The Tragedy of the Anticommons
I assume that most of you are familiar with the concept of the Tragedy of the Commons. If you aren’t, well, that was a Wikipedia link right there.
However, fewer are familiar with the Tragedy of the Anticommons, a term coined by Michael Heller. Where the Tragedy of the Commons is created by too little ownership, the Tragedy of the Anticommons is created by too much.
For instance, the classical solution to the TotC is to divide up the commons between the herders using it, giving each of them ownership for a particular part. This gives each owner an incentive to enforce its sustainability. But what would happen if the commons were divided up to thousands of miniature pieces, say one square inch each? In order to herd your cattle, you’d have to acquire permission from hundreds of different owners. Not only would this be a massive undertaking by itself, any one of them could say no, potentially ruining your entire attempt.
This isn’t just a theoretical issue. In his book, Heller offers numerous examples, such as this one:
…gridlock prevents a promising treatment for Alzheimer’s diseases being tested. The head of research at a “Big Pharma” drugmaker told me that his lab scientists developed the potential cure (call it Compound X) years ago, but biotech competitors blocked its development. … the company developing Compound X needed to pay every owner of a patent relevant to its testing. Ignoring even one would invite an expensive and crippling lawsuit. Each patent holder viewed its own discovery as the crucial one and demanded a corresponding fee, until the demands exceeded the drug’s expected profits. None of the patent owners would yield first. …
This story does not have a happy ending. No valiant patent bundler came along. Because the head of research could not figure out how to pay off all the patent owners and still have a good chance of earning a profit, he shifted his priorities to less ambitious options. Funding went to spin-offs of existing drugs for which his firm already controlled the underlying patents. His lab reluctantly shelved Compound X even though he was certain the science was solid, the market huge, and the potential for easing human suffering beyond measure.
Patents aren’t the only field affected by this tragedy. America’s airports are unnecessarily congested because land owners block all attempts to build new airports. 90% of the US broadcast spectrum goes unused, because in order to build national coverage, you’d need to apply for permission in 734 separate areas. Re-releasing an important documentary required a 600 000 dollar donation and negotiations that stretched over 20 years, because there were so many different copyright owners for all the pictures and music used in the documentary.
So, what does all of this have to do with rationality, and why am I bringing it up here?
The interesting thing about the tragedy of the anticommons is that most people will be entirely blind to it. Patent owners block drug development, and the only people who’ll know are the owners in question, as well as the people who tried to develop the drug. A documentary doesn’t get re-released? A few people might wonder why the documentary they saw 20 years ago isn’t available on DVD anywhere, but aside for that, nobody’ll know. If you’re not even aware that a problem exists, then you can’t fix it.
In general, something not happening is much harder to spot than something happening. Heller remarks that even the term “underuse” hasn’t existed for very long:
According to the OED, underuse is a recent coinage. In its first recorded appearance, in 1960, the word was hedged about with an anxious hyphen and scare quotes: “There might, in some places, be considerable ‘under-use’ of [parking] meters.” By 1970, copy editors felt sufficiently comfortable to cast aside the quotes: “A country can never recover by persistently under-using its resources, as Britain has done for too long.” The hyphen began to disappear around 1975.
This gives an interesting case study for rationality. If ‘underuse’ didn’t become a word until 1960, that implies that people have been blind to its damages until then. Heller speculates:
In the OED, this new word means “to use something below the optimum” and “insufficient use”. The reference to an “optimum” suggests to me how underuse entered English. It was, I think, an unintended consequence of the increasing role of cost-benefit analysis in public policy debates. …In the old world of overuse versus ordinary use, our choices were binary and clear-cut: injury or health, waste or efficiency, bad or good. In the new world, we are looking for something more subtle—an “optimum” along a continuum. Looking for an optimal level of use has a surprising twist: it requires a concept of underuse and surreptitiously changes the long-standing meaning of overuse. Like Goldilocks, we are looking for something not too hot, not too cold, not too much or too little—just right. …
How can we know whether we are overusing, underusing, or optimally using resources? It’s not easy, and not just a matter of economic analysis. Consider, for example, the public health push to icnrease the use of “statins”, drugs such as Liptor that help lower cholesterol. Underuse of statins may mean too many heart attacks and strokes. But no one suggests that everyone should take statins: putting the drug int he water supply would be overuse. So what is the optimal level of use? … We estimate the cost of the drugs, we assign a dollar value to death and disease averted, and quantify the negative effects of increased use.
