In Justice at A Distance, Loren Lomasky and I defend a liberal position on immigration. Our proposal is much more permissive than any mainstream position I know. We do recognize, however, the power of the receiving state to scan for criminals and terrorists. In a recent post, Jason Brennan goes even further and rejects our proviso. He argues that closing international borders on the grounds that crime will increase is unjustified because the same position applied to internal migration is clearly wrong. There is no more justification to prevent poor Mexicans from coming to Arizona than there is to prevent poor Californians coming to Arizona, even if in both cases crime will predictably increase. This is a powerful point that forces me to rethink our position and fill some gaps. I do so tentatively, as the argument needs considerable work.

A libertarian theory of immigration should start with private property rights. The territory of the state, I now claim, derives from these property rights in land. Individuals have created a protection agency to which they have delegated limited powers. But owners have not delegated their private property rights beyond the land required by the state to function, so the state lacks the power to exclude foreigners (or anyone else) if owners wish to admit them into their property. Thus, if an owner in Tucson wants to hire a Mexican from Hermosillo to mow her lawn, she is free to do so and the state lacks the power to prevent the entry of the employee. Immigration controls infringe the right to contract, itself an incident of property rights.

However, the state does have the power to prevent the entry of persons who predictably intend to kill, rape. or steal. The individuals who created government have delegated to it the power to protect them, and this includes the power to prevent crimes. Notice that I am not suggesting that people can be arrested preventively, only that they can be stopped from moving to the scene of their planned crimes, from reaching their intended victims. Conceived in this way, Jason’s objection is less powerful, for I would apply the same rationale to domestic migration. If poor folk from inner-city Miami want to move to my Tallahassee neighborhood they can do so by buying or renting land here. But suppose that these persons are members of a gang that predictably plans to murder and assault the neighbors. I do not think they can be arrested for planning crimes, but I think they may (indeed, should)  be stopped from moving to the neighborhood. I think our position comports with the central libertarian vision of allowing immigrants the opportunity and the freedom to participate in the global economy and thus seek a better future for themselves and their families. Open borders protect free, voluntary exchanges. They do not protect predatory behavior.

Seen in this way, the problem is not the principle but its implementation. Defenders of unrestricted borders will say that presently there are no reliable methods to scan for criminals, so any such policy will inevitably violate the rights of those who do not have sinister plans but simply seek a peaceful, better life. This is especially so if the reason to stop immigrants is that crime will statistically increase. Such policy is unjust because it treats groups as presumptively criminal, and so, by failing to individualize, it violates the rights of the innocent. I agree with Jason on this point. But if we could devise a scanning procedure that could identify with sufficient precision the immigrants who intend to commit crimes, I think our proviso stands.