The Sound and Fury of Holmes and Sunstein’s “The Cost of Rights: Why Liberty Depends on Taxes”

Stephen Holmes and Cass R. Sunstein’s The Cost of Rights: Why Liberty Depends on Taxes presents itself as, among other things, a rebuttal to libertarianism. Its scope is, of course, broader. The authors also want to get the goat of progressives who would underestimate the economic burdens created by government. The book is well-written and entertaining but, ultimately, a good deal less groundbreaking, revolutionary, or even intriguing than Holmes and Sunstein seem to think.

Stripped of the countless supporting examples and anecdotes, the basic argument of The Cost of Rights is this:

  1. There’s a difference between natural rights and legal rights. Natural rights preexist government. They’re fixed and, well, natural. We can argue about what they are, but we can’t make up new ones. Natural rights can be discussed within a positive liberty vs. negative liberty context.

  2. Legal rights can’t preexist government, because they are a) constructs of the state and its laws and b) claims by individuals upon the government (for services, for protection, for enforcement, etc.). Legal rights can’t meaningfully be broken into negative and positive sorts because even those rights that say we have a “right to not be X,” actually mean “we have a right to demand that government actively protect us from X.” (E.g., a right not to be killed is really a right to be protected, by the government, from being killed.) So, in this sense, all legal rights are positive rights.

  3. Because all legal rights demand something from government and that something means government must expend resources (on enforcement officers, on buildings to station them, on infrastructure, an so on), legal rights depend on taxes.

  4. Most things we think of as legal rights (e.g., the right to contract, to protection from violence, to private property and its enforcement) are only meaningful (i.e., can only be relied on to exist and be protected) if government expends resources to protect them. This means that a) government must exist and b) government must collect resources (taxes).

The bulk of the book’s 250 pages is paragraph after paragraph in the form of, “We have a right to contract, but that demands courts and police and politicians drafting contract law and judges interpreting it. And all that costs money,” followed by another like, “We have a right to be protected against theft, but that means police and courts and…” One gets the point quite quickly.

An Infinite Regress?

My colleague, Tom G. Palmer, wrote a scathing review of the book (pdf). Among his many critiques are that The Cost of Rights amounts to an infinite regress.

For there ever to be a right of any sort, by Holmes and Sunstein’s own theory, there would have to be an infinite hierarchy of people threatening to punish those lower in the hierarchy. Since there is no infinite hierarchy, we are forced to conclude that Holmes and Sunstein have actually offered an impossibility theorem of rights in the logical form of modus tollens: If there are rights, then there must bean infinite hierarchy of power; there is not an infinite hierarchy of power; therefore there are no rights.

I find this particular problem with Holmes and Sunstein’s thesis less troubling than Tom does. We have a right not to be abused by the police. To which Holmes and Sunstein respond, “Okay, so that means we need tax dollars spent on people watching the police for bad conduct and enforcing the laws against the police.” Tom’s infinite regress criticism means saying, “Ah ha! But who watches the watchers? Thus we need another layer of people enforcing the laws against the people enforcing the laws against the police.” And so on, for ever and ever.

From the standpoint of strict logic, Tom’s rebuttal works. There is an infinite regress here. But it doesn’t bother me because, in terms of practical, in-the-world effects, Holmes and Sunstein have a reasonable response. Namely, with each additional layer, the protection gets better. Not all police will abuse us. If, say, 10% of police do abuse citizens and that’s without one layer of additional protection, then perhaps only 1% will do it with one layer of protection. We’re getting closer to absolute protection of our right not to be tortured by the police. A second layer might reduce the rate of abuse to 0.1%. A third to 0.01%, and so on. At some point, even though our protection isn’t absolute, it’s close enough. Thus we avoid the infinite regress.

… Signifying Nothing?

So if the infinite regress left me unvexed, and if I approach the argument in the very abstract form presented in the four points above (and accept without question the natural/legal rights distinction and the conclusions drawn from it), what’s the takeaway from The Cost of Rights? Unfortunately, not a lot. Because Holmes and Sunstein haven’t really offered an argument against much of anything, especially not mainstream libertarianism.

In responding to The Cost of Rights, the libertarian can adopt one of two tactics. The first is to say, “Yep, we do need people to protect our rights, but why do they have to be the government?” In other words, to raise the anarcho-capitalist explanations of how rights protections and the rule of law could happen without the state. The book never addresses this possibility. It just assumes — without acknowledging the assumption — that “protection” and “enforcement” can only ever be the provence of the state. They may be right — many libertarians are not anarchists — but the idea of the state as the only effective rights protector is by no means a priori true.

The second tactic is to say, “So what?” Just because we need the state to protect rights doesn’t mean we need this huge state we have now. Minarchist libertarians, for example, say that yes we need the state to protect our rights and yes we need to pay the state to do that through taxation (or voluntary payments), but the state should be limited to protecting our natural rights and enforcing contracts and providing defense and that’s it. In this sense, the overarching, adumbrated argument of The Cost of Rights given above is, when it comes down to it, compatible with minarchism.

In fact, when Holmes and Sunstein attack “libertarians” and “limited-government” folks (they use the terms interchangeably), it’s clear that what they actually mean is “anarchists.” In other words, they think libertarians want no state and attacks them for holding that obviously (to them) silly position.

The Cost of Rights fails to genuinely wrestle with the actual arguments of its opponents.