Market Failure 4: Collective Action Problems (Tragedies and Catch 22s)

James K

James is a government policy analyst, and lives in Wellington, New Zealand. His interests including wargaming, computer gaming (especially RPGs and strategy games), Dungeons & Dragons and scepticism. No part of any of his posts or comments should be construed as the position of any part of the New Zealand government, or indeed any agency he may be associated with.

Related Post Roulette

67 Responses

  1. gingergene says:

    I’m confused. If in the Public Good Game the resource contributed are doubled, then why are there any common resources to distribute when none of the players contribute any of their resources? Or are there 20 common resources to be distributed no matter what?Report

    • North in reply to gingergene says:

      In that case the “payoff of 5” that the players receive is their starting 5 resources out of which they contributed no resources to the common pool. Think of “payoff” as being the total resources they have on hand at the end of the game rather than what they get back from putting into the pool.Report

      • Vikram Bath in reply to North says:

        What North said. If no one contributes to the common good, everyone still gets to take home what they started off with: 5 unites each.

        It’s worth looking at the example where one person does choose to contribute everything to the common good. Then, hier five is doubled to 10 and that 10 is then redistributed across all players including him. So, her take is then 10/4 = 2.5. The other players meanwhile have the 5 they started off with, and they get the 2.5 from the publicly-minded person. So, they go home with 7.5. So, the person who did the “right” thing for the group is worse off than everyone else and worse off than if she had just held on to her 5 to begin with.

        That Catch-22 James leads off with can’t be topped.Report

      • gingergene in reply to North says:

        Ah, I see. Poor reading comprehension on my part. Thanks.Report

        • Yhwh in reply to gingergene says:

          I think it’s a failure of the writing. Everyone starts with 5, if nobody contributes, everyone ends with 5, so there is no profit. In the equilibrium of the game everyone has 5, nobody contributes, and everyone’s profit is 0. If only 1 person contributes 1 to the pool, then 3 people get a 0.5 profit and the person who contributed takes an 0.5 loss from starting position.Report

          • James K in reply to Yhwh says:


            Yes, sorry I should have been clearer on the definitions. When this game is studied in t he lab, the 5 units of currency are provided by the researchers, so any money the players have at the end is a net plus. If you were studying the situation when each player is paying out of their own pocket, you’d use a different baseline.Report

  2. Oscar Gordon says:

    So hunting seasons are bad?Report

    • North in reply to Oscar Gordon says:

      Aren’t they access restriction imposed by the state? I a big fan of the state having a central roll in barring trajedy of the commons’ situations so I’m all for hunting seasons, fishing seasons, quotas etc.Report

      • Saul Degraw in reply to North says:

        Much better to have a central roll than an act of Enclosure that takes people’s land and gives it to the rich,Report

        • Kolohe in reply to Saul Degraw says:

          Either case has different problems baked in.Report

          • Saul Degraw in reply to Kolohe says:

            I don’t doubt this. I just prefer the problems with a central roll over the problems with a privatizing act of Enclosure that benefits the already rich and capital-heavy.Report

            • Mmmmmm….a central roll…with butter 🙂Report

            • More seriously, I guess it depends on what you mean by central role. If we’re talking something like a pure (or approaching pure) command economy, then I’d say even a series of private Enclosure Acts is better, although still unjust. If we’re talking about coordinating use rights or imposing quotas, etc.–which I presume is close to what you actually meant–I’m more on board. But whether the optimal locus for the role is “central” or more “local” might depend on the issue or the resource at stake.Report

          • North in reply to Kolohe says:

            Agreed, I always come at it from an Eastern Canadian background. Generally we witnessed laisse faire economics annihilate the eastern fisheries with government taking a more interventionist position only at a very late juncture. With some commons the resource is so vast that a state makes the most economic sense for enforcing quotas or restrictions. For a private entity to try and claim, apportion and enforce ownership rights over the grand banks fisheries would be flat out uneconomical. Tunas are busy merrily swimming down the path to commercial extinction for basically the same reasons as much of their habitat lies in international waters; a true stateless environment.

