Austin Music Lawyer
We practice in transactions at the nexus of the traditional music industry, content based technology and public policy.
Lawyers often say that bad facts make bad law – meaning that unusual or unlikely details of a case can shape precedent in unpredictable ways. But bad facts can also make for bad contracts, to judge by the contractual restrictions on re-recording that major labels may be adopting in the wake of the success of Taylor Swift‘s “Taylor’s Version” of her albums.
Entertainment Law & Finance, May 2023
StubHub brought Silicon Valley lobbying tactics to the Georgia Legislature with HB 398 and SB 183 earlier this year but thankfully they didn't put one over on the Members of the Legislature. The stated purpose of these Georgia bills is to remove resale restrictions on concert or athletic tickets and limit the resale to reseller market places like Stub Hub. This would, among other things, effectively have prevented artists from taking tickets off manifest for resale through their fan clubs, including at face price.
HB 398 and SB 183 are dead, but this outbreak of sanity and common sense was helped along by the testimony of Mala Sharma of Georgia Music Partners and David Lowery (who lives in Athens and teaches at the University of Georgia at Athens Terry College of Business). The discussion at the legislature regarding property rights and tickets was brought back to focus on the initial sale and the rights of the performing artist or sports team to control transferability in the first place as an essential property right.
Resale Royalties for Artists and Venues
Mala and David mentioned a very interesting concept in their testimony: Should resale royalties be paid to artists and venues when tickets are resold? To my knowledge, this has never been done, but such a resale royalty might encourage artists or sports teams to permit transferability for some or all of their tickets. It would also help to value that property right. So how would that work?
David Lowery made this statement in his written testimony:
Paying it Forward: Resale Royalties for Scalpers: StubHub may refuse to police itself but the State of Georgia can recover some of the value of StubHub’s free riding by establishing a resale royalty to be distributed to artists and venues for transactions occurring in Georgia. This is a very intriguing idea that would essentially force scalpers to return some of the value they have extracted from the artist’s brand. I speak of the resale royalty as returned to artists and venues, but I am program-agnostic. The payment should also be returned to the performers, universities, or taxpayer funded venues around the state.
The resale royalty could be a way to continue to support communities that were hard hit by COVID and venues that survived based on the Save Our Stages funding. It would be better to find this support from free-riders than from hard working Georgia taxpayers.
The "free riding" reference is to StubHub making a market for ticket resellers who price tickets far, far above the price set by artists, and then capturing the value of the artist and venue's marketing efforts and brand value without ever contributing to the brand investment or marketing spend. A resale royalty paid by scalpers would help rebalance the investment by taking a share of the reseller's gross income.
How would that work? Here's some draft language that might be a starting place referring the collection and distribution to Georgia's Department of Revenue, which is sort of the State-level IRS. I chose 5% as the base rate, but that could be more or less.
A ticket marketplace or other commercial reseller shall remit and pay a minimum resale royalty on all sales of tickets for which the reseller is required to pay sales tax in this State. Such minimum royalty shall be calculated as the greater of an amount equal to five percent (5%) of the total price paid by the seller of the ticket, inclusive of fees paid to the ticket marketplace, or the royalty uplift. The royalty uplift is calculated by dividing the total price by the face price of the ticket concerned, and multiplying the quotient by five percent (5%). By way of example and without limitation, if the face price is $100 and the total price is $500, the quotient is 5 and the minimum resale royalty is 25% of $500.
The minimum resale royalty shall be paid by the ticket marketplace reseller to the Department of Revenue accompanied by an itemized statement setting forth the athletic team, performing artist, or charitable organization promoting or performing at the ticketed event as reflected on the face of the ticket. The Department of Revenue shall pay 50% of the minimum resale royalty to such athletic team, performing artist, or charitable organization promoting or performing at the ticketed event and 50% to the venue at which the event occurs. The Department of Revenue shall promulgate appropriate forms and regulations to implement such payments.
Private individuals selling or gifting tickets outside of a ticket marketplace or other commercial reseller shall not be required to pay the resale royalty.
It's important to note that this approach does not interfere with the property rights concerned in the transaction. First, the artist (or venue) selling the ticket may still set the terms of the ticket sale including both the price and the transferability. If the artist does not want the ticket to be transferable, they still have the right to set that prohibition. If they do allow the ticket to be transferrable, then a resale royalty rewards them for that decision to give up their right to make the ticket nontransferable. This may actually motivate more artists to permit transferability. Because the resale royalty is conceived as a minimum, the artist could always elect a higher resale royalty than 5% as a condition of transferability.
Neither does the resale royalty interfere with property rights of the ticket buyer or seller once the artist permits transferability.
Because the ticket marketplace already pays a sales tax to the Department of Revenue (or comparable state agency) the resale royalty would simply be an additional payment at a slight incremental cost. Of course, processing the resale royalty would simply be a cost of doing business for the marketplace (like StubHub), so no service charge or ticket fee could apply to reduce the royalty payment. The Department of Revenue is in a better position to track down the artists and venues, particularly artists who reside in Georgia and venues located in Georgia since they already pay income or other taxes to the Department of Revenue. Plus the Department of Revenue handles unclaimed property in Georgia so could retain any unpaid royalties.
This seems like a novel approach to protecting the investment of the artist and the venues, including taxpayer financed community centers and arenas in tertiary markets.
Anti-Money Laundering Rules
David also highlighted the need for anti-corruption measures to discourage money laundering through ticket resellers:
Anticorruption Protections: When observing the amount of money and the number of high value transactions fixed by StubHub and other online marketplaces—effectively in cash—I am struck by the potential for bad actors to use the platform to hide cash transactions. I see no reason why StubHub should be treated differently than a bank in reporting these transactions to authorities such as the Georgia Bureau of Investigation or the Department of Revenue.
I would encourage the Committee to work with these agencies to determine the need for more detailed reporting of the origin and destination of higher dollar transactions, such as $10,000 in a single deal or series of deals closed by StubHub and its progeny.
One idea would be for a ticket marketplace or other commercial reseller to report to the Department of Revenue (or state sales tax agency) all transactions by any one buyer or seller of tickets that exceed $10,000 in any 12 month period. That reporting could include the name, address, and bank or credit card information for each such seller or buyer and any other information required by the Department of Revenue.
Securities Law Issues
It is also worth noting that some marketmakers allow for the sale of a future contract for tickets that have not gone on sale as yet. The future contract (like an option or a commodities future) allows someone to purchase the right to buy a ticket once the tickets are offered for sale. This seems to implicate securities law issues, broker-dealer regulations and potentially the general solicitation rule.
This subject did not come up in the legislation itself, unsurprisingly, but it did come up in the discussion during the bill’s markup before it was tabled.
There does seem to be a question as to what to call this future contract and if any federal or State securities laws are implicated.
Ticket resellers appear to be behind a number of state bills that were introduced and that are winding their way toward resolutions as state legislatures reach the conclusion of their 2023 general session (e.g., “sine die” which has already occurred in Georgia and is about to occur in Texas, see attached grid).
Hypebot, May 16, 2022 Available here.
Newsletter of the International Federation of Musicians, June 7, 2021
Op Ed about co-writing a study for WIPO on streaming remuneration.
Chris recently filed written evidence in the UK in the inquiry held by the Digital, Culture, Media and Sport Committee in the UK Parliament. The DCMS Committee has been kind enough to post the written evidence online, currently at this link.
First appeared in Morning Consult 8/28/20
Google’s appeal to the U.S. Supreme Court of two Federal Circuit decisions in Oracle’s favor is turning into the most consequential copyright case of the court’s term — if not the decade. The appeal turns in part on whether the Supreme Court will uphold the Federal Circuit’s definition of fair use for creators and reject Google’s dubious assertion of “industrial strength” fair use.
I co-wrote an amicus brief on the fair use question on behalf of independent songwriters supporting Oracle in the appeal. Our conclusion was that the Supreme Court should affirm the Federal Circuit’s extensive analysis and hold for Oracle because Google masks its monopoly commercial interest in industrial-strength fair use that actually violates fair use principles.
The story begins 15 years ago. Google had a strategic problem. The company had focused on dominating the desktop search market. Google needed an industrial-strength booster for its business because smartphones, especially the iPhone, were relentlessly eating its corporate lunch. Google bought Android Inc. in 2005 to extend its dominance over search — some might say its monopoly — to these mobile platforms. It worked — Android’s market share has hovered around 85 percent for many years, with well over 2 billion Android devices.
But how Google acquired that industrial boost for Android is the core issue in the Oracle case. After acquiring Android, Google tried to make a license deal for Sun Microsystems’ Java operating system (later acquired by Oracle). Google didn’t like Sun’s deal. So Google simply took a verbatim chunk of the Java declaring code, and walled off Android from Java. That’s why Google got sued and that’s why the case is before the court. Google has been making excuses for that industrial-strength taking ever since.
Why would a public company engage in an overt taking of Oracle’s code? The same reason Willie Sutton robbed banks. Because that’s where the money is. There are untold riches in running the Internet of Other People’s Things.
Google chose to take rather than innovate. Google’s supporters released a study of the self-described “fair use industries” — an Orwellian oxymoron, but one that Google firmly embraces. Google’s taking is not transformative but it is industrial strength.
We have seen this movie before. It’s called the value gap. It’s called a YouTube class-action brought by an independent composer. It’s called Google Books. It’s called 4 billion takedown notices for copyright infringement. It’s called selling advertising on pirate sites like Megaupload (as alleged in the Megaupload indictment). It’s called business as usual for Google by distorting exceptions to the rights of authors for Google’s enormous commercial benefit. Google now positions itself to the Supreme Court as a champion of innovation, but creators standing with Oracle know that for Google, “innovation” has become an empty vessel that it fills with whatever shibboleth it can carelessly manipulate to excuse its latest outrage.
Let’s remember that the core public policy justification for the fair use defense is to advance the public interest. As the leading fair use commentator Judge Pierre Leval teaches, that’s why fair use analysis is devoted to determining “whether, and how powerfully, a finding of fair use would serve or disserve the objectives of the copyright.” You can support robust fair use without supporting Google’s position.
Google would have the court believe that its fair use defense absolves it from liability for the industrial-strength taking of Oracle’s copyright — because somehow the public interest was furthered by “promoting software innovation,” often called “permissionless innovation” (a phrase straight out of Orwell’s Newspeak). Google would have the court conflate Google’s vast commercial private interest with the public objectives of copyright. Because the internet.
How the Supreme Court rules on Google’s fair use issue will have wide-ranging implications across all works of authorship if for no other reason than Google will dine out for years to come on a ruling in its favor. Photographers, authors, illustrators, documentarians — all will be on the menu.
Despite Google’s protestations that it is really just protecting innovation, what is good for Google is not synonymous with what is good for the public interest — any more than “what’s good for General Motors is good for America,” or more appropriately, “what’s good for General Bullmoose is good for the USA.”
Cite as 29 Tex. Ent. & Sports Law Journal 4 (No. 1, 2020)
Even though they have a long history, Presidential Signing Statements are not exactly front and center in every civics class or constitutional public law class in America. You may be hearing about them for the first time now. But that doesn’t mean they have not been an important part of Constitutional law-making and jurisprudence.
Presidential Signing Statements were first used by President James Monroe in 1822 in the form of a “special message” to the Senate. Presidents Andrew Jackson, John Tyler and Ulysses Grant also issued signing statements, but they were used infrequently until the 20th Century. Then their use picked up quite a bit starting with President Theodore Roosevelt and continuing to the present day. So the use of Signing Statements is quite bipartisan. While Signing Statements may not themselves have any actionable legal effect, they should not be ignored, either.
The MMA Presidential Signing Statement
Not surprisingly, there is a Presidential Signing Statement accompanying the Music Modernization Act (“MMA”) specifically relating to Title I and at that specifically relating to the MLC board appointments. The relevant language is:
One provision, section 102, authorizes the board of directors of the designated mechanical licensing collective to adopt bylaws for the selection of new directors subsequent to the initial designation of the collective and its directors by the Register of Copyrights and with the approval of the Librarian of Congress (Librarian). Because the directors are inferior officers under the Appointments Clause of the Constitution, the Librarian must approve each subsequent selection of a new director. I expect that the Register of Copyrights will work with the collective, once it has been designated, to ensure that the Librarian retains the ultimate authority, as required by the Constitution, to appoint and remove all directors.
Let’s explore why we should care about this guidance.
According to Digital Music News, there have been changes at the Mechanical Licensing Collective, Inc. (“MLCI”) the private non-profit permitted under Title I of the MMA:
[I]t appears that two separate MLC board members are jumping ship. The details are just emerging and remain unconfirmed, though it appears that two members — one representing indie songwriters and the other on the publishing side — are out of the organization.
Because the board composition of MLCI is preemptively set by the U.S. Copyright Act along with many other aspects of MLCI’s operating mandate, the question of replacing board members may be arising sooner than anyone expected. As MLCI is a creature of statute, it should not be controversial that law-makers play an ongoing role in its governance.
The Copyright Office Weighs In
The Copyright Office addressed board appointments for MLCI in its first request for information for the designation of the Mechanical Licensing Collective (83 CFR 65747, 65750 (December 21, 2018) available at https://www.govinfo.gov/content/pkg/FR-2018-12-21/pdf/2018-27743.pdf):
The MLC board is authorized to adopt bylaws for the selection of new directors subsequent to the initial designation of the MLC. The Presidential Signing Statement accompanying enactment of the MMA states that directors of the MLC are inferior officers under the Appointments Clause of the Constitution, and that the Librarian of Congress must approve each subsequent selection of a new director. It also suggests that the Register work with the MLC, once designated, to address issues related to board succession.
When you consider that MLCI is, for all practical purposes, a kind of hybrid quasi-governmental organization (or what the Brits might call a “quango”), the stated position of the President, the Librarian of Congress and the Copyright Office should not be surprising.
Why the Controversy?
As the Songwriters Guild of America notes in comments to the Copyright Office in part relating to the Presidential Signing Statement (my emphasis):
Further, it seems of particular importance that the Executive Branch also regards the careful, post-designation oversight of the Mechanical Collective board and committee members by the Librarian of Congress and the Register as a crucial prerequisite to ensuring that conflicts of interest and bias among such members not poison the ability of the Collective to fulfill its statutory obligations for fairness, transparency and accountability.
The Presidential Signing Statement, in fact, asserts unequivocally that “I expect that the Register of Copyrights will work with the collective, once it has been designated, to ensure that the Librarian retains the ultimate authority, as required by the Constitution, to appoint and remove all directors.”
SGA regards it as a significant red flag that the NMPA-MLC submission to the Copyright Office devotes the equivalent of ten full pages of text principally in attempting to refute this governmental oversight authority, and regards the expression of such a position by NMPA/MLC as arguably indicative of an organization more inclined towards opaque, insider management control than one devoted to fairness, transparency and accountability.
So the Presidential Signing Statement to the MMA is obviously of great import given the amount of ink that has been spilled on the subject. Let’s spill some more.
How might this oversight be given effect and will it be in the public record or an informal process behind closed doors? Presumably it should be done in the normal course by a cooperative and voluntary collaboration between the MLC and ultimately the Librarian. Minutes of such collaboration could easily be placed in the Federal Register or some other public record on the Copyright Office website. Failing that collaboration, it could be done by either the Department of Justice (unlikely) or by individuals (more likely) asking an Article III court to rule on the issue.
Of course, the issue should not delay the Copyright Royalty Judges from proceeding with their assessment determination to fund the MLC pursuant to the controversial voluntary settlement or otherwise. One could imagine an oversight role for the CRJs given that Congress charged them with watching the purse strings and the quantitative implies the qualitative. The CRJs have until until July 2020 to rule on the initial administrative assessment and appeal seems less likely today given the voluntary settlement and the elimination of any potential objectors.
