I have a feeling I’m about to wander off the reservation here.
I say this because what I’m about to propose is essentially a modification of a potential legislative proposal that rumor has it the NMPA is floating. That proposal seems to be generating some negative backlash in songwriter/publisher community (whether it deserves it or not.) Also, it will at first seem to go against some of my long held principles on market pricing, “orphan works” type proposals and the establishment of any further federal copyright exemptions and safe harbors. So hear me out while I make the case for a win-win compromise.
What is it that digital services want?
Digital services have two major issues with the current law and regulations on “streaming mechanicals,” those small royalties paid to songwriters and publishers on each stream of a song.
Interactive streaming services get a federal “compulsory license” to use any songwriter’s work on their services as long as they follow a relatively simple procedure (with some exclusions). A pre-condition to services using this protocol is that the law requires them to send a “Notice of Intent” (NOI) to owner or agent for owner of the song. The NOI is intended to make sure the songwriter knows their song is being used, and the service must demonstrate that it knows who to pay.
Implicit in the logic of the law and the related regulations is that the federal government did not want the interactive streaming services building up a black box of “unmatched” songwriter royalties. Nor does the law necessarily require that the streaming services would have access to every single song in existence.
However, commercial pressures, coddling from music distributors and lax enforcement of songwriters’ rights (until the recent class action lawsuits), has encouraged these services to use as many songs as possible with full knowledge that they didn’t have licenses for every song and were thus committing mass copyright infringement. Streaming services literally very likely have billions of dollars in willful copyright infringement liabilities on their books whether or not they acknowledge the liability on their books or to their shareholders.
In order to extricate themselves from this situation, I hear that interactive streaming services have generally proposed three key solutions that can be generalized as follows. (See Spotify comment to Copyright Office).
- A global rights database with information on every song ever written.
- A “safe harbor” or “grace period” that allows services to use songs when they can’t find the owners of the works in this global rights database.
- A webcasting style NOI process (like Section 114), whereby each streaming service notifies the copyright office of their intention to stream music. They don’t have to hunt down every single songwriter.
Does this seem reasonable? Let’s dig into the details and look at the problems.
Songwriters’ objections to these proposals
The first objection from songwriters goes something like this. Interactive streaming services knew the law before they got into the streaming business. Some of them (like Google) bought companies with song data (like Rightsflow) or developed their own (like YouTube’s Content Management System).
There is a real moral hazard in rewarding services’ bad behavior by forcing songwriters to do the heavy lifting on a rights database and then granting the offending party a very valuable safe harbor. In the view of most songwriters, we are the ones being abused and are the ones deserving of concessions.
The second objection is painfully obvious to songwriters, but seems to be totally lost on the streaming services, academics and many in the music industry: An up to date global registry of all songs is impossible. Last I checked 75,000 new albums were released every year in the United States. It’s likely that hundreds of thousands if not millions of new works are created every year worldwide. How many a day? An hour? It’s really great that the NMPA may “certify” the database but there’s really no need for their members to indemnify services that could easily have done the research themselves.
Consider also the spontaneous, informal and collaborative nature of all songwriting. Many new works are created by aspiring non-professionals with little knowledge of publishing rights. How on earth do you ever get this ownership and songwriting splits into a database in a timely fashion? Certainly there is ownership registration software that works with music creating production programs that is beginning to capture more of this information, but it will never capture all of it. So why would the government want to punish this creativity by making sure it is uncompensated through yet another safe harbor?
As a result of the inevitable incomplete nature of the database, it will never fix the problem it is designed to fix. There will still be mass infringement of works by streaming services and they will be vulnerable to more lawsuits. This is especially true for new works, niche works, and works containing samples. For this very reason there is a sneaking suspicion among songwriters that streaming services are aware of this problem and are simply using the lack of a unicorn database as a delaying tactic. Further you don’t have to put on a tin foil hat to realize that an incomplete global database is just gonna create a bigger black box of unpaid royalties that someone other than the songwriter will get to keep.
