This is a long post that sketches a system in which we can have about the right amount of digital goods at the right price, and pay the people who make them properly. IP engineers, lawyers, and economists, have at it: time to stop rearranging deck chairs and steer the ship.  The central underappreciated insight in my view is that the digital content technology system cannot be fixed by torturing dead-tree rules to fit it: technology rights must be technologically administered.
We should think about this in the larger context of infrastructure investment. In the last century and before, this country was able to channel enormous resources and courage to build stuff that shaped the quality of life for the better and that also paid off in enormous economic gains. I’m thinking of railroads, (less fondly of highways), museums, universities, water systems, the electric grid…
Not all of that infrastructure was physical: we also invested in all the knowledge in the libraries of those universities and the know-how in the heads of all the people who attended them, weights and measures standards so any 10-32 machine screw will fit the hole tapped with any 10-32 tap, all the music, books and movies, and more. “More” especially included a legal system that worked well enough to get those books written and songs played, at least as long as they had to be packaged in something that could be locked up in a room.
Almost all of it is worn-out, obsolete, undermaintained, and failing in so many ways. The roads are potholed and congested, the electric grid doesn’t connect to where we need to collect energy and too stupid to do what it should, the universities are turning away people who want to learn (my main spring course roster is 20% waitlist and no, a bigger room won’t solve the problem)… But the most important of these failures, triggered by our ability to encode information as bits instead of ink, is the collapse of the information distribution system, and Aaron Swartz is an appropriate symbol of almost everything that’s wrong with it. Today I got an overdue notice from the library for a paper book. What’s next, a horsedrawn carriage ride to return it: why, in 2013, should I have to get a book back on a physical shelf for someone else to be reading it??!! Why should someone in East Gulch, with its one-day-a-week, one-room library if it has one, not be able to read it at all? Why should musicians not be able to collect a nickel for their work without selling seats for a live performance?
Now is the time to build an information infrastructure of laws and conventions that will create more value and take fewer casualties. First, not just a hat tip, but a sweeping bow to Lawrence Lessig [link fixed 20/I/13], Terry Fisher, Neil Netanel, and the other thinkers who have been exploring this territory over the last decades.
First, the general principles:
(1) Honesty: When people decide to use information, they have the right to know what that use really cost (not what it cost to make the first copy, interesting though that might be). When people generate information, they have the right to know what value their work created or is likely to create.
(2) Privacy: People have a right to use information privately, especially with regard to government.
(3) Efficiency: Independent choices by actors in a free market, correctly informed by prices about costs and benefits (including external costs and benefits) , will almost always beat administrative or authoritarian choices.
(4) Justice: Creators have the right to be paid for their work in amounts commensurate with the value they create. Consumers have the duty to pay for the costs their choices impose on others. Everyone has the right to free speech as broadly understood under the First Amendment.
(5) We’re all in this together: information is the collective patrimony of society. Taxes paid reflect collective benefits but not necessarily exactly the benefits received by each person.
(6) The criteria for a digital medium property regime is that it be better than what exists, and good enough. It will have irremediable flaws and deficiencies.
Short version and implications: Creators (if they wish) need to be paid proportionally to the use made of their work; consumers of digital content need to have access to digital content at marginal cost, which is in almost every case zero beyond the consumer’s time and wear and tear on her hard drive. The big fact about consuming digital content is that doing so leaves no less for everyone else; the corollary of this fact is that charging any more than nothing to consumers is wasteful.
Here is a sketch of the only regime known to me that respects these principles. Nothing in it exceeds demonstrated capacities of existing technology. Note that it can coexist with conventional copyright for authors who prefer the latter.
A. Beginning in 20xx, there shall be appropriated from general funds to the copyright office the sum of $00 per year for administrative costs and implementation of the program described herein. Beginning in 20yy, there shall be appropriated from general funds to the copyright office the sum of $000 annually to be disbursed as “use-scaled” royalties as provided in the following titles.
B. Any legal person wishing to receive use-scaled royalties for creative work in digital form as author shall place with the copyright office a fair copy of the text, image, recording, etc. and identification sufficient to direct annual royalty payments. The copyright office shall retain copies and distribute royalties for 00 years after a work is first registered, and inform registrants of attempts to register works already in the system.