Since the Title I proponents drafted the bill to require a certain number of board seats to be filled by certain categories of persons approved by Congress in a Madisonian balance of power, the Presidential Signing Statement seems well grounded and furthers the Congressional mandate.
Yet there is this conflict over the Presidential Signing Statement. What are the implications?
A Page of History is Worth A Volume of Logic
The President’s relationship to legislation is binary—sign it or veto it. Presidential Signing Statements are historically used as an alternative to the exercise of the President’s veto power and there’s the rub.
Signing Statements effectively give the President the last word on legislation as the President signs a bill into law. Two competing policies are at work in Presidential Signing Statements—the veto power (set forth in the presentment clause, Article I, Sec. 7, clause 2), and the separation of powers.
Unlike some governors, the President does not enjoy the “line item veto” which permits an executive to blue pencil the bits she doesn’t like in legislation presented for signature. (But they tried–Line Item Veto Act ruled unconstitutional violation of presentment clause in Clinton v. City of New York, 524 U.S. 417 (1998).) The President can’t rewrite the laws passed by Congress, but must veto the bill altogether. Attempting to both reject a provision of a new law as unconstitutional, announce the President’s intention not to enforce that provision AND sign the bill without vetoing it is where presidents typically run into trouble.
Broadly speaking, Presidential Signing Statements can either be a President’s controversial objection to a bill or prospective interpretive guidance. Signing Statements that create controversy are usually a refusal by the President to enforce the law the President just signed because the President doesn’t like it but doesn’t want to veto it. Or to declare that the President thinks the law is unconstitutional and will not enforce it for that reason—but signed it anyway.
The President can also use the Signing Statement to define or interpret a key term in legislation in a particular way that benefits the President’s policy goals or political allies. President Truman, for example, interpreted a statutory definition in a way that benefited organized labor which was later enforced by courts in line with the Signing Statement. President Carter used funds for the benefit of Vietnam resisters in defiance of Congress, but courts later upheld the practice—in cases defended by the Carter Justice Department. The practice of using Presidential Signing Statements is now routine and has been criticized to no avail for every administration in the 21st Century including Bush II, Obama and now Trump.
Since the 1980s, it has become common for Presidents to issue dozens if not hundreds of Presidential Signing Statements during their Administration. So it should come as no surprise if the Department of Justice drafted up the statement for the MMA prior to it being presented to the President to be signed into law. (See the American Presidency Project archives https://www.presidency.ucsb.edu/documents/presidential-documents-archive-guidebook/presidential-signing-statements-hoover-1929-obama)
Defiance or Collaboration?
What does this mean for the MMA? The President certainly did not call out the statutorily required board membership of the MLC as an unconstitutional overreach that he would not enforce. To the contrary, the MMA Signing Statement expresses the President’s desire that the legislation comply with the requirements of the Constitution.
Moreover, the MMA Presidential Signing Statement is not a declaration about what the President will or won’t enforce but rather interprets a particular section of a long and winding piece of legislation. (Title I principally amended Section 115 of the Copyright Act—now longer than the entire 1909 Copyright Act.) This kind of interpretation seems to be consistent with the practices of prior Presidents of both parties, not an end-run around either the veto power or separation of powers.
Failing to acknowledge the admonition of the signing statement would seem an unnecessary collision both with long-standing jurisprudence and with a sensible recommendation from the President of how the Librarian, the Copyright Office and the Justice Department expect to approach the issue in collaboration with the MLCI. That’s possibly why the Copyright Office restated the Signing Statement in the RFP.
Title I of the MMA is a highly technical amendment to a highly technical statute. A little interpretive guidance is probably a good thing. Collaboration certainly makes more sense than defiance.
[Posted 10/23/19 in NY Daily News]
The meltdown of WeWork’s CEO and Mark Zuckerberg’s bizarre threat to sue the U.S. offer a teachable moment: When you concentrate vast and unaccountable control over major companies in founders — no matter how creative or capable — bad things happen.
In Silicon Valley, the problem starts with “supervoting” stock structures that let the CEO (mostly) boy wonders raise mountains of cash from star-struck investors without giving up meaningful control over the company. The trick is a gimmicky “dual-share” stock structure in which the insiders’ own shares have powerful voting rights but ordinary investors are stuck on the sidelines. SEC Commissioner Robert Jackson has warned this dual class in effect creates “corporate royalty.”George Orwell would probably say it’s just another example of the “Some Animals Are More Equal Than Others” corruption that eventually poisons pretty much every revolution.
In the aftermath of Facebook’s recent $5 billion FTC settlement and the staggering penalties against Google’s repeated malfeasance in Europe, Congress is looking for meaningful ways to rein in Big Tech; Democratic presidential candidates, led by Elizabeth Warren, are doing the same. A good place to start would be reforming the supervoting capital structure. By creating executive-kings that can’t be outvoted or overruled by shareholders, super-voting stock is where many of the worst excesses begin.
How did we get here? At Google, insiders like Larry Page, Sergey Brin and Eric Schmidt own supervoting stock that gives them control over 60% of the total shareholder vote. As a result, their insider block wins all shareholder votes, including those to appoint the board of directors who are supposed to mind the store.
Google’s charter gives supervoting “Class B” shares 10 votes per share. Class B shares are not available to the public, which can only invest in Class A (one vote per share) or Class C (no votes). The result is that the Class B royalty dictates policy, and shareholder meetings are one-way affairs that in which they tell everyone else how things are going to be and vote down any move to change the status quo.
Facebook is similar except all that power is concentrated in a single person — Zuckerberg. In late 2013, before Facebook joined the S&P 500, Zuckerberg exercised an option to buy 60 million supervoting Facebook shares, making his stranglehold on the company essentially permanent.
One could say that if investors want to buy into the Google or Facebook gravy trains with full disclosure, let them. But the rest of us bear the externalities of the river of corporate pollution that flows out of mismanaged companies where a new out-of-control horror story seems to emerge every week. It doesn’t take a genius to predict where no accountability and moral hazard on steroids will take you.
It’s not just Google and Facebook, WeWork and Cloudflare; like so many trends in Silicon Valley, founders of private startups and newly public companies copy what their role models get away with. Spotify, Uber, Linkedin, Lyft, Snapchat: the list of supervoting share issuers goes on and on.
One can understand that founders need relatively flexible control to get off the ground and develop their businesses. They need to be nimble as the landscape changes and competitors threaten. But once they elect to take advantage of the public stock markets, unaccountable absolutist rule simply has to end.
At a minimum, Congress should make it easier for boards to take away supervoting shares from irresponsible executives — as WeWork’s board finally did when the company’s IPO fell apart amid stories about insider excesses and a balance sheet that did not appear to add up.
If our representatives can stomach the fight, an even better path would be to ban the supervoting device altogether once companies go public. Make giving up all supervoting shares a condition of getting approval for an IPO. True equality in voting rights would empower shareholders and strengthen boards of directors to run these companies in ethical and responsible ways — and would help us all get started cleaning up the Big Tech mess.
Castle is a lawyer in Austin, Tex.
If you want to participate in the negotiation of the “assessment” for the operating costs of the Mechanical Licensing Collective, the clock is ticking—July 23, 2019 is your deadline to file.
There was no advance commitment or agreement on the budget or business plan for the Mechanical Licensing Collective (MLC) under Title I of the Music Modernization Act. Recall that a major selling point of the MLC was that the services using the compulsory license for streaming mechanicals would pay for the startup and operating costs of the MLC, including the elusive global rights database that has been tried so many times before at great expense.
It now appears that the clock is ticking on a voluntary agreement before the parties have to go before the Copyright Royalty Judges to be told what the budget (or the “assessment”) is to be. The Copyright Royalty Board has beat the July 8 deadline for noticing the proceeding and has posted the notice and the rules for the hearing.
The “Notice announcing commencement of Initial Administrative Assessment proceeding and requesting Petitions to Participate” can be found here:
The regulations require the participation of the MLC and the Digital Licensee Coordinator (DLC) in the proceeding and permit the participation of copyright owners, digital music providers, and significant nonblanket licensees. 37 CFR 355.2(c)–(d).
The Judges hereby announce commencement of the proceeding, direct the MLC and the DLC to file Petitions to Participate, and request Petitions to Participate from any other eligible participant with a significant interest in the determination of the Initial Administrative Assessment…
Any participant that is an individual may represent herself or himself. All other participants must be represented by counsel….
Petitions to Participate and the filing fee are due on or before July 23, 2019.
The CRJ’s rules relating to the proceeding can be found here and have some relevant language relating to who can participate in the negotiation of the administrative assessment negotiations in addition to the MLC and DLC:
[T]he Judges believe that the views of other participants may be helpful, and perhaps essential, for the Judges to determine whether good cause exists to exercise their discretion to reject a settlement. The Judges, therefore, have modified [the regulations for the settlement negotiations and proceeding] to clarify that participants other than the MLC and DLC may participate in settlement negotiations and may comment on any resulting settlement.
This proceeding is relevant for both copyright owners and digital services not represented by the DLC. Copyright owners will have to pay to ingest their metadata into the new global rights database when it is built. As I noted in a 2018 Newsmax post, all digital services will have to bear an allocation of the startup and operating costs before they can take advantage of the compulsory license—in addition to an unknown membership fee in the DLC. And then there’s the songwriter royalties, of course..
Both costs are unknown until there’s a voluntary agreement between the MLC and the DLC (all others being excluded from those negotiations) or the CRJs becomes final with all appeals exhausted.
Chris published Highlights of Managing Change Under the Music Modernization Act’s Music Licensing Collective in the current issue of the Texas Entertainment & Sports Law Section Journal. A version of this article appears as How Will the Music Modernization Act’s Mechanical Licensing Collective Work? in 34 Entertainment Law & Finance 1 (No. 9, Dec. 2018.)
by Chris Castle
(Published by Hypebot as Arithmetic Online: The Ethical Pool Solution to the Streaming Royalty Calculation)
“Sick of my money funding crap.”
A Fan’s Tweet
Subscription services are one of the few secular trends in the current economy that is not yet reactive to trade wars or interest rates. Subscription services are found in many areas of the economy, but music drives some of the big ones like Spotify, Amazon and especially the razor-and-razorblades plays like Apple.
But per-stream royalties do not come close to making up for the CD and download royalties they cannibalize. Not only do subscription retail rates need to increase, but it's also time for a major change in the way artist's streaming royalties are calculated from what is essentially a market share approach to one that is more fair. I call my version of this the “Ethical Pool.” I’ll only address subscription revenue here, but a similar approach could be devised for ad-supported or other business models.
Artists’ dismal streaming royalties on music subscription services are largely based on a simple calculation: A per-stream payment derived from a share of the service’s revenue prorated by number of streams. Artists get a portion of a service’s monthly revenue (at least the revenue the service discloses) based on a ratio of your plays to all the plays. Your plays will always be a lot smaller than the total plays. (This is essentially what Sharky Laguana referred to as the “Big Pool.”)
Sounds simple, but mixed with the near-payola of Spotify’s playlist culture and Pandora’s “steering” deals, it’s really not. Negotiating leverage allows big stakeholders to tweak the basic calculation with floors, advances (aka breakage), nonrecoupable payments and other twists and turns to avoid a pure revenue share.
It also must be said that stock analysts and venture investors always—always—blame “high” royalties for loss-making in music services. This misapprehension ignores high overhead such as Spotify’s 10 floors of 4 World Trade Center or high bonus payments such as Daniel Ek’s $1,000,000 bonus paid for failing to accomplish half of his incentive goals stated in the Spotify SEC documents.
Of course all these machinations happen behind the scenes. Fans are not aware and it's virtually impossible for any label or publisher to tell an artist or songwriter what their per-stream rate is or is going to be.
Fans Don’t Like It
So neither fans nor artists are happy with the current revenue share model. Many artists will tell you that the playlist culture and revenue share model are actually destructive.
Dedicated fans often don’t like it either (after they understand it) because it gives the lie to supporting your favorite artist by streaming their music. Artists don’t like it because unless you have a massive pop or hip hop hit, all you can aspire to is a royalty rate that starts in the third decimal place from the right if not the fourth. This is compounded for songwriters. (See Universal Music Publishing’s Jody Gerson on streaming royalties for songwriters.)
Simply put, if a fan pays their subscription and listens to 20 artists in a month, that fan likely believes that their subscription is shared by those 20 artists and not by 200,000 artists, 99.99% of whom that fan never listened to and probably never will, similar to Sharky’s “Subscriber Pool.”
This is why some artists like Sharky Laguana (and their managers) have begun arguing for replacing the status quo with “user-centric” royalties that more directly correlate fan listening to artist payments. I have a version of this idea I call the “Ethical Pool.”
How Did We Get Here?
How in the world did we get to the status quo? The revenue share concept started in the earliest days of commercial music platforms. These services didn’t want to pay the customary “penny rate” (as is typical for compilation records, for example), because a fixed penny rate might result in the service owing more than they made--particularly if they wanted to give the music away for free to compete with massive pirate sites.
Paying more than you make doesn’t fit very well with a pitch for a Web 2.0, advertising driven model: All you can eat of all the world’s music for free or very little, or “Own Nothing, Have Everything,” for example. It also works poorly if you think that artists should be grateful to make any money at all rather than be pirated.
Revenue share deals for big stakeholders have some bells and whistles that leverage can get you, like per-subscriber minimums, conversion goals, top up fees, limits on free trials, cutbacks on “off the top” revenue reductions, and the percentage of revenue in the pool (50%—60%-ish). Even so, the basic royalty calculation in a revenue share model is essentially this equation calculated on a monthly basis:
(Net Revenue * [Your Streams/All Streams])
Or ([Net Revenue/All Streams] * Your Streams)
In other words all the money is shared by all the artists.
Sounds fair, right?
Wrong. First, all artists may be equal, but on streaming services, some are more equal than others. Regardless of the downside protection like per-subscriber or per-stream minima, the revenue share model has an inherent bias for the most popular getting the most money out of the “Big Pool.” (This is true without taking into account the unmatched.)
And of course it must be said that the more of those artists are signed to any one label, the bigger that label’s take is of the Big Pool. So the bigger the label, the more they like streaming.
Conversely, the smaller the label the lower the take. This is destructive for small labels or independent artists. That’s why you see some artists complaining bitterly about a royalty rate that doesn’t have a positive integer until you get three or four decimal places to the right. Why drive fans away from higher margin CDs, vinyl or permanent downloads to a revenue share disaster on streaming?
Yet it increasingly seems that we are all stuck with the nonsensical streaming revenue share model.
Do Fans Think It’s Wrong?
There’s nothing particularly nefarious about this—them’s the rules and rev share deals have been in place for many years, mostly because the idea got started when the main business of the recorded music business was selling high margin goods like CDs or even downloads. Low margin streaming didn’t matter much until the last couple years.
It was only a question of time until that high margin business died due to the industry’s willingness to accept fluctuating micropennies as compensation for the low-to-no margin streaming business. (I say “no margin business” because the costs of accounting for streaming royalties may well exceed the margin—or even the payable royalty—on a per-stream basis when all transaction costs are considered as Professor Coase might observe.)
So understand—the revenue share model is essentially a market share distribution. Which is fine, except that in many cases, and I would argue a growing number of cases, when the fans find about about it, the fans don’t like it. They pay their monthly subscription fee and they think their money goes to the artists they actually listen to during the month. Which is not untrue, but it is not paid in the ratio that the fan might believe. Fans could easily get confused about this and the Spotifys of this world are not rushing to correct that confusion.
