A Puzzle About Rights?

The only legitimate function of government is the protection of rights. I take this as a central tenet of most libertarian thought.

Person A's rights establish moral limits on the ways others can treat her. Rights do this by establishing duties -- to say that A has a right to phi is to say other agents have a duty not to stop A from phi-ing.

I will assume that if Person B has a duty to phi, then he has a reason to phi. Indeed, I will assume he has a special kind of reason to phi: a moral reason. Moral reasons, as I understand them, have categorical force. If an agent has a moral reason to phi, that reason applies unconditionally, regardless of whatever else is true about the agent.

Other reasons do not have categorical force. My reason to eat this peach ceases to exist if I no longer desire to eat the peach. My reason to stop smoking would cease to exist if I discovered that, through some quirk of genetics, smoking was actually beneficial to my health.

The duties that correspond to rights have categorical force. Person B has a reason (his duty) to respect the rights of person A, even if Person B would very much like to violate A's rights (e.g. to steal his stuff.)

There is much that is overlooked in this analysis, but it will suffice to make my point. Consider two possible worlds:

World 1: Anarcho-capitalism. People make use of private protection agencies to protect their rights. This process is imperfect, and rights are violated quite often.

World 2: Minimal state. The state claims a monopoly on the retaliatory use of physical force. It enforces that monopoly quite harshly. Independents who pursue vendettas against suspected murderers are dealt with in a harsh manner. However, the state is fairly good at its job, and rights-violations do not often occur.

I will assume that anarcho-capitalists will hold that when the state persecuted independents in World 2, it is violating their rights. The state has no legitimate claim to its monopoly.

The first question is: which world gets closer to the libertarian ideal, whatever it is?

The second question requires some setup.

Suppose you find yourself in charge of the state in World 2. You come to accept the anarcho capitalist argument that it is a violation of individual rights to forbid would-be independents who would like to start up their own protection agencies from doing so. Thus, you consider abandoning this restriction, and moving toward World 1.

However, you know more rights violations will occur in World 1. You also know that in World 1, you won't be the one violating anyone's rights. Your hands will be clean, so to speak.

Should you move from World 2 to World 1? If rights have categorical force, then perhaps you should: after all, your sincere desire that no one's rights ever be violated cannot justify violating anyone's rights. However, there is something strange, in libertarian terms, about deliberately creating a world in which you know rights are more likely to be violated.

One might argue that since, other things equal, your rights are more likely to be violated in World 1 than in World 2, it would be rational to resist the move from 2 to 1. But this only shows that it is not always rational to respect the rights of others.

Thus, the puzzle remains: should you irrationally move your society toward World 1, so that you no longer have to violate anyone's rights, or rationally maintain the status quo in which you, personally, have to violate the rights of a few?