            So yeah I think the state has an important role in preventing the tragedy of the commons. But I grant that without market feedback mechanisms they’re always going to be chunkier and less “efficient”.Report

      • Oscar Gordon in reply to North says:

        From the OP:
        Equipment or season-length restrictions can turn into an arms-race between regulators and the regulated with farcical results.

        This is how sport hunting & fishing is done across the country & elsewhere. Seems to work well enough, so I’m curious of the farce aspect.Report

        • North in reply to Oscar Gordon says:

          I think your question answers itself. “Sport” fishing doesn’t adhere to the same economic laws because it’s functionally a service or an experience industry rather than a resource extraction industry. The sports fisherman’s marginal utility is not directly connected to the number of fish they collect* the way a commercial fisherman’s is.

          *I mean if they don’t catch any fish they’re going to be pissed, but after they’ve caught a certain relatively modest number they’re not getting a lot more enjoyment or benefit.Report

          • Yhwh in reply to North says:

            Sport fishing doesn’t have to ban highly technological rods or gear, because the sport fisher is still not trying to take all the fish in the pond. Commercial fishing needs to ban things like drift nets because otherwise one person can come in and take up not just 100% of the fish stock but a lot of endangered species and other bycatch. That does immense ecological damage.Report

        • James K in reply to Oscar Gordon says:


          I remover hearing of a commercial fishery (I think was in Alaska) where fishing was time limited. Fishers responded by investing in bigger boats, so the season was further shortened. The whole thing ended in a one-day fishing season.

          Sport hunting / fishing tends to have a lower impact on the resource, and equipment arms races are less likely to occur with hobbyists. Also, there can be other reason for imposing a limited seasons, like protecting juvenile animals from hunters.Report

          • Yhwh in reply to James K says:

            The sane response in that situation would be catch limitations attached to the license and allotted from an overall pool of how much it was deemed safe to take from the water. I would hazard a guess that the fishery in question didn’t have licensing limitations and allowed a functionally unlimited number of people to participate in the fishing season for some reason.Report

  3. Saul Degraw says:


    Would I be wrong to say that the English Acts of Enclosure were early attempts at trying to forestall the Tragedy of the Commons?
    They also had the effect of taking away game and farming land from the people and giving it to the already wealthy and landed aristocracy.
    How do you avoid the Tragedy of the Commons without also enriching the already rich?

    Do you see the same issue with when biodiverse ecologies are replaced with monocultures because of market demand? For example,
    rain forests becoming just palm trees because of the unending market demand for palm oil. How do you prevent private companies
    from taking their land and turning into monocultures? Wouldn’t you need “common land”* to encourage biodiversity?

    *I consider a public/state park to be a variant of common land.Report

    • North in reply to Saul Degraw says:

      The law locks up the man or woman
      Who steals the goose from off the common,
      But lets the greater felon loose
      Who steals the common from off the goose.Report

    • Murali in reply to Saul Degraw says:

      Isn’t state ownership different from being in the commons? After all, when a park is in the commons, no one can tell anyone to get out or not dig up the grass etc. But when the state owns it, they can close the park for maintenance, or tell people to stay off certain patches of grass or not pee in the lake etc. The ability to exclude is an important feature and what remains is divvying up those exclusion rights. Depending on how we want things to work out, and depending on what the situation on the ground is, the set of people who are said to own something can be larger or smaller.Report

    • Brandon Berg in reply to Saul Degraw says:

      How do you avoid the Tragedy of the Commons without also enriching the already rich?


      • LeeEsq in reply to Brandon Berg says:

        You would still need money or at least resources to get a loan to participate in the auction.Report

        • Troublesome Frog in reply to LeeEsq says:

          In theory, you could just allocate the resource randomly. Or give it to the poorest of your society as a form of welfare. Once a single owner is responsible for it, the problem goes away, regardless of how it ended up in the hands of that owner.

