I took in a pretty full day of panels at Day 1 of the Rethink Music conference (my preview post from last night is here). While tomorrow’s lineup promises lots of great stuff on copyright law and policy, today’s schedule was pretty business-oriented. As such, I’ll hold off my explicitly legal analysis until tomorrow night. For now, I want to answer a question that was asked at pretty much every panel: what’s the value of a song in today’s online environment?
The short answer: zero. A single song is worthless — but that’s not a bad thing. Let me explain.
There was much consternation among the audience and many of the panelists — specifically, the ones who work for the large institutional players in the music industry — that there is an entire generation that doesn’t understand the “value” of pieces of recorded music. They lamented that music has been chopped up into $.99 individual songs, $9.99 digital albums, or, even worse, subscription services like MOG that offer on-demand access to large libraries of music — though with many unfortunate limitations, as explained below — for $10 a month. Referring to these service, one audience member asked in dismay: “does that mean that a song is now worth 1/100th of a cent?”
Well, no: actually, it’s worth less than that. I’m happy to report that in 2011, individual songs have become worthless to me and many in my generation. That’s because pretty much every song is available anywhere, any time, for free, either streaming on YouTube or Grooveshark, or for unauthorized download on a filesharing site — indeed, most songs are available in all three places.
But it doesn’t follow from this that no one will pay for recorded music online; far from it. I can’t know for sure, but I believe that the true value of recorded music lies in abundance, and I believe that people will pay for abundance. Indeed, abundance is why I pay for unlimited high-speed Internet access. I don’t place any value on individual tweets, or individual photos I upload to Facebook, or searches for hotels for a possible trip to Toronto. Instead, because those things are free, I swim in a sea of tweets and photos and travel searches, and I leave it to those sites to monetize my abundant usage. To use another example that is admittedly somewhat different from the case of music, I don’t really know what value I place on a rerun of Seinfeld on TBS, or a new Colbert Report on Comedy Central, or an episode of Mad Men on AMC. I suppose in theory there is a value, but no one thinks of things that way. What we really value about cable is abundant access to oceans of content — most of it bad, some of it great — and that’s why we pay for cable and DVRs that let us watch as much of it as we want, whenever we’re in the mood.
Back to music in 2011. I place a very high value on abundant access to downloadable, freely transferable music, that I can play however I want, whenever I want, in whatever software I want, on whatever device I want. That is because I value abundance, and, unlike video, I value the freedom to consume music in ways that I want, not in many of the overly restrictive ways that record companies want. I would pay a lot of money for that, but no one yet offers it. Services like MOG, which require users to stream from proprietary software either online or on portable devices, and don’t allow direct downloads of any MP3s, are so close yet so far. MOG is interesting, and the founder clearly knows what he’s doing and has good ideas, but this service does not yet offer what people like me value (I don’t just mean this figuratively; I did the free trial, and decided not to become a subscriber). I value being able to play my music library in iTunes, or to set a song to a slideshow of my pictures, or to play it in my iPhone’s terrific native music player, and, though I’m not a professional musician, I also value being able to occasionally edit the songs for personal, fair uses. MOG lets me do none of these. It’s objectively true that, nearly 12 years after Napster’s introduction, illegal filesharing sites still offer the best online product there is: they combine the unrestricted uses of mp3s sold on iTunes or Amazon with the attitude of abundant access that MOG has. Make that combination legal, show me where to send my monthly fee, and I’ll be the first to sign-up — and I don’t think I’d be the only one.
There was some discussion at one of the afternoon panels about why the current generation of all-you-can-eat, on-demand services like MOG have not really taken off yet. It seems to me that the reason this has not happened is not a failure of marketing nor of “consumer education,” one of the day’s favorite phrases. It is because they do not yet fit perfectly with what consumers value. When that finally happens, I have a feeling the services that provide what I describe above will be quite popular with not just consumers but also with artists — because artists will finally reap the benefits of the public happily paying to be a part of the remarkable abundance of recorded music that so many have worked so hard and long to produce.