Here’s the other fact about that rev share equation: over time, the quotient is almost certain to produce an ever-declining per-stream royalty. Why?
Simple.
If the month-over-month rate of change in revenue (the numerator) is less than the month-over-month rate of change in the total number of streams or sound recordings streamed on the service (the denominator), the per-stream rate will decline over those months. This is because there will be more recordings in later months sharing a pot of money that hasn’t increased as rapidly as the number of streams.
As the number of recordings released will always increase over time for a service that licenses the total output of all major and indie labels (and independent artists), it is likely that the total number of recordings streamed will increase at a rate that exceeds the rate of change of the net revenue to be allocated. If there are more recordings, it is also likely that there will be more streams.
So streaming royalties in the Big Pool model will likely (and some might say necessarily will) decline over time. That’s demonstrated by declining royalties documented in The Trichordist’s “Streaming Price Bible”among other evidence.
Thus the fan’s dissatisfaction with the use of their money is already rising and is likely to continue to rise further over time.
User-Centric Royalties and the Ethical Pool
How to fix this? One idea would be to give fans what they want, or at least try harder. A first step would be to let fans tell the platform that they want their subscription fee to go to the artists that the fan listens to and no one else. This is sometimes called “user-centric” royalties, but I call this the “Ethical Pool”.
When the fan signs up for a service, let the fan check a box that says “Ethical Pool.” That would inform the service that the fan wants their subscription fee to go solely to the artists they listen to. This is a key point—allowing the fan to make the choice addresses how to comply with contracts that require “Big Pool” accountings or count Ethical Pool plays for allocation of the Big Pool.
Artists also would be able to opt into this method by checking a corresponding box indicating that they only want their recordings made available to fans electing the Ethical Pool. The artist gets to make that decision. Of course, the artist would then have to give up any claim to a share of the “Big Pool."
Existing subscribers could be informed in track metadata that an artist they wanted to listen to had elected the Ethical Pool. The subscriber could have to switch to the Ethical Pool method in order to listen to the track. That election could be postponed for a few free listens which is much less of an issue for artists who are making less than a half cent per stream.
The basic revenue share calculation still gets made in the background, but the only streams that are included in the calculation are those that the fan actually listened to. If the fan doesn’t check the box, then their subscription payment goes into the market share distribution as is the current practice, but their musical selection is limited to “Other than Ethical Pool” artists.
That’s really all there is to it. The Ethical Pool lives side by side with the current Big Pool market share model. If an Ethical Pool artist is signed, the label’s royalty payments would be made in the normal course. The main difference is that when a subscriber checks the box for the Ethical Pool, that subscriber’s monthly fee would not go into the market share calculation and would only be paid to the artists who had also checked the box on their end.
One other thing—the subscription service could also offer a “pay what you feel” element that would allow a fan to pay more than the service subscription price as, for example, an in-app purchase, or—clasping pearls—allow artists to put a Patreon-type link to their tracks that would allow fans to communicate directly with the artist since the artist drove the fan to the service in the first place. I’ve suggested this idea to senior executives at Apple and Spotify but got no interest in trying.
The Ethical Pool is real truth in advertising to fans and at least a hope of artists reaping the benefit of the fans they drive to a service. There are potentially some significant legal hurdles in separating the royalty payouts, but there are ways around them.
I think the Ethical Pool is an idea worth trying.
[Published by Newsmax, August 20, 2018]
If you read the current version of the Music Modernization Act, you may find that it’s more about government mandates that entrench incumbents than a streamlined blanket compulsory license that helps startups climb the ladder. In the weeds of MMA, we find startups dealt out of governance by rule makers and forced as a rule taker to ante up payments by their competitors in a game that the bill makes into the only game in town.
Billboard reports that Republican Senators John Cornyn and Ted Cruz of Texas suggested a fix for this flaw—allow private market competition alongside the MMA’s government mandate.
Let’s review why this fix is necessary and how it could balance the roles of rule makers and takers.
It’s necessary because the problem doesn’t come from songwriters. It comes from the real rule makers—Amazon, Apple, Facebook, Google and Spotify. And startups know which side butters their bread.
Public discussion of MMA has focused on the song collective and the compulsory blanket license for songs, but the mandated digital services collective is more troubling given the size of the players involved. MMA calls the services’ collective the “digital licensee coordinator,” or the DLC. Rule taker startups are governed by the rule maker DLC, but have no say in the DLC’s selection.
Like Microsoft’s anonymous amici, startups know their place —especially against Google, Amazon, and Facebook, whose monopoly bear hug on startups includes hosting, advertising and driving traffic.
The MMA authorizes these aggressive incumbents to effectively decide the price to startups for the “modernized” blanket license. Why? Because the MMA requires users of the license to pay for the lion’s share of the “administrative assessment,” the licensees’ collectivized administrative cost payment that the CBO estimates will be over $222 million for eight years.
Senators Cruz and Cornyn object to MMA’s delegation of the government’s authority to one collective for songwriters and one collective for digital services—with an antitrust exemption.
Cruz and Cornyn's point is a valid one. Why should the government only permit one game in town? Rather than have the DLC run by the usual suspect monopolists, why not allow competition?
This is important–if startups can’t afford to buy-in to the license, it does them no good, and their biggest competitors decide the price of that license through the DLC.
Both collectives are to be approved by the Register of Copyrights, but MMA puts the Register in the unenviable position of being constrained to appoint certain types of entities by statutory mandate.
Here’s the mirror image language from the MMA instructing the Register on selecting both the song collective and the DLC:
“[The Register must choose an entity that] is endorsed by and enjoys substantial support from [digital music providers/copyright owners of musical works] that together represent the greatest share of the [licensee/licensor] market for uses of [musical/such] works in covered activities, as measured over the preceding 3 full calendar years;”
In other words, startups, who are most in need of the statutory blanket license, need not apply to be the DLC.
That clause alone justifies the Cornyn/Cruz solution.
“Modernization” should make licensing easier: level the playing field for startups and protect them from famously predatory competitor incumbents, as well as copyright infringement lawsuits from the rule takers.
These are all good reasons for the private market solution. Competition at least gives startups hope for the pursuit of fair treatment.
A wise member of the Texas Congressional delegation once told me Big Tech gets to climb the ladder to the American Dream like everyone else. What they don’t get to do is pull the ladder up behind them after getting to the top.
Expanding competition rather than mandating an exclusive collective could keep that ladder available to everyone.
Chris Castle is the founder of the Austin, Texas-based Christian L. Castle Attorneys in Austin, Texas. Chris has testified on artist rights issues at the UK Parliament, spoken at Congressional seminars and lectures at law schools and business schools in the US and Canada on music-tech issues and artist rights.
[First published in June 2018 in issue 2:2018 of The Works, the Magazine of the British Academy of Songwriters, Composers and Authors at p. 18]
The three-part “Music Modernization Act” (“MMA”) is winding through the U.S. Congress. If passed, MMA will be the biggest change in U.S. copyright law in 100 years. But that doesn’t mean it’s good.
Two of the three parts of MMA apply to sound recordings and a third part applies to songs. (Just to make it extra confusing, the song part was previously introduced as a stand-alone bill under the same title, but has now been superseded by the three-part bill.) The sound recording parts are well-known ground and either codify current practices or has been thoroughly litigated. It is the song part that is controversial.
Crucially, the song part of the MMA applies to all songs ever written or that ever will be written that are exploited in the United States—not only US works. So the scope is global and applies to your current and future catalog. Unlike many collecting societies, there is no “opt out.”
MMA passed the lower body of the bicameral U.S. legislature and has been introduced in the Senate (the upper body) as of May 10, 2018. (This roughly corresponds to a first reading in the House of Lords.) As a tribute to the lobbyists, it must be said that the bill passed the lower body unanimously on a vote of 415-0. A cynic might say that’s because none of the Big Tech lobbyists opposed it (yet) and no one had read the bill.
MMA addresses US-centric issues: Effecting direct payment to producers of royalties for the limited performance right in sound recordings; closing the loophole that digital services leverage to avoid paying those royalties on recordings made prior to February 15, 1972; and creating a new collective for songs called the “Mechanical Licensing Collective” or “MLC”.
The MLC part is by far the most detailed and controversial part (some 90 pages of text) and is the subject of this post. In fairness, I am an MLC skeptic based on operational challenges. There is also a legitimacy issue--by comparison to the many recent hearings, public comments and consultations on other copyright topics, there has been essentially no public input on MLC. Thus the songwriter groups who were allowed to participate in the drafting bear an ominous burden of responsibility.
So don’t be surprised if you are hearing of the MMA for the first time in this post.
While there is a broad industry coalition supporting each of the three parts of the MMA, that support is a mile wide and an inch deep. There is something in it for everyone at the moment—but if support for all three parts waivers as it may yet do in the Senate, the coalition will likely coalesce around their particular issue and who knows what happens then.
What the MLC Does
There is an extraordinary amount of complexity in the MLC legislation. Space prohibits a comprehensive treatment here. Realize that the statute is just the start—implementing regulations will be promulgated by the Copyright Office if the bill becomes law. There will likely be a greater opportunity for the world’s songwriters to comment on those regulations than they have had for the statute itself. One thing is clear—any meaningful changes to the MLC’s operations will require an act of Congress that is unlikely to happen in the best of circumstances. So this needs to be right.
The MLC law performs two general functions: It creates a new “reachback” safe harbor protecting certain infringing uses of songs by tech companies such as Spotify as of January 1, 2018 (that’s right—beforethe MMA becomes effective) and also creates a new compulsory mechanical blanket license for songs used in interactive streaming and limited downloads afterMMA becomes effective. Realize that the bill was initially introduced on December 21, 2017—eleven days before the proposed reachback period commenced. A public copy of the bill was not available until after the reachback nominally started. Big Tech’s support of MMA demonstrates the value of their new reachback safe harbor, beyond even the pre-72 loophole they have exploited.
Reachback Safe Harbor
MMA’s reachback safe harbor is earthshattering. It eliminates the ability of copyright owners—especially small copyright owners--to sue streaming service infringers like Spotify for statutory damages, attorneys’ fees or injunctions if the service takes the blanket license. Damages are limited to the paltry streaming mechanical royalty only. The reachback bars claims for infringement lawsuits filedafter January 1, 2018 even if the claim arose beforethat date regardless of the statute of limitations. The reachback also precedes the enactment of the MMA.
In America, statutory damages and attorneys’ fees are the “big stick” that every copyright owner had enjoyed for over 100 years. That stick has been a major issue for tech companies, their paid academics and nonprofits in every major Internet piracy copyright case. The reachback is an historic capitulation that expands safe harbors just when the value gap effort was bearing fruit—bizarrely, the new reachback was announced just weeks before CISAC’s own extensive value gap economic study.
My bet is Big Tech will use this triumphant surrender as precedent against other copyright categories and in other countries and to fight any value gap legislation. The reachback has been criticized as unconstitutional by leading copyright lawyers and will almost certainly be challenged. That challenge will no doubt come from a copyright owner who is not an MLC board member, however.
You will note that none of the messaging about the bill addresses the reachback—but it is whyWixen Music Publishing sued Spotifyon December 29, 2017. As Robert Levine noted in Billboard, “A bill that restricts lawsuits after January 1, 2018 should not have been introduced just before Christmas….”
If MMA becomes law, suits like the Spotify class actions are gone which appears to be the plan. There arguably is little in the MLC part incentivizing services to report more accurately and they control the MLC’s budget, so it is unlikely that surrender accomplished anything.
Mechanical Licensing Collective
The MLC part of MMA establishes a new bureaucracy for a new blanket mechanical license. I will discuss a handful of operational issues, but first understand this: ASCAP and BMI already license the performance right for the same songs, to the same services, for the same performances under antitrust supervision of the government. The MLC will license the corresponding “streaming mechanical” for the same work but gets an antitrust exemption.
You might say, why aren’t the PROs doing a one-stop license for all rights? The government prohibits at least ASCAP doing so under the 77 year old consent decree. So the MLC gets an antitrust exemption on streaming mechanicals, but ASCAP (and essentially BMI) operate under government supervision for the same song, transaction and service.
This is even more Kafka-esque because the new head of the government antitrust division has expressly saidthat he wants to review the ASCAP and BMI consent decrees and possibly end them. The only two concessions that ASCAP and BMI get in MMA are related to these consent decrees—if terminated, they will have gotten nothing. Shouldn’t we at least find out if the consent decrees end beforebetting the farm on the government mandating a global rights database--that previously failed in the private market?
Designation of MLC: The head of the U.S. Copyright Office (called the “Register”) designates the non-profit corporation to be the MLC—but—the Register is limited in that selection to a nonprofit created by copyright owners, that “is endorsed by and enjoys substantial support from musical work copyright owners that together represent the greatest percentage of the licensor market for uses of such works in covered activities,as measured over the preceding 3 full calendar years”; can demonstrate to the Register that is has the ability to do the work. Because the Copyright Office knows state of the art technology, right? You see where this is going.
MLC Business Plan and Budget: None announced for a millennial change in the law. Remember—MMA passed unanimously in the House and no one brought this up.
Funding of MLC Operating Costs: Paid by participating services through an assessment by the Copyright Royalty Judges (“CRJs”). Unclear what qualifications the CRJs have to evaluate technology startups, or what happens if a digital service goes bankrupt (currently iHeart Media is in bankruptcy) and cannot pay its assessment (but probably will be made up from the black box).
Digital Licensee Coordinator: Little discussed is the effect on startups of the MLC assessment structure and control of the services side of the MLC by the biggest corporations in commercial history (Amazon, Apple, Google/YouTube) and the dominant streaming service Spotify. What if startups cannot afford their assessment? Are they denied the blanket?
Rate Standard/Ice in Winter: A new government rate standard will supposedly result in higher royalties for songwriters (misleadingly called “willing buyer/willing seller”--there’s been no free market in 100 years). Services can request that the CRJs give a reduction in royalty rates based on the CRJs’ assessment of the operating costs for the MLC.
Overbudget: Board can pay overbudget with black box money with no liability for board absent gross negligence. Amounts deducted may be replenished in future assessments, but may not be. No protection if not replenished.
Black Box: Unclaimed royalties are held for three years and then allocated based on market share (subject to permitted deductions). This is a major bone of contention.
Governance: 14 voting member board of directors, 10 publishers and 4 “professional songwriters.” Other inferior boards, e.g., for dispute resolution, have more songwriters. The European Songwriter & Composers Alliance has criticized the structureas out of step with industry practice of at least equal representation of songwriters.
Nonvoting Board Members: Lobbyists for the publishers, songwriters and digital services appear to have nonvoting board memberships. Note that this means that the digital services representative (the “digital licensee coordinator”) will likely have a right to information and to attend board meetings. This may have an effect on attorney-client information, litigation or royalty setting strategy. There is no prohibition on fees paid to lobbyists.
Global Rights Database: A new “GRD” to be created by MLC, probably subcontracted. Candidates: Harry Fox Agency, BMI, SOCAN, CMRRA/SX Works.
Claiming/Registration: Unclear how song metadata get into the new GRD, but apparently through an undefined claiming process that suggests some form of registration or other formality, at least with the MLC if not with the Copyright Office. Even if MLC doesn’t charge, there will still be a transaction cost on copyright owners.
Unfunded Mandates: Apparently all songwriters bear their cost of registering their works with the MLC, with no compensation to songwriters for complying with this mandate. Royalties for valid compulsory licenses already in effect (such as the millions of notices sent by MRI, the Harry Fox Agency, MediaNet and others) are to be sent to the MLC after the “license availability date”. Unregistered songwriter royalties presumably are black boxed.
