Prisoners, Traffic, and Their Associated Dilemmas
In my last post (“Why You Shouldn’t Give to that Charity You Like”), I discussed the problems with giving to a charity just because you like it or its cause.
Now, it may not seem like a serious problem that giving to a charity just because you like it reduces the effectiveness of charitable giving in general; especially if you think that your freedom to choose where to give is extremely important.
For the consequentialist who believes we are required to give to charity so long as there are individuals worse off than we are, this is a pressing problem. For those who are giving it means prolonging the need to give; for those who are receiving it means watering-down charitable giving’s ability to help all the people who need help. At some point in the future I’ll use this as a launching pad to explore just what I think our obligations are, but for the moment we’ll put this aside.
Instead I’ll focus on another part of the problem: by each doing what is best for ourself, all benefit less than we could. Or, to put it another way, when everyone focuses on me, the one who pays is we. The effect of donating to charities that we like is a collective action problem.
Collective action problems are the many-person variety of the famous “Prisoner’s Dilemma:”
Cain and Abel are arrested on suspicion of a crime. The police offer each a deal: cooperate and you will go free while the other serves the full time. If neither Cain nor Abel confesses, they will both get 1 year in jail. If Cain blames Abel, he’ll go free while Abel serves 15 years. If Abel blames Cain, he’ll go free while Cain serves 15 years. If they both blame the other, they’ll both serve 5 years. They are in separate interrogation rooms and cannot communicate their intentions to the other person.
If Cain and Abel want the best possible outcome (just 1 year in jail) they should both keep their mouths shut, right?
However, since they can’t talk, it’s hard to know what the other will do. The worst possible outcome for Abel would be for Cain to confess and Abel not to — then he would serve 15 years while Cain got off. But if Abel confesses, he can prevent his going to prison for 15 years. All it will cost him is sending Cain to prison.
Meanwhile in the other room, Cain’s thinking the reverse — if Abel confesses but he doesn’t, then Cain would serve 15 years while Abel got off; if Cain confesses then he can prevent himself from going to prison for 15 years — no matter what Abel does.
Cain’s problem is this: if both he and Abel do what is best for themselves (Each Blames the Other) then they will only achieve the third-best possible outcome (Both Serve 5 Years).
Collective action problems are many-person variants of the prisoner’s dilemma.
Imagine that you live in the suburbs of LA, with a job in the city center. Because no transit routes will take you to work directly, you decide the best option for your commute is to drive yourself. Most everyone else reaches a similar conclusion — driving their personal car directly to work is the fastest way to commute. As a result, the freeway traffic increases and so do commute times. Yet, if everyone had decided that they should take the bus to work, there would be fewer cars on the freeway and the commute by car could have been shorter .
This is a collective problem because you cannot change the situation just by switching from your car to the bus .
In cases of collective action problems, what is best for each (You Take Your Car) results in a worse outcome for all (traffic makes for Longer Commutes). In order to accomplish the best outcome for all (traffic’s decrease makes for Shorter Commutes), each has to settle to do what is second-best (Take the Bus) for them. The fastest possible commute is achieved by having Everyone Else Take the Bus and You Take Your Car; however, such a scenario is unlikely unless you are a dictator, so you must settle for Taking the Bus and getting everyone else to as well.
Collective action problems, where all would be better off if we each made a sacrifice, but where each who make the sacrifice would do so in vain unless all  made the same sacrifice, permeate societies.
For example, we would be far less willing to sacrifice some of our freedom to enact laws if others did not make the same sacrifice (here all are better off by prohibiting some harmful act, even if a few individuals would have enjoyed it). Another name for them is “coordination dilemmas,” because to get the best result for all would require some kind of coordinating force, be it moral, physical, or other.
So back to charity…
When we do what seems best for each of us (by donating to a charity we like because we want to donate to it) we all are made worse-off (because the money is dispersed amongst different organizations and causes and it might be that none achieve the critical mass they need to eradicate what they address) by the continuance of suffering. But if all did what was second-best (or third, fourth, fifth, etc) for each and gave up the privilege of donating to causes just because we liked them, all might be made better-off by achieving the critical mass needed to eradicate what the most important charity attempts to address.
In “Why You Shouldn’t Give to the Charity You Like,” I hinted at different possible solutions. One of these was GiveWell, which attempts to take advantage of moral persuasion to coordinate giving. Many other solutions would be more heavy-handed, and I intend to cover our moral bases before even attempting to address appropriate policies.
What you should take away from this – right now – is that while it may be nice to have the freedom to donate to whatever cause you feel like donating to, doing so may actually be prolonging the need for charity or the suffering that charity is supposed to address. Instead, if you (and everyone else who can give) relinquish some of that freedom and do what is second-best (or third, or fourth, or fifth) for you individually, everyone may be better off.