Immigration

A Libertarian Argument

The traditional libertarian position, the position I argued for in my first book (Chapter 14), is support for open borders. That was also the traditional American policy. For the first century of the country’s history anyone who could get here was welcome to come. Limits on Chinese immigration and on immigrants carrying contagious diseases were imposed in the late nineteenth century but broader restrictions only came in in the nineteen-twenties and did not, at the time, apply to immigrants from other parts of the New World.

Murray Rothbard switched his position from opposing restrictions on immigration to supporting them as part of his adoption of a paleolibertarian strategy of alliance with the right, joined by Hans Hoppe and others.¹ In trying to understand their argument I have used two sources, an article by Hoppe and a webbed debate between Dave Smith, a prominent figure in the libertarian party and the Mises Caucus that currently controls it, and Spike Cohen, the most recent vice presidential nominee of the LP. Cohen supported open borders, Smith opposed them, using arguments largely borrowed from Hoppe.

The argument offered by Hoppe and his supporters to justify immigration restrictions follows three related lines. The first starts with the idea that in the ideal libertarian society all property would be private and an individual, firm or voluntary community free to exclude or admit anyone. Since what we actually have is a society in which much property belongs to government, the nearest we can come to that is having the government control who can come. Since, Smith argues, a considerable majority of the population opposes open borders, the government should restrict immigration on their behalf.

One problem with this line of argument is that  it proves too much. In the ideal libertarian society an individual, firm or voluntary community would be free to forbid prostitution or gambling on its property, charge an annual fee for use of its property, do very nearly all of the things that libertarians currently regard as rights violations when the government does them. Since what we actually have is a society in which much property belongs to government …  .

The alternative approach, individualist rather than collectivist, is to argue that since government property was either seized from its rightful owners or bought or built with stolen money, the government does not have the same rights with regard to it that an individual owner has with regard to his property. That leaves open the question of what rights individuals have with regard to property claimed by governments but at least it does not provide the government a blank check to impose any rules the majority favors on everyone.

That brings us to the second line of argument, that in America as it now is some of the interactions with immigrants will be involuntary. Immigrants will collect welfare payments and send their children to public schools paid for by the taxpayers. Anti-discrimination law might force employers to hire immigrants, landlords to rent to them, even if they didn’t want to. Immigrant voters, if there were enough of them, could vote to tax other people and spend the money on themselves. Hence, it is implied, government restrictions on immigration are necessary to protect the current inhabitants from forced interactions with immigrants.

With no restrictions on immigration, individual employers are still for the most part free to employ or not employ immigrants, individual property owners to sell or not sell to them, landlords to rent or not to rent to them. In the society as it now exists transactions between current Americans and new immigrants are for the most part voluntary, just as they would be in a fully libertarian society, although not entirely so. Government restrictions on immigration do what private restrictions in a stateless society could not do, prevent other people from interacting with immigrants.

The existence of legal support for involuntary interactions is a problem but not a problem with immigration in particular; individuals are at risk of involuntary interactions with members of the existing population as well. It is an argument for repealing those legal rules that are inconsistent with the general principle that sales, employment, rental should happen if and only if both parties consent. Restricting immigration violates the rights of all inhabitants who want voluntary interactions with potential immigrants and of the potential immigrants as well. It makes little sense to prevent a large number of voluntary interactions in order to avoid a smaller number of involuntary ones.

What  makes more sense is to explore ways other than immigration restrictions in which involuntary interactions with immigrants can be reduced even if not eliminated. Open borders do not imply instant citizenship. While there were no restrictions on immigration in the early history of the U.S. there were restrictions on naturalization. Such restrictions could be retained in an open borders system; libertarian theory does not imply that everyone who comes can vote.

Carrying the argument a little further, welfare law could exclude non-citizens, although if it did it would be only just to also exclude them from having to pay the taxes that fund the welfare they were not eligible for. So far as the public schools are concerned, libertarians, at least the same ones who argue for immigration restrictions, support decentralization. A legal regime with open borders could give every school district the option of serving or not serving non-citizen immigrants. Again, justice suggests that if a school district rejects the children of non-citizens the taxes that fund the schools, including property taxes on property they occupy, should not be owed by the parents or their landlords. That would come as close to mimicking what would happen in a stateless libertarian society, where all schools were private, as is practical in the existing system.

Hoppe’s proposal along these lines was that any immigrant should be allowed in if a citizen is willing to sponsor him, where the sponsor would then be responsible for any costs the immigrant imposed on others, paying fines for any crimes he committed, damage payments for any torts, presumably also paying the cost of sending the immigrant’s children to a public school. The argument, presumably, is that the sponsor, by letting the immigrant in, is an indirect cause of all such costs. It is not an absurd argument but the notion of indirect liability that it depends on has implications that I do not think either Hoppe or his supporters would accept. If I sell you a gun, I am an indirect cause of any crimes you commit with it. Should I be permitted to do so only if I agree to be liable for the cost of such crimes? If I sell you a car … . The normal legal rule in a free society is that individuals are responsible for their own offenses. There is no obvious reason why the rule for immigrants should be different.