C. Any device with access to the internet shall be equipped with software distributed and certified by the copyright office that will, whenever a file is opened:
a. “Hashcode” or otherwise extract sufficient information to identify the work contained .
b. Transmit the code created under C.a , but not who or what opened the file, to the copyright office. One such transmission of code is a use of that work.
c. Transmission of the file containing a work between devices or persons shall not be a use.
D. The copyright office shall accumulate annually a count of uses for each registered work. Annual royalties shall be distributed in proportion to the fraction of uses recorded for each work, divided by the total of uses for all registered works. Total royalties may be separated into such separate funds for different media [for example, video/movie royalties may result in different per-use payments than music or text] as Congress may from time to time determine.
The copyright office may use statistical sampling techniques to determine the allocation of royalties where appropriate, balancing precision and accuracy of result against administrative cost.
E. The copyright office shall not trace the origin of, nor keep any record of, the individual or device that opened a file thus identified, nor cooperate with any law enforcement or security agency with regard to its records of use.
F. The copyright office shall perform the legal, computer science, and related research required to implement the actions described herein internally and/or by contract as the Register of Copyright may determine.
G. Any person may copy, distribute, read/listen to/watch/mashup/etc., any work registered for use-based royalties without limit of time or place.
H. No provision herein shall replace or nullify any rights granted authors and their assigns under existing copyright law, except that a work in which copyright is claimed shall not be admissible for registration for use-based royalties. Rightful owners of existing copyright who register their works for use-based royalties transfer their rights to the public domain as per G above.
I. Rights to use-based royalties are transferable among legal persons with notice to the copyright office.
Several problems:
1. You can’t “transmit the hashcode” if you’re using the file on a disconnected
device, for example an iPod or a Kindle, not connected to any network
(and possibly never connected to the network).
2. If/when you transmit the hashcode to the government, “they” will be able to
trace the packet back to its source. So “they” will know, if they care,
which devices were using which files when. Which blows a hole in privacy.
3. Hashcodes must presumably be short (else you couldn’t send them frequently),
and fairly easy to compute (else you couldn’t afford to do it each time you
open the file). So here’s my plan for retirement: reverse-engineer the algorithm;
pick the top 10 books and top 10 songs and compute their hashcodes; construct
documents with the same hashcodes (might take a while); register those fakes
and wait for the royalty checks to start pouring in.
4. Presumably changing a few bits in an audio file would change the hashcode.
So you could actually defeat the accounting pretty easily (though doing so
doesn’t save you any money).
In response to 3, as well as to Olof:
I was discussing essentially the same proposal with a friend of mine a year or so ago (although only for books and music; I was inclined to think that the high production costs of movies might shift the production-cost/enforcement-cost tradeoff enough to make this scheme inappropriate for them), and considered the problem of fraudulent self-hits. At the time, I felt that this meant that we couldn’t have the privacy aspect of the scheme: every hashcode would include in its transmission an ID for the consumer who accessed the digital content, which could be used to detect self-hits. Now, however, I think that simply having a moderately-effective screen to catch such fraud, combined with large-enough penalties for the fraudsters, would suffice in practice.
In response to 4:
Your parenthetical addendum goes right to the whole point of the system: since the consumer has no incentive to cheat, we get rid of the entire unholy mess of DRM at one stroke.
The only remaining issue is how to bell the cat….
It isn’t clear that making a work with a particular hashcode constitutes “fraud” -
and it’s even less clear that you could ever prove in court that any particular
hashcode collision was deliberate, rather than accidental. Hashcode collisions
will happen, and you’d better have a plan for dealing with them.
One approach would be to check each new work when it’s registered, and refuse
those which clash with already-registered works. But then you’ve introduced
a mechanism for censorship - you don’t like Krugman’s op-ed this week ? tell
him there’s a hashcode collision and it can’t be published for gain.
On the second point, while you’ve removed the incentive for consumers to cheat,
you’ve left massive incentives for lawmakers to cheat, either by shrinking the
pot of money arbitrarily, or by specifying the division of the pie in ways
which favor wealthy corporations over the less powerful, as with the Mickey
Mouse copyright extensions.
Although Hashcode collisions are rare but real they are also easily fixable. Tell Krugman that his column contains a collision and can’t be published? He republishes after replacing a contraction, or just adds a space between new words, voila new hascode
The weakness of your suggestion is that you have now given government the power
to delay publication and demand changes. That is open to abuse.
The hashcode scheme also doesn’t deal with plagiarism and mashups, though I guess
the idea is that in registering for the new-style hashcode-based royalties you
would be implicitly granting for anyone to do whatever they liked with your
IP (which of course would make it very different from copyleft).
