By Marc D. Ostrow
On Friday, February 6, the Copyright Office issued a 202 page comprehensive report (plus appendices) on the music licensing business, “Copyright and the Music Marketplace.” The Report is the culmination of a nearly year-long process of soliciting and evaluating input from interested parties on how to fix what everybody agrees is a broken system.
Anyone with an interest in the music business should read the full report – or at least the 11-page executive summary. But in case even that’s too much, here’s all you need to know, in layman’s terms and with analysis, in little more than half the length of the executive summary:
The Report starts with four guiding principles:
– Music creators should be fairly compensated for their creations
– The licensing process should be more efficient
– Market participants should have access to authoritative data to identify and license sound recordings and musical works
– Usage and payment information should be transparent and accessible to rights holders.
Like Mom and apple pie – it’s kind of hard to argue with these. But before we get to the Report’s recommendations as to how to implement these principles, including four subsidiary principles, we need some background on the current music licensing framework. So instead of the Report’s 50-page primer (which is quite readable and mostly correct), here’s a roughly three-page summary of the current music licensing landscape, rocky as it is.
The Report is primarily concerned with the distribution of recorded music, whether through sales of physical product like CDs and downloads or public performances, whether over the radio or by streaming services on the Internet. This means that unless it’s a recording of public domain music, like Beethoven, most recordings consist of two distinct copyrights: (1) the copyright in the musical work, which is typically controlled by one or more music publishers; and (2) the copyright in the recording of that work, which is typically controlled by a record label. This is best illustrated with “cover” records. For example, I prefer the Carole King version of “You’ve Got a Friend” to James Taylor’s. Same song, two different recordings; two separate copyrights for each recording.
Let’s deal with the songwriter/publisher side first. ASCAP, BMI and SESAC are performing rights organizations (PROs) that license the public performing right (and only that right) in musical compositions (i.e., songs, but not the recordings of them) when they are performed live in stadiums, concert halls and clubs, broadcast on radio and TV or streamed over the Internet. PROs typically issue “blanket licenses” to users, meaning for a set fee (either a flat fee or percentage of the user’s revenue, depending upon the license), the user has an all-you-can-eat buffet of the music in that PRO’s repertoire allowing the user, such as a radio station, to play any song in the PRO’s catalog as often as it likes. The PROs pay 50% of the licensing revenue to the writers and 50% to the music publishers after deducting their operating costs.
ASCAP and BMI, according to the Report, represent more than 90% of the domestic music market while SESAC and another recently-formed entity represent most of the remainder. ASCAP and BMI (but not SESAC) have been operating under Department of Justice Consent Decrees since World War II. And they haven’t been amended since the dawn of the Internet. Think about that. These decrees were instituted to settle alleged anti-trust violations when 78s were the dominant recording format. Under DOJ regulations in place since 1979, most consent decrees are supposed to terminate within 10 years – not 75!
The Consent Decrees for ASCAP and BMI are overseen by two different federal judges in the New York City. When either PRO can’t reach an agreement as to a license fee either with an individual user (e.g., Pandora) or an entire industry (e.g., radio), the parties may have a “Rate Court” proceeding before the judge. Like all federal litigation, a Rate Court case is very time consuming and costly. Both Consent Decrees state that the judge must determine a “reasonable” fee, which has been interpreted to approximate what a willing buyer and a willing seller would pay for a license in a free, open market.
Most important about these Consent Decrees is that they require ASCAP and BMI to grant a license to anyone who requests one, making the process a de facto compulsory license regime. What’s more, users often pay nothing – sometimes for months or even years at a time – while the parties either negotiate or litigate what a “reasonable” fee should be. Songwriters and publishers have long maintained that users, availing themselves of a compulsory license with the ability to use the “product” while negotiating a fee, are at a significant bargaining advantage.
Still sticking with songs (as opposed to recordings), when a song is covered by another artist, the Copyright Act provides the label with a compulsory license whereby the label pays a statutory rate to the owner of the song. This is how Carole King the songwriter gets paid for James Taylor’s cover recording. The statutory rate is currently set every five years by the Copyright Royalty Board (CRB) in Washington, DC. This three-judge panel sets the fee, not based upon a market rate standard, but in accordance with a separate statutory provision requiring a “fair return” to the work’s creator, while balancing certain public policies, such as maximizing availability of works and minimizing a disruptive impact on businesses and industry practices. The Report indicates that this standard results in lower rates than a fair market standard. Although designed to be solely a license for cover recordings with first recording rights reserved to the copyright owner, most recording contracts have provisions tying the release and payment of all songs to the statutory scheme (often at a lower payment rate). Songwriters and publishers have long maintained that this compulsory scheme, as with performing rights, provides artificially low rates.