Driving faster gets you home sooner but increases your chance of crashing. Is the trade-off worthwhile? To answer that question, you need to know how to value life. If life were beyond value, we would require perfect auto safety, cars would be infinitely expensive, and car use would drop to nothing. But if there is too little safety regulation, too many will die. With auto safety, society faces aother Goldilocks’ quest: we strive to ensure that, all things considered, cars kill the optimal amount of people. It sounds callous, but that’s what an optimum is all about.
… The possibility of underuse reorients policymaking from relatively simple either-or choices to the more contentious trade-offs that make up modern regulation of risk.
There are several lessons one could draw here.
One is that we’re biased to be blind to underuse, whereas overuse is much easier to spot.
Another is the more general case: we should be careful to look for hidden effects in any policies we institute or actions we take. Even if there seems to be no damage, there may actually be. How can we detect such hidden effects? Cost-benefit calculations looking for the optimal level of use are one way, though that’s time-consuming and will only help detect under/overuse.
Not to mention that it’s hard, requiring us to set a value on lives and other moral questions. That brings us to the second lesson. Heller’s analysis implies that for a long time, people were actually blind to the possibility of underuse because they were reluctant to really tackle the hard problems. Refuse to assign a value on life? Then you can’t engage in cost-benefit analysis… and as a result, you’ll stay blind to the whole concept of underuse. If you’d looked at things objectively, you’d have seen that we need to give lives a value in order to make decisions in society. By refusing to do so, you’ll stay blind to a huge class of problems ever after, simply because you didn’t want to objectively ponder hard questions.
That reinforces a message Eliezer’s been talking about. I doubt anybody could have foreseen that by refusing to put a value on life, they’d fail to discover the concept of underuse, and thereby have difficulty noticing the risk of patents blocking the development of new drugs. If you let your beliefs get in the way of your rationality in even one thing, it may end up hurting you in entirely unexpected ways. Don’t do it.
(There are also some other lessons, which I realized after typing out this post… can you come up with them yourself?)
I don’t agree that by failing to put a value on life you necessarily also fail to discover the concept of underuse. Doesn’t it follow immediately from the fact that you can have a positive externality that you would necessarily also have underuse?
See also Lawrence Lessig’s Free Culture for more on tragedy of the anticommons effects in intellectual property
ETA: Speaking of which, does anyone know what the intellectual licensing situation is with sites like this? May I, for example, state categorically that all my submissions are licensed under creative commons-by-nc-sa?
See About Less Wrong:
Also,
One of these days the Creative Commons people will get around to writing a variant that takes into account the moderator’s need to edit malformed/misspelled [X]HTML.
To answer your question: yes, you can license it cc-by-nc-sa, but it might conflict unhelpfully with the moderators’ reserved rights. IANAL, but a cursory google shows the matter hasn’t noticeably been tested in court. It might be possible that if a moderator modified your post to correct some spelling or markup, they’d be infringing upon the cc license.
A work can be released under more than one license. Releasing something to the general public under CC doesn’t interfere with the (implicit, unnamed) license granted to Less Wrong.
The drug patent story is completely unbelievable, it’s the type of yarn that an incompetent bureaucrat would spin to explain not making money for his company. Everybody in the biotech business would bend over backwards to increase the use of their IP—if you don’t use it, the only people making money off it are IP lawyers. Everybody knows there are ways of circumventing any IP position, it’s just a matter of time and money—bluffing in this game could mean you will be cut out of deals. There are strategic relationships, swaps, judicious use of trade secrets, and because money is to be made it is not too difficult to get multiple partners to work together, or at least not as difficult as when egos or tribal feelings are involved.