I find it so interesting that the migration from intellectual property consumption via discrete purchases toward service subscriptions started with and remains most advanced for bandwidth-hogging movies (Netflix), while music and books lag behind (even if you get those and store them in the cloud, nearly everyone buys individual books and songs rather than subscribing to an “unlimited” service). Clearly industry idiosyncrasies are more of a bar on progress than technology…
What portion of the industry and legal idiosyncrasies that result in motion pictures being more adaptable to evolving technologies and consumption habits is traceable to its being a much a much younger medium? Music is as old as human memory, and we’ve been providing it with some form of copyright protection for centuries. Both the music economy and the legal framework that governs it existed in some form during a time in which ideas like “the cloud” were utterly incomprehensible. Movies, on the other hand, are relatively new, and the ability to easily transmit/copy movies is even newer. Perhaps the motion picture economy wasn’t as set in its ways when the digital revolution came along, and the law governing motion pictures hadn’t had as much time to get so tangled as to preclude easy adaptation to new technologies (as Jason below suggests has been the case with music).
Sound recordings were not protected under federal copyright law until 1972, and that was only added after the technology to reproduce sound recordings was widespread.
The music industry peaked relatively recently – in the early 2000s – when boy band and other pop stars were regularly moving millions of units annually. It’s been a rapid change from opulence to penniless and the industry did not – and does not – know what to do (Spotify is a great and current example).
The reason the “unlimited subscription” model hit the film industry first is because of the first-sale doctrine and the large file size of digital movie files. Music files could be downloaded over dial-up connections in the late-90s but it was unfeasible to download a 500Mb DivX rip of a DVD.
It’s also a question of consumer habit.
The movie industry has had the concept of subscription access (i.e. blockbuster) in place for decades. Even in the days of singles clubs and things like that, the idea of having access to literally millions of songs is still somewhat new to most consumers.
Dustin, you’re right, and it should be noted that there’s lots of law here too shaping the “industry idiosyncrasies.” The best example of this is that there are at least two copyrights in any given song — one for the musical composition, and one for the sound recording itself — so it takes double the licenses to try to launch a Netflix for music. Worse, the legal rights for each kind of copyright turn out to be substantially different — for instance, the holder of the sound recording copyright is not entitled to a royalty from radio play, but the composition holder is — so, naturally, different actors tend to hold each kind of copyright, and each right has quite different economic value. This makes transaction costs exceedingly high in this area, which clearly shapes industry practice. It’s a shame.
Hopefully, there’ll be more on this legal/policy framework tomorrow.
Wholeheartedly agree – can’t wait for the future you describe. One point though. A *digital copy* of a song has no value. I’ve never once paid for a digital copy of a song. However, for me, I will happily pay for albums and singles from my favourite artists in CD or vinyl form.
These do have real value because to me because you can do what you like with them and they will survive any hard drive crash, server failure – and gets around the fact that there is no guarantee whatsoever that the music you’re subscribing to Spotify/MOG/etc to listen to isn’t going to be randomly pulled from availability at any point in the future.
Some artists are also starting to see sense and include a free MP3 or FLAC download link with the physical copy – something to stick straight on your MP3 player, the best of both worlds.
For what it’s worth, I think this post is not just wrong but obviously so. I understand you’re trying to be provocative. You say a single song is “worthless” because you and many in your generation can get it for free on youtube, Grooveshark, or through unauthorized download. A couple obvious rejoinders among the many available:
1) The value of a good is not equivalent to what the least-interested buyer would pay. Suppose you’re not interested in buying my car at all–any car for you would be more trouble than it’s worth, and you wouldn’t even give me a penny. The fact that you would pay nothing for it doesn’t mean it’s worthless.