In fact I find pretty difficult to imagine which class of creators would find
this scheme attractive: the GPL/FSF people wouldn’t favor it because it allows
others to pick up the work and modify it without making the modifications public;
wealthy corporations wouldn’t choose it because they have the resources to
protect and profit from their IP under the current system. So I don’t think
it would end up being used for much stuff that’s valuable; and as a result,
I don’t think there would be strong enough political lobby to get it meaningful
funding.
#1: But, if I’m understanding this right (which I’m not entirely convinced of), you would only to transmit these at what, in today’s terms, would be the point of purchase. I suppose you could have an offline mode if movie viewing was on a per-view basis where the tokens would be queued until the next online event. Or say a movie could be viewed X times until a new token was required.
#2: True. But I think he’s talking more about forbidding the government from building a tracking database. I do see a potential problem with intermediaries as a packet traverses the network. Suppose, this is done over HTTP (bad idea, btw). You’d have to guard against 3rd party proxy & cache servers storing off traffic.
#3: You may be waiting a long time for those retirement checks since the hash code would be registered with the author of the original work.
You have to assume that each work would have a unique ID. So the ID would be more like a hash code + “index number”, mimicking how a hash table stores its elements. Which brings up a LOT of issues around reliability, redundancy, scalability, security (think DoS attacks on hash tables)… There are HUGE technical issues to work out. This system would be to culture what oil is to our present economy.
#2 “Forbidding the government” is not something any of us should trust, after the
various post-9/11 wiretapping abuses. Anytime you give them the technical
capability to intrude, you can bet they’ll do it, regardless of the law.
#3 You’re assuming that any hashcode collision is resolved in favor of the
earliest-registered work. But that isn’t going to work - at some point the
new Disney/Pixar movie will just happen to clash with an obscure academic
book, and legal chaos will follow.
#4 If you need an extra uniquifier to resolve hashcode collisions, then that
is an ID which a) has to be issued by the government, and b) has to be
entangled with the work in a cryptographically-secure way. a) means the
government can censor everything; b) is the existing DRM problem, which this
proposal was supposed to avoid.
As it happens, I’ve worked quite a bit in the hashtable business, so we can go
into the weeds on those details. If you don’t know the complete set of works
in advance (which obviously we can’t, since the scheme has to work for books which
haven’t been written yet), then collisions are inevitable.
#1 My understanding was that the intent was to allow free copying of works between
anybody. So the whole idea of “purchase” would go away, and the royalty would
be due on each “opening” of the work (with some tricky details about consuming
parts of a work, large vs small works etc).
Right. There’s no importance to possessing digital content, only having access to it. And the act of economic importance is engaging with a work, not having it on your hard drive or anywhere else
None of these objections really holds up.
1. For the disconnected devices (which Kindles and recent ipods are not) there’s still the machine that loads the file onto the disconnected device. Sure, you can jigger things so that you load a lot of devices and only report one use, but the incentive is not going to be strong. And if the device is merely not currently connected, the hash packets will wait for transmission, just the way that any other packets do.
2. This is what proxies are for. Every library in the country, for example, could operate a hash-forwarding service. (Yes, the device would need to be capable of using a proxy.)
3. Hashes are currently around 256 bits, which is about enough for one per particle in the universe. Kick that up to 1024, which would take all of 16 milliseconds to transmit on the slowest-available connections, and collisions are going to be vanishingly small. Practically speaking, only people looking for collisions will find them, and dinging a file for collision doesn’t actually prevent or delay publication, it merely prevents registration for this particular compensation scheme.
4. Sure, there will be lots of files around whose hashes aren’t registered with the compensation scheme. People who own copyright in those files can use the current methods to seek compensation for them (with a lot less hassle from the public, since anyone distributing an altered copy is pretty clearly doing something they likely shouldn’t.)
The scheme is counting heads, rather than desire. It’s better than the status quo (what isn’t?) but I assume is it an early alpha version.
Can you expand on this a little?
Should all work be valued by what current society deems popular? I like the idea except for the tracking. How does international access work? Are you going to have the UN take over?
If you put 20 chapters of a book into the repository separately, then do you get 20 uses instead of one?
…and does a half-hour symphony get ten clicks and a three minute song one? Probably. More to come on this.
