like per stream minimums, per subscriber minimums, subscriber conversion incentives, and other goodies.  You may have a bigger share of the monthly revenue pie (so the service retains less), but the high level reality comes down to this:  The bigger your market share, the more money you make, and the smaller your market share, the less money you make.  It is essentially a zero sum game.

There is another mathematical reality to this calculation–the per-stream rate tends to decline over time.  Why?  Because if the rate of change in the revenue pie (the “A” value below) does not increase faster than the rate of change in the number of recordings being streamed (the “C” value below), the per stream rate will always decline.  Or more accurately, the nominal per stream rate since it is not actually fixed (outside of the downside protections).  And Spotify is never going to exert pricing power as long as it is in a growth mode, so the “A” value is very unlikely to ever increase enough to counter the increase in the “C” value.

The calculation looks something like this (although the algebra allows different steps to get to the same result):

Critics will say there is no per-stream rate in streaming licensing contracts.  That’s true, but there is a nominal per stream rate that can be approximated, which is about the only way to compare services to each other, so it is a relevant calculation.  (See The Trichordist Streaming Price Bible).

So you can see that if you have a lot of popular recordings that are in genres that get streamed a lot, you will be very happy with the streaming revenue share calculation.  A concrete example might be hip hop compared to contemporary classical.  The ability to fund both an on and off-platform marketing spend that outstrips the competition clearly has a pronounced effect on revenue, also.  Playlist placement may make a difference, but whoever knows how particular tracks get included in playlists, please come forward.  So even within a genre, the better heeled have a better chance of dominating the royalty allocation and the less well-heeled will see their income fluctuate for Kafka-esque reasons beyond their control.

In the calculation illustrated above, that would mean that if your “B” value is larger than anyone else’s B, then you will dominate that month’s allocation.  The more downside protection you have negotiated in your license, then it is likely that your nominal share of the pie will be even bigger if that downside protection is baked into your royalty calculation.  (It’s also possible that one party’s downside protection actually reduces the monthly pie for those who don’t have it, but that’s a timing question that is difficult to know the answer to without a royalty compliance examination and even then may be inconclusive.)

From a user’s perspective, the big pool royalty allocation means that their subscription payment goes into the big pool and is allocated based on how that user and all other users stream during the month nearly regardless of which artists the user actually listens to.  The impact any one user makes on the allocation of revenue to a particular artist varies inversely to the number of active users on the service in a particular month.

Streaming revenue, then, has become a hyper-efficient market share distribution based on factors like who has the bigger catalog being streamed on the service, which at the margins probably means who has the bigger catalog full stop.   This is why the fan says she is “sick of my money funding crap.”  “Crap” obviously is in the eye of the beholder, but what she is really saying is that she’s tired of having nearly all of her subscription payment being paid for recordings she never streamed (and maybe would never stream).

User Centric and the Ethical Pool

User centric royalties would capture the user’s revenue (let’s use subscriptions as an example) and allocate it solely to the same user’s streams.

Let us accept as a given that changing the entire royalty system would require the consent of everyone involved who has a license (which is a lot of copyright owners).  That consent is unlikely to be given the larger the market share of the copyright owner.  Plus a single holdout could stop the whole thing.

Let us also accept as a given that changing the entire royalty system would be very complex and costly.

But–here’s another given.  The lower the per stream rates go, the lower the royalty payments go, and the greater the incentive for independent artists to get out of the streaming game altogether.  (Not to mention getting angry.)  When your royalty payment is minuscule it is easy to just say no.  Your motivation for staying in the service certainly isn’t based on the financials and is more likely to be fear of missing out or other preference curve distortion.

Therefore, rather than asking other copyright owners to opt-in to a user-centric model (which many probably won’t do), artists who are poorly served should consider opting out of the big pool.  (It must be said that U.S. songwriters gave up the right to opt out of audio-only mechanical licenses long ago, and even perpetuated this travesty in the recent Title I of the Music Modernization Act.  Why they did not fight back when they had the services on the ropes through class action litigation is anyone’s guess–but maybe it had something to do with money other than royalty rates.)

If those artists want to stay on the service, then the service could essentially duplicate their big pool accounting ruleset and create a separate pool of revenue–the “Ethical Pool.”  This would require allowing users to still pay their subscription fee, but direct where their money would be allocated for royalty purposes.  (Users would probably have to opt out of the big pool service since their fees would no longer pay for it, but hardcore fans probably wouldn’t mind and might even view it as a feature.)  That could be accomplished by a button in the user’s account settings and artists seeking to participate in the Ethical Pool could have a green check or some other icon indicating that the artist was in the Ethical Pool so would the fan please make that change.

This might take a few months to reach a critical mass of users and artists opting in, but when your royalty starts three or four decimal places to the right anyway, how much worse could it be?  Properly promoted, the changeover could be relatively simple and simplicity for the existing and new users should be the guiding light.

No deal points would change for the “big pool” copyright owners so their consent would not be required for the Ethical Pool.  The only change would be the rough equivalent of artists dropping out of the service and their fans leaving with them.  That risk was part of the original negotiation.  For the dominant players, it is unlikely that they will see any real fall-off in revenue, and if anything their hyper-efficient market share allocation may even increase slightly (because the C value would decrease by the number of streams that transition to in the Ethical Pool).

The Ethical Pool will require an investment by the service to create this parallel infrastructure, but if the big services don’t do it, it’s only a matter of time until insult exceeds injury and artists start leaving.  Or someone launches a service that caters to artists who would be Ethical Pool candidates.

I don’t think this is the end of the story, but it is a workable interim step on the path to sustainability that is more important now than ever.  The Ethical Pool would at least make the money clean and remove that feeling that the fan is actually hurting the artist when streaming.

The arc of the moral universe is long, but it bends towards justice.  However–it doesn’t bend by itself.  At this moment, the arithmetic on the Internet is starting to bend that arc.  The way it bends is up to us.