Hoppe’s proposal also makes little sense in other ways. Unless he intends immigrants to function like slaves or indentured servants, working for a single employer or those he lends them out to, they will be engaged, like other people, in a multitude of voluntary transactions with lots of different people. It makes no sense for all of those transactions to hinge on the permission of a single sponsor who could withdraw that permission any time he chose or, if he cannot, is liable for acts over which he has no control.

The third line of argument is that, as long as some property is owned by the government, the government is entitled to control its use. As Dave Smith points out, an adult man does not have the right to go into the girls' room of a public school; government property is not and should not be a commons. Hence the government may, if it chooses, refuse to allow immigrants it has not approved to use public property. Since public property includes almost the entire highway system, that makes it hard for an immigrant not approved by the government to do anything in the country much beyond employment in a farm on the border.

Here again, the argument proves too much. The government’s control of the public school is accepted not because the government is the rightful owner but because the control is for the purpose of the school, the same purpose it would have if private. Most people, almost certainly most libertarians, would be outraged if the school announced that the girls’ room was only available to girls whose parents promised not to own firearms, or not to publicly criticize the mayor, or not to do something else they were legally entitled to do. In that case the control over public property would be being used not for the purpose of that property but as a way of coercing people.

Similarly here. If roads were private, their owners might require some form of driver’s license in order to make their roads safer. If the roads are public, it is not unreasonable for them to require a driver’s license and refuse to allow an immigrant without one to drive. But that does not justify forbidding a legal driver from carrying the immigrant. What Smith’s argument is proposing is the government using its control over public property not to serve the purposes of that property but to prevent voluntary transactions between its citizens and foreigners.

Here again, the argument proves too much. A private owner is entitled to refuse to allow anyone not vaccinated onto his property. Hence the government is entitled to impose a vaccine mandate, enforced by forbidding anyone not vaccinated from using public property. It is entitled to effectively lock urban residents into their homes by forbidding them from using the public roads or sidewalks. It is entitled to ban drug use, prostitution, very nearly any of the activities libertarians believe it is not entitled to ban, by denying the people who do those things access to any public property.² The collectivist exception swallows the libertarian rule.

A response some supporters of Hoppe’s position offer is that citizens, unlike non-citizen immigrants, are part owners of government property, entitled to use it even if they use drugs or are unvaccinated or whatever. That is not how partial ownership works in other contexts. As an Apple stockholder I am a part owner of the company, entitled to receive a fraction of whatever dividends the corporation chooses to pay out and to vote at the annual meeting. That does not prevent Apple from forbidding me to trespass on their property.

If one does accept the argument, one of things I should be free to do on government roads is use them to transport immigrants to my property or anywhere else I am entitled to go.

I conclude that libertarians ought to support open borders. They may want to qualify that by including the condition that immigrants not receive all the rights of citizens, such as the right to vote or to receive government benefits, until they have been naturalized. Requirements for naturalization are a prudential issue, not a moral one.

Throughout this argument I have been responding to libertarian supporters of immigration restrictions in terms of libertarian rights theory, that being the basis on which they make their arguments. For reasons I have discussed elsewhere, that is not my preferred approach:

… one reason to base my arguments on consequences rather than justice is that people have widely varying ideas about what is just but generally agree that making people happy and prosperous is a good thing. If I argue against heroin laws on the grounds that they violate the addicts’ rights, I will convince only other libertarians. If I argue that drug laws, by making drugs enormously more expensive, are the chief cause of drug-related crime, and that the poor quality control typical of an illegal market is the main source of drug-related deaths, I may convince even people who do not believe that drug addicts have rights.

A second reason to use practical rather than ethical arguments is that I know a great deal more about what works than about what is just. This is in part a matter of specialization; I have spent more time studying economics than moral philosophy. But I do not think that is all it is. One reason I have spent more time studying economics is that I think more is known about the consequences of institutions than about what is or is not just, that economics is a much better developed science than moral philosophy. (Chapter 42 of The Machinery of Freedom, “Where I stand.”)

The Consequentialist Case for Easy Immigration

That suggests a different question than the one I have been discussing, although a related one, not whether immigration restrictions are consistent with libertarian rights theory but whether they have good or bad consequences.

If the comparison is to what I have proposed, open borders under rules where immigrants are not entitled to collect welfare, the change would have good consequences. What if, as some argue, that is not a politically viable option? One obvious response is that neither is the shift to open borders. Libertarians aside, talk about open borders is almost entirely by people on the right attributing the position to people on the left, not by significant figures on the left advocating it.

