In his famous 1968 article in the journal Science, Garrett Hardin illustrated his notion of the "tragedy of the commons" by suggesting, "Picture a pasture open to all. It is to be expected that each herdsman will try to keep as many cattle as possible on the commons. But of course, argues Hardin, all other herdsmen will have the same goal. The result is overgrazing which destroys the nurturing pasture and starves all the cows. "Therein is the tragedy," asserts Hardin. "Freedom in a commons brings ruin to all." Only centuries of "tribal wars, poaching, and disease" kept the tragedy at bay.
Hardin's tragedy is based on the logic of ruin embodied in the game theory concept of a prisoner's dilemma. In prisoner's dilemma two prisoners are questioned separately and if neither confesses then both will go free. However, if one confesses, he will receive a lesser sentence than the other who remains silent. If both confess, then both are severely punished. The prisoners' optimal strategy is to remain silent and both go free. However, not knowing what the other will do, the best individual strategy is to confess, which results in the worst outcome, punishment for both prisoners.
Hardin overlooked the fact that herdsmen are not like isolated prisoners; they can talk to each other. And as economics Nobelist Elinor Ostrom and her colleagues make clear in a recent Sciencearticle, "Lab Experiments for the Study of Social-Ecological Systems," talking makes a big difference.
To highlight the crucial importance of communication in preventing the tragedy of the commons, Ostrom and her colleagues devised a series of laboratory games. In each game, five players seek to harvest resource tokens from a computer grid using their on-screen avatars in the course of six four-minute play sessions. The initial 210 resource tokens reproduce if left alone long enough, offering the patient players a possible optimal harvest of 665 tokens in each period.
The researchers divided the play periods into two sets of three sessions each. At the beginning of some four-minute session periods, players could not communicate with nor engage in costly punishment of other players. In that case, the researchers report that "the resource was consistently depleted within about 90 seconds, confirming that without communication or punishment, the 'tragedy of the commons' prediction of Hardin is supported." The result was that players got to stare at an empty screen for two and half minutes. Even more dismaying, players learned that forbearance was a fool's game, so resource depletion sped up in each subsequent session. The results, respectively, were that groups of players collected an average of 292 tokens, 265 tokens, and 274 tokens, far less than the 665 that would be optimally available in each session.
But what happened when players were allowed in the second set of three sessions to either communicate, punish, or both? For example, when communication and punishment was allowed in the last three sessions, the groups that averaged 292 tokens boosted their harvest to an average of 451 per session, for an overall average take of 743 tokens. Just as hoped, making and keeping credible threats prevents free-riding and increases cooperation.
Lots of earlier research has found that allowing players to punish free riders can increase overall payoffs. However, Ostrom and her colleagues found that without communication, punishment alone can be counterproductive. For example, when the games began with punishment but no communication followed by three rounds of neither, the players harvested an average of 280 tokens in the punishment rounds, and 256 in the no communication or punishment rounds, for a pitiful overall average of 536 tokens. Even worse, a series of sessions that began with neither communication nor punishment followed just by punishment, resulted in harvesting an average of 265 in each of the first three sessions, falling to 247 per session in the punishment rounds, resulting in an average take of only 512 tokens.
Punishment alone made things worse. Why? The researchers suggest that the problem is that "receiving a sanction does not carry a clear message." Does the sanction relate to the amount, the speed, or the pattern of harvesting? "When communication is possible, punished participants correct their harvesting rate by slowing down," note the researchers.
In fact, the researchers found that when first three sessions involved both communication and punishment followed by three sessions with neither, the average harvests were considerably higher than when neither communication nor punishment were permitted. In that case, harvests average 402 in the first three sessions and 331 in the following three sessions, resulting in an overall average of 733. This is slightly lower than the results when the first sessions do not permit communication or punishment were followed by sessions allowing both, 743 versus 733, respectively. In both cases, it is clear that communication improves overall harvests of a common resource when combined with punishment.
But an intriguing result of the research is that communication by itself boosts overall gains even more. When researchers allowed five sets of players to begin their first three sessions by communicating, the average harvest was 441 tokens, the highest average of any opening set of sessions. When the players were not allowed to communicate nor punish in the following three sessions, average harvests fell slightly to 415 tokens. The overall average harvest was 856 tokens, more than 100 tokens better than the next best result of 743 tokens.
Interestingly, the researchers did not run six full sessions that included just communication versus six full sessions combining communication with punishment. Assuming the average harvests remained the same for each treatment, communication plus punishment would edge out just communication by 902 tokens to 882 tokens. If the lowest result from games that forbade both communication and punishment were assumed for all six sessions, the average harvest would be just 530 tokens. In other words, communication alone boosts overall harvests by 66 percent.
The good news is that research shows that just talking can make Hardin's logic of ruin anything but inevitable. In fact, historical research shows that Hardin's overgrazed meadows are rare. For example, the tragedy of the commons didn't occur in Medieval England because local herdsmen negotiated a set of rules (communication) and established enforcement mechanisms (punishment) to allocate access to scarce pasturage among themselves.
Elinor Ostrom and her colleagues have repeatedly found the same thing in their field work. All over the world, local people talked among themselves and worked out serviceable rules for protecting and benefiting from common pool resources, like streams, forests, and fisheries. Take the famous case of the harbor gangs among lobster fishers in Maine. Although the state government says that anyone is legally permitted to catch lobsters commercially, the harbor gangs restrict access by outsiders by cutting the buoy lines to the traps set by interlopers. This informal management results in a more sustainable fishery and boosts the incomes of the local fishers. A study published in theProceedings of the National Academy of Sciences last fall found that local communities are much better at managing forest commons than are governments. In contrast to government management, local communal ownership boosted incomes and forest sustainability.
Ostrom previously noted that large studies from "around the world challenge the presumption that governments always do a better job than users in organizing and protecting important resources." In fact, a 2002 study correctly noted, "The overall state of the world's fisheries is much worse today than 45 years ago, even though most fisheries have come under government regulation in this period." By preventing local people from talking among themselves, it is often the case that governments actually create prisoner's dilemmas over resources that result in the tragedy of the commons.
Ronald Bailey is Reason's science correspondent. His book Liberation Biology: The Scientific and Moral Case for the Biotech Revolution is available from Prometheus Books.