Finally–safe harbors have been a disaster for songwriters and performers. The biggest music streaming service in the world is YouTube. YouTube is by a far the worst paying streaming services. This is precisely because the DMCA safe harbor allows them to operate like a mob protection racket: “Some nice songs you got there, sign this usurious contract, cause we’d hate to see our users upload your songs, and then you don’t get any royalties.”
Another safe harbor for streaming services? No thanks.
If the management at these streaming services were truly exercising their fiduciary responsibility to their own shareholders they would be OPPOSED to YouTube/Google’s extra-legal interpretation of the DMCA safe harbor. How will any of these streaming services achieve profitability when they have to (unfairly) compete with YouTube/Google.
This is an area where the interests of songwriters and interactive streaming services are aligned. If streaming services took the long view, they would join songwriters in demanding congress/FTC/DOJ force YouTube/Google to end this shakedown practice. They wouldn’t demand their own safe harbor. Another safe harbor just invites a new round of lawsuits, political cronyism and abuse.
What do songwriters want?
Songwriters wants are simple: fair pay and control of their works. That’s pretty much what any property owner wants.
When it comes to streaming services it generally implies two things,
- Higher royalties; and
- A right for the songwriter to withdraw their songs from digital services that pay too little.
As it stands performers and labels have the right to NOT license their work to interactive services if they don’t like the terms of direct licenses. In contrast songwriters have no right to withdraw their works from compulsory licenses. Further the federal government caps songwriters share of streaming revenues at 10.5%.
You think the minimum wage sucks? Try working under a maximum wage.
This is patently absurd if you think about it. Would the federal government mandate that all bands sell their t-shirts for $5? Would the federal government mandate that Whole Foods cap pay to their employees at 10.5% pro-rata share of their gross revenues? Fundamentally this is an extremely regressive practice. I find it particularly amusing that some of the strongest supporters of the regressive status quo are Congressional Democrats.
The Known Unknowns
While higher revenue and the right to withdraw a work seem pretty simple on the surface, closer inspection reveals problems.
Raise songwriters royalties how high? What is fair compensation when there has never been a free market for mechanicals royalties. From 1909 to 1976 the statutory mechanical royalty for phonographs was stuck at 2 cents a copy. This wage and price control has cast a long dark depressive shadow on all mechanical rate setting proceedings after 1976.
But it goes deeper than that. Should all songs and songwriters receive the same rate per spin? Always? Do individual songwriters value each of their own songs the same? Do consumers place the same value on each song? How about streaming services? Do they value songs by certain artists over others? Why are songs treated as a commodity like light Brent crude or pork bellies?
My guilty pleasure is Taylor Swift’s “Shake it Off.” I paid $1.29 for the song. I thought my band Camper Van Beethoven should cover it. Would I have paid $5.99? Probably not (I’d have streamed it on YouTube). Would I pay $5.99 a song for an unreleased Captain Beefheart song? Probably. Would I pay $50 dollars for a sequel to Sleep’s Dopesmoker? I don’t know. Maybe. But the compulsory license mandates flat rate pricing.
Economists generally agree that flat pricing is wasteful practice. Flat price streaming (both rates to songwriters and consumer) is a kind of POTS economics in an Ebay world. Companies as diverse as Amazon and Delta Airlines employ dynamic pricing algorithms that change prices instantaneously. Why can’t songwriters have more control over the pricing of their work? Why is the music business stuck with a 1970s Columbia House Record Club model? “Get 19 albums for a dollar, when you sign up for a year!’ I’m pretty sure we can do better than that.
Second complication, if a songwriter wishes to withdraw their catalogue (or a work) from a federal compulsory license, who do they notify? Is the withdrawal immediate? Do they need to notify every service? Do they notify the Copyright Office? Aren’t we back to the same impossible database problem explored in the previous section?
