Copyright law is living in the past
Let musicians embrace quotation and reimagination
Today we have a guest post from Chris Dalla Riva, who writes the Substack newsletter Can’t Get Much Higher about the intersection of music and data. The regular Friday mailbag will instead run on Saturday this week.
According to Billboard, Luke Combs’ “Fast Car” was the eighth biggest song of 2023. Needless to say, the success of the song has likely made a lot of people happy. It’s likely made Luke Combs happy. It’s likely made his label happy. It’s also likely made the millions of listeners happy. But the song’s success has also made another person very happy: Tracy Chapman.
“Fast Car” was written and originally performed by Tracy Chapman. Her version was also popular. It spent 21 weeks on the Billboard Hot 100 in 1988, peaking at number six in late-August. In a July 2023 statement to Billboard, the reclusive Chapman expressed her gratitude for the song’s renewed success:
I never expected to find myself on the country charts, but I’m honored to be there. I’m happy for Luke and his success and grateful that new fans have found and embraced ‘Fast Car.’
I don’t doubt that Chapman is happy about “Fast Car” reaching a new generation, but I’m sure that she’s also happy because it’s been a financial windfall for her. Billboard estimated that between the song’s release on March 17 and June 8, it generated $500,000 in global publishing royalties. Given that the popularity of the song has barely diminished since then, Chapman has likely walked away with a ton of money.
So, how did this cover come about? Combs’ manager Chris Kappy noted that Combs’ and his team got express permission from Chapman and hers: “This is her song and we were going to live within any parameters she had for her song.” What’s interesting is that had Chapman not wanted Combs to record the song, there would be no way for her to stop him.
Unlike other forms of intellectual property, musical compositions are subject to a compulsory license. This means that as long as the song you are covering has been previously released, your recording is relatively faithful to the original,1 you serve a notice of intention to the copyright owner, and pay the government-set royalty rate to the copyright owner, nobody can stop you from releasing your song.2
Let’s change this situation slightly.
Imagine you are a young rapper. You decide you want to sample the hi-hat heard in the first second of Tracy Chapman’s recording. You are going to take that hi-hat and build a new song around it. There’s an issue, though. Unlike musical compositions, sound recordings are not subject to a compulsory license. You need to get permission from both the owner of the sound recording copyright, namely Chapman’s label Elektra Records, along with permission from the owner of the underlying composition’s copyright, namely Chapman herself.
This might be problematic. First, the financial terms set forth by Chapman or her label might prove unaffordable to a young artist like you. Second, either Chapman or her label can deny you permission to use “Fast Car”. In other words, it might be easier for you to record an entire cover of “Fast Car” than it would be to use one second of Chapman’s recording of it.
In short, music licensing is a strange, often counterintuitive world. I don’t claim to have solutions to all of the issues that have cropped up in that world over the years, but I want to lay out a case for why sound recordings should be treated similarly to musical compositions and subject to compulsory licenses in certain cases. To understand my case, we need to explore some history.
Why are Compositions Subject to Compulsory Licenses?
The Founding Fathers were concerned enough with intellectual property that they expressly granted Congress the power to enact laws surrounding such in Article I of the Constitution: “The Congress shall have Power To … promote the Progress of Science and useful Arts, by securing for limited Times to Authors and Inventors the exclusive Right to their respective Writings and Discoveries.” Once the Constitution was ratified, Congress took little time to wield that power.
In 1790, the first copyright law was enacted, which established a limited copyright term of 14 years for books, maps, and charts with the ability to renew for an additional 14 years. Protections for musical compositions were added in 1831. Though the law was amended to cover public performance in 1897, radical changes did not occur until 1909.
In the early 1900s, controversy arose with the proliferation of player pianos. Player pianos were able to play songs without human aid by way of a perforated roll. The perforations in those rolls were instructions for which key to be pressed at which time. As the roll rotated mechanically, the piano was able to play the song. Musicians were concerned with this for two reasons:
Player pianos would stop people from needing to hire live musicians
Player pianos would stop people from buying sheet music, which generated royalties
These issues were compounded by the fact that player piano companies didn’t pay royalties to the songwriters for using their songs. Their argument was that piano rolls were not copies of sheet music. They were just part of the machine that reproduced music. The Supreme Court agreed. In the 1908 case White-Smith Music Publishing Co. v. Apollo Co., the court ruled that player piano companies did not owe royalties to songwriters.
The validity of this ruling did not last long. With lobbying from artists, Congress went to work on the 1909 revision of the Copyright Act, which would overturn White-Smith Music Publishing Co. v. Apollo Co. and extend copyright coverage to mechanical reproductions of songs. Furthermore, this act established the compulsory license for those reproductions. There are two competing histories for why Congress chose to establish the compulsory license.