The concept of eminent domain exists specifically to deal with this problem, and it is at least a few hundred years old. Eminent domain is used to force land owners to give up or allow the usage of land when necessary to allow for the construction of projects like roads and sewers, and it has also been applied in limited cases to patents. However, expanding the use of eminent domain is likely to be politically untenable, because for every time it’s used there must be an unwilling counterparty who will make noise.
Eminent domain is a dirty, dirty hack. “We can’t figure out how to make property work, sod it, lets turn off property and just take it at gunpoint”.
There is a much cleaner solution that works without breaking property: buying options. You want to build a road, but you don’t want to he held hostage. So you play a game rather like connect-4 by buying (for trivial sums) contractual obligations to sell to you at a fixed price within some sensible time limit. Once you have any workable unbroken line of optioned land between A and B, you exercise the options and buy it. All a hold-out does is cut himself out of the deal, since you can route around him.
That.… seems like it’s actually a really good idea. (It doesn’t fix the issues the original posting describes, about the bajillion patents/way to finely divided pastures, but as a replacement, in our current world, for eminent domain, that sounds brilliant. At least to me it does.)
Actually, it does, slightly extended.
Consider: you want to herd cattle. You buy a herding lease from “unified pastures, inc”. They have gone and bought many, many options to graze cattle, and plotted out a cattle range from a connected group (with perhaps areas to fence off as hold-outs). Once they have a big enough range, and customers, they’ll exercise the options and again, the hold-outs are just cut out of the deal.
Okay. But what about those patent mess things?
Repeal the concept of patents. They aren’t property, they’re anti-property, a variety of special authorization to extort, no different in principle from the pre-Gandhi salt monopoly in India.
I wouldn’t say you need to repeal patents entirely. Just limit them better, and enforce those limitations. Same with copyrights.
It seems obvious that current patent laws are too strong/bad for business, but the concept of patents does serve the useful purpose of encouraging innovation, even as they prevent innovators from actually going through with it.
There has to be a spot where the laws reach an optimum between encouragement and prevention, and it would be very surprising if it’s at zero.
There is a suggestion in this article about how prices for patents could be set by the market without the anti-competitive downsides of current patent law. Like most of Landsburg’s stuff, I’m not sure I agree with all of it, there are some flaws which seem likely to undermine the main case, but it’s certainly not a ridiculous suggestion.
The Georgist idea seems to have a place here.
For both land and patents, let the owners float the properties in the market and price it themselves. They will be taxed based on the value that they themselves place on the patent/property.
If the usage is truly efficient, they wouldn’t mind paying the tax. If it is not, then they themselves will mark down their property/patent value, thus allowing others to buy them out.
Ofcourse, the hidden assumption here is that the government is the real property owner (of all properties in its borders) at all time, which in a de-facto sense is true today, even if it may not be true de-jure. (If the government came to grab your property, would you really have any option against them?)
So how do you create an incentive for a company to do private research and development?
This sounds like a great idea for many cases, but don’t you think that the transactions costs will sometimes be a barrier?
Even in a world where people would be universally willing to do this, it may fail in the case where a unified plot of land is required for a building or project of whatever kind (as in the recent Kelo vs. New London eminent domain case).
Also, “routing around” a particular house may be difficult; it’s made impossible if a group of people in the way band together and collectively hold out to take the builder’s surplus.
As is all too common, though, with eminent domain, often the government is (surprise!) not very good at playing real estate developer, and the land that is seized is not put to very good use. A great example of this is with the property that was taken in the Kelo case, which to this day remains undeveloped.
Ironically, the government wasn’t actually the real estate developer in this case. In fact, this was the point of the Kelo case; government has long had the power to take land for public use, and the case was about whether the government could act as something of a coercive coordinator for collecting land for private development.
Of course, had the case not taken several years to resolve, who knows what might have happened.
There are plenty of these anti-commons cases in real life, however; in my Law and Economics class we discussed a similar one in Russia, where nobody occupied the stores in a particular street because the requisite property rights were too difficult to collect, so everybody was forced into stalls right in front of the stores.
It wouldn’t fail where you need a unified plot. It would just require you to spray out a wide cloud of options until some fully-connected subset of them could contain the required shape.
