First principles, redux

I have been reliably informed that my recent essay First principles of political philosophy is too complicated, too prolix and too irrelevant. So let me make another try.

Politics and ethics are fundamentally about negotiating about preferences, which may be in competition or may be compatible. Unlike science, politics is not about finding the truth. (Once we have negotiated the fundamentals, we must, of course, use scientific reasoning to figure out how to most efficiently achieve those fundamentals.)

All politics and ethics can be reduced to game theory. For simplicity, we reduce an ethical dilemma to a set of choices between two people or classes of people. We list one person's choices across the top, we list the other person's choices along one side, and for each pair of choices, we specify the outcome according to the people's individual subjective preferences. When we look at ethical and political problems in this manner, the best outcome is sometimes obvious, but often it is ambiguous.

There are three main kinds of games: win-win, win-lose, and Prisoner's Dilemma. Win-win games are trivial: each person is individually motivated to choose the outcome that's best for everyone, because that outcome is also the best for him- or herself. Win-lose (or zero-sum) situations are harder, but still fairly simple: they are always solved by who has the greater power of coercion.

Because coercion is possible, all politics and ethics are about how we employ coercion, to what ends, and by what methods. In modern society, we typically place all our coercive "eggs" in one basket: The government, the institution that has a monopoly on violent coercion. Many political decisions can therefore be simplified into determining what we want the government to do. In win-lose games, if one side can get the government to choose that side, they will always win.

Which leaves Prisoner's Dilemma games. Understanding the Prisoner's Dilemma is absolutely necessary for understanding sophisticated politics and ethics. All nontrivial ethical situations that cannot be reduced to a win-lose game can reduced to a Prisoner's Dilemma game.

	Cooperate	Betray
Cooperate	($3, $3)	($0, $6)
Betray	($6, $0)	($1, $1)

Here's the problem: If you choose to cooperate, then if I choose to cooperate, I will get $3; if I choose to betray, I will get $6. If you choose to betray, then if I choose to cooperate, I will get $0; if I choose to betray, I will get $1. No matter what you do, I'm better off choosing to betray. And likewise for you: no matter what I do, you are better off choosing to betray. However, if we both choose to cooperate, we are both better off than if we both choose to betray.

An example of the Prisoner's Dilemma is property rights. If we both respect each other's property (i.e. we choose not to steal when we have the opportunity to do so), then we are both better off. But if you respect my property, I'm even better off stealing yours (and likewise for you): I get my property and yours too. But if we both steal from each other, we don't have any property rights at all, and cannot benefit from just leaving our property lying around without it being stolen.

Many kinds of mutual promises are Prisoner's Dilemma games. We both make a promise to each other as to how to act in the future. When the future rolls around, if we both keep our promises, we both benefit. However, I will benefit more if you keep your promise and I break mine. If we both break our promises, neither of us would lose as much as we would have lost if one kept his promise and the other broke his.

One way to reliably achieve mutual benefit in Prisoner's Dilemma games is to coerce everyone to cooperate to their mutual benefit, so no one can choose the unstable "cooperate/betray" outcome. There are other ways than specifically governmental coercion to ensure the "cooperate/cooperate" outcome, but they are so distant from our present-day social psychology as to be unrealistic.

However, if there is some body that has a monopoly on violent coercion, they can use that coercion to coerce some people to always cooperate, and fail to coerce others to do the same. In which case, one person, or side, or party, or class must cooperate, but the other, absent coercion, will betray.

The vast majority of social constructions, from Capitalism to Religion, consist of justifying the coercion of one class of people to always cooperate, while permitting, encouraging or even to some extent coercing another class to always betray. Give your money to the priests or you will go to hell... or be burned at the stake as a heretic. The laity cooperates and tithes, the priesthood betrays (they cannot cooperate; there is no heaven they can assist the laity to enter); the laity is therefore coercively exploited.

Capitalism works in much the same way: The workers must work for the owners or starve: They are coerced into cooperating. The owners simply accumulate the workers' surplus value and share nothing; they are permitted to betray. Absent coercion, capitalism would be impossible; we would have anarcho-syndicalism, where people could "own" only what they themselves actually used to produce wealth. It is only by violent coercion that capitalist owners can prevent workers from employing the means of production to create surplus value for themselves, rather than for the owners.