Voluntary Licenses: Services and copyright owners may enter voluntary licenses outside of the MLC (voluntary license payments likely won’t get black boxed). These voluntary licenses presumably include catalog licenses as well as modified compulsory licenses (such as the typical HFA license) whenever made. Voluntary licenses suggest that the MLC will only administer songs that are subject to the blanket. However, since there is no opt out, services could refuse to renew a catalog license and then take that catalog under the blanket without paying a minimum guarantee. Eliminating statutory damages, attorneys’ fees and minimum guarantees would be the trifecta for services.
Accountings: Payment obligations of digital services are clearly spelled out more or less consistent with current practice, but payment obligations of MLC to copyright owners are unclear, other than black box.
Audits: Only the MLC may audit the services that pay its bills. Copyright owners may audit the MLC only. However, audits must be conducted by certified public accountants and those auditors are obligated to look for overpayments—which probably violates their duty of loyalty. As Warner Music Group’s Ron Wilcox testified to the CRJs, “Because royalty audits require extensive technical and industry-specific expertise, in WMG’s experience a CPA certification is not generally a requirement for conducting such audits. To my knowledge, some of the most experienced and knowledgeable royalty auditors in the music industry are not CPAs.” Requiring CPAs definitely raises the cost of audit on copyright owners, and demand of overpayment reimbursement definitely adds a touch of in terrorem.
In summary, the blanket license solves some market problems, but the MLC may create more. The MLC law has been resoundingly criticized as unconstitutional by the top copyright litigators (particularly the reachback safe harbor) and as unfair by many songwriter groups. At a minimum it’s a major capitulation.
However, I would count on it becoming law.
[This article first appeared in the American Bar Association Entertainment & Sports Lawyer, Spring 2017]
There is a fundamental rule of music licensing—if you don’t have a license from the copyright owner, don’t use the music. But in the new new thing of “permissionless innovation,”[3] the “disruptors” want to use the music anyway. Nowhere is this battle more apparent than the newest new new thing—mass filing of “address unknown” compulsory license notices for songs.
You’re probably familiar with U.S. compulsory mechanical licenses[4] for songs mandated by Section 115[5] of the Copyright Act.[6] We think of the compulsory (or “statutory”) license[7] as requiring music users to pay mechanical royalties after serving a “notice of intention” (or “NOI”)[8] on the song copyright owner and complying with other statutory requirements[9]—but it may come as a surprise that some of the biggest contemporary online music users serve millions of NOIs for interactive streaming to both avoid paying statutory royalties and wrap themselves in the liability insulation of the statutory license.[10] And because these users claim they cannot find the address of the song owner, millions of NOIs are served on the Copyright Office and not the song owner.
According to the independent source Rightscorp, [11] a company that has been tracking and indexing all those NOIs as published by the Copyright Office,[12] over 25 million “address unknown” NOIs have been served on the Copyright Office between April 2016 and January 18, 2017, or an average rate of approximately three million per month.
NOI Table
Top Three Services Filing NOIs April, 2016—January 2017[13] and Number of NOIs Per Service
Amazon Digital Services LLC: 19,421,902
Google, Inc.: 4,625,521
Pandora Media, Inc: 1,193,346
But according to a recent story on the subject in Billboard[14]:
"’At this point [June 2016], 500,000 new [songs] are coming online every month [much lower than the reported numerical average to date], and maybe about 400,000 of them are by indie songwriters [which may include covers], many of whom who don’t understand publishing,’ Bill Colitre, vp/general counsel for Music Reports, a key facilitator in helping services to pay publishers, tells Billboard. ‘For the long tail, music publishing data from indie artists often doesn’t exist’ when their music is distributed to digital services.”
Conversely, neither digital retailers, i.e., music users, nor aggregators appear to be able (or perhaps willing) to collect publishing information for new releases or long tail for unknown reasons.[15] Of course, as we will see below, Google has collected publishing information for a decade through its Content ID product, and Music Reports itself sells its Songdex product[16] that contains a significant amount of song data and is widely relied upon by its music user principals.
Whether their motivation is avoiding liability, avoiding royalties, or both, this means that Amazon, Google, Pandora and others[17] pay no statutory royalties on any of the song copyrights in their millions[18] of “address unknown” NOIs until the song copyright owner becomes “identifiable” in the “public records” of the U.S. Copyright Office, a process that arguably defeats Congress’s entire purpose of the NOI in the first place.
This problem will not be solved by maintaining an ownership database in the cloud that would allow users to exploit songs or family pictures until the work is registered.[19] An ownership database does not solve the problem of users who have no penalty for failing to use the database or for willful blindness.[20]
As I think the reader will see, by capitalizing on a perceived loophole in the U.S. Copyright Act, the users may well have gotten themselves absolutely nowhere, the government may have participated in yet another unconstitutional taking,[21] and songwriters, as usual, are left out in the cold to spend precious resources correcting the mistakes of giant multinational corporations.
That’s A Nice Song You Got There—Shame If No One Could Find It
Songwriters have three common reactions to the scale of the “address unknown” NOIs. First, they assume the songs subject to these NOIs must be in the “long tail”. This does not appear to be true, as there clearly are some high value new releases. Yet the industry has handled this “problem” for decades without resorting to mass NOIs.[22]
Songwriters ask how services fail to identify owners when songwriters and their publishers take care to register their songs in the databases readily made available by ASCAP, BMI, GMR and SESAC. [23] Songwriters are surprised to learn that music users need only search the public records of the Copyright Office and not even the databases of the user’s own agent.
But the biggest shock is usually from the sheer number of filings and the realization that these services are getting a free ride from exploiting a loophole.
And because these services do not render accounting statements as I will argue the law clearly requires[24], there is essentially no way for a song copyright owner to know what they are owed in the case of mistake or prospective payment.
Alternatively, if music users unilaterally decide to pay royalties retroactively, it will be even more important that proper statutory accounting statements be rendered for each song. Since users elected the “address unknown” process to serve NOIs on the Copyright Office, it only makes sense that these monthly and annual accounting statements also are served on the Copyright Office.
It is worth noting that the U.S. compulsory license does not accord songwriters an audit right, another loophole in the law that has never been corrected.[25] If these loopholes are combined at scale, then Amazon, Google and Pandora—companies with a combined market capitalization of nearly $1,000,000,000---can exploit millions of songs, pay no royalties, have at least some protection from infringement claims and cannot be audited.
Now that’s a hack. Meet the new boss, worse than the old boss.
Unlike the typical “pending and unmatched” or “black box” distribution, the compulsory license accounting requirements should substantially reduce unmatched exploitations. Since it appears that no statements have been rendered under “address unknown” NOIs for 2016 as of this writing, I will argue that song owners are entitled to send a termination notice to the music service for failure to account regardless of whether the copyright owner is identifiable in the Copyright Office’s public records. I will also argue that if that failure is not lawfully remedied, those purported licenses terminate “automatically”.[26]
How did this mess occur? It all starts with a shard of a statute that arguably was never intended for compulsory licensing.
The Statutory Origins of Mass NOIs
As of April 19, 2016, the U.S. Copyright Office began posting on its website[27] copies of millions of “address unknown” NOIs served on the Copyright Office by Google, Amazon, Pandora, iHeart and other services. These very large music users are taking advantage of two little known and previously little used sections of Section 115 of the 1976 Copyright Act that both define how an NOI is to be sent and also limit when statutory mechanical royalties are payable. Both code sections were enacted decades before the interactive “streaming mechanical” was conceived.
Section 115 (b)(1)[28] (and related regulations[29]) covers how NOIs may be sent to the song copyright owner and is the origin of the “address unknown” NOI:
Any person who wishes to obtain a compulsory license under this section shall, before or within thirty days after making, and before distributing any phonorecords of the work, serve notice of intention to do so on the copyright owner. If the registration or other public records of the Copyright Office do not identify the copyright owner and include an address at which notice can be served, it shall be sufficient to file the notice of intention in the Copyright Office. The notice shall comply, in form, content, and manner of service, with requirements that the Register of Copyrights shall prescribe by regulation.
Section 115 (c)(1)[30] provides when royalties are payable (or not) under an “address unknown” NOI:
To be entitled to receive royalties under a compulsory license, the copyright owner must be identified in the registration or other public records of the Copyright Office. The owner is entitled to royalties for phonorecords made and distributed after being so identified, but is not entitled to recover for any phonorecords previously made and distributed.
The Copyright Act curiously omits any guidance regarding actual knowledge of the music user or its agent regardless of what is in the notoriously incomplete Copyright Office records. For example if the user or agent maintained a voluminous database of song information, can that data simply be ignored?
The Act also does not address knowledge that could reasonably be available to the music user, such as information readily available at no cost in the PRO Databases. It is important to note that music users are simultaneously accounting under blanket licenses to the U.S. performing rights organizations for the performing rights of the same uses of the same songs by the same service. So the PRO Databases are readily available to users.
The Source of the Problem
Users may argue that regardless of what they knew or should have known, if a song copyright owner is not “identifiable” in the “public records” of the Copyright Office,[31] the music user can serve NOIs on the Copyright Office.[32] Once service is effective on properly served “address unknown” NOIs,[33] the music user then is entitled to claim all of the protections from liability for copyright infringement and against audits as a statutory licensee--with the added benefit of avoiding any mechanical royalties.
How will the music user know which NOIs to serve on the Copyright Office? Until music users publicly release that information, we have no way of knowing with certainty.[34]
However, given the patterns in filing that have developed (see NOI table above), one might get the impression that some music users are not checking if the song owner is identifiable anywhere. Instead, it appears at least possible that the users may be simply sending NOIs for all songs they use.
This is troubling because song owners provide multiple ways for licensees to reach them including online databases maintained by the various performing rights organizations such as ASCAP,[35] BMI,[36] Global Music Rights[37] and SESAC[38] that cover over one million songs.[39] But Section (c)(1) requires that users search the database that is the least relevant, the least up to date, the most anachronistic and most difficult to use: The public records of the Copyright Office, which includes the “Public Catalog.”[40]
The Public Catalog has recently taken on a heightened level of importance. One music user responds to address change requests by simply telling the song owner that the music user “now” receives their data from the Copyright Office Public Catalog. This user implies song owner registration is required by Section 115 of the Copyright Act. They also tell the song copyright owner to update their registration with the U.S. Library of Congress—not the Copyright Office.
The clear implication is that all song copyrights must be registered which is simply untrue, however advisable registration may be. As the Copyright Office clearly states, “No publication or registration or other action in the Copyright Office is required to secure copyright.”[41] Registration is not required in order to enjoy copyright protection or the rights of a copyright owner generally other than some litigation-related benefits. As the Copyright Office statement implies, this is old news. [42]
Also note that Congress could easily have required registration in Section 115(c)(1), but did not. Congress contemplated all public records, whether existing in 1978, currently existing[43] or coming into existence in the future. Given that the Copyright Office is subject to Freedom of Information Act requests, “public records” may be very broad indeed.
Since no copyright registration is required, it is not surprising that the copyright owner’s entitlement to receive all benefits of Section 115 is not conditioned on registration[44] including the right to send a termination notice.[45]
Another issue is less obvious—the “data” from the Copyright Office Public Catalog expressly excludes pre-1978 works and is thus inherently unsuitable for purposes of “address unknown” NOIs.
Pre-1978 Public Records
The landing page[46] of the Public Catalog clearly states[47] that pre-1978 records are only available on paper from the Copyright Office.
But note--Section 115 and the accompanying regulations make no such distinction regarding pre-78 works. So if the only effort the music user is making is to search the post-1978 catalog, any purported entitlement to an address unknown NOI for a pre-78 song may well fail.
Setting aside the international treaty implications[48] and the potential lack of pre-78 works in the Public Catalog, it is unlikely if not impossible for a pre-78 song owner to be found in the digitized Copyright Office records. There is no method[49] for copyright owners to “update” or even initially record their contact information unless they either record a document listing all their registered works by title and registration number, or they file a supplementary registration to amend an already completed registration—which is costly.
Which means the song owner must have already registered the works concerned, which is not required in order to enjoy the rights of a copyright owner.
Line of Least Resistance Leads Them On[50]
But why would music users point song owners at the Library of Congress? Perhaps because the Library of Congress sells an electronic database of the post-1978 Copyright Office registration and recordation records. If you can find the link[51] to purchase these databases on the Library of Congress website you are a better researcher than I (or the reference desk at the Library of Congress which couldn’t find it either).
The Library sells two databases that I found: The “Retrospective: 1978-2014” for $50,225 and the “2015 Subscription” for $28,700. These LOC Databases may be the reference data upon which the “address unknown” mass NOI filings are based. These LOC Databases might explain why at least one music user is pointing song copyright owners to the Library of Congress to update their contact information.
Music users (or perhaps their agents) who can afford to purchase these LOC Databases probably could also afford to hire a copyright research service to examine the pre-78 card catalog.[52] But as of this writing, it appears that at least some pre-78 songs may be missed.
While we do not know with certainty how the services or their agents conduct research, we can extrapolate how a list might be compiled.
How Are Song Titles Determined for Mass NOIs?
The process could be as simple as users asking their agent what information is in their agent’s databases, i.e., information of which agent or principal would have actual knowledge. If that is the method used by all music users, then the number of unknown titles should be relatively constant across services and it appears not to be as reflected in the NOI table above.
It may be the case that a user ingests the sound recording metadata and utilizes the sound recording title as the song title.
This might explain why “Fragile” performed by Sting becomes “Fragile (Live)” in Google’s mass NOI filing. If the music user then looks for a song title of “Fragile (Live)” in the public records of the Copyright Office, that title is unlikely ever to be found because the song was registered as “Fragile.”
Given the obscurity-in-plain-sight surrounding mass NOI filings, it is safe to say that unless the song copyright owner is extraordinarily alert and has the computing power to decompress very large NOI files posted on the Copyright Office website, she may never know that her song is being commercially exploited royalty-free.
That’s right—users get all of the benefits[53] and none of the burdens, thanks largely to a haystack of the users own creation.
Building a Haystack of Needles
Sifting through millions of NOIs for your songs is a labor of Hercules for the independent songwriter and even for an indie publisher. Amazon, Google, Pandora and iHeart, among others—but notably, not Apple so far--have built a haystack of needles worthy of the Augean stables. Rightscorp is developing[54] a searchable and indexed database that will allow songwriters to search the mass NOI filings, as may others, but neither the Library of Congress nor the Copyright Office provide any simple way for songwriters to conduct that search as of this writing.
Crucially, it is important that any searchable database come from an independent source as the process is fraught with obvious moral hazard. Neither the services filing the mass NOIs nor their agents should be providing search functions to songwriters as this really would be like asking the fox to file an after action report for the fox’s attack on the chicken coop.
Rendering Statements of Account
It seems improbable that users filed tens of millions of NOIs free of errors. Even a 1% error rate is 250,000 improper NOIs. What is clearer is that no monthly or annual statements of account[55] have been rendered to date and for that reason alone any purported license based on an “address unknown” NOI may be subject to statutory termination.[56] Even if no royalty is payable or only payable prospectively from an unknown time, the statutory obligation to render statements crucially still applies to music users. How will the songwriter ever know what royalties are payable otherwise? This is likely why Congress did not distinguish accounting obligations for “address unknown” NOIs from “known known” NOIs.
Since these statutory users chose to serve NOIs on the Copyright Office, those users have nowhere else to send the required statements but to the same place they sent their NOIs—the Copyright Office. This would be entirely reasonable and consistent with the longstanding requirements that statements be sent to the same address as the NOI.
It is also important to note that the “identification” requirement only applies to NOIs and not to the song copyright owner’s right to terminate the statutory license for failure to account. Note that the termination right is not for failing to render statements to the copyright owner who the music user has decided they cannot identify, but rather for failing to render statements[57] to the Copyright Office that the music user has decided they can identify. This is not a question of having rendered the required statements (and certifications) to the wrong person; in this case, the statements have not been rendered at all.
