As I write this article, I am listening to the band TesseracT on Spotify. The service is quite a bargain for me — I can listen to virtually any record ever made, for just $9.99 a month. But it's not such a great deal for TesseracT, which gets just $0.00348 for each song of theirs I play. If I listened to their album "Sonder" on continuous repeat 24 hours a day for the next year straight, they would receive just $355.90.

Streaming services like Spotify have wreaked havoc on the livelihood of musicians around the world. In time they may even destroy the artistic ecosystem on which they depend. There is one obvious solution: full communism for music.

Of course, it's not just streaming that is destroying the livelihood of American musicians. As David Dayen reports for The American Prospect, the consolidation of radio, record label, and ticket sale companies has also bitten hard into artist incomes. The radio behemoth Liberty Media, the three big labels, and Live Nation all have enormous market power relative to all but the biggest stars, and ruthlessly use it to cut down the share of money going to musicians. The pandemic made things even worse by erasing live shows overnight — basically the last thing supporting the musical middle class.

But streaming is still probably the most disruptive thing that has happened to music in general over the last 20 years. It has basically killed CD and mp3 sales, and now makes up about four-fifths of the global music industry's money outside of live concerts. Revenues are actually increasing strongly now thanks to paid streaming subscriptions — but as Dayen details, that isn't compensating for the destruction of previous revenue models except for the biggest stars.

Streaming is a classic natural monopoly, in which fixed costs (the price of servers, web developers, and especially rights to music) vastly exceed the marginal costs (the price of sending one song over the internet). It follows that under a typical market dynamic, one or a few companies will eventually vacuum up the vast majority of market share. Once a company is established and large, it will naturally find it easier to put more and more music in its catalog, making it steadily more appealing for more customers — and gaining greater bargaining power over artists. Indeed, this is basically what has happened. YouTube, Spotify, and the French company Deezer totally dominate the streaming market, and numerous competitors have already been bought out or gone out of business. Worse, a private music streaming monopoly is plainly already tending towards an artistic monoculture, where a few big names take home most of the money and everyone else gets scraps.

For a natural monopoly, there are two classic policy solutions: regulation or nationalization. Power utilities around the country, for instance, are almost universally either owned by various levels of government, or are heavily regulated by government oversight boards. Regulation could work, but would require a lot of complicated oversight of private owners, who will always be attempting to squirm out from under the rules or capture the board to boost their profits.

Either option would require competent governance. But it would be cleaner and simpler for the U.S. government to simply buy Spotify outright and convert it into a public service. Its whole market capitalization is just $50 billion — a pittance in a $4-trillion budget — and has just a few thousand employees. If the American state can run the vast complex of dams and power plants in the Tennessee Valley Authority, surely it can run one moderately large web service.

Now, one could set up in many ways. But let me suggest a sketch of a business structure. First, all U.S. copyright holders will be required to license their content for a reasonable fee. Second, YouTube (and any other American company that streams music) will be required to set up copyright claim systems that actually work, instead of the current system that allows for rampant copyright infringement, constantly dings people for legal "fair use," and is also easily abused. This would likely drive the company out of streaming altogether.

Third, artists would be paid progressive streaming rates: (let's say) a penny per play for the first million streams per month, $0.0075 for the next 500,000, $0.005 for the next 500,000, and gradually decreasing from there. These numbers are just a rough guess, but the point is to redirect the flow of streaming income more towards middling artists, thus pushing against the winner-take-all dynamic that prevails on Spotify. The only people who make real money from streaming at present are superstars like Taylor Swift who get billions of plays — and even for them it's a small share of what they get from concerts, endorsement deals, merchandise, album sales, and so on.

Fourth, there will be no recommendation algorithm. Listeners will have to search out their own music, and create their own playlists (which they will be able to share). This will push back against the commodification of music, which is degrading the art form. Spotify's algorithmic recommendations tend to turn music into an aural wallpaper that one barely listens to — I'm sure I'm not the only one with long Spotify playlists full of artists I barely recognize. And once more there will be an important role for music journalists and critics to find and raise up new artists.

Fifth, only artists will receive particular data about their listeners — will not collect any personal information itself. This is the reverse of current practice, in which artists get nothing and Spotify is keeping tabs on every song you listen to. That is both invasive for listeners and a huge headache for artists, who lose out on valuable information on their fans that could be used for sales and planning concerts.

Finally, listeners would still need to pay up — but not much. Currently Spotify's headline subscription rate is $9.99 per month, though it has various discounts for students and families. The entire global recorded music industry took in about $21.6 billion in 2020, of which Spotify took in about a third. Given historic and projected industry growth rates, probably all that would need to be done to make the service pay for itself over the medium term would be to nudge up the price a bit, and let subscriptions continue to roll in. Or we might choose to subsidize it with a few billion dollars in public funds.

I'm sure to many readers this proposal sounds rather extreme (and we might haggle over details). No doubt Republicans would have a conniption fit over the prospect of Cardi B earning millions through a public service. But if we want music to remain a thriving part of American culture, something must be done. Dayen's recommendations to bust up the radio, record label, and ticket sale oligopolies are well worth doing, but so long as streaming remains under the control of private capitalists, modest musicians will struggle.

So why not let good old Uncle Sam run internet music delivery as a public trust?