This statutory compulsory license (meaning music publishers and songwriters are subject to an “offer” they can’t refuse) is called a “mechanical” license due to the mechanical reproduction of the music and is a term dating back to the days of piano rolls when the license provision was first enacted. But the mechanical license applies solely to audio-only recordings – there is no compulsory license for film, TV, videos, games and other AV uses. Although many music publishers issue mechanical licenses directly, a licensing collective, the Harry Fox Agency (HFA), issues these licenses for probably more than half of the market. However, unlike the performing rights licenses issued by PROs, there are no “blanket” mechanical licenses and they are issued on a work-by-work basis, something that online music services find particularly inconvenient and impractical.
As for audio-visual uses, a “synchronization” (or “synch”) license is required from both the owners of the song and the recording of that song. So, if you want to use Tony Bennett and Lady Gaga’s recording of “Cheek to Cheek” in a movie, you need to get permission from Irving Berlin’s music publisher and also permission from the artists’ label for that particular recording of the standard. Synchronization licenses, unlike mechanical licenses, are typically negotiated and issued directly by the copyright owners, the labels and publishers.
The Report states that between public performance and mechanical income, about 75% of a songwriter’s (and therefore a music publisher’s) income is subject to government regulation (compare that to a novelist whose income isn’t regulated at all). So, that means that the majority of a songwriter’s income can be determined by four judges – one in New York and three in DC. By contrast, a label’s income (and therefore a recording artist’s income) consists mostly of sales of recordings (e.g., CDs and downloads) and licensing of those recordings, such as “synchronization” usage as discussed above. There are no compulsory licenses or consent decrees for these uses so it’s a pure, free market negotiation between labels and users for these rights. And music publishers, who can negotiate synch licenses in a free market unshackled by consent decrees and compulsory licenses, are usually able to get about the same fee for their rights as the label gets for theirs.
But not all restrictions disadvantage the songwriter. With respect to performances, the United States, except in very limited circumstances discussed below, does not grant a public performing right in a sound recording. For example, when Sinatra’s recording of “New York, New York” is played on oldies radio (or over loudspeakers at Yankees games), the songwriters, Kander & Ebb, and their music publisher, get paid through their PRO. What do Sinatra’s heirs and his label get? Nothing! As the Report points out, the United States is one of less than a handful of industrialized nations, including Iran and North Korea, which do not have a public performing right in a sound recording for radio.
Why? There are historical reasons in that the radio stations felt that they were providing the labels with promotion for the sale of recordings. Also, every Congressional district has at least one or more radio and/or TV stations. As the Report points out, with the recent shift in consumer preferences from purchases (e.g., CDs and downloads) to streaming (e.g. YouTube), the promotional value of radio probably isn’t what it used to be.
However, because of laws enacted in the 1990s, there is a limited public performing right in a sound recording for digital transmissions, basically, streaming over the Internet, whether through YouTube, Spotify, Pandora or another service. And there is a compulsory license for non-interactive streaming services, which like the mechanical license, has a rate that’s determined by the CRB. The royalties for the compulsory streaming licenses are administered by a collective that’s similar to the PROs, SoundExchange, which distributes this income to labels (50%), featured artists (45%) and side artists (5%). As for “interactive services” (and the Report spills much ink over the lengthy statutory provisions about what is and is not “interactive”), these license fees are determined in market negotiations by the parties.
Our discussion began with the notion that there are two copyrights in a recording: one in the underlying song and one in the actual recording or “master.” However, for historical reasons, recordings that were made prior to 1972 are not covered by the federal Copyright Act, unlike the songs embodied in them. Rather, these recordings, which are still purchased and performed all the time, are governed by state law.
Recent well-publicized lawsuits in New York and California have determined that, at least in those two states (and likely in many others), there is a state-based public performance right in a sound recording, the contours of which remain largely unknown. For example, it’s possible that in some states, this performing right for pre-1972 recordings could be even broader than the one granted under federal law for later recordings in that there conceivably could be a performing right in the older recordings played over the radio under various state, but not federal laws. This could lead to a quagmire of uncertain and inconsistent treatment.