I would argue that music users ought to serve the lawfully required accounting statements on the Copyright Office because the music user chose to avail themselves of the benefits of Section (c)(1). Allowing the music user to avoid complying with the lawful accounting requirements in addition to avoiding payment does not have a statutory basis and arguably seems clearly outside the intention of Congress.
Indeed, if the Library of Congress fails to require these accountings for millions of NOIs, valuable property rights of potentially hundreds of thousands, if not millions, of the world’s songwriters may be foreclosed.
Eyesight to the Willfully Blind
Note that Section 115 does not address what happens if the music user (or its agent) in fact knows the identity of the song copyright owner at the time of serving the “address unknown” NOI.
Actual knowledge is particularly relevant in the case of companies like Google. Google purchased the mechanical rights licensing company Rightsflow for the very reason that Rightsflow’s database provided valuable rights ownership information for Google to use in its business.[58] Google has also operated its Content ID platform on YouTube[59] for many years through which Google collected vast amounts of song ownership information directly from and at great transaction cost to rights owners. It is difficult to understand how Google in particular does not have actual knowledge of the contact information of millions of song copyright owners to whom it sends statements and payments for other services under other licenses.
Actual knowledge is also relevant in the case of Music Reports, Inc., that apparently is the agent[60] that many of these statutory license users evidently engaged to administer the mass NOI filings. Music Reports not only has developed and marketed its “Songdex” product based on the millions of song owners it can identify, but also has applied for a patent[61] for its mechanical royalty licensing business process. Music user principals of MRI would seem to have access to a vast database of highly reliable song ownership knowledge from a highly credible agent.
Yet it seems implausible and inconsistent with other statutory provisions of the Copyright Act[62] that Congress intended to protect music users who have actual knowledge of the identity of a song owner.[63]
But how will a songwriter even know if their song is implicated?
Which Songs Are Affected?
As noted above, “long tail” deep catalog and new releases seem likely to be affected by “address unknown” NOIs, albeit for different reasons. Deep catalog may be affected because no one registered the song titles or the works were registered before 1978 and the music user did not research the ownership in the paper Copyright Office public records.
New releases may be implicated because the Copyright Office itself has yet to process the registration—and I will discuss the Library of Congress’s limitation on “cataloged registrations” below. Presumably, a filed but yet to be conformed copyright registration would also not be “cataloged”. The Copyright Office acknowledges on its copyright registration portal that the processing time for e-filings is six to ten months, and for paper filings ten to 15 months.[64] This loophole would thus destroy the songwriter’s peak earning power on new releases.
How Mass NOIs Could Be Misidentified
The LOC Database suggests some reasons for potential mismatches:
The only information file available which contains such copyright information as author, title, copyright claimant name, and registration number. Represents cataloged registrations and relevant documents entered into the U.S. Copyright Office database since 1978.[65]
It appears that not all registrations are cataloged and not all documents are entered into the LOC Database. Relying solely on the LOC Databases might result in gaps for “address unknown” NOIs.
Because the LOC Database by definition excludes pre-78 works, these excluded works could be another source of error. The plain language of the Copyright Act includes those pre-78 songs for purposes of an “address unknown” service. How pre-78 copyrights are treated in the mass NOI filings is unclear, but it is worth noting that “Surfer Girl” by the Beach Boys is included in one of Google’s filings[66], which is clearly a pre-78 song.
Who is responsible for cross-checking accuracy? Probably no one. The Copyright Office expressly disclaims any responsibility for incorrect NOIs and warns everyone involved that incorrect notices may only be challenged in a court,[67] in this case by whichever songwriter or publisher who is willing to litigate with the biggest corporations in the world.
That actually leaves it to Congress to take a leadership role in reviewing their library, their statute or any misapplication of these rules.
The Copyright Office Address Unknown Posting
The Copyright Offices posts[68] “address unknown” NOIs on a rolling basis. In order for songwriters to know if their songs are in these NOIs they have to wait for the Copyright Office to post the files, decompress them, sort them, and try to find their own songs by searching the resulting massive Excel file—assuming the songwriter has the skill and computing power available. As of this writing there are approximately 150 NOI filings, but each filing can contain tens of thousands of songs titles for which the music user claims the protections of the statutory license.
This process is not realistic and seems inconsistent with the intentions of Congress.
What is to Be Done?
There are a few ways that mass NOIs can be dealt with. As we review each potential course of action, the same themes will recur: Someone in the government needs to take responsibility for verifying these NOIs are filed as required by law, and the “address unknown” NOI process as currently practiced places an unfair burden on songwriters.
1. Recordation Filing: The Copyright Office will likely accept a simultaneous electronic and paper recordation of a certification of a song copyright owner with a list of song titles. The electronic filing should provide immediate notice to music users. This approach is costly, however, and may be ill suited to individual song copyright owners or independent publishers.
2. Dramatico Musical Works: It appears that the Copyright Office is accepting filings for dramatico-musical works which are not subject to compulsory licenses.[69] (Dramatico-musical works include musicals, for example.) Owners of dramatico-musical works may wish to take ameliorative action to stop the infringing use of their copyrights.
3. Pre-78 Songs: It appears that at least some music users may be ignoring the paper records of the Copyright Office and filing NOIs for song copyrights that may well be identifiable in the pre-78 public records.
4. Improper Filing: However cumbersome, songwriters have a reasonable expectation that the Copyright Office should be able to confirm if the NOIs comply with the statutory requirements. Noncompliant NOIs should be barred.
5. Failure to File and Certify Statements of Account: Regardless of whether royalties are due, music users are arguably required to file monthly and annual statements of account. This is particularly reasonable given the scale of the mass NOI filings, the likelihood of error and the statutory requirements. To my knowledge, no statements of account have been filed as of this writing. This would indicate that all NOIs are subject to termination.
6. Direct Licenses: Based on a sample of songs that I consider likely to be subject to a direct license with a major publisher, it seems possible that “address unknown” NOIs may be getting filed on songs that are directly licensed. Publishers with direct licenses may wish to confirm if they are receiving payments for any directly licensed songs or if users are not paying based on the “address unknown” NOI.
7. Revenue Share Calculations: If songwriters or publishers receive a pro-rata revenue share based on the total number of songs performed during an accounting period, it would be well to determine if non-royalty bearing songs subject to an “address unknown” NOI are being included in the ratio.
Of course, the easiest fix is for the music user to not exploit music without a license from the song owner. That approach has worked well in the past, so perhaps it could work for these music users, too.
[1] Copyright 2017, Christian L. Castle.
[2] Founder, Christian L. Castle, Attorneys, Austin, Texas (www.christiancastle.com).
[3] From Keep the Internet Open by Vinton G. Cerf, Google’s Chief Internet Evangelist, New York Times (May 24, 2012) available at http://www.nytimes.com/2012/05/25/opinion/keep-the-internet-open.html
[4] The statutory mechanical license was established in the United States by the 1909 revision to the Copyright Act, in part to obviate the holding in White-Smith Music Publishing Company v. Apollo Company, 209 U.S. 1 (1908). The Supreme Court held in White-Smith that piano roles were not reproductions protected by the then-current Copyright Act, but also invited the Congress to amend the Copyright Act if the Court’s holding was not welcome: “It may be true that the use of these perforated rolls, in the absence of statutory protection, enables the manufacturers thereof to enjoy the use of musical compositions for which they pay no value. But such considerations properly address themselves to the legislative, and not to the judicial, branch of the government. As the act of Congress now stands we believe it does not include these records as copies or publications of the copyrighted music involved in these cases.” See also Stern v. Rosey, 17 App. DC 562 (D.C. Cir 1901) (holding the “peculiar use” of musical compositions in early wax cylinder phonographs not protected).
[5] 17 U.S.C. § 115.
[6] The Copyright Act of 1976, Pub. L. No. 94-553, 90 Stat. 2541 (Oct. 19, 1976).
[7] A statutory license or “compulsory license” is “a codified licensing scheme whereby copyright owners are required to license their works to a specified class of users at a government-fixed price and under government-set terms and conditions.” Satellite Home Viewer Extension Act: Hearing Before the S. Committee on the Judiciary,108th Cong. (2004) (statement of David O. Carson, General Counsel, U.S. Copyright Office) (May 12, 2004). “[C]ompulsory licensing . . . break[s] from the traditional copyright regime of individual contracts enforced in individual lawsuits.” See Cablevision Sys. Dev. Co. v. Motion Picture Ass’n of Am., Inc., 836 F.2d 599, 608 (D.C. Cir. 1988) (describing limited license for cable operators under 17 U.S.C. § 111). A compulsory license “is a limited exception to the copyright holder’s exclusive right . . . As such, it must be construed narrowly. . . .” Fame Publishing Co. v. Alabama Custom Tape, Inc., 507 F.2d 667, 670 (5th Cir. 1975) (referring to compulsory licenses such as the compulsory mechanical license in the Copyright Act of 1909). Compulsory licenses are generally adopted by Congress only reluctantly, in the face of a marketplace failure. For example, Congress adopted the Section 111 cable compulsory license “to address a market imperfection” due to “transaction costs accompanying the usual scheme of private negotiation. . . .” Cablevision at 602. “Congress’ broad purpose was thus to approximate ideal market conditions more closely . . . the compulsory license would allow the retransmission of signals for which cable systems would not negotiate because of high transaction costs.” Id. at 603.
[8] 17 U.S.C. § 115(b).
[9] 17 U.S.C. § 115(a).
[10] The Copyright Office Copyright and the Music Marketplace report (hereafter “Licensing Study”) notes that many song copyright owners view the entire compulsory licensing system as unfair. Licensing Study at 108, available at https://www.copyright.gov/policy/musiclicensingstudy/copyright-and-the-music-marketplace.pdf (“Many stakeholders are of the view that the section 115 license is unfair to copyright owners. As one submission summed it up: ‘The notifications, statements of account, license terms, lack of compliance, lack of audit provisions, lack of accountability, lack of transparency, ‘one size fits all’ royalty rates and inability to effectively enforce the terms of the license demonstrate a complete breakdown in the statutory licensing system from start to finish.’”)
[11] Mass NOI Update: Christopher Sabec and Rightscorp Tackle the Songwriters’ Copyright Office Problem, Music Tech Solutions (January 26, 2017) available at http://musictech.solutions/2017/01/26/mass-noi-update-christopher-sabec-and-rightscorp-tackle-the-copyright-office-problem/ (hereafter, “Sabec Interview”) andhttp://www.rightscorp.com.
[12] Copyright Office Mass NOI filing page https://www.copyright.gov/licensing/115/noi-submissions.html
[13] Source: Rightscorp, Inc.
[14] Christman, Say You Want a Revolution? U.S. Copyright Office Modernizes Key Part of Digital Licensing, Billboard (June 24, 2016) available at http://www.billboard.com/articles/news/7416438/us-copyright-office-music-reports-compulsory-licensing-digital-notice-of-intent (hereafter, “Christman”).
[15] According to panelists at 2016 SXSW, many retailers refuse to collect publishing information, and aggregators do not collect it or collect it sporadically. Castle, Is It Possible for Songwriter Metadata to Be Delivered to Retailers?, MusicTech.Solutions (March 16, 2016) available at https://musictech.solutions/2016/03/16/is-it-possible-for-songwriter-metadata-to-be-delivered-to-retailers/
[16] According to Music Reports, its Songdex product contains “detailed relational data on tens of millions of songs, recordings and their owners, covering virtually all of the commercially significant music in existence”. Available at https://musicreports.com/#songdex
[17] According to Christopher Sabec, CEO of Rightscorp, as of January 27, 2017 his company has determined that the top three mass NOI filers are Amazon Digital Services LLC (19,421,902 NOIs), Google, Inc. (4,625,521) and Pandora Media, Inc. (1,193,346). There is an unknown overlap among these filings and they are not unique.
[18] There is an as yet unknown degree of overlap between the services filing, so the total number does not necessarily mean that there are 25 million different songs involved, but rather 25 million notices served on the Copyright Office.
[19] See, e.g., Samuelson et al, Copyright Principles Project: Directions for Reform, 25 Berkeley Technology Law Journal 1 (2010) at 10.
[20] David Lowery, Getting Copyrights Right, Politico (May 13, 2013) available at http://www.politico.com/story/2013/05/building-a-real-copyright-consensus-091231
[21] See Professor Richard A. Epstein, Takings (1985); Songwriters of North America, Michelle Lewis, Thomas Kelly and Pamela Sheyne v. the Department of Justice, Loretta Lynch and Renata Hesse (U.S.D.C. Dist. Col. 1:16-cv-01830) at 22 (“Plaintiffs’ constitutional claim is based on theories that the 100% Mandate violates plaintiffs’ rights of procedural and substantive due process, and takes their property without compensation.”).
[22] The Copyright Act of 1909, Pub. L. 60-349, 35 Stat. 1075 (March 4, 1909), revised to January 1, 1973, § 1(e).
[23] By comparison, the European Union has developed a series of sector-specific guidelines for reasonably diligent searches for orphan works Memorandum of Understanding for Diligent Search Guidelines for Orphan Works as part of the European Digital Libraries Initiative available at http://www.ifap.ru/ofdocs/rest/rest0001.pdf (hereafter “EU Search Criteria”).
[24] See 17 U.S.C. § 115(c)(5) and 37 C.F.R. §§ 210.16 and 210.17.
[25] See Licensing Study supra note 10.
[26] 17 U.S.C. §115(c)(6).
[27] Copyright Office Mass NOI filing page https://www.copyright.gov/licensing/115/noi-submissions.html
[28] 17 U.S.C. §115(b)(1) (emphasis added).
[29] 37 C.F.R. § 201.18(f)(3).
[30] 17 U.S.C. § 115(c)(1) (emphasis added).
[31] It is worth noting that the current Section 115(c)(1) appears to have been a compromise for abandoning the old “notice of use” requirement in Section 1(e) of the 1909 revision of the Copyright Act. The “notice of use” was an obligation on the song copyright owner to file a notice in the Copyright Office when the song has had a “first use”. (“[I]t shall be the duty of the copyright owner, if he uses the musical composition himself for the manufacture of parts of instruments serving to reproduce mechanically the musical work, or licenses others to do so, to file notice thereof in the copyright office, and failure to file such notice shall be a complete defense to any suit, action, or proceeding for any infringement of such copyright. “) As noted in the House Report for the 1976 revision of the Copyright Act, “[t]his requirement has resulted in a technical loss of rights in some cases, and serves little or no purpose where the registration and assignment records of the Copyright Office already show the facts of ownership. Section 115(c)(1) therefore drops any formal ‘notice of use’ requirements and merely provides that ‘[the copyright owner must be identified in the records of the Copyright Office] in order to be entitled to receive royalties under a compulsory license’….” Notes of the Committee on the Judiciary, House Report No. 94-1476, Royalty Payable Under Compulsory License. A topic for another day might be the extent to which this trade off of the 1909 “notice of use” for a 1976 “identification” requirement as a precondition for enjoying the rights of a copyright owner violates the Berne Convention’s prohibition on formalities.
[32] 17 U.S.C. §115(b)(1).
[33] 37 C.F.R. § 201.18(g) (“Notices shall be deemed filed as of the date the [Copyright] Office receives both the Notice and the fee, if applicable.”)
[34] But see Christman, “If a direct deal hasn’t been cut with the publisher but the streaming service, or its agent -- often companies like Music Reports Inc., the Harry Fox Agency’s Slingshot operation, and Musicnet -- know all the rights owners of a song, it has to file a notice of intent with those rights owners. If it doesn’t know all the rights owners, the service can search the Copyright Office's database to see if that song and its owners are registered and, if so, can retrieve addresses and issue a notice of intent. If it can’t find the song or the owners, then the service has to file a notice of intent for the compulsory license for that song with the U.S. Copyright Office.”