2) Even accepting the premise that the value of a single song depends alone on what you and the relevant subset of your generation that are like you would pay, the value still isn’t zero. You “pay” even for your free music, even at the margin (to say nothing of internet access costs, etc.), in at least two pays. The first is search costs: if you want to listen to just one song, you are going to have to find it. Many will be available in all three places you mention, but some won’t, and all will take at least some effort–if only the effort of typing the title into google. That’s almost nothing, but that search time is lost time, time you could have been searching for something else. The second is liability risk: by downloading an unauthorized copy (one of your options, though query whether the other two are immune here), you expose yourself to the very small risk that you may be the next Joel Tenenbaum. But even a very small risk is not nothing when you consider the crushing statutory damages available. We’ll see what the First Circuit does in your case on that score.
In any event, you obviously have an agenda in this area, which is fine, but it’s important to contain your normative vision so that it doesn’t obscure your descriptive account.
Steve, thanks for the comment. Quick responses:
Your first point is obviously right, but I didn’t mean to imply that the “I” was limited to “me.” I meant to refer to music consumers broadly — sorry if this was the less than clear — and I think the general statistics on sales and filesharing, plus surveys of attitudes about willingness to pay (according to UK Music, about 85% of young Brits who illegally download say they would be interested in paying for an unlimited, all-you-can-eat download service).
Your second, point, though, I think misses the relevant inquiry. My point is about perception of value among consumers, not value in any formal economic sense. I highly doubt that many Internet users, who find and engage with music at their leisure, think in any precise way about their time value, nor do they do a cost-benefit analysis of getting sued. Instead, what I mean is the more simple question: would I pay for this? Is this individual song or album worth my hard-earned dollars? And the answer is “no” for most recorded music sold in the “old” way these days — hence the drop in revenue for the Record Industry from over $14 billion a decade ago to over $6 billion last year. But, as the UK Music survey indicates, the answer is “yes” to the question: do you place value on having access to downloads of every song, ever? People will pay for that — if only that were an option.
Among other things, you appear to be saying that people who are willing to download unauthorized copies for free are unwilling to pay for copies. Presumably, if these same people could have access to every song ever recorded for free, they wouldn’t pay for that either.
Music recordings with modern technology present a really fascinating issue of economics and property law. If we accept that (1) people aren’t going to pay more than they have to for something, regardless of what we might say about their moral duties; (2) modern conditions allow users substantial access to music without paying for it, through either ex ante fees or the expected costs of ex post legal action; and (3) musicians need or want money for their work, such that good music will be under-produced if listeners can get away with not paying; then basic microeconomics suggests that what was 12 years ago a well-functioning private market has transformed into a massive collective action problem.
I don’t think legal action is a plausibly effective means of dealing with free riders anymore. I also don’t think it’s culturally plausible that “better” paid distribution applications will lead a lot of current free riders to start paying, because it sounds like free download sites are already so good that paid sites could offer only a small quality premium at best. No matter how much an individual likes music, unless he is incredibly wealthy it would be privately irrational for him to pay for music with the aim of incentivizing the production of more music (this is the collective action issue: if most people don’t contribute, those who do are internalizing only a tiny fraction of the benefit of their payments, which tends to dissuade rational actors). In fact, I expect free riding to increase as the proportion of the population that is web-savvy increases (and as stealing music becomes normalized throughout different levels of our society).
Unless I’m wrong about the inability of the industry to deter file-sharing substantially through lawsuits, the implausibility of paid sites that are so much better than free sites that many users will be willing to pay, or the claim that music will be under-produced if musicians don’t get money for it, then I can see only two possible solutions. The first is for labels and artists to cross-subsidize music recording, for instance through touring or merchandising. But even if this is economically feasible, many worthwhile artists don’t want to tour (e.g. the Beatles); it would be nice (and socially optimal) if the internet created a world where artists could connect to listeners without touring, rather than one in which they must tour or get day jobs.
The second solution relies on the recognition that music recordings have gone from being a standard private good two and more decades ago (when the current heads of the industry were getting started) to being a classic public good today. Recordings are now nonrival (in that any number of people can have the same mp3 with no marginal cost) and nonexcludable (in that IP rights owners and the government cannot, at least given political realities, prevent non-rights holders from unlawfully accessing sound recordings). The classic solution to the collective action problems associated with public goods is to have the government fund production of the good with coercively obtained tax money and then make it freely available to everyone. This could probably be done very well with music and the internet, and in a way that would still resemble a private market from the artists’ point of view by giving more funding to more popular artists and thus subsidizing music that taxpayers actually care to listen to (i.e., it would not be like an NEA for rap).