The first is focused on regulating monopoly power. At the time, a company named Aeolian was a large manufacturer of player pianos and other musical paraphernalia. They had begun purchasing the rights to popular songs in an effort to corner the player piano market. In an attempt to reign in Aeolian’s power, Congress established the compulsory mechanical license so that Aeolian’s purchases would be for naught.
Other historians argue that while this is a nice narrative, it overstates Congress’s concerns about Aeolian. In a 2001 paper, Jane C. Ginsburg argued that the compulsory license was about readjusting the balance between copyright owners and technology firms “by imposing a … scheme that permitted continued distribution of the new technology, while assuring payment to copyright owners.”
Reasoning aside, the establishment of the compulsory license for mechanical reproductions of compositions laid the foundation for the music industry in the first half of the 20th century. As the industry grew, the cover song grew with it. Artists would frequently release albums of cover songs dedicated to a single artist or songwriter. Artists would also release competing versions of the same songs, like how in 1946 three different versions of the song “To Each His Own” topped Billboard’s National Best Selling Retail Records Chart. In the same way that music was passed orally through the folk tradition for thousands of years, the compulsory license made it possible for that tradition to continue in the age of recorded music.
Why aren’t Recordings Subject to Compulsory Licenses?
Just as compulsory licenses for compositions were created to solve a practical legal issue, the reason that compulsory licenses don’t exist for the usage of sound recordings — the reason it is almost certainly more time-consuming and expensive to sample one second of “Fast Car” than it is to cover it — comes down to Congress trying to solve a different issue.
Though compositions could be copyrighted throughout the entirety of the 20th century, recordings could not. Congress extended copyright protection to sound recordings in the 1970s. Unlike compositions, Congress did not establish a compulsory license for sound recordings because the music industry successfully argued that the most common reason to copy a sound recording was to pirate it.
The problem with that argument was that at the same time Congress was passing the law, nascent hip-hop artists and disco DJs were proving that there were ways you could use sound recordings to make novel creations.
The rappers and DJs of the early 1970s were revolutionary figures. These people looked at a device meant to playback recorded music and decided that it could be an instrument. They weren’t the first people to explore this idea, but they pushed into uncharted territory.
Disco DJs would remix popular songs to keep an endless stream of music going for hours. Hip-hop artists would chop up previously-recorded rhythmic breaks and rap over them. The newly updated copyright law did not reflect these new uses of sound recordings. It was going to be up to the courts over the next few decades to decide what was legitimate.
Across many cases — though the two most important being Grand Upright Music, Ltd. v. Warner Bros. Records Inc. and Bridgeport Music, Inc. v. Dimension Films —courts ruled that the market should determine what the going rate to use a sound recording should be. In other words, if you have not expressly agreed with the copyright holders on financial terms to use their recordings, you’re out of luck, or in Bridgeport’s more succinct parlance, “Get a license or do not sample.”
Why should Sound Recordings be Subject to Compulsory Licenses?
One of the key ideas when thinking about intellectual property is the incentives/access tradeoff. This idea highlights two competing forces. First, by granting people a limited monopoly over their creative ideas, we incentivize them to create. That’s important because creative ideas can benefit large parts of society.
At the same time, that limited monopoly cannot be so powerful that those that aren’t the creator can never access the creation once it has been released. Jacob Victor outlines this dichotomy in his 2020 paper “Reconceptualizing Compulsory Copyright Licenses”: “[C]opyright entitlements are expected … to optimally embody a tradeoff between incentives and access, allowing authors to profit off their works through the market while still ensuring that the public and other creators are not unduly restricted from accessing them.”
Our current system, which requires negotiating a license for the usage of any sound recording (e.g., a rapper sampling one second of Tracey Chapman’s recording of “Fast Car”), both disincentivizes creation and limits access. Let me explain how.
Bad Metadata: To properly license a sound recording, you need to get approval from both the owner of the sound recording and composition. This requires being able to identify who those entities are. This can be difficult or impossible. In 2019, The Verge noted, “Missing, bad, or inconsistent song metadata is a crisis that has left, by some estimations, billions on the table that never gets paid to the artists who earned that money.” This not only makes getting artists paid hard. It makes licensing their recordings hard.
Finding the Copyright Holder: Even if the sound recording you want to use has proper metadata, it can be hard to locate and contact the relevant parties. Plus, should you find them, the copyright holder has no obligation to respond.
Negotiation with the Copyright Holder: Assuming you are able to get in touch with the relevant parties, you still need to negotiate a license. There is often a huge imbalance in power during these negotiations because the sound recording and composition owners are often notable artists, record labels, songwriters, or producers. Prospective licensors are not necessarily of the same stature.
When you consider the combined costs involved with properly licensing a sound recording, along with the fact that most artists have little to no legal training and that a copyright holder can outright deny a license, it becomes clear that our current laws can inhibit creativity.
