CMU Trends Labels & Publishers Legal

Trends: As the GRD runs aground, what next?

By | Published on Wednesday 15 October 2014

Crumpled Copyright

The music publishing sector’s Global Repertoire Database ran aground earlier this summer, with key partner in the project PRS For Music declaring it was “disappointed the GRD cannot move forward as planned”. But why was the GRD project initiated in the first place, what challenges was it attempting to tackle, and what now?

Data is a big word in the music business in 2014, though the industry actually faces two separate data challenges, and it is worth distinguishing between them.

Firstly there is fan data. The single biggest opportunity for the music business in the 21st Century is direct-to-fan, the fact that artists and their business partners can now have a direct relationship with core fanbase through online channels. They can learn about those fans, work out what they will pay for, and create and sell direct products and services based on that information.

But the challenges and opportunities that stem from fan data are for another day. Today our focus is the music industry’s second data challenge, that of copyright ownership data. The data that tells us who owns and controls music copyright.

Unlike trademarks and patents, in most countries copyright is an unregistered intellectual property right. The copyright comes into being as soon as the protected creative work is created, without any registration, providing said ‘work’ fills various criteria set out in copyright law.

Because there is no registration, there are default ownership rules that tell us who owns any one copyright when it is created. Default owners vary from copyright system to copyright system and, crucially, according to the kind of copyright.

So in the UK, the default owner of what the music industry calls the ‘publishing rights’ is the creator, ie the lyricist or composer. Whereas with the sound recording copyright, the default owner is the funder (what the Copyright Act unhelpfully calls the ‘producer’), ie whoever pays for the recording to take place, which, with a signed artist, would likely be a record company.

Anyone who has dabbled in music licensing knows that copyright ownership is a complex business; though, actually, at a basic level it’s pretty simple. In the main, the criteria for what gets copyright protection are pretty clear, and ownership rules are pretty straightforward (once you get over any confusion as to what the law means by ‘producer’).

The problems come about because, beyond prescribing default owners, copyright law is incredibly flexible when it comes to ownership. This is sensible law, but it’s the flexibility that creates the complexity.

First, copyright law allows co-ownership of copyright, so if ten people collaborate on a song, the copyright is co-owned by all of them. The co-owners can agree between themselves on what stake each contributor gets, though if they don’t, and a dispute arises, a judge will decide based on level of contribution.

Second, most copyright systems allow transfer of ownership, of both existing and future copyrights, so that default copyright owners can transfer – or ‘assign’ – their rights to others, usually for money. At it’s core this is what a publishing contract is about, a songwriter transfers his or her copyrights, in existing and/or future songs, in return for upfront money and other services.

Copyright law has very few opinions about assignment, other than that it is allowed to happen. Copyrights can be assigned in part or in full. Assignment can be for the life of the copyright or for a set period agreed by contract. And assignment may relate to one territory or the world.

The creator may retain certain controls over their work even after assignment (statute provides so called ‘moral rights’ for the creator, though most of these controls usually originate in the assignment contract). And the creator would usually be due a cut of any profits generated by an assigned copyright, but again, normally according to terms set out in contract rather than statute.

Because most elements of assignment are governed by contract rather than legislation, terms of ownership can also be altered in the future if creators and the entities they assign rights to reach new agreements down the line. In the record industry, where copyrights generally start with the label (because it is the default owner of the sound recording right), successful artists may negotiate new agreements with the record company, which is eager to keep them on its roster, which gives the artist more control over, then co-ownership, and ultimately complete ownership of the copyright in their recordings.

Though, even when creators – whether they be songwriters, composers or recording artists – retain or reclaim ownership of the copyright in their work, they will likely appoint a corporate entity to manage those rights, via what is usually called an administration deal in music publishing and a distribution or licensing deal in the record industry. Again, copyright law has no real opinion on the nature of these deals, but once done the corporate entity often acts as if it was the copyright owner for the duration of the contract.

All of which means, even if we know the default owner of a copyright work, or about any initial assignment contracts relating to the work, the question remains: what did the most recent agreement between the individual creator and corporate rights owner say? And if the creator is the copyright owner, has he or she appointed a corporate rights owner to administer his or her rights, on a regional or worldwide basis?