It might fail in the case where a very large number of people collaborate, or a very scarce resource is a show-stopper. But in that case, is it truly unfair? Or are they just setting a market rate?
Well, as long as you buy enough options, you have a reasonable chance of finding such a unified plot in what’s available to you, right?
One problem with eminent domain is that it doesn’t insure optimum use because the only feedback mechanism is through democratic processes, so it’s just a public choice problem then. With eminent domain you will probably destroy underuse and just replace it with overuse.
Tyler Cowen reviews The Gridlock Economy here. Alex Tabarrok reviews Against Intellectual Monopoly here. I was sympathetic to the point the latter was making, but I liked Alex’s contrasting point about finding an optimum.
This is degenerating into mere politics. The part that was interesting to LW was:
Do you mean the assignment of a value of life or the general principle of assignment of values to everything? In either case, both of those seem sorta like no-brainers (which, I imagine, is why no one is discussing them).
It seems to me that the most relevant thing in this post was the idea of a bias against recognizing underuse in general. It actually reminds me of when I was introduced to Robin’s idea of the danger of excess medical care in that most people (myself included, at the time) had a bias against recognizing the harms done by extra treatment.
You’d be surprised. There are plenty of people who refuse to assign a value on life, feeling that doing so would somehow diminish a life’s importance, or be immoral. (This type of person seems to be overrepresented in the humanities...)
Well, it also seems like a no-brainer to me the Breatharianism is insane, but I know people certainly subscribe to it. What I meant by that was more that it seems well-established among LessWrong readers.
At least the seasoned Less Wrong readers. But we should never write our posts only to the established audience—newcomers to the site should also find something of value.
I don’t agree that by failing to put a value on life you necessarily also fail to discover the concept of underuse. Doesn’t it follow immediately from the fact that you can have a positive externality that you would necessarily also have underuse?
Didn’t say it was an absolute—yes, you can certainly discover the concept of underuse even if you refuse to put a value on life. But your odds of doing so are lower than if you would have.
I’ll agree that they are lower, but I am not sure that they are significantly lower. It seems to me that ANY positive externality would be evidence for underuse and you can think of a large number of them without ever putting a value on life.
That said, I do think that it is obviously important to put a value on life so that you can do cost-benefit analyses.
Something interesting about the tragedy of the anticommons is that it is created entirely by government. Intellectual property, spectrum allocation, and zoning/land use regulations would not exist in an anarchic world. I’ve heard incredibly cogent and compelling arguments against each type of government-created “market,” and it’s a damn shame that people look at you like you’re a crazed lunatic when you suggest that either a) people shouldn’t be granted monopolies on ideas and mathematical equations (i.e., computer code); b) the government should not be regulating the spectrum at all (god forbid we’d live in a world where all communication were as plentiful as wi-fi is today!); or b) maybe people should be allowed to build gigantic skyscrapers wherever the hell they want without parking. The last one is particularly pressing (in my opinion), though it’s upsetting that the closest that people get to recognizing this particular tragedy of the anticommons is supporting New Urbanism, which is at best a bastardized version of the most truly efficient urban allocation mechanism – market urbanism.
“Something interesting about the tragedy of the anticommons is that it is created entirely by government.” Watch out for the affective death spiral. If we had no zoning regulation and no restrictions on contractual disposition of land property rights, we’d get a crazy-quilt of restrictive covenants.
Covenants can only exist if there’s a party to enforce them. And that party has to be paid, since preventing people from doing something costs money. Unfortunately, often the government takes that job upon itself (as in Houston).
A more realistic and free market-oriented approach would be to force people who want access to their neighbor’s property to...buy it! Sure, it would encourage people to own larger plots of land, but given the inherent perils of homeownership, I’m not entirely sure that converting a nation of homeowners to a nation of renters would be such a bad idea.
It sounds to me like you are ignoring the Tragedy of the Commons there, though. The purported reason for each of these government interventions is to enforce property rights where they don’t exist. I think the whole point of this post about the tragedy of the anticommons is to illustrate that you are finding an optimum, not a single limit.