No. A database of songs or songwriter/publishers who have “opted out” would be smaller. Sure it would be a dynamic database, but if the right to opt out is contingent on registration with a public database, it would by definition be 100% accurate. Especially since songwriters recapture a valuable right by doing so. This is a real incentive, (unlike this proposal which seeks to “punish” rights holders by withholding puny streaming royalty checks that aren’t simply worth going to the bank to cash).
The idea of this compromise is that we are fundamentally flummoxed by the question “What is this song worth?” So if we are going to come out of our trenches and try to reach a compromise we must set up a mechanism that helps us discover the answer to the question. Keep in mind there is not a positivist answer to this question. It is constructivist by nature. The answer is whatever the participants agree is fair. As Ari Emmanuel once said, “Fair is where we end up.” It is whatever the licensor and licensee deem fair on that day at that time.
Let me say right at the start that no one is gonna be completely happy with this solution. There are elements here to which I would normally be ideologically opposed.
This compromise is largely based on something called Extended Collective Licensing that is used in some Nordic countries. Other elements are pulled from proposals the NMPA seems to be floating. This is an extremely rough draft, and I’m making substantial concessions to streaming services, not because I necessarily want to give them concessions, but because this is what I think an acceptable compromise would look like
Anyone with a valid compulsory license gets to keep it (although the “address unknown” mass NOIs have to be dealt with separately). Going forward for new licenses and works:
- Do away with the 115 NOI system for streaming mechanicals.
- Establish a quasi governmental body like SoundExchange to license and administer a compulsory streaming mechanical. The CRB would continue to set rates. But this should eventually sunset. For simplicity sake let’s call this entity SongExchange.
- However songwriters/publishers could serve notices and opt out songs or catalogue from SongExchange and the compulsory license.
- In order to exercise this right all songwriters and publishers would have to agree before opting out,
- Substantial advance notice must be given to SongExchange and streaming services.
- A necessary condition of opting out of SongExchange would be to register in a machine readable public database and opt out information once registered must be kept current.
- This database will belong to the public and shall not be privatized.
- If the owner of the song fails to keep information current it reverts back to SongExchange for administration.
- SongExchange would administer the database and administrative fees would be split between songwriters and services.
- Opted out songs could be directly administered by owners or assigned to third parties for “extended” collective negotiation.
- Songwriters/publishers may form new non-profits or co-ops to administer these rights.
- Alternately songwriters/publishers may assign these rights to the existing non-profit PROs, ASCAP and BMI. The DOJ consent decrees should be modified or dropped in order to make this possible.
- Songwriters/Publishers may assign these rights to private PROs like SESAC or GMR.
- Like labor unions, professional sports leagues and farm co-ops, these third parties should be given explicit exemption from the Sherman act and other anti-trust restrictions.
- The CRB must be allowed to consider as evidence rates set by songwriters and composers that opted out of the SongExchange (“race to the middle”)
- SongExchange must hire an independent 3rd party to conduct a royalty compliance examination of the services and the services must accept those audits, and any group of songwriters can conduct a group audit under the same conditions.
- SongExchange may impose penalties on services that have excessive unmatched and unidentified compositions. These penalties may be used for overhead.
- SongExchange will hold unclaimed royalties for a period of three years. Song Exchange will post a list of unclaimed works. After three years the unclaimed royalties are transferred to the applicable Secretary of State as unclaimed property to be held as other unclaimed property (like unclaimed utility deposits or bank accounts).
This proposal is obviously disruptive to the status quo. I’m not opposed to phasing it in over a few years. Here’s what I think this proposal does that is not possible in the current system:
- It simplifies the licensing of songs by eliminating the per-song NOI system;
- Provides an accurate database to assist in licensing songs;
- Is pro-competitive and eliminates need for ATR supervision;
- Allows true price discovery;
- Encourages competition among rights holders and third party licensing authorities; and
- Allows songwriters to collectively bargain if they wish.
It should also be added, that unlike any other proposal offered so far, this is the only proposal that can achieve a 100% licensing rate by streaming services. All other proposals (even those floated by the streaming services themselves) would leave streaming services open to copyright infringement lawsuits.