          One example is water in California. We have an oddball way of allocating it by historical accident which currently makes no sense, but it if turned out to be easy to trade water credits (within the limits of water transportability, of course), that initial stupid allocation doesn’t affect the ultimate efficiency of the allocation. If you have a huge water credit and don’t need all of it, you just sell it and it ends up somewhere where it can be put to better use.Report

          • @troublesome-frog

            Or auction the asset to the highest bidder and use the proceeds for welfare.Report

            • DavidTC in reply to James K says:

              I’ve suggested that instead of carbon cap and trade, what we should do is divide up the amount of carbon we’re willing to put in the air by every single adult citizen, and then let them indicate on their taxes if they want it auctioned by the government the next year, and they get the money back on their taxes that year.

              Or, alternately, they can keep ‘their’ allowed carbon emissions(1), and get a certificate back with their tax refund saying they’re keeping theirs.

              It gets even more interesting if environmental groups start *buying* those certificates. Not via the auction system, although obviously they could do that already…but if the auction was constantly giving, say, ~$50 a person-unit, the Sierra Club might offer instead, for example, a year membership and a tote bag if you give them your carbon emission permission, which they will do nothing with. (Think of it sorta as non-profits offering a lower bid, to get people who might not be willing to get *nothing* for their carbon emissions.)

              1) Please note this would have nothing to do with the actual amount of carbon *people* emit, which generally is not regulated.Report

      • Brandon Berg: Auction.

        You would need a mechanism that prevents those who didn’t win the auction from going ahead and using the resources anyway. (But that is at least sometimes achievable.)Report

    • James K in reply to Saul Degraw says:


      One of the reasons why I’m not keen on “Tragedy of the Commons” as a name for Open Access problems. Common ownership isn’t really the problem.

      Consider the issue of an overused village common. Enclosing it and giving it to a wealthy landowner certainly fixes the inefficiency, but it may not be to your tastes from a distributional standpoint.

      Other alternative include:
      1) Enclose and auction off the commons and giving the proceeds to the villagers who used the common.
      2) Enclose the common and transfer it to a newly-formed corporation. The villagers who use the commons would be given equal shares in the corporation.
      3) Introduce a permit system restricting on how much each family can use the commons.
      4) Institute a village council empowered to restrict use of the commons to a sustainable level.
      5) The villagers could establish norms as to what an appropriate use of the commons is, and use shaming or other social pressure on those who abuse the commons.

      You’ll note that some of these solutions involve common property and some don’t, but all of them close access to the commons in some way.Report

  4. Francis says:

    Lots of public goods are rivalrous. Off the top of my head: roads, schools (also excludable), parks (also excludable), libraries (ditto).

    Do public utilities provide public goods? Are things like water, power, and sewer just monopoly goods (that can have enormous externalities) or are they properly conceived of as public goods?Report

    • Brandon Berg in reply to Francis says:

      “Public good” has a specific technical meaning, namely that it be neither excludable nor rivalrous. It’s not just another way to say “Thing I want the government to pay for.”

      Also, note that private roads, schools, parks, and libraries do exist, and function quite well.Report

      • Saul Degraw in reply to Brandon Berg says:

        Yes. All these things exist and some do function quite well. That does not negate the need or morality/ethics and good for
        public schools, public libraries, public roads, public parks, etc.

        There are also public schools, libraries, roads, parks that exist and function quite well and are major attractions and destinations such as the New York Public Library, Central Park, Golden Gate Park, Cal, UCLA, UVA, UWashington, etc.

        There is a small number of radicals that would ban all private schools and the like. Liberals are merely arguing that there is
        good in also having public schools, libraries, and parks.Report

        • Brandon Berg in reply to Saul Degraw says:

          That does not negate the need or morality/ethics and good for
          public schools, public libraries, public roads, public parks, etc.

          That’s not a real thing.

          I know you loves you some redistribution, so I’m not going to argue that, but that’s entirely orthogonal to the question of how these things are provided. There really is no compelling reason to have them provided directly by government. Government gives people money/vouchers to buy privately provided food, clothes, medical care, housing, and so on and so forth, and no one really questions this, because that’s the way it’s always been done.