I expect that a lot of people know more about these issues than I do and can make better suggestions, but based on Jason’s account of the conference, it sounds like a lot of people in the industry can’t. And, to be clear, I agree with Steve that better-quality paid sites are not a plausible solution: so long as free sites are available, the people who download for free are not going to start paying.
Mark-
Expectedly great comment — but it’s also importantly deficient, because it’s stuck in the world of rational-actor economics. We know that doesn’t fly in lots of places, but nowhere more so than on the Internet.
The comment is expectedly great because, in a few paragraphs, you’ve identified the central issue and offered a terrific analysis of the nature of the problem and what may be the best-case solution: a move from the current system to a true alternative compensation scheme where artists would be compensated in proportion to their popularity, like the one described by Professor Terry Fisher here. These proposals have gotten more traction in other countries, especially Brazil, so we may have a few samples of how well this system works in a few years.
In the medium-term, though, such a radical change is fairly unthinkable in the U.S.. The all-you-can-eat download solution that I describe here is more feasible because all it would take would be a willingness of the four largest recording companies to give it a try. As you recognize, though, to some extent its success may rely on what you call “privately irrational” conduct — paying $20 a month for a service that’s not THAT much better than going to see the sea of illegal music. But, because the “economics is all” mindset doesn’t hold true in this area, I don’t think that’s a huge problem.
“Privately irrational” conduct abounds on the Internet: see, e.g., the Internet itself, whose entire existence depends on something called the Internet Engineering Task Force, which is “a loosely self-organized group of people who contribute to the engineering and evolution of Internet technologies” that receive no compensation for making sure the Net just works. That’s why, as Yochai Benkler quickly summarized in a conference white paper, voluntary payment models have been remarkably successful in this space. That should make the heads of economists spin, but it’s true: it turns out that, even in our capitalist society, people give money for music they love voluntarily when they feel they are receiving value and the money is going to the artist. (If you’re curious about more ways people’s behavior on the Internet seems to diverge from the rational-actor model, read his book The Wealth of Networks this summer. Dense but brilliant.)
So those are the keys: give people value, and compensate artists. The service doesn’t have to be totally, 100% rational from an economics standpoint, and, to be sure, it won’t get 100% adoption. There will be free riders. But people love music, and consumers want to see artists they love succeed, so I think enough consumers would pay as long as they believe they are being treated fairly by being offered a really great product. No one can be sure that I’m right, but it sure seems to me that we ought to at least try something like this and see if it works.
To be fair, the general behavior of most music fans in our generation, as described in your post, matches my economic model very closely: people can gets tons of music for free, so they do. Beyond that, I don’t want to be married to the “rational actor” approach; people will expend their resources, including time and money, where doing so will make them happy, and so the question we’re dealing with is to some extent one of social psychology rather than economics (economics deals with how people act to achieve their goals, but has nothing to say on where those goals come from). People don’t generally derive utility from transferring money to someone else when they don’t have to, but this is hardly a universal rule and it isn’t an economic conclusion at all (though it is often an economic premise).
I hope to read The Wealth of Networks and have enormous respect for Benkler, and I know he’d find my freshman microeconomics here woefully simplistic. But arguing is a good way to learn, and I do have some responses to the white paper. I don’t know anything about the less popular artists he discusses, but I don’t think the Radiohead and NIN examples are generalizable in the way you want them to be.
