It goes like this. Exercising rights is always costly --- in lost opportunities to do other things, if not in money. So is the preventative maintenance rights require, and without which they are not credible. This maintenance is glaringly obvious for property rights (police, fire departments, courts, prisons), but really exists for all of them. (If we're to have a right to free speech, there needs to be institutional machinery to support that right, and "watchdogs must be paid," as Holmes and Sunstein put it, not altogether felicitously.) If rights are to be equal, or even if all citizens are to enjoy a certain minimum of rights, then these costs cannot be individually borne, on, say, a pay-as-you-go basis. Rather, the cost of rights must be distributed across the citizenry. That means both taxation (to pay for rights) and state action (to implement them). Ergo, "liberty depends on taxes."
Is there more to the book, then, than this wordy gloss on the title? Perhaps surprisingly, yes. The authors fortify the basic argument with a great many examples, blocking all available avenues of escape. (At least, they dealt with all the objections which occurred to me.) They also go on to apply the argument to many simmering (American) political issues --- whether "rights have gone too far," separation of church and state, environmental regulation, welfare, and the like. Their discussion of welfare is typical: they reach a standard liberal conclusion --- welfare is good (though not all welfare programs are) --- for very un-liberal reasons: "they lend legitimacy both to the property rights of the wealthy and to the state apparatus that enforces them" (p. 208). This is of a piece with their whole approach, which is refreshingly cold-blooded. Whatever rights may be justified as points of pure ethics, the rights which citizens enjoy in practice are the result of more or less explicit political bargains, about who gets how much of what, in exchange for what else. Inescapably, there are tradeoffs between different rights, and between rights and other goals. What rights we actually enjoy depends on both our collective resources and our willingness to apply those resources to collective ends.
Does this devalue rights? Only if you really can't bear to think of how your sausages were made. Ideally, of course, the rights enjoyed by citizens would be arrived at by democratic deliberation, sensitive to the claims of justice, equity, and universal human dignity. Some parts of the world can be recognized in that picture, at least on good days; for much of the world it's not even a mocking caricature. The task, as always, is to make reality conform better to our ideals.
I'm pretty sure that Holmes and Sunstein have (so to speak) gotten rights right; the point is such a basic one that, once you see it, it's hard to miss. That in itself isn't enough to improve the quality of our public deliberations (or haggling). That will happen when their notions have so thoroughly diffused into what passes for common sense as to pickle the brains of "practical men, who believe themselves to be quite exempt from any intellectual influences". The Cost of Rights is clear, narrowly-focused, uncompromising and provocative, so there is actually a good chance that this will happen.