          But since we have a long history of having government provide education directly, then to a certain kind of person it’s unthinkable that we should do it any other way. Except at the university level. There it’s okay. Again, because that’s how we’ve always done it.

          There are also public schools, libraries, roads, parks that exist and function quite well and are major attractions and destinations such as the New York Public Library, Central Park, Golden Gate Park, Cal, UCLA, UVA, UWashington, etc.

          That’s all well and good, but totally irrelevant to the point at hand, which is that these things are not public goods. The fact that government provides something does not make it a public good.

          Technically private provision does not make something not a public good—they can be provided on a philanthropic basis, for example—but in general the fact that something can be provided privately and supported by usage fees is a sign that it’s probably not a public good.Report

          • Saul Degraw in reply to Brandon Berg says:

            Ah, turning liberals into conservatives! What a great rhetorical tool.

            Yes, “the way thing’s have always been done” is not great logic some of the time but that doesn’t mean it is always bad logic or thought. There could be very good reasons why something was always done a particular way. In this
            case, public provided parks, libraries, schools, etc. do provide the most universal access.

            They are also relatively historically recent in my opinion.Report

      • Francis in reply to Brandon Berg says:

        If being non-rivalrous is an essential element of public goods, then there are very few public goods.

        For example, the judicial system is both rivalrous (courtroom time has to be allocated somehow) and excludable (there exist mandatory private judiciaries, eg arbitration).

        Is a police force rivalrous? Seems to me that it is — patrolling one neighborhood deprives the other neighborhood of the benefits of the patrol.

        Perhaps it is useful to have a definition of goods in which the sole qualifying good is national defense. But I have not been persuaded.Report

        • Brandon Berg in reply to Francis says:

          Innovation is a public good (copyright and patents make it excludable, although it becomes public again when it enters the public domain). In-the-clear broadcast media is a public good, although encryption can make it rivalrous. Clean air is a public good. The classic example is a lighthouse.

          Also, some goods have a public-good component without being completely public. Vaccines, for example, are private goods, but the herd immunity that they create is a public good.Report

          • Francis in reply to Brandon Berg says:

            Just for the sake of argument,

            IP protection around invention is an exception so large as to swallow the rule. There is a time value to invention and a lot of people think that 20 years is way too long.

            Broadcast media is very much rivalrous, or FCC auctions wouldn’t be necessary.

            Air that is so clean (and water so pure) that no one is sickened appears to fit the definition of public good (and justify the existence of the EPA!), but do we risk getting into definitional fights over the use of the commons vs the provision of public goods, as opposed to debating the science?

            My very limited point, fwiw, is that I think that imposing such a strict definition on public good may impede rather than promote useful conversation.Report

            • Guy in reply to Francis says:

              IP protection around invention is an exception so large as to swallow the rule. There is a time value to invention and a lot of people think that 20 years is way too long.


              Innovation is a public good (neither excludable nor rivalrous), so we create the IP system to make it temporarily excludable. This supplements an application of the solution proposed by the OP, which is government funding of innovation via, for example, the NSF.

              Basically, following the warning in the OP about governments not knowing exactly what the optimal solution to the problem is, we make some of what would have been a public good into a market good. The principal is sound, although the practice is a bit screwed up because times have changed and regulations have been captured since the system was set up.

              Also, broadcast media is not rivalrous. Anyone can pick up your signal. However, it is excludable, because any band you broadcast on will interfere with a rival company’s attempt to broadcast on that same band. Thus they will not rationally compete with you if you manage to capture all the bandwidth.Report

              • Guy in reply to Guy says:

                The thing with broadcast media is that you can’t (naturally) stop people from producing devices that will pick up your signals.Report

              • Michael Cain in reply to Guy says:

                But will they be able to do anything with it? Back in the days when content was simple analog, there were limits to how much it could be concealed and still be useful. Now that things are digital, there are a number of strong encryption schemes with fancy key management and device enabling that make decryption in real time not impossible, but prohibitively expensive. Unless you use one of the approved devices for decrypting the signal.