And there are some further complications too…

There are three music copyrights
When a licensee wants permission to use a record, they are usually looking to license the use of three separate copyrights: there is a copyright in the lyrics, another in the music (which together are what we earlier referred to as the ‘publishing rights’), and a third in the sound recording itself.

All three of these copyrights may be co-owned by multiple parties; and all three copyrights may have been assigned to other entities, maybe multiple other entities, and original assignment agreements may have been superseded by new contracts.

A copyright consists of a number of actual rights (or controls)
Not only is the music rights industry actually dealing in three distinct copyrights, ownership and control may also be different for each element of each copyright. The law sets out what rights or controls a copyright owner enjoys over the creative works they own. In UK law these rights are referred to as ‘acts restricted by the copyright’.

The list of restricted acts varies from copyright system to copyright system (and sometimes according to the kind of copyright), though commonly includes: the exclusive right to make actual copies of a work, the exclusive right to perform the work in public, the exclusive right to communicate the work, and to exclusive right to make adaptations. Quite how well each restricted act is defined varies from system to system.

The basic principle of copyright law is that only the copyright owner can do any of these things by default, and if anyone else 
wants to copy, perform, communicate or adapt a protected work, they must get permission from the copyright owner first. The copyright owner charges for permission, which is how copyrights make money.

It’s not uncommon in the music industry for different people, or divisions, or entities, to control each of the elements of the copyright. So there can be different people and processes involved depending on whether you want to copy, perform, communicate or adapt a work. This is most obvious in the way music publishers often appoint different collecting societies to represent their ‘mechanical’ (copy) rights and their performing rights.

This means that not only do licensees need to work out who controls or owns the lyrical, musical and sound recording rights in a track they want to use, but they may need to work out if they are planning on copying, performing, communicating or adapting a track, and based on that deal with different or multiple licensing agents.

This has caused particular problems in the streaming sector – because if you stream recorded music, does that count as making a copy, instigating a public performance or communicating the work? Or is it a combination of all three? Copyright law has no specific opinion on this (yet), so instead you have various stakeholders each arguing for their own share. Even though you might be talking about one copyright with one ultimate beneficiary.

The statutory economic rights of recording artists
Although the complexity of copyright ownership mainly stems from copyright law’s flexibility on the issue, one final complication comes from a bit of copyright statute that is pretty inflexible: the statutory economic right of recording artists.

The copyrights in sound recordings are commonly owned, outright for life of copyright, by the record companies who organise the recordings to take place, usually because they are the default owner under law, and if not through an assignment contract.

Featured artists, so the people directly signed to the label that owns the copyright, will usually be due a cut of any profit (ie income after costs) generated by the sound recordings released in their name. However session musicians and record producers will usually have no right under contract to a share of profits generated by the sound recording, instead receiving a one-off payment for their time and contribution.

Many copyright systems now provide a statutory economic right for recording artists (featured artists, session musicians and, depending on their contribution, record producers) which gives them an automatic right to a cut of the money generated by certain specific uses (usually public performance) of the recordings on which they appear. This economic right does not affect actual copyright ownership, exists oblivious of any contractual arrangements between a label and artists involved in a recording, and cannot be waived by said contracts. The management and payment of these royalties (often referred to as ‘neighbouring rights income’) is usually handled by the record industry’s collecting societies.

So, to conclude, if a licensee wants to make use of a track, they need to think about the separate copyrights in the lyrics, music and recording. All of which may be co-owned, and controlled by different entities in different territories, some for life of copyright, some for a set period of time, so owners may be different today than last year. And the controller may differ depending on whether you want to copy, perform, communicate or adapt the track, even though you may want to do all those things, or you may not actually know which of those applies to what you plan to do. And on top of that, someone needs to work out whether recording artists involved in the track are due any statutory royalties, and if so how they are being paid.

To repeat, it’s not copyright law that is complex, it’s the flexibility of copyright law, and the way the music industry has capitalised on that flexibility, that has resulted in a very complex music licensing environment. Though no one is really suggesting that the flexibility itself is the problem, indeed it’s one of the good things about copyright.

