Likewise, Pandora has taken pains to note that heavily streamed artists can make plenty of money through their service. Last October, co-founder Tim Westergren blogged that the company pays upward of $10,000 a year apiece for playing recordings by more than 2,000 artists. That's not saying that the artists will see that much money, naturally, since the rights-holders are the ones who get paid; for a label, though, $10,000 is like selling 1,000 CDs.
But then there's the question of publishing, which is what’s bugging Lowery—the way songwriters get paid. Performers benefit from having their recordings played that aren't directly monetary: glory, promotion, name recognition. Songwriters generally don't, so they get a rate determined by law when their work is purchased or played. Though middlemen such as performance rights agencies and publishing companies take their cut, somebody like Lowery can still see a significant trickle of money from co-writing a minor hit 20 years ago.
But the momentum of online streaming shows no signs of slowing—in the first half of this year, it was up 24 percent from last year—and that's why songwriters are worried. Statutory publishing rates for streaming audio are currently low enough that the amount Pandora lays out for publishing amounts to only 4 percent of its annual revenue, or less than one-hundredth of a cent per stream, according to this editorial by two members of Congress. Pandora's hoping to lower that rate even further; earlier this summer, the members of Pink Floyd wrote an indignant USA Today opinion piece about Pandora's shady maneuvering. (It didn’t mention that the band had released its catalogue to Spotify a week or so earlier.)
There are plenty of high-profile holdouts whose music isn't on Spotify: AC/DC, Tool, Garth Brooks, Led Zeppelin, the Beatles—all artists for whom Spotify arguably can't do much good anyway. Nobody's going to hear the Beatles' music for the first time and fall in love with it by stumbling onto somebody's Spotify playlist, and Led Zep doesn't need to convince young music fans to come see them play live. But the rates Spotify pays to the kind of big names that attract subscribers are often much higher, as Sasha Frere-Jones notes.
At the other end of the scale, there are some independent labels (like the Chicago-based indie-rock label Drag City) and artists who have opted out of Spotify, which, as Thom Yorke correctly observed, doesn't generally pay them much anyway. English singer-songwriter Sam Duckworth complained last month that 4,685 plays on Spotify netted him a bit under 20 pounds, "the equivalent to me selling two albums at a show ... I think it's fair to say that at least two of those almost 5,000 listeners would have bought the album from me if they knew the financial disparity from streaming."
That might well be the case. If we're getting into hypotheticals, though, how many of those listeners might have bought the album, or come to one of Duckworth's shows, because they heard the stream and were impressed by what they heard? And how does that compare to what would have happened if one of Duckworth's songs had been broadcast once on the radio and heard by 4,685 people? More broadly, when does it make financial sense for musicians to restrict access to their recordings, and when is it simply a matter of asserting more control over their art?
There's no way to know the answers to those questions, but Nigel Godrich's explanation that pulling Atoms for Peace's recordings from Spotify was "about standing up for other artists' rights" doesn't entirely hold water. For less-than-famous performers, recording royalties have never been a way to get rich, or even to make a living; they're a way to build up enough of a reputation to make a living through other means, and maybe accrue a little bit of cash. It's songwriters—whether or not they're performers too—who need to be wary about what's going to happen as radio and recording sales gradually give way to streaming.