How to protect against gaming this system? — if I want to raise some money for myself, all I need to do is read my own book, or hire a botnet to read it for me.
This particular problem wouldn’t exist if the money came from individual content consumers rather than from a public pool.
Make the payments (gifts?) voluntary, and make the whole system opt-in for content owners.
https://thesamefacts.com/2006/01/uncategorized/paying-for-music/
It’s quite possibly more high-minded than practical, but I do find the direct democratization appealing.
In response to 3 of Richard Cownie, as well as to Olof:
I was discussing essentially the same proposal with a friend of mine a year or so ago (although only for books and music; I was inclined to think that the high production costs of movies might shift the production-cost/enforcement-cost tradeoff enough to make this scheme inappropriate for them), and considered the problem of fraudulent self-hits. At the time, I felt that this meant that we couldn’t have the privacy aspect of the scheme: every hashcode would include in its transmission an ID for the consumer who accessed the digital content, which could be used to detect self-hits. Now, however, I think that simply having a moderately-effective screen to catch such fraud, combined with large-enough penalties for the fraudsters, would suffice in practice.
In response to 4 of Richard Cownie:
Your parenthetical addendum goes right to the whole point of the system: since the consumer has no incentive to cheat, we get rid of the entire unholy mess of DRM at one stroke.
The only remaining issue is how to bell the cat….
PS: Trying to post this as a “Reply” to Richard Cownie’s comment seems to result in an error — is that because I’m using Safari?
Hmm, “seems to” is apparently the operative phrase here….
It’s a bit confusing to call the agency running the scheme the “copyright office” since it’s an alternative to traditional copyright which will continue in parallel. “Copyleft” has already been taken for renunciation of copyright of the Creative Commons type. So I can’t think offhand of a nice catchy term. “Social copyright”?
Clause H implies that traditional copyright has to be claimed instead of being automatic as now. A needed reform anyway. So we need a (traditional) Copyright Office as a registry, as with patents.
Moral rights: in Continental IP law, these include the right of the author to be identified. Also, more questionably, the right to integrity of the work, which can prevent spoofs and reworkings. I don’t want any money for reuse of my blog posts, but I’d be very annoyed if they were simply plagiarized. Should you have to register to claim such rights?
A pity that commenters have seized on the technical hashcode issue. The real problem is of course that the scheme is socialist. Not a killer for me, but how can Middle America be persuaded?
“The real problem is that the scheme is socialist”
Well, yes. It’s obviously not something that could pass through the current
Republican House majority, which is going to be deeply unsympathetic to the
IP rights of creative types (other than big corporations), and deeply hostile
to anything that can be classed as “spending”. So I was assuming that the
discussion was about what we would hypothetically want in a perfect world,
ignoring the politics.
However, it’s possible that the political landscape might change after future
elections. The technical issues will remain.
A thought-provoking post. Thanks for putting it out there for discussion. My 2 cents, offered in the spirit of constructive criticism:
A. Given (5), this seems to suggest taxes as the source for $00x. I think I see a practicality issue here convincing a public that can’t agree to socialize the costs of something as critically necessary to all of us (collectively as well as individually) as health care to agree to do so with something far less obviously necessary such as information and entertainment.
B. (and H.) I like the opt-in concept.
C. Aside from the practical and constitutional issues involved in requiring all digital consumption devices to report their consumption to the government, this one (perhaps counter-intuitively) has compliance issues with (2). Hashcodes have their place in protecting privacy (for instance to hide the plaintext of a secret password while allowing the receiving party to confirm the password by comparing the received hash against the password’s hash), but in this case the hash hides nothing because the thing being hashed isn’t secret — it’s more like using tinyurl to shorten a link. It would be trivial for anyone to discover what the hashcode is for a particular piece of work. Given that all internet transmissions necessarily contain the originating IP address along with the contents of the transmission, and this information is to be collected at a central repository controlled by the government, this proposal makes it trivially easy to track what content is being consumed by IP address. Under IPv4 that doesn’t necessarily track to an individual device with 100% confidence, though apparently high enough that the courts currently accept IP evidence in copyright cases, but under the coming IPv6 protocol every device will have it’s own unique IP address and there will be near-100% confidence in tracking internet transmissions to the individual device. I think the privacy implications of easy government tracking of all consumed digital content to the device level are pretty obvious.
