By Louis Putterman

Environmentalism has made major strides in capturing many people’s imaginations over the last few decades.  From the concern of just a few activists, agreement that the natural environment is worth preserving extends at least to a willingness to recycle containers and unwanted paper, awareness of the dangers posed by global warming, and worry about melting polar ice caps, rising sea levels, and more violent storms.  But when it comes to significant trade-offs between our convenience (the door to door luxury of the private automobile) or livelihoods (running industries on cheap fossil fuels), on the one hand, and actions to reduce the environmental costs that our grandchildren and great-grandchildren will bear, on the other, how much sacrifice are we really ready to make?

A paper published in Nature this summer suggests the optimistic conclusion that most people are in fact willing to sacrifice to assure a human future on our planet, but it also contains the important proviso that this willingness may be conditional on assurances that others will do the same.  Moreover, the authors show that successful outcomes might depend on empowering majorities to reign in the excesses of resource over-exploiters.  This puts institutional arrangements on a par with emotional or ethical valuation of the future as a factor that could prove decisive in determining whether our planet is still fit for human habitation generations from now.

Hauser, Rand, Peysakhovich and Nowak* conducted experiments with short-term workers hired on Amazon Mechanical Turk.  In the main treatments, participants were assigned to groups of 5, and each was told that his or her group had 100 units of a resource from which each could take whatever he/she wished provided that the total didn’t exceed 100, with none receiving anything if that total was exceeded.  Being prevented from communicating with one another, a safe strategy was for each to take 20 units.  However, there was a twist: the participants were also told that their 100 unit resource pool would be replenished and passed along to another group of 5 participants, but only if the total they took was not more than 50 units, rather than 100.  For example, if the current 5 individuals each took 10, they each earned the money value of their 10 units, and the experimenters then replenished the remaining 50 unit pool to become a fresh 100 unit pool for a new group of 5.  If the resource was replenished for that next group and its members in turn took 50 or less, the pool was replenished for yet another group, and so on (with a small chance of random termination taking the place of a time discount). The upshot is that while it was a safe selfish strategy to take 20, the corresponding strategy for players who care about the prospects facing future groups was to take only 10 each.

Remarkably, the majority of participants restrained their takings to 10 or less despite the fact that they would never meet those in future groups and could not be personally credited by anyone for the sacrifice. This provides a basis for concluding that the Hauser et al. subjects at least cared enough about future individuals to make sacrifices in order to conserve resources for them. There was a problem, though: in the basic set-up just described, having most people make the sacrifice isn’t enough to conserve the resource.  If 3 out of 5 limit their takings to 10 but the other two take 20, for example, then total takings will be 70.  This is below 100 and therefore goes through in terms of earnings of this group, but it’s above 50 and so it causes the resource to “crash” and be unavailable to future groups.

Hauser et al. then experiment with an alternative treatment in which the amount each member of a 5 person group is given, out of that group’s pool of 100 units, is determined by the median amount that those 5 individually propose to take.  For example, if members propose to take 8, 10, 11, 20 and 20, the median is 11 so each is given 11, 55 in total are taken, and the pool does not replenish for a further generation.  If the members propose 10, 10, 10, 20 and 20, as in the earlier example, the median is 10, so 50 are taken, and the pool is replenished.  A binding median proposal is a standard way of representing majority rule in laboratory decision experiments.

In the median decider treatment, for which large numbers of additional participants were recruited, most individuals in most groups again cared enough about future players to take 10 or less, but under the new institution, this ended up causing pools to be conserved and replenished the large majority of the time, instead of crashing due to one or two selfish grabbers.  In fact, the fraction of players who proposed 10 when they knew that the median proposal would be binding on all was significantly larger than the fraction of players who took 10 when each made an independent decision.  As the authors emphasized, this can be interpreted as indicating that a substantial number of individuals are prepared to make a sacrifice (take 10 instead of taking 20) when they are assured that it won’t be in vain because the choice will be binding on everyone.  Apparently similar individuals didn’t make the sacrifice in the first treatment because they suspected (usually with good reason) that one or two selfish individuals would cause it to be in vain.

The application to world environmental problems is in some ways straightforward, in others more complex.⁺  The straightforward part is that if this small stakes representation of our global environmental dilemma can be extrapolated to a larger scale, it gives hope that the present generation might preserve a livable world for our children and grandchildren, but warns us that we’re likely to succeed in this quest only if we adopt rules that are binding on everyone.  The complicated part is that the institutions for making such rules binding on the world as a whole don’t yet exist, as the experience with the Kyoto Protocol makes clear.  Moreover, even where rule regimes do exist, as on a national level in the United States, political outcomes have thus far fallen far short of what’s needed to match the good results of Hauser et al.’s research subjects.  Is this because most people are not sufficiently concerned about the future when there are much larger stakes than in the experiment, ones perceived as affecting standard of living and convenience in more substantial ways?  Or is it because our political process breeds outcomes skewed in ways ruled out in the experiment, for instance due to the influence of wealth on the political process?

These issues lie close to the theme in my book. Put briefly, we have good news and we have bad news.  The good news is that most people are not as selfish and as narrowly rational as traditional economic textbooks used to suppose, and that this opens up space for successful cooperation, including cooperative efforts to address shared social dilemmas.  The bad news is that institutions matter, and that there remains much to be done to improve existing institutions if we’re to harness “the better angels of our nature” to achieve a fairer, more life-enriching, and more sustainable world.

* Hauser et al., Cooperating with the Future, Nature 511, 220-223, June 25, 2014.

+ Putterman, A Caring Majority Secures the Future, Nature 511, 165-166, June 25, 2014.

Originally published here.

April 17 2016