True, the continued separation of the music publishing sector and the record industry is arguably counter-productive, given most music rights companies operate (usually separately) 
in both, and the fact that licensing (as opposed to selling CDs) is an increasingly important part of what both publishers and labels do. The split is particularly inefficient in the collective licensing domain (as is the additional split in music publisher collective licensing of mechanical and performing rights).

However, the single biggest issue when it comes to copyright ownership complexities is one of data; or, rather, the single simplest solution to overcome complex ownership and control arrangements is a big fat database.

Because in most countries copyright is never registered with a central authority, there is no central repository of ownership information. No one-stop portal that will tell you who controls the various copyrights in a recording of a song with lyrics, what stake each co-owner owns, who has the right to license in what country and in what scenario, and which recording artists are beneficiaries of the statutory economic right.

Of course, the collecting societies are sitting on some of this data. In certain scenarios – most commonly public performance – the vast majority of music publishers and record labels (such a vast majority to the point where licensees assume it is all music publishers and record labels) choose to allow one central organisation to licence on their behalf, ie a collecting society, often on an exclusive basis per territory.

The society then usually offers licensees so called blanket licences, giving them access to their collective members’ entire catalogue of songs or recordings on a per-play royalty or profit share basis. The society does the deal (subject to collective licensing regulations) and then collects any monies due, distributing them back to rights owners based on usage.

In order to distribute the monies, the societies need to know who owns what copyrights on what terms in what countries. Recording industry societies also need to know what musicians are due a share of public performance income under statute. And so each collecting society has, over the years, built up databases of copyright ownership information, covering the works owned by their member publishers or record companies.

On the recordings side, in addition to the databases controlled by the collecting societies, you have the International Standard Recording Code, or ISRC, programme, an initiative led by the International Federation Of The Phonographic Industry, which in theory makes it easier to identify the title and ownership of any one sound recording. In the online space the IFPI has also been involved in the GRid and DDEX projects, aiming to foster identity, ownership and metadata standards across digital music.

So, the music rights industry is not without any copyright ownership data, there are various repositories of information around the world. Though none of these are complete. And they vary greatly in quality of data. And the information is not always made available to the wider world, at all, or in a user-friendly fashion.

In response to discussions kick started by the European Commission, the music publishing sector began work in 2009 on what became known as the Global Repertoire Database, a project that would basically join together the existing databases of all the music publishing collecting societies in the world, instigate some data standards, and provide a one-stop repository of song ownership information.

Of course, that would still only cover the lyrical and musical copyrights, because the GRD never intended to document sound recording copyright ownership. Though if the GRD was successfully built, and if the ISRC code of any one recording could be linked to the GRD code of the song that featured in it, and if musician data could then be linked in somehow, that could be the uber music rights database that is really needed. And if the record industry got into the habit of always embedding ISRC codes into the metadata of digital music files, that there is a workable system.

Bringing GRD and ISRC data together would have required serious collaboration between the music publishing and recorded music sectors, which is always tricky, even though the three biggest players in music publishing are also the three biggest players in recorded music (Universal, Sony and Warner).

But even trying to cross that hurdle in the 
near future isn’t looking likely anymore, because earlier this summer it was confirmed that the GRD project was in the process of collapsing. Despite being initially backed by all the big publishers, publishing collecting 
societies and various digital music companies, word has it that disagreements between key stakeholders over funding of the initiative forced the whole venture to stall.

This was confirmed by PRS, which told reporters: “We are disappointed the GRD cannot move forward as planned, though we remain committed to the principles of a single point of works registration and reconciliation of ownership shares under broad stakeholder governance. These principles remain key to the efficient processing of multi-territory licences and accurate distribution of royalties across all usages of creative works. We are actively studying alternative ways of achieving these goals, taking full advantage of the learning gained from the GRD project to date”.

The collapse of the GRD project is a massive setback in the mission to create an uber music rights database. Though why does that even matter? So copyright ownership is complex, who cares?

Well, some people are happy with the complexities. Not least, the rapidly growing rights administration sector, which works for rights owners, artists and licensees.

