I’m reading David Estlund’s new book Utophophobia right now for an symposium at the PPE Society. Estlund’s book got me thinking more about non-ideal theory and the moral status of the state in non-ideal circumstances. While ideal theory asks, “What institutions would we have if everyone were perfectly good and just?,” non-ideal theory asks, “What institutions should we have in the real world, where many people are evil, where people will take advantage of the rules, and where people’s willingness to comply is not a given?”. Non-ideal theory is generally much harder work than ideal theory, because ideal theory imagines away most of the interesting problems and most of the complexity of actual social life. Ideal theory is the arithmetic to non-ideal theory’s differential equations.

Many people believe that under ideal circumstances, we would not have any kind of state. (Gregory Kavka disagrees. I never found his argument persuasive, but see Chris Freiman for a careful refutation.) If everyone were just and good, then the kinds of problems the state could solve would not arise. We wouldn’t need to enforce rights because by hypothesis people would respect each other’s rights. We wouldn’t need the state to collect and redistribute income because people would voluntarily give what they should, using non-coercive coordination methods, and no one would take advantage of others’ generosity or sense of justice. Just as a perfectly just society would have no need for criminal justice courts, it would have no need for a state, period. It would be anarchic. Not Rothbard-style anarchy, with private enforcement agencies, but Mickey Mouse Clubhouse anarchy, with no enforcement mechanisms at all.

But then this leads to a kind of puzzle. Justice requires cooperative anarchy. In an ideal society, it would be wrong to subject me to the authority of the state, to have police officers and politicians lord over me, to tax my income, and so on, because these institutions would not be necessary to people to do what they ought to do.

However, even in unjust circumstances like ours, where the world is filled with moral degenerates, a very large number of people are sufficiently good and just that anarchy would work with them. For instance, while I’m not a saint, if everyone were as conscientious and committed to fair play as I am, and we all knew that, we could easily dispense with governments and instantiate cooperative anarchy. Why, then, should good people be subject to the state’s power? Why shouldn’t the state only apply coercive force to wrongdoers and degenerates and leave the good people alone?

Consider the problem of moral extortion. Make a list of all the things it is generally permissible for you to do. This might include marry an an adult who wants to marry you, read some book, draw a picture of a prophet, buy yourself a new guitar, listen to some music, move to a new town, write a philosophy paper, or whatnot. Let X stand for all the things it is generally and prima facie morally permissible for you to do.

Now imagine someone comes along and makes a credible threat: “If you commit [some particular act in set X], I will respond by doing something awful.” Ask: Are you now obligated not to do that act, simply because someone else has threatened to do something wrong in response?

I’ve asked people about this before, when writing When All Else Fails. Most people initially say no. For instance, suppose a jealous ex-girlfriend threatens to kill people if you marry your new girlfriend. Most say that while this is awful, you are not obligated to stay single. However, most people also change their minds when the stakes become high enough. If, say, Thanos threatens to kill half the universe unless you stop listening to country music, they say you are obligated to give in to his threat. At some point, the heckler gets a moral veto. At some point, the extortionist succeeds in changing the moral landscape, rendering certain previously permissible acts impermissible.

Similarly, the heckler/extortionist can also render certain previously impermissible acts permissible, or at least excusable. If Thanos threatens to kill half the universe unless you steal a car, then you are excused in stealing a car. Stealing isn’t justified, per se, but it at least excusable.

What if we change the story a bit? Instead of having an identifiable person actively threatening such extortion, imagine instead you simply know that if you do some otherwise permissible act X or if you fail to perform some otherwise impermissible act Y, then some significant number of people will react by doing evil things, but you cannot easily identify ahead of time who will do it and what exactly they will do. Again, I suspect most people would say that when the stakes get high enough–when the injustice others will commit becomes sufficiently severe–this can change the moral valence of your actions. It can render X impermissible and at least excuse Y.

Now back to the state. If everyone were as morally good as, say, Chris Freiman, we would live under cooperative anarchy. Unfortunately, some people are significantly morally worse than Chris is. They are willing to exploit others, prey upon others’ kindness, attack and harm them, and/or unwilling to contribute voluntarily to public goods or charitable endeavors.

It’s one thing to justify state coercion against such rotten people. (Presumably, we need a theory of when preemptive coercion is justifiable against them.) But it’s another to defend state coercion against good enough people, such as Chris. What explains that? Some people would predictably do the right thing and avoid doing unjust things, even in the absence of coercive reinforcement. Further, justice requires cooperative anarchy and forbids the state; at most, the state is permitted as a response to injustice.

Here, it seems, the state (or any other kind of pre-emptive coercive enforcement agency) is excusable because of moral extortion. We know ahead of time that some people out there will act like hecklers, some big, some small. We cannot quite tell who will commit injustice and who will not, but we know ahead of time that a significant number of people will. So, enforcement agencies like the state are, perhaps, an excusable pre-emptive response to the threat such people face.

This idea is familiar, of course. But it’s worth noting what the force of the argument is. It looks like the argument on behalf of the state, or any other kind of pre-emptive coercive enforcement agency, takes the same form as the argument for giving in to moral extortion. It would normally be wrong and inexcusable to subject a good person like Chris to threats, to tax his income, or whatnot. However, there are lots of bad people out there who implicitly threaten to act unjustly. We cannot predict ahead of time who the unjust actors will be and what exactly they will do, but we know that a significant enough number of people will do quite bad things. So, the argument goes, it becomes excusable in the face of these “threats” to instantiate the state.

Now, I’m not convinced this kind of argument succeeds at all. However, it seems to me that arguments for the state take this kind of form. The state is not in the first place an instrument of justice, but at most an excusable institution instantiated in the fact of moral extortion.