But suppose that, as a result of either deliberate policy or the breakdown of enforcement, it became in practice easy for people to immigrate to the US without any special government permission, as it was for most people for most of the history of the US. Would the consequences be good or bad?

The answer is not clear. The straightforward economic argument to show that the consequences for both the present inhabitants and the immigrants would be on net positive, essentially the same as the argument for free trade, depends on a laissez-faire economy, does not hold given the existence of transfer payments, as  many have pointed out. There is the risk of people coming in order to live on welfare, making the current inhabitants paying for that welfare worse off.

On the other hand, while welfare payments may be sufficient to support a better life than some immigrants are used to, the income that can be earned even by relatively unskilled workers in the US can support a better life still, with the prospect of improved skills and employment over time. People willing to leave the life they are used to in order to seek something better in a foreign country are more likely than random people — the present inhabitants, for example — to choose the latter alternative. I am not aware of any evidence that large numbers of immigrants, legal or illegal, presently exist on welfare, and I observe around me in California large numbers of Hispanic immigrants, whether legal or illegal I don’t know, trimming trees, doing yard work, house cleaning, and the like. A common pattern seems to be an English speaking immigrant, probably legal, employing a crew of workers, mostly not English speaking, possibly illegal.

The discussion so far assumes a welfare state and asks what the consequences of relatively free immigration will be. It’s worth reversing the question. Given relatively free immigration, what are the consequences for the welfare state?

Current welfare laws and public schooling, the other area where there is the potential for substantial costs via non-voluntary transactions, are mostly controlled and funded at the state and local level, the main exception being the Social Security system. If a lot of immigrants choose to live on welfare and select their residence accordingly that would make generous welfare programs expensive for the localities that have them, as New York City is discovering, especially expensive if combined with legal restrictions that block immigrants from legal employment. That should be a powerful political incentive to reduce or eliminate welfare transfers or restrict their availability, spend less on public schools, make it easier for immigrants to work. Those are simultaneously benefits from open immigration and changes that weaken the argument against it.

What about the exception? Social security is a net cost for the young, benefit for the old. Immigrants are mostly young; those who are old do not have the record of past contributions needed to collect current payments. The net effect of more immigrants joining the system would be to postpone the collapse of the present Social Security system, benefiting the present elderly.

Libertarians and the Theory of the Second Best

Suppose the argument I have just sketched is wrong, suppose that, under current political institutions, easier immigration will make the present inhabitants of the country worse off. Should libertarians then support immigration restrictions?

This raises the general question of how to deal with what economist call the Theory of the Second Best. Suppose there are two policies, A and B, each of which makes us worse off. Further suppose that having both A and B, while worse than having neither, is better than having A alone. An example would be for A to be a government policy of bailing out failing banks, B a set of bank regulations designed to make it less likely for banks to fail. In the present discussion, A is a welfare state, B restrictions on immigration. If A exists, should libertarians support B?

The answer, I think, is that a libertarian economist who believes that situation exists should hold two different positions on the question. His view qua economist should be that, given the existence of A, he favors B. His view qua libertarian should be that he favors the abolition of both A and B. His reason for favoring B, after all, has something to do with his being an economist, nothing to do with his being a libertarian.

Hence my position qua libertarian, what I believe the Libertarian Party’s position should be, is to support the abolition of both welfare state policies and immigration restrictions, a change that would both reduce rights violations and make present and future residents of the country on net better off.³

That is not the position that Hoppe and his supporters within the Libertarian Party appear to be arguing for, possibly, if my ungenerous interpretation of their behavior is correct, because it is not the position of the conservatives they want to ally with.

1. Hoppe is a major influence on the Mises Caucus, the group currently controlling the Libertarian Party. Although the platform of the Mises Caucus rejects open borders the LP  platform still supports them, probably because the platform has not been revised since the Mises takeover. That may change at the national convention this May.

2. Dave Smith responded to the argument about using control of government property against drug users, raised by Spike Cohen in their debate, by saying that it was legitimate to forbid addicts from shooting up on government property. He did not consider that the argument he was offering, applied to that case, would let the government ban the possession or transportation of illegal drugs on government property, including the entire highway system.

3. The change would make some individuals worse off, both some who benefit from income redistribution and some whose loss due to having to compete with immigrant labor would outweigh their gains due to consuming goods or services less expensive because of immigrant labor. What I mean by “on net better off” is an improvement in economic efficiency, roughly defined as a change that would increase total utility if we did interpersonal utility comparisons on the assumption that everyone had the same marginal utility of income. For details on problems with that definition of improvement and arguments for using it despite the problems see Chapter 15 of my webbed Price Theory or Chapter 2 of my Law’s Order.