Plenty of entrepreneurs have recognised that there is money to be made by helping other people navigate the complexities of copyright ownership.

There are now agencies working for rights owners who are keen to ensure they are getting every penny they are due, artists who suspect they are not receiving all their neighbouring rights income, and licensees who can’t face the thought of identifying and dealing with multiple stakeholders to use a single recording of a single song.

These companies would still have a role to play even without the complexities – as efficient administrators and effective deal makers – though with copyright ownership information sometimes so hard to find, those with the knowledge and patience to navigate the complex licensing ecosystem can charge a fee for that service.

Cynics might also argue that collecting societies have something to gain from the complexities, which possibly meant it was unwise to have them as the principle funders of the GRD.

In most cases where collective licensing applies, rights owners choose (ie they are not obliged) to license through a collecting society, rather than dealing directly with licensees, usually for practical or efficiency reasons. Collective licensing makes most sense where you have a group of licensees who make frequent use of many different songs and recordings (high-users) but who can only afford to pay relatively low per-play royalties (low-payers).

Radio is a classic area where the music industry licenses collectively. Rather than every radio station having to do a deal with every rights owner for every song and every recording – creating legal and administrative costs that would eat into the relatively modest revenues that the radio sector is able to share with the music rights community – the labels and publishers license collectively through their respective societies.

Of course, if you apply the ‘high user + low payer = license collectively’ rule, logic dictates that the music industry should license streaming services through the collective licensing system too. The music publishers initially did, and in many cases still do. However, the record industry has mainly licensed digital directly (the exception being Pandora-type services, where US law obliges the labels to license via SoundExchange). There are two main reasons for the direct dealing.

First, collective licensing is usually subject to extra regulation (designed to overcome the competition law concerns of collective dealing), which gives the rights owner less power at the negotiating table. And second, collecting societies traditionally only ever licensed businesses in their home territories, but streaming services usually want multi-territory and possibly global licenses. Although collecting societies in Europe have experimented with multi-territory licensing, such licenses are arguably easier to negotiate directly.

While we once thought the labels might eventually move over to a collective licensing model for digital (it would certainly solve some of the digital gripes in the artist community), what we are seeing is the opposite: that the 
big music publishers are slowly moving towards doing direct deals with the streaming services (though this shift has been delayed in the US because of a dispute over collective licensing rules).

If eventually all streaming music services are fully licensed through direct deals, what about radio? There has been so much consolidation in both the radio and music industries in the last 25 years, that the number of standalone direct deals that would be required for stations to enjoy the level of access to music they are used to are not now so numerous; especially if the aggregators who represent the indies in the digital space turned their attention to other media.

However, even if we see a rise of direct dealing in the digital space, and subsequently in those areas where collective licensing has traditionally been the norm, there will remain the challenge of collecting and distributing royalties. And the collecting societies have an important data role to play here, even if they are not actually negotiating licensing deals, particularly in music publishing where collaboration and, therefore, co-ownership of copyright is much more common.

That is to say, even if the music publishers licensed Spotify et al directly, it would make sense for the collecting societies to continue to collect and distribute royalties, because they are sitting on the data that enables monies to be split up between different co-owners of the same songs.

Unless, of course, there was a working, efficient Global Repertoire Database, via which streaming services could automatically calculate who was due what, report to rights owners and make payments. A working GRD, therefore, might result in the collecting societies having less power and reach.

As (mainly) not-for-profit membership organisations that exist for the benefit of their songwriter and publisher members, any loss in power and influence for the collecting societies shouldn’t matter. But human nature means that the people running these societies are unlikely to give up that power and influence without a fight.

But let’s be clear, many more people lose out in a world with no central registry of copyright ownership data, which is why few people in the music industry disagree that this is an issue that needs to be addressed. Even if the same people are not sure how to build a central music rights database, let alone pay for it.

Licensees lose out simply because the law requires them to navigate all of this at their own expense, running the risk of liability for copyright infringement if they fail to identify and contract with the right copyright owners. Collective licensing, where it applies, simplifies this to a point, though most licensees feel that collective licensing weakens their hand at the negotiating table, even though collective licensing regulations are meant to ensure that doesn’t happen.