[35] ASCAP’s ACE Repertory search is available at https://www.ascap.com/repertory
[36] BMI’s BMI Repertoire search is available at http://repertoire.bmi.com/startpage.asp
[37] Global Music Rights search is available from the GMR homepage www.globalmusicrights.com
[38] SESAC’s Repertory search is available at https://www.sesac.com/Repertory/Terms.aspx
[39] Herein the “PRO Databases”.
[40] U.S. Copyright Office, Circular 23, at 1 available at https://www.copyright.gov/circs/circ23.pdf: “Together, the copyright card catalog and the online files of the Copyright Office provide an index to copyright registrations in the United States from 1870 to the present. The copyright card catalog contains approximately 45 million cards covering the period 1870 through 1977. Registrations for all works dating from January 1, 1978, to the present are searchable in the online catalog, available at www.copyright.gov/records.”
[41] U.S. Copyright Office, Circular 1, at 3.
[42] Berne Accession available at http://www.wipo.int/treaties/en/notifications/berne/treaty_berne_121.html
[43] In addition to the registration and recordation records, the Copyright Office maintains the “CO-10 Address File” with the addresses of “[c]opyright claimants whose address has been requested by a member of the public.” 63 F.R. 51609 (Sept. 28, 1998).
[44] 17 U.S.C. § 115(c)(5).
[45] 17 U.S.C. § 115(c)(6).
[46] http://cocatalog.loc.gov/cgi-bin/Pwebrecon.cgi?DB=local&PAGE=First
[47] See id. “Works registered prior to 1978 may be found only in the Copyright Public Records Reading Room.”
[48] See, e.g., Denniston, International Copyright Protection: How Does It Work? available at http://www.bradley.com/insights/publications/2012/03/international-copyright-protection-how-does-it-w__ (“The central feature of the Berne Convention is that it prohibits member countries from imposing “formalities” on copyright protection, in the sense that the enjoyment and exercise of copyright cannot be subject to any formality except in the country of origin. For over a hundred years, the United States resisted joining the Berne Union, in part because of the desire to maintain the formalities U.S. law required. In order to be eligible to join the Berne Union, Congress had to amend the Copyright Act to dispose of the many formalities the Act required. Therefore, while the United States Copyright Act can impose a requirement that the owner of a United States work must register the copyright with the Copyright Office before filing an infringement suit in federal court, it cannot impose that same obligation on foreign nationals.”)
[49] Changing Your Address with the Copyright Office, available at https://www.copyright.gov/fls/sl30a.pdf
[50] Isn’t That So written by Jesse Winchester.
[51] “Copyright Cataloging: Monographs, Documents, and Serials (database)” available athttps://www.loc.gov/cds/products/product.php?productID=23 (hereafter “LOC Database”).
[52] Wall Street Journal Editorial Board, A Copyright Coup in Washington (Nov. 2, 2016) available at https://www.wsj.com/articles/a-copyright-coup-in-washington-1478127088
[53] Licensing Study 108-110.
[54] See Sabec Interview (“[Rightscorp intends] to create a technological solution to this technological problem. We have already created a searchable database and can assist rights holders in determining the extent of their exposure.”)
[55] See 37 C.F.R. § § 210.16 and 210.17.
[56] 17 U.S.C. § 115(c)(6).
[57] See 37 C.F.R. § § 210.16 and 210.17.
[58] Smith, Google Acquires Music Royalty Manager RightsFlow (Wall Street Journal, Dec. 9, 2011).
[59] See How Content ID Works available at https://support.google.com/youtube/answer/2797370?hl=en
[60] We leave aside the degree to which the agent’s knowledge can be imputed to the principals, but the issue seems ripe, fertile and of particular importance given the scale of the mass NOIs. See, e.g., Restatement (Third) of Agency, Section 5.03 (“For purposes of determining a principal’s legal relations with a third party, notice of a fact that an agent knows or has reason to know is imputed to the principal if knowledge of the fact is material to the agent's duties to the principal….”)
[61] U.S. Patent Application No. 20160180481, “Methods And Systems For Identifying Musical Compositions In A Sound Recording And Licensing The Same” (June 23, 2016) available at http://appft.uspto.gov/netacgi/nph-Parser?Sect1=PTO2&Sect2=HITOFF&p=1&u=%2Fnetahtml%2FPTO%2Fsearch-bool.html&r=1&f=G&l=50&co1=AND&d=PG01&s1=%22Music+reports%22&OS=%22Music+reports%22&RS=%22Music+reports%22
[62] See, e.g., 17 U.S.C. § 512(c)(1)(A)(i) and (ii) ([a service provider entitled to the safe harbor] does not have actual knowledge that the material or an activity using the material on the system or network is infringing; (ii) in the absence of such actual knowledge, is not aware of facts or circumstances from which infringing activity is apparent)
[63] See, e.g., EU Search Criteria.
[64] https://www.copyright.gov/registration/
[65] Copyright Cataloging: Monographs, Documents, and Serials (database), description https://www.loc.gov/cds/products/product.php?productID=23
[66] See https://thetrichordist.com/#jp-carousel-19404
[67] 37 C.F.R. § 201.18(g) states (emphasis added): “[T]he Copyright Office does not review Notices for legal sufficiency or interpret the content of any Notice filed with the Copyright Office under this section. Furthermore, the Copyright Office does not screen Notices for errors or discrepancies and it does not generally correspond with a prospective licensee about the sufficiency of a Notice. If any issue (other than an issue related to fees) arises as to whether a Notice filed in the Copyright Office is sufficient as a matter of law under this section, that issue shall be determined not by the Copyright Office, but shall be subject to a determination of legal sufficiency by a court of competent jurisdiction. Prospective licensees are therefore cautioned to review and scrutinize Notices to assure their legal sufficiency before filing them in the Copyright Office.”
[68] Copyright Office Mass NOI filing page https://www.copyright.gov/licensing/115/noi-submissions.html
[69] 17 U.S.C. § 115 (“In the case of nondramatic musical works, the exclusive rights provided by clauses (1) and (3) of section 106, to make and to distribute phonorecords of such works, are subject to compulsory licensing under the conditions specified by this section” emphasis added.)
From the Huffington Post, October 16, 2016
Two vastly wealthy multinational media companies are exploiting a copyright law loophole to sell the world’s music without paying royalties to the world’s songwriters. Why? Because Google and Amazon—purveyors of Big Data—claim they “can’t” find contact information for song owners in a Google search. So these two companies are exploiting songs without paying royalties by filing millions of notices with the Copyright Office at a huge cost in filing fees that only megacorporations can afford.
That’s right—Google and Amazon are falling over themselves to use their market power to stiff songwriters yet again. And as I will show, it is not just obscure songs that are affected. New releases, including one example from Sting, are also targets suggesting significant revenue loss to songwriters.
Songwriters are the most regulated workers in America. The government sets wage and price controls on most uses of songs and practically everything else about a songwriter’s business—except fulfilling government’s basic role of keeping them safe from piracy and multinational monopolists gaming the system. Congress needs to stop this latest scam.
The U.S. Compulsory License
The latest loophole that Google and Amazon are hacking is uniquely American—the compulsory license for songs. No other country has one. Most songwriters would prefer that the U.S. repeal this legacy anachronism from 1909 that keeps the government’s boot on their throats.
In order to get the government’s license, services only need notify the songwriter (or their publisher) that the service intends to use the song under the compulsory license. Of course, sending this notice of their intention to use the song (called an “NOI”) requires knowing who to send it to, and that is the “hack” that Google and Amazon are exploiting now. Others services surely will follow their market leadership if Congress fails to act.
The hack uses market power to manipulate a loophole in how those NOIs are sent. Common sense tells you that to send a notice, you must know who to send it to, even for a song. But does common sense also tell you that if you don’t know, the law should allow you to exploit the songs without compensation? Particularly if you’re the biggest purveyor of data in human history?
The legacy compulsory license allows services to exploit songs if they decide they can’t find the songwriter—and not pay royalties until the songwriter finds them.
That’s right—Google and Amazon trade on a loophole that allows them to serve NOIs on the U.S. Copyright Office if the song owner cannot be found in the public records of the Copyright Office regardless of what other information is readily available to these services, including their own. And once Google or Amazon serve that “address unknown” NOI, they don’t have to pay royalties and they cannot be sued for copyright infringement—so the millions in filing fees they will spend at the Copyright Office is a kind of insurance premium. This excerpt from the Copyright Act states the rule:
Why Can’t Google Search?
The “address unknown” NOI starts from this premise: Google is supposed to search for the song owner’s contact to send NOIs.
That’s right—Google is supposed to search. Think about that. This 1976 rule was never intended to apply to a music user with Google’s search monopoly. Yet, if Google “can’t” find the song owner after a search, then Google can serve an “address unknown” NOI to the Copyright Office and then exploit the song for free until the songwriter can be “identified” in the Copyright Office records—which may be never.
That registration by songwriters—while prudent—is costly and entirely voluntary. Forcing songwriters to register essentially turns the system into a version of YouTube’s “opt out” debacle, and probably violates international copyright treaties.
But the idea that Google can’t find someone is a remarkable thought. Gmail alone has over one billion users. Google knows everything about everyone and makes billions of dollars from reselling and manipulating that information. Not to mention the fact that Google bought the music licensing service Rightsflow—itself an NOI mill. Not to mention ten years of information Google has scraped from Content ID on YouTube or sheet music on Google Books.
Amazon also has a phenomenal amount of information about music ownership. As one of the biggest CD and DVD retailers, Amazon certainly has a head start in song research.
However—it appears that Google and Amazon are not using their own data for NOIs. Instead, they apparently are buying databases from the Library of Congress that tell them whether a song is registered for copyright or otherwise recorded in the digitized Copyright Office files (which songwriters are not obligated to do in order to get the benefits of the compulsory license). Those Library of Congress databases at best only cover copyrights after 1978 for technical reasons, so tens of thousands of jazz, blues and classical compositions created before 1978 are not included, as well as songs from outside the US before or after 1978.
Why buy this data when these giant corporations already have so much information at their fingertips? Because the point for the services is not to find out who actually owns the songs, the point is to find out if the Copyright Office has a record of who owns the songs based on the Library of Congress data.
That is the hack.
Kafka-esque Moral Hazard
In other words—the government allows Google to claim they can’t find the songwriter even if Google’s own data would reveal their identity just because the song owner isn’t included in the Library of Congress database at the time Google searches. And there’s the “gotcha”.
Kafka’s next book is in there somewhere.
Offering all the world’s music all at once presents a licensing problem that no system will be able to solve due to the sheer numerosity and disaggregation of the creative process. How many songs will be written by the time you finish reading this post and how would you find out who wrote them?
So it should not be surprising that the market has offered a few ways to solve for this problem: Direct licenses (bypassing the NOI altogether) and NOI clearance companies that specialize in maintaining song owner information to send out mass mailings of NOIs (sometimes called “carpet bombing NOIs”).
These are two significant methods available to Google and Amazon and my guess is that these monoliths employ both methods for their interactive streaming services (the kind of service that competes with Apple and Spotify).
What’s the Alternative?
If Google and Amazon cannot find the song owner under their direct licenses or through an NOI company, how can they find the song owner? The easy answer is don’t use the song. But that approach is counter to offering all the world’s music at scale by creating supply that is not responsive to demand.
Deciding which songs are right for “address unknown” NOIs requires some Silicon Valley style hocus pocus. Remember—it’s not that Google can’t find the song owner. The loophole requires that they can’t find the copyright owner in the pubic records of the Copyright Office, even if Google has actual knowledge of their whereabouts.
Then you have to believe that Google knows where to get the information for which direct licenses they want, they know how to carpet bomb NOIs, they have a decade of information in Content ID, but when it comes to some songs, Google has to turn to the Library of Congress? And Google’s only choice is to serve “address unknown” NOIs on the Copyright Office?
Once served, the Copyright Office posts these mass filings on their website in large Excel files so that songwriters can sift through the haystack to find their needles. This hit and miss and self-serving process is fraught with moral hazard and should not be the law in 2016.
This is what the filing looks like—but realize that “1 NOI” means “1 NOI With An Excel file with over 40,000 songs on it”.
Sting Songs Give Some Examples
A spot check of a couple of Google’s filings reveals that Google is not getting it right. Let’s use three Sting songs for an example.
Sting’s recent release “50,000” (coincidentally a tribute to David Bowie and Prince) is on Google’s “address unknown” NOI list. That song is probably subject to a direct license, but the song copyright registration may not yet have been processed. There’s almost always a delay in processing copyright registrations, so new releases will rarely appear in the Library of Congress database day and date with the song’s release. Google will not be paying royalties on Sting’s song, but will be exploiting it.
That’s right—a song that is a tribute to an artist rights advocate like Prince is itself being ripped off.
Google has also filed an “address unknown” NOI for a song entitled “Fragile (Live)”. My bet is that “Fragile (Live)” is “Fragile”, the well known hit song and anthem of the environmental movement.
This likely means that someone at Google seems to think—or wants to think—that “Fragile (Live)” is a different song than “Fragile”, probably because there is a sound recording registered for “Fragile (Live)” in the sound recording metadata but no song registered by that name in the Library of Congress database. And why would there be if it is the same song? We humans have a way to catch this kind of mistake.
It’s called listening.
This pattern repeats with “Brand New Day (Cornelius Mix)”, also included on Google’s “address unknown” NOI. Again, a version of the sound recording, not the song. The song remains the same.
It is highly likely that the songs “Fragile” and “Brand New Day” were registered with the Copyright Office long ago. That’s probably why the “Live” and remixed versions of the sound recordings show up in Google’s NOI filing for the songs and the original versions do not.
In this case, not only are these songs likely covered under a direct license with Sting’s publisher, but even if they are not, the song owner’s information is identified in the public records of the Copyright Office. The loophole does not apply, but Google takes it anyway and the cost of checking up on a multinational media company falls on the songwriter.
And given that it’s Google, the songwriter will probably have to sue them to a final non-appealable judgment in order to fix the mistake that should never have been allowed to happen in the first place.
The government’s compulsory license has become distorted by rent-seeking behavior by multinational media corporations. It should be stopped or substantially modified. If Google is allowed to use this loophole to profit at the expense of songwriters from its considerable influence peddling and litigiousness, that will be crony capitalism writ large.
(From the Huffington Post, July 12, 2016)
Starting in 2014, the Obama Administration conducted a prolonged and expensive process of soliciting public suggestions to update the archaic antitrust consent decrees that regulate songwriters represented by the ASCAP and BMI performing rights organizations. Most of these suggestions came from songwriters themselves, but a handful came from digital music services.
Spoiler alert: The ones from digital services and—apparently—front groups for digital services—were the ones that mattered. Especially front groups funded by former clients of Obama Administration officials who were pretty clearly violating a revolving door ban in Executive Branch ethics rules while working on amending the ancient consent decrees. A unilateral amendment that has transmogrified into legislation by any other name.
It has recently come to light as reported out of closed door meetings with at least one conflicted official at the Department of Justice that the Obama Administration has decided to ignore all of the ideas presented by songwriters, publishers and performing rights organizations and focus on the one thing that is almost guaranteed to destroy the PRO system in the U.S.
The Obama White House intends to adopt the punitive policy of “100% licensing” to the great benefit of Google once again. That would be Google, the former client of the conflicted Justice Department official.
What Does “100% Licensing” Have in Common with a DeLorean?
