The philosopher Michael Huemer is usually favorable to the free market, and he is also a strong defender of anarchism. Although I disagree with some of the arguments in his defense of anarchism, The Problem of Political Authority, it is an excellent book.
In a recent blog post, he surprisingly suggests that taxation may in some cases be justified. He offers two examples, Pigouvian taxes on negative externalities and Georgist land taxes. In what follows, I’ll concentrate on the first of these.
He offers the following argument for Pigouvian taxes (named for the Cambridge economist A.C. Pigou).
Sometimes, people do stuff that harms other people, and the people who are harmed don’t consent to the harm. I.e., people “produce negative externalities”, as the economists say….
Example: Pollution. Whenever you drive your car, you release a little bit of pollution into the air, which imposes a tiny expected harm on a huge number of other people and animals, including future generations. I bet you don’t get their consent, either.
On some absolute deontological views, you always need consent before imposing (certain kinds of) harm on others. But that’s impractical. You can’t get the consent of everyone in the world, including the future generations who will be affected by your pollution. So we’d have to say either
(a) “You can’t pollute at all.” This requires shutting down modern civilization. Or
(b) “Pollution isn’t the right kind of harm” (it’s not aggressive, people don’t have rights against pollution, or something like that). But this would mean that it would be fine to completely destroy the atmosphere with pollution (if someone had the ability to do that).
(a) and (b) are both bad. We shouldn’t completely prohibit all pollution, nor should we take no action at all against pollution. While complete destruction of the atmosphere may not be on the table (yet), we would surely have too much pollution if we didn’t do anything at all to polluters.
This point of course applies to other kinds of externalities. If people get to impose negative externalities for free, there will be too many negative externalities. Lots of activities will get done that impose greater total costs than their total benefits. And almost everyone is going to lose out overall from all the negative externalities.
Solution: Pigouvian taxes. These are taxes on externality-producing activities. They’re supposed to be set so that the tax is about equal to the amount of external harm produced by the activity. This deters people from doing the activity, if and only if the total cost created by it exceeds the total benefit. Example: A carbon tax that would approximate to the amount of harm caused by greenhouse gas emissions. This would deter the least economically valuable carbon-emitting activities, while still allowing the most valuable ones.
Huemer now asks, How does this argument justify taxation?
Wait—you can see what the utilitarian rationale is for Pigouvian taxes, but why isn’t it still theft (even if a beneficial theft)? My thinking is that the person creating the negative externality actually owes compensation for doing so. Extracting owed compensation from someone isn’t theft. So this form of taxation isn’t theft. But note that the government would be obligated to use the money collected to actually compensate the people who are harmed by the taxed activity.
I do not think that this argument succeeds. Huemer mentions “absolute deontological” views, which hold that you must get someone’s consent before harming them and suggests this has unacceptable consequences as regards pollution. I don’t deny that’s a possible view, but a more common deontological position is that you can’t violate someone’s rights without his consent. Not every harm counts as a rights violation, at least if you characterize “harming” as “making worse off.” For example, if you are hired for a job that I wanted and would have gotten had you not been chosen, you have harmed me but haven’t violated my rights.
If you adopt the more common deontological position, then the problem Huemer raises does not come up. Pollution from your car requires consent only from those whose rights would be violated by what you do. But hasn’t Huemer already considered and rejected this response? He says that if you hold pollution isn’t aggressive and doesn’t violate rights, then you would have to accept that it would be all right for someone to destroy the atmosphere, if he could do so.
Huemer has wrongly taken for granted an all-or-nothing position, a common failing among libertarians. Either any amount of pollution violates deontological constraints, or none does. But why assume this? The scope of property rights is to an extent conventional within a society, and it is irrational to delimit property rights in either of the ways Huemer indicates. In a classic paper, “Law, Property Rights, and Air Pollution,” Murray Rothbard shows how pollution problems have been dealt with in the common law. Surely this is a better rights-based approach to pollution than either of the views based on rights that Huemer mentions.
Suppose that we adopt the view of pollution rights I suggest. Should we then say that it’s all right for someone to pollute, so long as he compensates those whose rights he violates? That seems wrong; in general, you aren’t at liberty to violate someone’s rights so long as you make him no worse off than before your rights violation. The rights holder usually can demand more than this or refuse consent altogether to the rights violation.
As I mentioned, Huemer has another case in which taxation might be justified. He says,
I think Henry George may be right. Henry George thought that (a) everyone is entitled to the value that they themselves produce, but (b) they’re not entitled in the same way to value produced by nature. If you happen to be the first person to claim some valuable natural object, that doesn’t really give you a greater claim to its value than other people who arrived later.
So when you build a building on some land, you should own the value that you added via your labor. But you don’t have any special claim to the value that the land had prior to your arrival. Instead, George thought, everyone should get an equal portion of the value of all land & natural resources. (We can simplify this to just “land”, since natural resources are generally part of the land.)
Huemer goes on to indicate that a land tax on the “intrinsic value” of the land, paid to those who aren’t using it, can be seen as a form of compensation to the nonusers. I’m not going discuss this further, as the initial Georgist starting point is implausible and Huemer hasn’t argued for it. In the view I take to be correct, which I haven’t argued for either, you initially acquire physical resources, not “value.” The value of what you acquire depends on the preferences of people in society but isn’t itself an acquisition. Further, on Huemer’s assumption that you don’t own the intrinsic value of the land, why does “society” own it in a way that requires compensation for what has been “taken” by the land user?
Huemer’s ingenious arguments leave us where we started: taxation is theft.