Radiohead initially released In Rainbows only through its website. Everyone who would have bought a CD thus bought it online (if they knew how), and it is unsurprising that those people were willing to pay money (generally less than a typical CD price, I assume) for the download, since it was an incredibly novel and exciting release format and a massively popular band (with all the money going straight to that band, which is surely attractive to fans). These effects would probably wear off after a few more similar releases and may not apply at all to more marginal groups (but this is a psychological claim that I cannot prove without experiments). (Anyone know why they did not offer the pay-what-you-want-option for The King of Limbs?) At any rate, Radiohead charged a minimum of almost a dollar (45p in 2007) as a “transaction fee” for people who chose to pay nothing, meaning that there was not actually an option to get the album for free from their site. Anyone who wanted it for free — which I assume was millions of people — presumably got it for free from the usual sources. So, unless I’m missing something, we’d need some sort of big poll of Radiohead fans to determine whether any substantial proportion of the people who would have done free illegal downloads if the CD had been released in the usual way chose to pay for the download. Otherwise, Radiohead’s data show only that many of the people who were willing to pay for a CD were also willing to pay somewhere between the minimum transaction fee and what the sticker price for a CD would be (or, actually, that is the best the data could be expected to show; Benkler suggests Radiohead haven’t actually released the data at all, so we don’t know!) . That is interesting and impressive (again, assuming the secret data show that it occurred), but it does not necessarily tell us anything about illegal downloaders, the group of fans we’re most concerned with, the fastest-growing group, and the group most relevant to the future of the industry.
(Here, for what it’s worth, is a newspaper poll of Radiohead fans asking them what they paid; but it seems to only include fans who downloaded from the official site.)
The argument is similar with NIN: according to Benkler, Reznor offered a free stream and free parts of tracks, but charged money for downloads of full tracks, the really valuable thing. A lot of people bought them and he made a lot of money; but a lot of people would have bought them for money in CD form and on iTunes, and if he made more money this way it was by cutting the label and distributors out of the deal, not (so far as we can tell) by getting people who would have otherwise downloaded illegally for free to use his site. If a major artist would offer a record for download from an official site with an option of “absolutely free,” and if the site were slightly more efficient than the usual file-sharing sites, and if he would then release the download data, we could probably draw reliable assumptions. But I don’t think the Radiohead and NIN natural experiments are very helpful for our debate.
Also, for what it’s worth, Benkler’s arguments don’t sound like they support a $20/month all-you-can-eat subscription service, at least unless the company is paying out your money on a pro rata basis to the artists you actually listen to. If the site just spends its revenues on generic licenses from the labels, then a Radiohead fan will be giving 99.999% of his fee to other artists (e.g., Garth Brooks and Willow Smith), or rather to other artists and to record label execs, and won’t have the motivation of supporting his favorite artists that is central to Benkler’s argument.
I hate to appear to disagree with you about everything, but I’ll just say this: The Wealth of Networks is a standard rational choice account. Benkler’s most recent work on cooperation is different, but the book is standard, if excellent. His point is that not all producers of expression rely on exclusive rights. To take a simple example, academics freely share papers and books because their money comes not from royalties but from universities, and their standing in the community depends on people knowing and respecting their work. The amazing thing about the internet is that it enables people to work on small chunks of big projects together, without much infrastructure cost. If I enjoy contributing to wikipedia, and I can do so in whatever size chunk I want without buying anything special to do so, then I will, thanks to the internet.
This is standard rational choice. His Coase’s Penguin paper, adapted for the book, should make that clear enough. Coase’s great contribution in The Nature of the Firm is the point that some goods are more efficiently produced by firms than by individuals contracting in the market. Benkler’s is the point that, in light of the infrastructure of the internet, sometimes peer/commons-based production will be more efficient than either individual or firm-based production. And Benkler, like Coase before him, explains the conditions that make one model more attractive than others. One of the things that makes the book so valuable is that it relies on the same traditional economic reasoning to lead to such surprising conclusions.
its really interesiting and a good thing…
Response Post: Rethink Music Conference, Day 2: It All Comes Back to Statutory Damages | Just Enrichment
Response Post: Joel Fights Back » Jason Harrow: Day 2: It’s all about statutory damages
Response Post: When Should Speech Equal Money? | Just Enrichment
Response Post: Why Free Speech? Answers from Law & Economics: Speech as a Public Good | Just Enrichment