Simply put, 100% licensing derives from the English common law—even older than the consent decrees. It refers to the ability of a co-owner of an undivided interest in real property to grant a nonexclusive license to allow a third party to use the whole parcel without the consent of her other co-owners. A co-owner relying on this rule also assumes the obligation of accounting to her other co-owners and may not license at a rate that constitutes economic waste of the property. So yes—the Obama Administration wants to lead songwriters back to the future.
The Obama Administration seeks to apply this theory to song copyrights through the consent decrees. After all the hopeful aspirations that the legacy consent decrees were going to be fixed by the Obama Administration based on the public comments solicited by the Justice Department, it now appears that at this late stage of the Administration’s term, the can will just get kicked down the road even further.
Still confused about 100% licensing? If you’re confused, it’s partly because the concept applied to songwriters subject to the consent decrees is inapt and illogical. Take these examples:
Example 1: Two ASCAP and BMI Co-Writers: If I am an ASCAP songwriter and you are a BMI songwriter and we co-write a song, we have to decided on our shares. Let’s say we decide to share the song 50/50. That means 50% of the song that you write is subject to the BMI blanket license because you have an agreement with BMI that BMI represents your share of all songs you write. My 50% is likewise represented by ASCAP for the same reasons. We may have a simple song split agreement between us, typically a one-page form where we acknowledge each other’s shares and that we will each administer our respective shares. We have to agree on any licenses, such as synchronization licenses, and our PROs collect all the performances.
Under the current regime, because ASCAP and BMI are the only two PROs that are subject to the consent decrees out of the handful of PROs that exist in the U.S., the government determines all aspects of how songwriters can license their songs. The government controls songwriters through the consent decrees that are the longest in U.S. history (in place since 1941) and also through the rate courts, an expensive and cumbersome process of a single federal judge deciding what the ASCAP rates should be and a second federal judge deciding what the BMI rates should be.
Services have for decades obtained licenses from both ASCAP and BMI and usually one or more of the competitors to ASCAP and BMI (who are themselves very much competitors) such as SESAC and more recently a new PRO called Global Music Rights.
Under 100% licensing, a digital music service can go to either ASCAP or BMI to license 100% of our song, even though we agreed between us that our shares would be licensed through our respective PRO blanket licenses. If the digital service went to ASCAP first, then ASCAP would have to license the BMI share and vice versa.
Doesn’t seem legal? But wait, there’s more.
Example 2: One ASCAP, One BMI and One GMR Songwriter Same example, except add a third writer who is represented by GMR. Global Music Rights is not subject to the ASCAP or BMI consent decrees.
In this case, because the GMR writer made the creative choice to write with you and me, she is now subject to either the ASCAP or the BMI consent decree, just as if she had signed to ASCAP or BMI. But she can’t tell which one, because the digital service can decide which one she will be forced to license through—even though she never signed to ASCAP or BMI, and may have made that decision because she did not want to be subject to the government’s regulation of songwriters through the already unpopular—and about to get a whole new level of unpopular—government regulation of songwriters.
If this seems complex, it’s really exponentially more complex in practice for the billions and billions of streams every month, not to mention a “regulatory surprise” that imposes a cost not included in the judge-made rates set under the consent decrees.
And that’s right—the Obama Administration is about to rule that the consent decrees now extend to all songwriters who write with an ASCAP or BMI co-writer regardless of what their private contract says with their PRO or with their co-writers. This all is for the benefit of digital services, especially Google. That would be Google who is locked in a fight with GMR over YouTube.
And that’s also right—the most valuable company in the world that is being sued by countries for violating antitrust law needs to be protected from the anticompetitive urges of songwriters. At least according to Google’s former lawyer now driving the attack on songwriters in the Obama Justice Department.
Legislation by Any Other Name
If the consent decrees are starting to look like legislation, you’re not the only one who thinks so. Initial reaction from Congress is negative, such as this from Congressman Doug Collins:
“The Department of Justice’s position is arrogance at its worst. The decision fails to address the vitally important issue of terminating or reforming outdated consent decrees, and instead broadly expands its interpretation of existing consent decrees. Under this expanded interpretation, the PROs must adopt “100 percent licensing” even if the PRO does not represent all joint owners of a musical work. This is a departure from common practice in the music industry and one that could have drastic consequences. The decision won’t fix anything, but it will create new problems for the music industry. I fear these problems could undermine creative collaborations, harm songwriters, and tip music industry contracts and negotiations on their head.”
The Obama Administration’s relatively new position appears to have been based on extraordinarily bad advice—advice that is so bad it looks punitive. This in part because in order to get to the punchline, the Administration has to ignore the implications to international trade, replace a voluntary licensing doctrine with a government mandate, ignore written agreements between generations of songwriters, and impose untold transaction costs on songwriters without requiring an increase in royalty rates to permit cost recovery.
The Four Preconditions
It is true that the ancient common law rule rule has been applied to copyright in the U.S. from time to time, but it is actually quite rare because of four preconditions.
Limitation to U.S. Law
First, to the extent the rule obtains at all, it is a U.S. creature. Applying this rule to copyrights originating in countries other than the U.S. when the rule is not recognized in those other nations raises the real possibility that the proposed application by the Department of Justice is unlawful. In fact, it may actually be a treaty violation that could cause the United States to be hailed into a WTO arbitration. (See Fairness in Music Licensing Act where that exact thing happened, the U.S. lost, and U.S. taxpayers are subsidizing foreign songwriters.)
Limited to Voluntary Licensing
The ancient 100% licensing rule also involves voluntary licensing by the co-owner. To my knowledge, it has never been applied to a government mandated license in copyright, real property or otherwise.
Interferes with Private Contracts
The rights of the co-owner typically will originate with some agreement or purchase agreement that grants to the co-owner the right to the use of the whole of the property even though they only own a partial interest. In order to be effective, the co-owner license must not violate an agreement to the contrary between or among the co-owners.
At a minimum, songwriters often avail themselves of “song split agreements” to document their percentage ownership. Since song split agreements typically provide for each writer to administer their respective shares of copyright, it is likely that there are hundreds of thousands, if not millions, of song split agreements covering songs available under ASCAP and BMI blanket licenses.
Not only are there likely to be written agreements covering these songs, the fact that each songwriter has registered their works with their respective PROs of which they are writer members is pretty easily interpreted as an “implied in fact” contract from the mere uncontested registration of song shares with multiple PROs. As the U.S. Supreme Court noted in Baltimore & Ohio R. Co. v. United States, 261 U.S. 592 (1923):
[A]n agreement “implied in fact” founded upon a meeting of minds, which, although not embodied in an express contract, is inferred, as a fact, from conduct of the parties showing, in the light of the surrounding circumstances, their tacit understanding.
What could be clearer than the uncontested act of PRO registration?
Negotiation for All Costs of the License
Perhaps most importantly and most relevantly for this post, the co-owner’s license is presumed to be negotiated at a rate that will take into account the cost of the license to the granting co-owner. This is another place that the Obama Administration’s proposed rule disintegrates or even becomes punitive.
Under the DOJ’s proposed 100% licensing rule, the applicable rate payable to PROs under the consent decrees is a rate for the use of the music licensed. No rate court took into account the “surprise” cost of administering songs in a 100% licensing world now being created by the DOJ from whole cloth.
Cost Recovery
Since the cost of administering these licenses was never included in the rate, any fee charged for 100% licensing by PROs would simply offset the costs for the convenience of the licensee music user and not be a rate for the benefit of songwriters, it seems proper that any music user seeking to trade on this theory should pay the freight.
In other words, ASCAP and BMI songwriters should be able to charge a fee for the convenience of 100% licensing that should be outside of the consent decrees and rate courts altogether. If not, the new transaction cost of administering 100% licensing when deducted from the already minuscule rate court license fees may well cause the music user to be in a better position than she would be in if the PRO had fully performed under the consent decree’s terms.
This is particularly true regarding the rates that were mandated by the two rate court judges without an opportunity for songwriters to be heard regarding these additional “surprise”regulatory costs now contemplated by the Administration.
If the Obama Administration wants to lame duck their way out of amending (or terminating) these ancient consent decrees, they could at least do songwriters the courtesy of requiring their revolving door appointees to tell songwriters to their faces at a public hearing before the Senate Judiciary Committee.
Instead, the Administration encouraged songwriters to wrack their brains on ways to improve the ancient rules, then pulled the rug out in favor of what Rep. Collins called “arrogance at its worst.”
From Hypebot, May 25, 2016 (reprint from MusicTech.Solutions) by Chris Castle
Like any large organization, Google has competing bureaucracies and therefore its wholly-owned subsidiary YouTube does as well. (Google is now the largest media company in the world.) YouTube’s organizational independence is additionally blurred because it is the #2 producer of revenue inside Google relative to search and advertising sales.
There seems to be a three-legged stool of competing interests in dealing with YouTube which we can describe with generalized labels–the “engineers”, the “policy people” (essentially Fred Von Lohmann) who are mostly lobbyists and lawyers, and the “business people” starting with Robert Kyncl at least at the moment. It’s unclear who has the upper hand in this triumvirate, but it’s pretty clear that the business people do not control their destiny.
That leaves jump ball for control of YouTube’s deals between the engineers and the policy people who seem to compete with coming up with the solution that is the worst for anyone with a passing acquaintance with private property rights in general, and artist rights in particular. I say artist rights because it’s not just copyright that is the problem with YouTube–it’s also right of publicity, control over derivative works, translations, moral rights, misappropriation, and other consent rights one would expect any artist would have. Plus copyright. Artists certainly do get some of these rights pretty much for the asking from the big bad record companies in the form of marketing restrictions, for example. Hence, “artist rights.”
YouTube’s ineffective negotiating power with Big Google is particularly confusing because YouTube is both a search engine and an advertising publisher. (Let’s call the larger Google “Big Google”.)
We sometimes forget that YouTube is the largest video search engine in the world. Once the European Commission gets through fining Google for predatory business practices with Google search, and finishes up the Commission’s separate prosecution of Google for predatory business practices with Android, the Commission may then have the appetite to bring a case against Google for some of the same predatory practices as applied to YouTube. I’m going to guess that Google’s dominance of video search is likely equal to or in excess of its dominant position in organic search and mobile meaning the EC’s scrutiny will be quite enhanced and (by that time) educated in Big Google’s ways.
Why it is that YouTube has such little clout internally is anyone’s guess. My bet is that if YouTube didn’t have to check with a host of bureaucrats at Big Google, it would be much, much easier to do business with YouTube.
What’s obvious is that the engineers and policy people do not understand a fundamental point about dealing with the creative community. They are every bit as much of ambassadors to the creative community–the entire creative community, not just the YouTubers who essentially are entirely dependent on YouTube for their success–as are the creative people or marketing folk at YouTube.
To state the obvious, unlike the YouTube lottery winners, professional artists who are not dependent on YouTube are not dependent on YouTube. If pushed, there very well may come a day that they move on. En masse.
That may happen sooner than you might think, despite YouTube’s monopoly on video search. YouTube is currently taking a beating from artists and songwriters. Note that the beating is administered to YouTube–not to the engineers and the policy people at Big Google. Or not yet, anyway. Most professional creators don’t know these bureaucrats exist. Those bureaucrats at Big Google are largely faceless (with the exception of Fred Von Lohmann) and take no heat when YouTube gets roasted alive by key opinion makers in the music business (such as Irving Azoff).
To see where this goes, we need only look to Pandora’s experiences with this kind of response to the Internet Radio Fairness Act of 2012. Pandora is still digging out of that hole some four years later.
Four years later.
I seriously doubt that the day to day business people at Pandora wanted to go through this misstep, and the stockholders definitely did not. IRFA sprang from the intellectual loins of Pandora’s “policy people” by all accounts, and the business people apparently didn’t really have much to say about it.
So how could we repair the problems with YouTube? I think that it’s going to be a heavy lift, but it would start with Big Google telling their engineers and policy types to back off. Then we’d at least have an idea of whether YouTube can ever be a good partner. I suspect we could have at least much better relations with an independent YouTube.
Google may be willing to bet that they can outspend and out lobby the creative community. I don’t know as I’d take that bet. While the government has had their boot on the throat of creators in the form of compulsory licenses, consent decrees, and the very unpopular DMCA safe harbors, they can’t make creators happy about it.
YouTube should try to shake off the control of their internal masters at Google. Then at least we’d know who we are dealing with.
Hypebot Guest Post by Chris Castle on Music Tech Solutions
I had the good fortune to participate in a SXSW panel about the mechanics of YouTube revenues. If I say so myself, it was a wonderful panel with some deep expertise (“Stop Complaining and Start Monetizing“). There was a real interest in the audience about the mechanics of the rights involved and the revenues paid.
If you have that same interest and you weren’t able to go to SXSW, here’s a basic chart of revenue splits that may help you:
YTP, YTPC= “YouTube Partner“, “YouTube Partner Channel”
SR= Sound Recording
WW= “Wild West” meaning no particular rule.
Notice that the basic categories are song, sound recording and video which track the main three copyright categories of musical work, sound recording and audiovisual work.
The percentages refer to shares of “Net Ad Revenue” often defined as:
“Net Ad Revenues” means all gross revenues recognized by YouTube attributable to any sponsorship of or advertising displayed on, incorporated in, streamed from and/or otherwise presented in or in conjunction with any User Video displayed on a Covered Service including, without limitation, banner advertisements, synchronized banner advertisements, co-ads, in-stream advertising, pre-roll advertising, post-roll advertising, video player branding, and companion ads, less ten percent (10%) of such gross revenues for operating costs, including bandwidth and third-party (affiliated or unaffiliated) advertising fees. Net Ad Revenues excludes any e-commerce referral fees received by YouTube from “buy buttons” or “buy links” on the Covered Services that facilitate recorded music “upsells” when a Publisher separately receives payment from a third party in connection with such an upsell (e.g., royalties for a CD or sheet music sale); provided, however, for the avoidance of doubt, that such exclusion does not extend to (a) advertising of the type described in the first sentence of this Section for recorded music products, the revenues from which shall be included in Net Ad Revenues; and (b) all other types of e-commerce referral fees and revenues, which shall be included in Net Ad Revenues.
One key component of your YouTube earnings is the “CPM” paid by advertisers to Google. Even if you have the right to audit YouTube (which few do), it is highly unlikely that you will ever be able to determine what the CPM is that Google uses to pay you on YouTube. Multichannel networks (“MCNs”) like Machinima have reportedly tied creators to CPMs that were well below market, particularly considering that the highest CPMs on YouTube are often associated with exactly the kind of talent most frequently signed to an MCN.
“Official” or “Premium” Videos
When a label uploads an “official” music video on YouTube or Vevo, the video has higher production values than UGC and is usually supported by a sustained marketing effort outside of YouTube that drives traffic to the site. If the premium video appears on Vevo, then 100% of the royalty is paid to the label, which in turn has licenses from the publishers for the song. If the video is on YouTube proper, then the label’s share is reduced by the publisher royalty, often around 15% of net ad revenue.
Claiming and YouTube’s Content Management System (“CMS”)
Because of a combination of YouTube’s monopoly position in the market, Google’s controversial reliance on the notice and takedown provisions of the Copyright Act and its sheer litigation muscle, YouTube will let anyone upload anything also known as “user generated content” or “UGC”. If you have access to YouTube’s “Content Management System” or “CMS” you have the chance to block UGC through YouTube’s “Content ID” fingerprinting tool.
Compared to the massive volume of videos uploaded to YouTube, a very, very small percentage of copyright owners have direct access to Content ID. According to YouTube:
YouTube only grants Content ID to copyright owners who meet specific criteria. To be approved, they must own exclusive rights to a substantial body of original material that is frequently uploaded by the YouTube user community.