                Sirius broadcasts on known frequencies using known modulation schemes, so it’s pretty easy to access the bit stream. Doing something useful with it is a much harder problem.Report

              • Guy in reply to Michael Cain says:

                Oh, yes, I’m just talking about unencrypted traffic. I should have made that clear.Report

            • Brandon Berg in reply to Francis says:

              @francis Regarding IP, what Guy said. Innovation is a public good. Public goods are problematic because they tend to be undersupplied by market actors, so patents and copyrights were created to solve that problem.

              Use of the broadcast spectrum is rival, but consumption of the broadcast is nonrival. When you tune your radio to a particular station, that doesn’t weaken the signal for everyone else.

              Correction: I said encryption can make broadcast media rivalrous. I meant excludable.

              My very limited point, fwiw, is that I think that imposing such a strict definition on public good may impede rather than promote useful conversation.

              Why? Public goods have certain characteristics that cause them to be undersupplied by market actors. Goods that don’t have these characteristics aren’t public goods, There are also club goods (excludable but nonrival) and commons (rival but non-excludable), but these are different from public goods, and have different problems with different solutions. It does no good to conflate them. If there aren’t a lot of public goods, then there aren’t a lot of public goods. If so, that’s a good thing, because neither markets nor governments are good at providing the optimal amount of them.Report

        • James K in reply to Francis says:


          The police and judiciary are not simply providing patrols or courts, they are providing a region of reduced violence – that is the real point of police and court systems. Adding one more person will not increase the cost of providing that bubble of relative safety, and in economics its cost at the margin that is important.Report

          • Brandon Berg in reply to James K says:

            Adding one more person doesn’t increase the cost, but only because the impact of one more person is too small to measure. If we add, say, a thousand people, that should have a measurable impact.Report

    • Autolukos in reply to Francis says:

      None of those qualify as public goods in the economists’ definition. Several have positive externalities (sewers have particularly large positive externalities), but all of them are both rivalrous and excludable.Report

    • Vikram Bath in reply to Francis says:

      Francis: Lots of public goods are rivalrous. Off the top of my head: roads, schools (also excludable), parks (also excludable), libraries (ditto).

      Why do you classify these as rivalrous? If roads are rivalrous, they are only weakly so. If I use the road, it only very slightly impairs your ability to use them. And even when a whole lot of people are using the roads at the same time, it rarely prevents you from using them at all.

      Rivalry does exist on a continuum, but roads definitely are on the non-rivalrous end whereas, say, pants are on the other.

      Similarly, I can go to a school, park, or library even if you were there before me.Report

      • Brandon Berg in reply to Vikram Bath says:

        Roads are very rivalrous at peak times.Report

      • Francis in reply to Vikram Bath says:


        Schools, parks, libraries, road and especially roads in Southern California fill up. They are limited resources which present significant allocation issues (frequently but not always solved by congestion).

        If a public good is defined as a good for which there are no allocation problems (which is another way of saying non-rivalrous), then none of those goods are public goods.

        If there are degrees of strength in public goods, with national defense at one end and Los Angeles freeway system at rush hour at the other, I have no quibble with that approach. But that’s not the definition given either by our gracious host or by B. Berg above.Report

    • James K in reply to Francis says:


      Schools, parks and libraries are not public goods. The argument for the government getting involved in them is more about positive externalities than Public Goods. Note that this does not give government any particular reason to run them directly.