The Report also contains a lengthy discussion of recent ASCAP and BMI Rate Court decisions, both of which held that publishers could not partially withdraw certain rights from ASCAP and BMI while leaving others. For example, Sony/ATV, one of the three major publishers, felt that it could negotiate better deals regarding digital performances than what it could get through ASCAP and BMI because of the constraints imposed on those PROs by the Consent Decrees. Reaching the same conclusion albeit under slightly different reasoning, both the ASCAP and BMI Rate Court judges determined that a publisher had to be either “all in” or “all out” and that it couldn’t cherry pick certain aspects of the performing right. These decisions figure prominently in the Report’s recommendations.
Why would a major publisher feel they could get a better deal by itself? As we’ve seen in the synch license arena, where there’s a free market, song copyright owners get paid about the same as recording copyright owners in most instances. Contrast that to the download situation where the publisher gets paid 9.1 cents for the download (the compulsory statutory rate) while the label gets about 70% of the sale price on iTunes (a market negotiation).
The Report also contains lengthy and detailed descriptions of the lack of uniformity in data associated with both musical works and sound recordings. Without going into detail about ISWCs, ISRCs, ISNIs and DDEX standards, suffice to say there is currently no consistent, uniform, international process for assigning codes to musical compositions, albums or individual tracks, writers or artists. And there’s no centralized database for this necessary information. This leads to inefficiencies and delayed licensing and payment for creators.
With the foregoing background, here are the Copyright Office’s four subsidiary principles regarding implementation of their four Guiding Principles:
– Government licensing should aspire to treat like uses of music alike
– Government supervision should enable voluntary transactions while supporting collective solutions
– Rate-setting and enforcement of anti-trust laws should be separately managed and addressed
– A single market-oriented rate-setting standard should apply to all music uses under statutory licenses
So now let’s look at the Report’s most significant recommendations to implement its eight principles:
– Regulate musical works and sound recordings in a more consistent manner. (As we’ve seen, song and master recording rights are often treated differently, with more restrictions on songwriters and publishers than on recording artists and labels.)
– Extend the public performance right for recordings to traditional “terrestrial” radio. (This fosters the first goal and the Report recommends that non-interactive radio be subject to the same compulsory license scheme as are non-interactive streams.)
– In keeping with similar treatment for similar rights, the Report also recommends full federal copyright protection for pre-1972 recordings. (Besides being fair to older artists, this avoids the potential legal chaos discussed above).
– The Copyright Office further suggests that all rate-setting for both recordings and the underlying musical works should (a) be subject to the same “willing-buyer / willing seller” or “fair market value” standard and (b) that all rate setting, even for music performance rights, should be done by the CRB. (This would remove rate-setting for music performance rights from a single, life-tenured federal judge in New York and place it before a tribunal with a specific mandate and expertise. It also fosters the goal of uniform treatment for songs and records.)
– The Report also states that the CRB should only meet as needed and that procedures for setting interim rates, as well as for the overall process, should be streamlined. (This should foster voluntary negotiations and make rate-setting proceedings faster and cheaper).
– The Report also suggests that detailed provisions, such as what constitutes an interactive streaming service, should be put into regulations rather than in the copyright statute, so that they can be more easily modified to adjust to changes in the marketplace.
– The Report stopped short of stating that the ASCAP and BMI Consent Decrees should be repealed. (This position is undoubtedly in deference to the Justice Department’s ongoing review of those decrees, but is clearly supportive of relaxing restrictions, as discussed below.)
– Allow for audit rights under the compulsory mechanical license and allow SoundExchange to terminate licensees who avail themselves of a compulsory license but do not pay. (These are obvious legal loopholes that need to be plugged. If creators are subjected to a compulsory licensing regime, they should at least have the ability to ensure they’re being properly paid and that deadbeats don’t keep the benefits of the license).
The Report also recommended that, as the Copyright Office had previously, licensing collectives be permitted to expand their role and become Music Rights Organizations (MROs) that would license both performing and mechanical rights and possibly other rights as well. ASCAP’s Consent Decree forbids it from licensing mechanicals and other rights and BMI has voluntarily refrained from doing so to date. However, the CEOs of both organizations have indicated that expansion of their licensing capabilities is in their business plans and users should welcome the availability of multi-use licenses.
For example, if ASCAP, BMI, SESAC, Harry Fox and Sound Exchange all became MROs and licensed performing rights and mechanical rights, there would be six MROs competing for business. The Report also recommended congressional overrule of the Rate Court decisions, to the extent of allowing publishers to withdraw digital rights for interactive streaming so that publishers are on parity with the labels in the ability to negotiate for these rights. Although not mentioned in the Report, I think that the MROs should also be able to license the posting of lyrics, as HFA currently offers this service. The PROs and HFA currently allow for a music publisher to issue a direct license and not go through the collective. This should be maintained to both ensure free competition and allow copyright owners to handle individual negotiations where warranted.