Still, you might argue that our current system is good for a few reasons. First, artists get to control their work. As someone who makes music, this argument is powerful. But when you consider that people can cover your work or make parodies of it without permission, that argument becomes less cogent.
In addition, copyright terms in the United States currently last a very long time, namely the life of the author plus 70 years. Because copyright outlives the original artist, artists sometimes aren’t even negotiating for a license with that person. They are negotiating with their heirs or a corporation. I find the “artists deserve to control their work” argument less powerful because of that.
Secondly, why change the system if copyright holders are making money? When you consider the search and transaction costs of licensing a sound recording, copyright holders likely aren’t making as much money as they could be. Furthermore, we can’t just consider the benefits that accrue to the copyright holder. We also need to consider both the licensor and the public. Maxwell Christiansen spells this out well in a 2017 paper:
The implication of the current licensing process … either discourages sample artists from sampling or from participating in the licensing process. Both avenues harm all parties involved … The right holders are not paid the royalties they deserve if the sample artist samples their song illegally, the sample artist either decides not to express his idea or takes a risk of being sued, and the public is deprived of the music that the sample artist would have created …
You can see what Christiansen is getting at when you consider a world where compositions were not subject to a compulsory license. Prince was notoriously protective of his work. But because his compositions were subject to a compulsory license, Sinéad O'Connor didn’t need his permission to cover his composition “Nothing Compares 2 U” in 1990.
The song ended up being a massive hit. It likely led to huge financial gains for both Prince and O’Connor. It also led to massive benefit to the public, who loved the song. If Prince had to approve covers of his work, it might not have gotten made, and all parties would be worse off.
How Could Compulsory Licenses for Sound Recordings Work?
I'm a musician and a writer, not a lawyer. There are many viable ways to implement compulsory licenses for sound recordings, but any new law should meet a few basic criteria if we want it to truly support artists and their work.
First, though unstated, I have largely been focused on sampling. For the sake of clarity, let’s define a sample as a piece of a sound recording less than 30 seconds that is repurposed in a new recording. A specific example of this would be M.I.A. and Diplo repurposing the beginning of The Clash’s “Straight to Hell” on the 2000s classic “Paper Planes”.
There are other ways to use sound recordings in new creations — like mashups and remixes — but those should probably be thought about separately. Furthermore, this compulsory sample license can only be used for other sound recordings and performances. It cannot be used for film, television, or commercials. That would still be covered separately under a synchronization license.
Given these facts, this compulsory sample licensing scheme should try not to be any more administratively complex than the scheme for compositions. The scheme must not be overly burdensome on artists who want to use multiple samples in their works or sample pieces that already contain samples. The system should clearly define when a sample must be licensed. Here are some ideas that I think could help achieve these tenants.
Administration: The Copyright Royalty Board should determine the fixed compulsory sample licensing rates as they currently do for compositions. As with compulsory licenses for compositions, a notice of intention must be filed with the relevant copyright holders or, if they cannot be located, the Copyright Office.
Rates: The rate should be based on how many seconds of the original recording that you use. Thus, you will pay more for using more. Measuring the length of the sample should be based on the original recording. If you take 5 seconds of a sound recording and manipulate it so it’s 15 seconds, it would only count as 5 seconds in the calculation. Samples below a certain length should not require a license. Samples longer than a certain threshold (e.g., 30 seconds), should not be covered by the compulsory sample license.
Sample Stacking: If a recording uses multiple samples, the lengths of samples will be added together, but the royalty will be split based on the percentage of overall time of each. For example, if sample A is 3 seconds and sample B is 7 seconds, the rate would be based on 10 seconds with 30% going to sample A and 70% going to sample B. If a recording uses a sample that contains other samples, the royalty will be split equally among all parties. The idea here is that if you are using multiple samples or sampling things that already contain samples, your recording is likely less reliant on any single recording.
Infringement: If an artist discovers an unlicensed use of one of their recordings, they can sue for infringement.
The history of popular music is a history of quotation and reimagination. It is “The House of the Rising Sun” evolving over decades before The Animals had a massive hit with it in 1964. It is The Beatles getting their chops by covering scores of compositions before composing their own. We need our laws to reflect that the ways we quote and transform music have changed.
The exact wording of this in the Copyright Act is a bit more specific: “A compulsory license includes the privilege of making a musical arrangement of the work to the extent necessary to conform it to the style or manner of interpretation of the performance involved, but the arrangement shall not change the basic melody or fundamental character of the work, and shall not be subject to protection as a derivative work under this title, except with the express consent of the copyright owner.”
I am ignoring some complexities here. Licenses for covers only have to be acquired if you are selling physical copies (e.g., CD, vinyl, cassette). With the passage of the 2018 Music Modernization Act, you no longer have to file a notice of intention for a digital only cover (i.e., only available on streaming). That piece of legislation laid out changes that require digital streaming services (e.g., Spotify, Apple) to get blanket licenses to cover compulsory licensing costs.