D. I see problems with requirement (6) in that it encourages the production of fluff over substance by rewarding producers based solely on popularity. It’s worse than the status quo because producers have no control over the price charged for their product, reducing the ability to recover the greater production costs of delivering higher-quality content. Since every producer always gets paid the same per-consumption, the creative focus will trend toward proven popular formulae and controlling production costs. D. also has problems with (6) in that it encourages click-fraud. There would need to be some way to prevent producers from taking measures to artificially drive up the numbers on consumption of their products. It would be trivially easy to use software to repeatedly transmit a work’s hashcode to the central repository.
E. We once had similar safeguards against wiretapping telephone technology, but they have been rendered moot for the most part. The federal government has already built and operates a huge infrastructure for the tracking of internet activity. I hold out very little hope for the long-term effectiveness of this aspect of the proposal.
F. Outsourcing privacy-sensitive data-processing activity to the private sector increases privacy risk.
G. Looks fine on it’s own, but unfortunately it won’t stand on it’s own without a lot of these other provisions.
H. Note that this would require the re-instatement of the requirement to register a copyright. As it stands right now, copyright is automatic.
I. Agree that transfer of rights should be accommodated in any such alternative to current copyright policy.
My thinking about alternatives to the current digital copyright status quo involve ideas like the copyleft concept championed by Larry Lessig and the Free Software Foundation and similar ideas proposed by other Pay What You Want pioneers. I think we’d get more traction toward the goals this proposal advances by raising awareness of and advocating for greater voluntary participation in PWYW digital distribution than by trying to “steer the ship” while moneyed interests are still fighting over control of the steering wheel.
The success of gpl-licensed software (every android device uses it, for just one example) and Radiohead’s In Rainbows PWYW release offer to show us a way to alternative protocols which facilitate a market in digital content while accommodating the realities of digital distribution, without the immediate necessity to re-legislate the status quo. If a substantial majority of digital content providers can be convinced to adopt an alternative voluntarily, it should be politically easy to drop conventional copyright protection from digital content and legislate whatever safeguards prove necessary to promote the efficacy of the alternative. Without the support of content providers, any alternative faces huge political barriers.
nit: I could find nothing remotely associated with Lawrence Lessig at the link bearing his name. Looks like a mistaken target unless I’m missing something.
link fixed, thanx, sorry
An excellent book! Thanks for fixing the link.
GPL/copyleft software is terrific. But it mostly works because the people involved
can get paid while working on it - either by corporations, or academic institutions.
Writing GPL’ed software - like writing academic works - is not a way to make money
directly, but it is a way to establish reputation and peer recognition, which then
lead to jobs which pay fairly well (and are enjoyable).
To make a no-payment-for-use scheme work more widely, you need to think carefully
about whether it will motivate not just the few Stallman’s, but also the people who have -
or want - a spouse and kids and a mortgage.
As a longtime open source developer, writer and audio drama actor/mixer, I would certainly not mind receiving compensation for all the work I give away, but I think this is a seriously flawed, highly gameable proposal. For one thing, I see no honest way to associate compensation to value. There are some things that are, frankly, worth more than others, and some that we use simply because they bear little or no cost. I don’t see how you address that fairly, without making things much worse than they are. There is no reliable correlation between value and numbers of bits of data.
And it is not clear to me that what you’re trying to solve is a sufficiently large problem that it’s worth putting all this power into the hands of greedy politicians. Those of us who do this free work do so because we value the results more than we value our time. Isn’t that the case for bloggers (and commenters) here? In some cases, to be sure, our work carries a value to commercial interests, such that they are willing to support maintainers. In many others, we do them just because. Is there really a shortage of creative people such that you think creating some compensation will bring more? Have you looked at the shear amount of fiction published for free online? The number of free and open-source software projects maintained at GitHub and elsewhere? Is it really hard to find free performances on Youtube?
What real problem is there, other than a wish that we could do what we love - and get paid for it as well, even if few people are willing to pay us of their own free will?
The Electronic Frontier Foundation made a similar proposal years ago for how to handle online music. It’s called “Voluntary Collective Licensing” and it’s modeled after how songwriters are paid today. See https://www.eff.org/pages/better-way-forward-voluntary-collective-licensing-music-file-sharing for the proposal.
