The Policy Insights Gained by Understanding Social Dilemmas
Understanding mental models is like a shortcut to becoming smarter about understanding the world. One of these is social dilemmas, which provide significant insight into politics and policy.
About a decade ago, I listened to a number of university lecture series and I was struck by how often the concept of the prisoner’s dilemma came up in various disciplines. I had encountered it already in my formal education when I took a class in evolutionary ecology. The above image shows what it is, but it takes a bit more to fully understands its implications.
The prisoner’s dilemma is a thought experiment that imagines two conspirators who have committed a crime. The police don’t have enough evidence to prove the whole case against them, so they interrogate each one. As you can see above, they each have the choice to betray their co-conspirator or to stay silent. The police are prepared to offer each of them a deal where they go free if they betray the other, but if they both betray each other there’s no deal for either of them.
The above boxes show the potential outcomes for the prisoner’s dilemma scenario. It’s clear that the best overall outcome for both conspirators is to both stay silent. However, when each of them is deciding whether to betray the other, their most rational choice is to do so. Each needs to think about what the other might do, and whatever that choice is, betrayal is the best option for each of them. Needless to say, this is very counter-intuitive.
Here’s how it breaks down:
A must think about what B might do. If B isn’t going to betray A, then A’s best choice is to betray B. If that happens, A goes free and B goes to jail for 3 years.
But if B is going to betray A, then A’s best choice is still to betray B! Otherwise, A spends 3 years in jail instead of 2.
Betrayal is always the best choice for A. B is in exactly the same situation, so B’s best choice is also to always betray A.
This is a simple example of game theory, which was conceived by John Nash, the mathematician who was played by Russel Crowe in the 2002 Oscar-winning film A Beautiful Mind. Accordingly, the mutual betrayal outcome of the prisoner’s dilemma is called the “Nash equilibrium” of the game.
The idea that the most rational choice for individuals is the worst choice overall has broad policy implications. The same dynamic exists for a fisherman who knows fish stocks are dwindling:
- He can choose to keep fishing, but he knows the fish stocks will crash soon, destroying his livelihood; or
- He can choose not to fish in hopes that the fish stocks will rebound. But if he does that, then he makes a huge personal sacrifice while the other fishermen keep catching fish, and maybe catch the share of fish he would otherwise have caught. The fish stocks crash anyway.
Either way, the fisherman‘s most rational choice is to keep fishing despite the dwindling stocks.
The fishermen could all get together and agree “OK, let’s not fish for a year or two to allow our fish stocks to replenish. This is better for our long-term livelihood.” This might at first seem like it would work, but it wouldn’t be stable.
For one, it would be difficult for hundreds or even thousands of fishermen to agree not to fish for a whole year, all making a huge sacrifice on their livelihood. But even of they seemed to all agree, some would likely start cheating. The cheaters are making a rational choice: the fish stocks might bounce back if only a few cheaters are catching fish, and the cheaters don’t have to sacrifice their livelihood for a year.
If there’s a few cheaters, then more cheaters will start fishing. They’ll see the early cheaters and think “why should I sacrifice my livelihood when they get to keep fishing?” Soon enough, all the fishermen will be fishing again and the stocks will head right back towards a crash.
The general term for these sorts of collective action problems is “social dilemma”. The word that’s usually used to describe the cheaters in the fishermen example is “defectors”, while those that don’t cheat are “cooperators”.
The policy implications of social dilemmas is huge. For example, it explains why it’s extremely difficult for boycotts to succeed. The same logic for the fishermen above also applies to boycotts. Anyone being asked to participate in a boycott will think “why should I make the personal sacrifice of not buying this product when others will just keep doing it, and my actions alone won’t make enough difference to the company’s profits to matter?”
Indeed, many people are perfectly aware that some of the products they buy were manufactured in ways that might go against their personal values (e.g. clothing made in sweatshops) but will go on buying those products anyway. It’s rational of them to do so, given the fact that making the personal sacrifice of buying a more expensive, or less desirable, alternative won’t change the behaviour of the company.
Similarly, the above logic applies to making personal sacrifices for the sake of the environment. Why bother avoiding driving, flying, and consuming meat (all personal choices that reduce one’s carbon footprint) if it won’t make an overall difference anyway? Others will keep doing those things en masse, and the planet will just keep on warming.
Social dilemmas also demonstrate why libertarianism can’t work. Libertarians don’t want government to ever regulate anything, and some don’t want government at all, full stop. But without government and government regulations, social dilemmas can’t be solved. Sometimes libertarians answer this by suggesting that in their ideal society, people would solve social dilemmas by forming associations with memberships, dues, and enforcement mechanisms. They wouldn’t need government or regulations.
But the associations libertarians conceive of are really just government by another name. Members are like citizens, dues and like taxes, and enforcement is like the police. The solution to social dilemmas is to inject some positive reinforcement for cooperators and/or punishment for defectors. Those mechanisms must come from the top down, one way or another.
There may be some narrow exceptions. The same logic that explains why boycotts are difficult to organize applies to voting. Any citizen’s vote is extremely unlikely to decide an election, and at best will represent a tiny fraction of the votes needed for any candidate to win. Voting takes time and effort. It’s therefore rational for people not to bother voting.
Yet many people regard voting as their civic duty, and do it despite its extreme uselessness to any individual voter, on a personal level. Other factors must explain why this is. One, of course, is that it’s a conspicuous activity. Voters must go out into public to do it. They also might receive a sticker showing they did it. They can take pride in doing it. Friends and family will also give voters accolades for fulfilling their civic duty, and might shame voters if they fail to do so.
The social pressure to vote is a cultural force, which somewhat succeeds at getting people to the polls. It’s imperfect, because voter turnout is often low, especially for more local elections. Still, it demonstrates the power of culture at tempering the negative effects of social dilemmas.
No matter the force that helps solve a social dilemma, it must change the expected outcome of the game for the player, such that cooperation has a expected outcome than defection.