If there are six competing MROs offering a variety of bundled licensing services, which would include the right to withdraw certain rights and directly license all rights, it would seem that the ASCAP and BMI Consent Decrees would not be needed (at least not in their present form) as there would be ample competition. As the Report indicated, there are currently only three major labels and three major publishers. They aren’t subject to Consent Decrees. While the US currently has three PROs, most other nations have only one, and that PRO often is able to bundle mechanical rights. The time has come to recognize that the public doesn’t need excessive government protection from the collective licensing by songwriters.
The Report also recommended that membership in MROs be mandatory and that there be a “general” MRO, the GMRO that would act as a stop-gap for certain unrepresented parties and would standardize data formats and create a global rights database for users. I believe neither mandatory membership in a MRO (given that membership in licensing collectives is currently voluntary), nor the creation of a GMRO, another level of governmental involvement, is necessary. First, if a MRO were able to offer more comprehensive services and there was competition for members, there would be enough incentive for all writers, publishers, artists and labels to join one.
Second, as the Report acknowledges, the various interested parties, including the PROs, have been working on various projects to facilitate the uniformity and transparency of data. If, for example, the PROs were to offer mechanical licensing, they would be strongly incentivized to synch their works registrations with recording and artist information. Similarly, if HFA were to offer performing rights, they would be incentivized to ensure that their recording information is coordinated with works information. Third, with MROs having both data for songs and recordings, they could create an aggregate portal for users to look up who controls which rights to songs and recordings. Finally, I also don’t think that a GMRO is necessary to address the problem of unlicensed or unaccounted for shares in works and other missing data. The MROs can license based upon partial representation and hold reserves until such time other interested parties properly register their works and shares.
The Report attempts to address the issue of transparency of licensing and royalty information. Standardizing works and recording codes will help. So will the elimination of the “pass through” mechanical license for downloads in that publishers have to be paid through the labels and not directly by the download services like iTunes. And while the issue was raised regarding equity stakes in and advances from, streaming services like Pandora, no real solutions regarding creators sharing in the wealth were offered. Similarly, the Report alluded to the “whack-a-mole” problem under the DMCA of dealing with rampant infringement on services like YouTube but did not offer any recommendations, an area where the balance between the services and creators, especially individual artists, should be adjusted .
Although the Copyright Office had previously suggested that the compulsory mechanical license be repealed, the Report stops short of advocating it. Instead, it suggests that publishers have limited opt-out rights for interactive streaming and downloads. It further recommends that mechanical licensing should be done on a blanket license basis, like the PROs. The Report’s recommendation that an artist may obtain a compulsory license for a cover recording released as a CD but not as a download makes no sense to me as it is a needless discrimination in format (e.g., LP versus cassettes in the analog world) rather than means of distribution (e.g., purchases versus performances).
I also believe that the song-by-song mechanical license should still be available as an option. For example, an artist making a self-produced recording that include covers should be able to obtain only the licenses needed. And those licenses should be available for both physical copies and downloads. Finally, I think that if the mechanical licensing regime remains compulsory, the CRB should set rates for different tiers of usage. Three should suffice. In the synch market, for example, a Rolling Stones song will command a higher fee than one by an unknown writer. The publisher can select which tier it wants its song priced at and if the user market balks, the publisher can then change to a lower tier.
In sum, the Report offers some solid recommendations as to changes to the legal and regulatory aspects of music licensing. Other suggestions such as creating a new agency, the GMRO, and mandating coding standards are probably unnecessary if private parties are better incentivized through revised laws and regulations. But the Report contains far more detail and nuances, both regarding the current licensing landscape and its recommendations, than can be covered in my brief summary. Songwriters and composers, whose income is currently regulated the most, would likely benefit most from the Report’s recommendations, although recording artists could also receive a significant boost to their income with the adoption of a performing right for radio and TV airplay.
Undoubtedly, major players in the user community, such as streaming services, will object to some of the proposed changes to the music licensing landscape, such as relaxing Consent Decree restrictions and having all compulsory licenses subject to a fair market standard. However, as the Report points out, music creators should not have to subsidize any particular business model. But as the Report also notes, it is ultimately up to Congress, rather than the Copyright Office or the Justice Department to make most of the needed changes. Given Congress’ recent history, it’s hard to be optimistic about legislative fixes happening anytime soon. But one can hope….