Note that in the proposed “Aaron’s Law” on this page, creators would have to opt-IN to this system. In that way it parallels the EFF “Voluntary”
licensing proposal. If creators are going to have to volunteer for this, though, there is no need to have a government agency do the work; the private sector can do just fine. The private sector is better at having its income and its outgo match, unlike the federal government, where copyright holders would quickly lobby up the federal handout $ without increasing the quality or quantity of their cultural contributions. I really don’t think Congress should be in charge of deciding how much money the whole society spends on culture. (What is currently funded by Congress, like Arts grants etc, is a tiny fraction of the total social spending on copyrighted works. And I don’t believe in central planning as a workable economic model.)
Also, a private sector version would work fine internationally — while the US-federal-centric system doesn’t seem to have considered what happens when a Canadian writes a book or publishes a song (and Americans use it).
Gaming the system by reporting lots of “uses” of one’s own works could be simply eliminated by sending in the usage information with the periodic payment by the user. E.g. when you pay $5 per month, you tell them what works you used that month. If you don’t pay, you don’t get a vote on how the payment is divided. This also tends to eliminate the problem of “what is a use?” that the naive proposal above waves aside by saying “whenever a file is opened”. (There are many files opened by automatic processes such as backups or hard-drive indexing; and many copyrighted works that are used but never “opened” in a computer GUI.) A user could voluntarily report that they used a single work this month, and all of that user’s payment for the month would go to the author of that work, minus overhead. There are ways to game this system too, but they are more benign.
The best part of a voluntary scheme run by the private sector is that it doesn’t require a corrupt Congress to pass an “Aaron’s Law”; we can just start doing it under current copyright law, using a license similar to the Creative Commons licenses, but with permission to copy conditioned on periodic usage reporting and payments.
The “periodic payment by the user” is what the proposal is trying to avoid - because
in a world of digital distribution, any level of user payment beyond the marginal cost
of a digital copy (very very close to zero) will result in inefficiency, in the sense
that some people who could gain utility from the work will be prevented from doing
so by the inability or unwillingness to pay.
I like the user-doesn’t-pay aspect of it, if we could make it work. And we ought to be aiming for
a world where as much knowledge as possible is freely available to billions.
I’ve dealt with USA music/songwriting royalties a little, and it seems a fairly well-
designed system. But I don’t think it achieves the optimal level of compensation:
a whole lot of people derive huge enjoyment from good music, yet very few people
manage to earn a living income from music royalties.
I’m reading your suggestion as indicating that you think there is no value in actually creating content - just the value of distributing it. But why should any random collection of bits receive just as much compensation as any other? Pricing in general is a difficult problem, but the least bad system appears to be user-pays, since each user can decide what it’s worth to him. Yes, that does mean that some people will not be able to enjoy something. If the creator wants his work to be freely available, that’s his choice.
If the user doesn’t pay, who does?
And yes, while very few people manage to earn a living from music royalties, why is that a problem? I’d love to earn a living from doing random things that nobody wants to pay for - but why should I have such a right?
Hmm, I posted a reply earlier but it got lost somehow.
First “value” is a terribly slippery term, because it seems to suggest that “value”
is inherent in the thing created, somehow. But when you try to pin that down,
what is it ? Under the Marxist “Labor Theory of Value”, it’s precisely the hours
of labor that went into creating the thing. But that doesn’t work. I saw a good
counter-example once: a luthier had moved from the Eastern Bloc (I think Poland)
to the USA, and he spent roughly 6 months creating a single - very beautiful - guitar.
Then he put a price on it: $30000, compensation for his 6 months of labor. No-one
bought it. The other handmade guitars in the shop were in the $3000-$6000 range.
No-one was buying a $30K guitar. So here’s the puzzle: what was the “value” of
the guitar ?
You can also dissect this joke: an assembly line stops, no-one knows how to fix
it, so eventually the managers call in old Joe who is retired. He walks up and
down the line, takes out a hammer, hits it in a particular spot, and the factory
is up and running again. They ask him how much they owe for his fix, and he says
$1000. “That’s a lot of money” says the manager “I need an itemized invoice”.
So he writes it out:
Hitting with hammer - $ 1-00
Knowing where to hit - $999-00
That’s the essence of IP. Knowing where to hit isn’t worth anything in itself;
it’s worth a lot when someone uses the knowledge. “Value” lives somewhere in the
interaction between a supplier of IP, and a consumer of IP.
I prefer the term “utility”, because it is less confusing: when the consumer
acquires the information (not exactly a transfer, because the creator still has
it all), the consumer’s utility is increased, and the creator’s utility is the
same. The increase in total utility might be what you think of as “value”.