The fact that all of these things mentioned here are created by government (and I am not sure that you’ve proven that tragedies of the anticommons can’t arise naturally) just gets to the point that you can easily over-correct for a failure of natural incentives, which means that you should probably be putting some thought into designing feedback mechanisms to naturally find the optima that you are looking for.
The purported reason for each of these government interventions is to enforce property rights where they don’t exist.
No doubt. I’m not ignoring the tragedy of the commons, though – I’m just saying that I have yet to see compelling evidence that it actually exists.
A free-for-all spectrum might have some interference, but then again, the concept of “interference” is a bit dated anyway, as modern technology can much more effectively filter it out than it could when the Titanic sunk (which was when we got a lot of these spectrum regulations in the first place).
A free-for-all in land development is not a great example of the tragedy of the commons, since it is not, strictly speaking, a commons. Sure, the “skyline” could be considered a commons, but then again, given astronomical land values in dense urban settings, it’s a bit unclear as to why a cluttered skyline would be such a bad thing. (The roads could be considered a commons, though given the deteriorating state of the environment and the very workably private mass transit companies that existed before the socialization of transport in the US, it’s difficult to argue that government-owned transport in the form of “road commons” is such a great idea.)
As for the IP argument, people have been arguing about this for a good long time. The best resource that I can recommend arguing against a tragedy of the commons is Against Intellectual Monopoly by Levine and Boldrin. For a brief summary of some of the argument against IP in pharmaceuticals (which is by far the field in which IP has the most support), see here.
Curious about the downvoting of parent. It’s concise, informative, well-argued (in a sense), and contains references for further reading. It therefore seems like a candidate for the top 20 or so comments we’ve had so far.
I may vote for Libertarian political candidates, but that doesn’t mean that I think every single damn website in the cosmos has to be about libertarianism.
If you’re sympathetic to libertarian ideas, I’m surprised you’re not interested in furthering discussion about libertarian approaches to spectrum allocation and land use and transportation policy. What mainstream websites/blogs/pundits do you know of who give anything more than a passing mention of these issues? I find that these issues in particular (IP not so much) are woefully neglected by...just about everyone. Even libertarian organizations (Reason magazine and foundation come to mind) often have very statist stances when it comes to land use and spectrum regulation.
“A fanatic is someone who can’t change his mind and won’t change the subject.” Not every conversation on the Internet has to be about libertarianism. Anyone who can’t accept that about any of their pet topics is a commenter we can’t afford to have.
Agreed. Discussions of dogmatisms such as libertarianism in any of its forms is boring, boring, boring and now what I’d like to read here.
Not that the original post here is an example of that, per se, although I think it comes close.
For the record, while I do have sympathies for libertarian ideas, I consider myself closer to the socialist end of the spectrum.
Notice you didn’t mention ‘human rationality’ at all.
From our vague semi-non-official guidelines:
Are you saying that there is no incidence of the tragedy of the commons at all, or just that these things are not tragedies of the commons? If it’s the latter, I think it’s pointless to argue the specifics of any particular examples when the broader point still stands. When there is a tragedy of the commons, one possible solution is to create property rights so that incentives align with social optima, but the problem of the tragedy of the anticommons can arise if the property rights you create are too strong.
In practice, there will be cases where you don’t want to try to re-align incentives. If you have a situation where you are going to be naturally close to the social optimum (maybe the spectrum or the skyline are good examples of this—I’m not familiar with these cases intimately), then unless you have a well-calibrated government you are more likely than not to over-shoot the social optimum. If you have something that’s seriously misaligned—maybe people burning huge amounts of neurotoxin-containing wood and wearing a mask or something—you might overshoot or undershoot the social optimum, but even a poorly-calibrated government might be able to get you closer.
I wouldn’t want to be so absolute and say there is no such thing as the tragedy of the commons, but I am saying that I think it’s vastly overblown, and that in most cases when it’s invoked, it isn’t actually present.
You’re right about that: Many people use the spectre of the tragedy of the commons ad nauseum to sound clever and insulate their own ideological attachments to property rights and private ownership. Usually, when there’s a commons, some sturdy, democratic, responsive trustee entity or market can take care of problems of overuse.