Participating in Content ID allows you to help YouTube create a vast and valuable library of reference versions ofyour works. (YouTube does not compensate you for participating in Content ID.) Rightsholders usually participate in Content ID for two reasons which are not mutually exclusive: Blocking or “monetizing”. Monetizing means that you give YouTube permission to sell advertising against your works. Naturally, YouTube hopes you will choose to monetize because over 90% of Google’s revenue comes from selling advertising online.
YouTube creates a reference version of your work in the form of a “fingerprint” (a psychoacoustic technique that has long been in use by the U.S. Navy among others to distinguish sound patterns–see Jonesy in The Hunt for Red October). A fingerprint is a mathematical rendering of the waveform of an audio file that essentially reduces a sound recording to a kind of hash that makes comparing fingerprints quicker and more accurate.
YouTube maps the reference fingerprint to other identifiers such as the International Standard Recording Code for sound recordings, song title, artist name and copyright owners for all of the above including song splits in many cases. When a work is in the Content ID system, YouTube will compare an uploaded video to the Content ID database reference fingerprint and most of the time will follow the rules established by the copyright owner to block or monetize (often called “match policies“). If the match is done before the UGC video is uploaded, then it won’t go live, and if the match is done after it is live, then the users will see one of YouTube’s controversial messages saying the file is blocked due to a claim by copyright owner X.
What this boils down to is that if you don’t have a label or publisher, you will need to go to a claiming service like Adshare, The Collective or Onramp in order to get access to CMS and Content ID in order to monetize your works outside of a YouTube Partner Channel (which is done through an Adsense account associated with your YouTube Partner account). If you have a label, publisher or claiming service, then all of these entities should have access to CMS and Content ID and will be able to claim your songs, sound recordings or videos and monetize them if you wish.
Deciding if Monetization is Right For You
If you’re familiar with term recording artist agreements or publishing agreements (or what is normally called a “record deal” or “publishing deal”), you’ll probably remember negotiating “marketing restrictions” involving the use of your recording or song in advertising. Those clauses usually restrict the use of your works in political ads, certain kinds of products (firearms, tobacco for example), or more artist-specific restrictions. There are also restrictions on the kinds of movies or television programs (even videogames) in which your works can be used.
If you allow your work to be used in UGC and you elect to monetize, you can just forget all that on YouTube. “UGC” includes just about anything you can imagine short of explicit pornography, but would include, for example, sex tourist home movies, jihadi recruiting videos (although “songs” are unlikely to appear there), hate speech and the like. All of those are on YouTube and frequently are not behind any kind of age restriction wall.
The ads that get served as preroll for these videos are themselves often unsavory. For example, Google serves ads for “dating” sites that are in categories frequently identified as thinly disguised human trafficking operations. There are ways to block these particular uses if you have access to CMS but due to YouTube’s “catch me if you can” business practices, you may have to spend the time to track down each use which otherwise can stay on YouTube for months or years.
Winning the Lottery
We often hear about “YouTube stars” with elite channels (1 million plus subscribers) who are very well compensated. The source of this high level of compensation is rarely limited to advertising revenue. Most of the time, their ad revenue is salted with a high number of payments for what are essentially sponsorships, endorsements or product placements, often called “brand integrations“.
In the music and movie businesses, the term “star” is usually reserved for a relatively small group of performers who have demonstrated ability over time to reach a large audience, often a global audience. YouTube “stars” may have large YouTube communities and may be able to introduce products to fans on YouTube, but whether that will hold up on YouTube over time or translate to other platforms remains to be seen in most cases.
It is also important to realize that advertising is a highly regulated business, particularly when it comes to false or deceptive advertising that is regulated by the Federal Trade Commission. Machinima has just entered a 20 year consent decree with the FTC to settle claims that it misled consumers by passing off paid endorsers as independent reviewers. Given that Machinima and other MCNs are supposed to protect their talent from such missteps suggests that YouTube stars may well have more to watch out for on YouTube than do recording artists or songwriters on record labels or music publishers.
Online Advertising in Decline
Whether it is ubiquitous ad blocking software, “do not track” settings on browsers, or distrust of advertisers, online advertising is in decline. Like a ship that is sinking very slowly, it is sometimes difficult to tell if you’re really lower in the water, or if that was just a wave. And remember, over 90% of Google’s revenues come from online advertising, moonshots notwithstanding.
If the online advertising ship really does sink, all the driverless cars, military robots and Google Glass will not save Google or YouTube. That’s something to keep in mind when you agree to participate in the YouTube monetization game.
From Texas Lawyer (January 28, 2016)
Two recent copyright infringement class action lawsuits against Spotify show the $8 billion music streaming service's Achilles heel—Spotify's largely secret failure to license songs.
According to Billboard, Spotify failed to license or pay royalties for between 10 to 25 percent of Spotify's 30 million song music offering and claims to be accruing a "reserve fund" of $17 million to $25 million for unpaid royalties. Spotify has so far refused to disclose how many songs are unlicensed or how much it owes for which songs.
Given the scale, it is very likely that Texas songwriters could be included in either potential class, particularly independent songwriters who have registered their U.S. song copyright.
Two different songwriters brought the class actions separately: David Lowery ($150 million) and Melissa Ferrick ($200 million).
Recall that sound recordings contain two separate copyrights: the song and the recording. Different parties can own each work. These class actions concern the song copyright.
Spotify defends itself by complaining that U.S. song licensing is uniquely complex. Outside the U.S., songs come in a single blanket license (e.g., the UK's PRS for Music). In the U.S., the bundle of rights is divided between the blanket public performance rights licensed by performing rights organizations like ASCAP and BMI under government orders, and compulsory mechanical rights under a statutory license. The government controls the price for both rights. Spotify allegedly failed to comply with the compulsory license laws despite warnings—resulting in the alleged copyright infringement forming the basis of the class actions.
This complexity is created by government regulation. ASCAP and BMI are prohibited from licensing the mechanical rights (for the exact same performances, songs and users). The federal government has long controlled licensing and prices for songs, so Spotify knew this when they entered the U.S. market. For example, the federal government set the price for the compulsory song license at $.02 in 1909 where it stayed for 67 years; the ASCAP consent decree started 75 years ago.
While the government actively regulates songwriting, it largely avoids enforcing compliance with its licenses, essentially leaving enforcement to songwriters and music publishers.
The class actions accuse Spotify of failing to comply with the two main requirements of statutory requirements for compulsory mechanical licenses. The user must: notify each copyright owner of the proposed use in advance; and account and pay government royalties for uses of the song.
The class actions allege that while Spotify may have licensed some of the songs it offers, it has not licensed potentially millions of songs. The cases allege that Spotify admitted these failures in various public statements.
It is reasonable to ask if a person who fails to comply with burdens of the compulsory license ought to continue to enjoy its benefits. The Copyright Act has no "use it or lose it" mechanism that would suspend or extinguish an abuser's right to the compulsory license—once again, enforcement is left to songwriters and publishers.
The advance notice requirement states that the user must either send the notice to the copyright owner or send it to the Copyright Office if the owner cannot be found in the records of the Copyright Office. Press reports suggest that Lowery could have been found with a "ten second" online search of Copyright Office records.
Spotify claims that it cannot locate the rights owners, but apparently does not acknowledge the unknown owner option. The class actions suggest that Spotify's failure to send notices for millions of songs suggests something more than a mere oversight.
One possible explanation for Spotify's predicament is a reluctance to pay the filing fee to register the notice with the Copyright Office who also publishes a list of all unknown owner notices on the Copyright Office website. Notice filing is a kind of insurance against copyright infringement claims until the owner comes forward, if ever. That filing fee works out to about $25 per song—which looks cheap compared to statutory damages.
It appears that even $25 per song does not trump the "optical" risk assessment—the likelihood of getting caught balanced by a bet that songwriters would never have the means to sue.
A similar problem arose in 2004 regarding unclaimed royalties at record companies and publishers. The New York Attorney General investigated the rightsowners and found that they needed to do a better job of paying royalties. That started with the rightsowners publishing a list of all the artists for whom they were holding royalties.
Plaintiff Lowery evidently knows the history—before filing his action, Lowery asked the current New York Attorney General to conduct a similar investigation into Spotify's undisclosed and unpaid royalties, but evidently got no reply after several months.
Press reports suggest that one solution might be to follow those 2004 guidelines starting with requiring Spotify to publish a public list of unlicensed song titles (from Spotify's sound recording metadata). Lowery's class action effectively asks the court to order this disclosure.
Spotify may well be doing it better than other services but it appears they all have the same problem—they apparently decided it was better to seek forgiveness than ask permission. Or comply with the law.
[From The Huffington Post, January 18, 2016]
Spotify's music offering may have appeared to be the long tail, but now it looks like the wrong tail. After two class-action lawsuits for unlicensed songs and copyright infringement were filed against the company by songwriters in as many weeks, Spotify is doing back flips to blame the victim for its admitted failures to lawfully obtain mechanical licenses. But at the end of the day, the decision is governance not guidance, and Spotify's governance is solely in the hands of its corporate directors and officers. Those directors and officers must answer to shareholders for weighing the marginal value of millions of unlicensed songs over the infringement liability (not to mention an even worse press).
Free Lunch or Free Will?
According to press reports, 10 percent to 25 percent of the songs on Spotify "are not properly licensed and/or not distributing royalty payments." Spotify also claims to have licensed approximately 30 million recordings (of 30 million songs, give or take for covers).
Based on these assumptions, that means there could be three million to 12 million songs that "are not properly licensed and/or not distributing royalty payments." This is not a few songs that fell through the cracks like new releases, a 1/16th of a song for a sample, the odd songwriter who cannot be found or who is non-responsive.
Millions of unlicensed songs doesn't look like an understandable accident, it looks more like an unacceptable policy. And that would be a policy that is exactly what the compulsory mechanical license was designed to prevent.
At the end of the day, the policy, i.e., the choice, to go forward without licenses, rests solely with Spotify. The well-financed company could have complied with the compulsory license -- enacted by the U.S. Congress for this exact situation -- but the songwriter class actions will likely seek to prove that Spotify chose not to do so. Whoever Spotify hired to undertake the mechanical process of mechanical licensing, millions of unlicensed songs strongly suggests that someone at Spotify decided to go forward without complying with the law. It appears that the thinking was that the upside value of having "all the world's music" was greater than the downside risk of getting caught. The marginal value of another few million songs was greater than actually complying with the law and paying songwriters.
This decision is what is called "business risk." Incredible as it may seem, this decision -- this willful decision -- to accept the business risk of using millions of unlicensed songs was apparently driven by a belief that in order to have an effective consumer offering, Spotify had to have tens of millions of tracks available to consumers.
It's Not About a Database
You'll hear a lot of hand wringing (from many quarters including Spotify) about the lack of a central database that is constantly updated with complete rights information just so digital services can look up who to license from and who to pay. This database has never existed, but you know what does exist? The U.S. Copyright Office online searchable database of copyright registrations.
In David Lowery's case, the songs he complains of in his class action all were registered with the Copyright Office. He published a list of them -- with registration numbers -- on his blog on October 20, 2015. He also complained to the New York Attorney General about the problem in November with no response.
Evidently, Spotify did not bother to look up Lowery's copyrights at the Copyright Office registration system as required by the compulsory license rules. That's behavior that is consistent with having millions of unlicensed songs. If the industry were to go to the trouble and expense of building this unicorn database, it would not have solved Spotify's problem with David Lowery because evidently nobody bothered to do the song research.
And this is the most elegant explanation of why Spotify is in the situation they are in -- they don't want to look because they don't want to pay the cost of complying with the law that provides them the great benefit of a compulsory license and they don't want to pay songwriter royalties. If Spotify really wants to pay songwriters, wouldn't they have done at least as much as the Copyright Office does with unknown writers -- publish a list online with all the information that they have (like artist name and song title).
Enter the Long Tail
This policy of using millions of unlicensed songs may well have been informed by the "long tail" theory and thought experiment posited by one Chris Anderson (in case you forgot him). You can read all about it in Anderson's counterintuitive utopian book The Long Tail: Why the Future of Business is Selling Less of Morewhich was based on a 2004 article in Wired.
I'd be very interested to know exactly where this consumer research is that shows the marginal value of an additional 12 million songs is so meaningful to a music service that it trumps the infringement exposure. I frankly have never seen it -- aside from Spotify's reliance on Anderson's version of the long tail.
Anderson goes down the wrong rabbit hole by relying on anecdotal observations of "Ben" an anonymized (or perhaps fictional) character (at pp. 3-5). "Ben" is a teenager from an affluent family in Silicon Valley who gets most of his music from "friends" and "Bit Torrent" (recall that Spotify's CEO was a developer of uTorrent, a key piece of the piracy picture acquired by Bit Torrent in 2006). So Anderson starts by analyzing a legal market with comparisons to the black market. That obviously wasn't going anywhere logical. Neither is any market of what the New York Times called "pixel-size niches".
Anderson's long-tail thought experiment has been criticized by a number of people such as Harvard Business School Professor Anita Elberse in the Harvard Business Review and most famously in the music business by Will Page, the former economist for PRS, the UK performing rights organization.
Any record company production manager could have chimed in -- and perhaps would have if it wasn't so obvious that it did not really bear much discussion. The corresponding transaction costs of a variety of functions including rendering royalty statements for minuscule unit sales were not worth keeping the title in the catalog. You know, kind of like sending a royalty statement for three streams. Preparing the statement may well cost more than the royalty even if the statement is itself digitally delivered. Not to mention taking the phone call from the angry songwriter who got a statement for $0.19.
Record companies are no strangers to the long tail -- that's often called classical and instrumental jazz. It is worth noting that record companies have for decades deleted titles that didn't sell enough to justify keeping the title in the company catalog. This is consistent with Professor Elberse's research demonstrating that "the tail increasingly consists of titles that rarely sell and that are produced by smaller-scale players." Professor Elberse assumed that there were no infringement costs associated with those "titles that rarely sell" thus exponentially increasing the cost of the tail, or as this particular tail is known in some circles, the wrong tail.
Is the Long Tail the Wrong Tail for Spotify?
Then-PRS economist Will Page reached a similar conclusion after analyzing PRS royalty payments in 2008. Those who have had about enough sanctimony from Spotify about how it is God's gift to fighting piracy will find this nugget of interest when wondering how much the marginal value of the last 12 million tracks that don't sell really is worth if they are all unlicensed:
Will Page, the former economist for PRS, found in his 2008 study that famously debunked Chris Anderson's absurd "long tail" theory, the "long tail" is pretty meaningless for music services:
[PRS] found that only 20% of tracks in our sample were 'active', that is to say they sold at least one copy, and hence, 80% of the tracks sold nothing at all. Moreover, approximately 80% of sales revenue came from around 3% of the active tracks. Factor in the dormant tail and you're looking at a 80/0.38% rule for all the inventory on the digital shelf.
Mr. Page joined Spotify in 2012, which was after Spotify's U.S. launch in 2011. If Spotify's policies led to its problems with songwriters, that is certainly not Mr. Page's fault, but Spotify could have just asked him about the cost/benefit analysis before continuing to take the business risk of failing to get compulsory licenses on what apparently is a gargantuan scale.
Was the marginal value of the long tail worth it for Spotify when compared to statutory damages? Commentators often mock statutory damages, especially for willful infringement, as being over the top. In the case of the compulsory mechanical license, you can look at this another way.
Congress did backflips to make the compulsory license easy to get. If a well funded company like Spotify (last valuation reportedly $8.4 billion) chooses to ignore Congress's efforts, then perhaps Congress wants to make sure that the marginal value of ignoring the compulsory license is always less than the statutory damages for choosing to do so.