But it doesn’t happen unless and until the information is used.
Second, I’m not at all suggesting that anyone has a right to earn a living from
songwriting. I’m merely suggesting that our current system leads to a society
with a huge numbers of lawyers and accountants, and rather few people making a
living as songwriters, even though few people derive pleasure from lawyers, and a
lot of people derive pleasure from good songs. I suspect the balance was
rather different in the days of Tin Pan Alley. And I think we might have a
happier (higher-utility) society with a system which skewed a little more
towards the rights of individual content creators.
I think we might partially agree. To me, “value” has nothing to do with any Marxist theory. Value, rather, is in the eye of the user. It is inherently subjective. If you consider something to be of high value, you’re going to be willing to pay for it rather than go without. As your joke notes, the managers should be willing to pay Joe’s price, because they value having the line running. Even though the time he spends is minimal, if the line breaks down, they’re going to call him again.
And yes, we have lots of lawyers and accountants, partly because people are willing to pay lawyers and accountants, but at least as much because a single lawyer or accountant can only handle so many clients. On the other hand, a single songwriter’s work can be enjoyed by millions of people - which means that you don’t need as many songwriters.
It’s the same comparison of pro athletes or movie stars, who can make so much money because so many people can watch them, while a teacher can only teach a couple hundred students in the course of a single year. So the former make a lot more money, but there are many more jobs for the latter.
https://www.eff.org/pages/making-p2p-legal is another EFF explanation that shows how previous rounds of innovation produced collective licensing systems (both voluntary and compulsory).
Re: “you don’t need as many songwriters”
We’re inching closer to agreement. But I don’t take it as a given that the current allocation of
effort between say, accountancy and songwriting, is optimal for society. People are in fact
very willing to pay to experience good music. And, as you correctly observe, technology now
allows many people to hear the music of a few songwriters. But to me, that doesn’t prove that
we’re getting the optimal level of songwriting effort and musical enjoyment. I think it’s
quite possible that having a larger population of professional (i.e. able to earn a living wage)
songwriters would lead to higher quality songs at the top of the heap. In the 1920s and 1930s,
we had a large number of professional songwriters, and emerging at the top of the heap we had
some of the greatest songwriters of all time, e.g. George Gershwin, Jerome Kern, Irving Berlin.
I’m a little skeptical about whether the current system matches up to that, and the mismatch
between technology and copyright/IP system might be one contributing factor.
If songwriting (or fine art, or political blogging, or whatever) is an activity that is useful
to society as a whole, then we should strive towards a system that allows many people to
earn a living wage, and thus allows a few extremely talented people to emerge from that mass
of professionals. If, on the other hand, you have a system where only a very few people
can make a living as songwriters, then people who are talented songwriters, but mediocre
accountants, will choose accountancy as the safer career. And I doubt the optimality of
that outcome.
The case of software development is interesting, as I mentioned earlier, because there are a
lot of “free” software projects, but there are also a huge number of (quite well-paid) jobs
for software developers. People can get experience in paid jobs, and use that experience
on “free” projects; they can use success on “free” projects as a credential for getting
paid jobs.
[Disclaimer: I’m a software developer, not a songwriter nor an accountant]
The question is, who is going to pay all of those songwriters? And who gets to decide that songwriting, etc. is “useful to society as a whole”? We have the NEA which attempts to some of this - but that just makes which arts get subsidized a political issue. I don’t know that I’d want a bunch of congresscritters deciding what counts as “good music” or worse, “music that is good for society.”
I do get what you say about people taking the safe choice. I used to do a lot of theater, and people would ask why I wasn’t doing it professionally. My answer, of course, was that I was rather fond of eating and having a roof over my head. There is no shortage of people willing to create for free - and that includes songwriting. If there is a problem, maybe it’s just that few people know about all the good amateur creating going on. I know that in the amateur audio drama world, we’re elated when a thousand people listen to one of our productions. I don’t think we’re really hurting for songwriters.
Yes, people will create for free. But they get better if they can do it full time.
And they can only do it full time if a) they’re independently wealthy to start with
or b) they can earn a living from it. See Malcolm Gladwell’s “Outliers” for some
examples. He’s particularly convincing on The Beatles, who were not much good
before they went to Hamburg, performed 8 hours a day for months on end, and came
back ready to conquer the world.