      Public utilities and roads are more like natural monopolies, which could justify government ownership.Report

  5. DensityDuck says:

    From several years back, an OG post about the Tragedy of the Commons. I’m linking it just to show the site history, I’m not tying it to specific things in this post.Report

  6. Patrick says:

    I think this is a very good comment thread upon which to plant this link.Report

    • North in reply to Patrick says:

      I don’t know why but on reading it my impression of the buyer was “pathetic, bahahah” which is odd considering how much dough he has and I don’t.Report

      • LeeEsq in reply to North says:

        Martin Shkerli has all the signs of a lonely man. I know this from personal experience as a lonely man with less money. Lonely men are always looking for a thing or action that will magically make them less lonely. I’ve done much less expensive things myself with a similar motivation though.Report

    • Glyph in reply to Patrick says:

      Pharma Bro ain’t nuthin’ to fish with. This story has been making me laugh all day.Report

    • Burt Likko in reply to Patrick says:

      Oh, my goodness, no, this is entirely inadequate. Such news deserves a post of its own.Report

    • LeeEsq in reply to Patrick says:

      It seems that the Wu Tang clan deserve just as much blame if not more for turning their album into a collector’s item rather than treating it as a normal album.Report

      • Glyph in reply to LeeEsq says:

        I don’t know if “blame” is the right word. Look, hip-hop has always been about getting paid. And as artists, they were also making a statement that their art has value. Lots of pieces of art are one-of-a-kind, and sell for millions of dollars to the super-rich. What the Wu did was unusual, and could have been a bad business decision for them, but it’s not inherently blameworthy.

        The fact, though, that the BUYER turned out to be someone who is notorious for having said “this thing that I have to sell is rare and has great value, and I will charge the maximum the market will bear for it, so I can get paid, y’all” is delicious, delicious irony. Yes, he was doing it with a lifesaving drug, and that’s different from a piece of art, but still.

        The buyer was ALWAYS gonna be some rich dude. They just didn’t realize it was gonna be THIS one.Report

  7. Jaybird says:

    As the problem is open access, the solution is to restrict access.

    Now we just have to figure out the best way to pick access restriction enforcers.

    I recommend my brother-in-law, my uncle, and my third cousin (twice removed) to be put in charge of making sure that nobody who shouldn’t be using this resource tries to steal from it (and, by extension, us as a society).Report

  8. Kazzy says:

    I was thinking about this idea of “cooperation” recently while riding the subway. Ideally, if everyone plays by “the rules”, riders will file all the way into the middle of the train and newly boarding passengers will stand away from the doors until all disembarking passengers have cleared the train. If EVERYONE does this, the system works better: more people on each train, easier loading and unloading of passengers, and a more efficient system. But all it takes is a few self-interested players and the system goes to shit. You have the guy who stands right in front of the door and shoves his way on the moment the doors open. You have the woman who boards the train and then stands right by the door, forcing others to squeeze around her to access the open middle. And sometimes that middle stays relatively empty even as people pack like sardines near the doors… because people are afraid they won’t be able to get off… because others aren’t cooperating.Report

    • Jaybird in reply to Kazzy says:

      A high-collaboration/high-trust system can be effed up PDQ by a handful of people attempting to game the system in a visible way.

      As such, it’s easy to move from high-whatever to middling- (or low-) whatever.

      How to move from lower to higher?Report

      • Glyph in reply to Jaybird says:

        I hate to say it, but lower-to-higher probably requires the judicious application of at least some authoritarianism and/or shaming.

        At my daughter’s preschool, convenient parking is kind of at a premium. So some people (and full disclosure: I was sometimes one of them) would often park on this one area that was clearly painted with diagonal yellow NO PARKING lines.

        It was never really clear to me why this area was painted this way, there was plenty of clearance for cars to get around it, and it represented enough “wasted” convenient parking space for about 4 cars.

        Problem is, pretty shortly after some of us started using this area this way, people started parking willy-nilly randomly anywhere they could get reasonably close to the building, in areas of the lot that ARE difficult to get another car around – and in fact they will now completely block one driveway/lane that really should remain clear, so now people have to drive a longer way around, to exit the lot.

        Maybe I’m ascribing too much responsibility to us early cheaters, and this would have happened anyway; but it does seem a bit like a ‘broken windows’ thing, and IMO the only thing that would easily correct it now would probably be the preschool having someone out there, monitoring and scolding.Report