Rights owners simply lose income. Particularly from low-user low-payer licensees. High-user, high-payer licensees will usually navigate the complex music copyright ecosystem, mainly because they really need the music and will be sued if they don’t do the required deals. But low-user low-payer licensees often don’t have the knowledge or resource, so might just go with no music at all, or use music without licence and not get caught. Although these licensees are low-users and low-payers, there are an awful lot of them. Music licensing complexities are stopping labels and publishers from capitalising on this market in the way the online photo agencies are.

Songwriters, composers and featured artists are also likely losing income, first because, under any assignment or licensing contact they have with a publisher or label, they will be due a cut of the monies the corporate rights owners generate. So if publishers and labels lose income, so do they. Recording artists, meanwhile, may also be losing neighbouring right royalties they are due because their involvement in any one recording is not documented with the relevant collecting society, or because usage where the statutory royalty is due is not being properly reported.

All rights owners, creators and artists, meanwhile, will be losing additional income simply because inefficiencies in the system, caused by the lack of a decent centralised ownership database, mean extra administrators or middle men are often required. This becomes even more of an issue as a greater amount of music right revenue is generated 
by services with incredibly high usage but very low per-play payments, ie the streaming platforms. These services generate a flood of usage data, which is much harder and more costly to monitor and process when the music industry lacks any standard system for documented copyright ownership and beneficiaries.

So, there are commercial incentives for the music rights sector to sort out its copyright ownership data problem. Though, it has to be said, the real impetus to find a solution will likely come from lawmakers. Indeed the Global Repertoire Database project began under pressure from the European Commission.

Licensees loudly bemoan the complexities of music licensing (even though the complexities mean that, once any one licensee has navigated the mess, they have a competitive advantage over newer entrants to the market).

Meanwhile the providers of user-upload and search engine platforms, who now routinely have to deal with a flood of takedown requests from copyright owners when they inadvertently host or link to infringing material, frequently complain that they receive erroneous takedown notices, something they can’t monitor because of the lack of a central copyright ownership database to check with.

And politicians around the world hear these complaints. So much so that whenever the music industry seeks new rules and regulations to crack down on piracy, lawmakers demand a simpler music licensing framework as a quid pro quo. The music rights sector loudly agrees that music licensing should be simpler. And to date legislators have been mainly placated by such loud proclamations. But they will eventually start demanding actual progress in this domain.

But what if the music rights sector fails to deliver? Well, it rarely comes up in political circles, and is almost a taboo topic in the music rights sector itself, but a small contingent in the legal sector have pondered whether compulsory copyright registration is, actually, the only real solution here. Even though that throws up all sorts of issues about international data sharing and the high administration costs of registering existing copyrights.

Rights owners would almost certainly fight any such change to copyright law, though such a fight would be easier to pursue if initiatives like the GRD didn’t fall over because of bickering between rights owners.

There may, however, be a simpler solution. Leave it to the start-ups. If rights owners and collecting societies across the music industry were to start routinely publishing catalogue information in a user-friendly and consistent fashion, then there are start-ups out there already working on technical solutions to this problem.

Given access to the right data, said start-ups may well develop the kind of database – combining GRD, ISRC and musician list data – that the music rights industry needs, and without the politics that will hinder any industry-led initiative involving numerous stakeholders with competing agendas.

Of course, the music rights sector may bother itself about allowing another batch of start-up entrepreneurs to become billionaires on the back of their assets, this time copyright ownership data. Though here there is big kickback. Why not let venture capital income build the database this industry needs, let the start-ups take all the risk, and then let the market select the best system the start-up community comes up with?

As it happens, the world’s biggest music publishers, Sony/ATV and Universal – in the wake of the collapse of GRD, and possibly aware that doing so will help their case for collective licensing rules to be changed in the US – have already made commitments to start publishing ownership data relating to their catalogues in a user-friendly fashion.

Quite how useful this new flow of data actually is and will be remains to be seen, though it could be the start of this practical, low-cost and risk-free solution to the music industry’s copyright ownership data problem.