Actually, pro sports are another interesting case. You think pro athletes
make a lot of money: they don’t, in general. Their industry is structured so
that large numbers of talented young people play and practice essentially
full-time, either in colleges or the minor leagues, for several years,
for very little pay, or for college scholarships. Out of that mass of talented
and hardworking people emerge a modest number of journeymen (e.g. average NFL
players who earn maybe $400K/year for 4 or 5 years) and a tiny handful of
superstars.
Take away the mass of people in minor leagues and college, who are working at it
full time but just barely earning a living, and surely the quality of play in the
NFL and NBA would be much less impressive.
Median NFL salary - $770K, median NFL career - 3.5 years.
Median MLB salary on median team (Pirates) - $900K
Median lawyer salary (from BLS) $113K
So even out of the small minority of full-time athletes who make it to
the NFL or MLB, median total career earnings are unspectacular (e.g. 3.5*770K = $2.7M)
compared to say a median lawyer earning 30 years * 113K = $3.4M.
I don’t know - I think earning 2.7 million in less than four years is pretty spectacular. And if they are at all responsible with that money, they could invest it and make even more, plus have time to earn more in other jobs. They don’t usually commit suicide when they retire.
First, most of the kids who commit to trying to become athletes have to be
pretty much full time from 16, maybe earlier. So it’s not really four years;
it’s really about age 16-26, which is not just 10 years, but a particularly
crucial 10 years which makes it hard to get qualified for other good jobs.
Second, the $2.7M in 4 years (probably not really that much, because marginal
players likely to wash out early are also those earning below-median pay) only applies
to the tiny fraction of full-time (mostly college) football players who make
it to the NFL. Each year it’s about 100K high school seniors, of whom 9000
go on to college football, or whom about 200 reach the NFL.
If you’re choosing between football and becoming a lawyer on entering college,
there’s about a 99% probability that you’ll make more money in the long
term as a lawyer.
“If you’re choosing between football and becoming a lawyer on entering college, there’s about a 99% probability that you’ll make more money in the long term as a lawyer.”
True. And likewise, if you’re choosing between songwriting and teaching, the odds are you’re better off as a teacher. In fact, if you choose between any two careers, where one of them pays extremely well for a very small number, and the other pays moderately, but there is plenty of demand, you’re better off with the latter. But that’s just the nature of such careers. People may well find the one to be more glamorous, but there’s simply no room for lots of people to make a living at it - because a single top performer can provide what is desired for so many people. The only way to change that is to create a massive inefficiency: force people to pay for something they don’t want and don’t need, just because somebody thinks it’s somehow better for society.
“That’s just the nature of such careers”. In the days before digital
distribution, it was inevitable that only a few books would get published and
only a few records would be pressed. So only a few content-creators were
likely to earn a living.
But it’s different now. Anybody can produce content, and get it to an
audience of 1B+ at low cost - the “Gangnam Style” video is an example.
That raises the possibility that tilting the IP/copyright laws
could actually change things so that more people could make a living, e.g.
producing stuff that appeals to a few people in Europe and a few people
in Korea and a few people in California. Even if the big rewards go to
the superstars with a huge audience.
Well, there are certainly problems with the copyright laws as they exist now, but I don’t see how you’re going to change them in such a way as to allow more people to make a living. The Gangnam style video was a fluke - the kind that services such Youtube make possible (give away free, and benefit from advertising revenues if it goes viral). All it proves is that people can make a living from songs if they happen to be popular.
“don’t see how you’re going to change them in such a way as to allow more people to make
a living”. Well, we’re discussing technical details of one particular approach that might
help.
You can say “Gangnam Style” was a fluke; or you can say that it was the kind of phenomenon
which 20 years ago would be have been utterly impossible, but now has a low, but non-zero
probability. Since Justin Bieber is in the same class - obscure performer achieves
rapid worldwide fame based on low-cost internet video - it isn’t a one-off. Though I’m
actually more interested in the new possibility that people with a niche product can
make a modest living by narrowcasting to small, but nationwide or global, audience.
Political news sites like talkpointsmemo.com or fivethirtyeight are an example, and the
kind of IP issue that affects them is that they break a story by hard - and expensive -
reporting, and then big-money news-aggregator sites like HuffPo pick it up without
recompense (sometimes even without attribution). That strikes me as a problem (roughly
the same IP problem as the music industry had with sampling and mashups), and one
which might be solved in part by changes to IP/copyright law, though I don’t have a
concrete proposal about what such changes would be.