OPR VIII.23: The Justification of Coercive Laws

Overview of §23

I’m going to structure my discussion of this section a little differently. I’m also going to be a bit polemical about it. Perhaps this treatment will galvanize some discussion about this section, which I believe is rather significant for Gaus’ argument.

In the preface to the book, Gaus decries the tendency of political philosophers to turn into hedgehogs, each championing one among many potential political creeds (liberal egalitarianism, libertarianism, Marxism, communitarianism, etc.). Gaus then announces his attention to take a different, “foxier” path, one that does not endorse a particular political philosophy, but asks how we should think about the moral order—including the political order—in a world lacking reasonable agreement about many basic moral questions. In doing so, he eschews any desire to champion, hedgehog-style, any single political philosophy—including libertarianism, a philosophy with which he is prominently associated (p. xv). I distinctly remember reading this passage months ago and saying to myself, the philosopher doth protest too much. And lo and behold, many months later, I am not surprised to see that Gaus’ in-depth examination of the nature of morality is yielding libertarian political conclusions.

I should be clear here—I don’t think there’s anything wrong with Gaus attempting to defend libertarianism in this book. If something like reflective equilibrium is a legitimate method for producing valid moral claims, one would expect Gaus, a libertarian, to endorse a libertarian political system, and to argue that the moral situation in which we find ourselves today calls for libertarian principles. If his analysis of morality had yielded the conclusion that communism was the only acceptable economic system, one would expect Gaus to go over the argument seven or eight hundred times until he had found the problem with it. Again, let me repeat—there is nothing illegitimate about this as a method of argument. But it is a little disingenuous of Gaus to act as though he is merely following the argument where it goes, and the fact that it leads to pro-libertarian conclusions is just some sort of happy accident.

Technically, of course, Gaus’ argument does not unambiguously endorse any particular political order. But Gaus’ analysis of morality most definitely tilts in a libertarian-friendly direction. Reasonable people who aim to settle upon a political order, says Gaus, will consider which principles they could reasonably accept as the basis for such an order. They will do so on a principle-by-principle basis; they can do this, because Gaus eschews any efforts at evaluating competing political orders as a whole. Of course, Gaus says, they will endorse all the standard things that libertarians like to endorse—basic rights to life, liberty, and property. But will they endorse anything else? It should be clear that there will be a very serious obstacle to their doing so. Endowing the political order with any further powers will only be morally legitimate if every reasonable moral citizen would agree to such endowment. And there is an important class of people—libertarians like Gaus himself—who can be expected to object to any such expansions of state authority. In effect, libertarians can reasonably count upon other citizens endorsing a set of principles which, considered all by themselves, constitute the libertarian ideal point. Not surprisingly, libertarians will be reluctant to endorse any move away from this ideal point, and their endorsement is necessary in order for any such move to be legitimate.

Now my statement of Gaus’ position is unfair without two critical caveats. First, Gaus does not say that a libertarian is free to reject any law that makes government more expansive than his own preferred libertarian system. Such rejection would not be reasonable. For it to be reasonable, the libertarian would have to believe that there is no possible law in the area in question which is preferable to no law at all. (Again, Gaus assumes an issue-by-issue evaluation of laws, although he concedes this might prove difficult to accomplish.) If any such laws exist, then the libertarian would be unreasonable to reject them. The problem—and I am somewhat surprised that Gaus does not acknowledge this explicitly—is that for most libertarians, in most areas of life the best law would in fact be no law at all. If this is the case, then libertarians would indeed by justified in rejecting any proposal for the government to do anything regarding such issues. Take public schools, for example. In order the justify the public provision of education, one would have to show that no one could reasonably reject the addition of such schools to a more minimalist state (one that guaranteed only basic negative liberties plus some right of political participation—the “liberty of the moderns” that Gaus mentions frequently). But surely there are many libertarians who would so object—indeed, who would find a system where all education is left in private hands to be preferable to any other system. If so, then no public system could be justified. (One could respond to this position by denying that such libertarians would count as reasonable moral agents. I cannot see Gaus endorsing this move.)

Gaus’ own examples do not make this result plain. He denies that he is setting up a “libertarian dictator,” who gets to veto any political order more expansive than his favorite (p. 499). But the libertarian dictator, on Gaus’ view, would be someone who rejects a move to a more expansive state on grounds of its coerciveness. (More on coercion shortly.) Gaus wants to say that one does not automatically win just because one is more sensitive to coercion than everyone else. But his examples presuppose that the coercion-sensitive person regards a number of possible authoritative laws as superior, or at least as good as, no authoritative law at all. Surely there are many areas of life where, for libertarians, this condition simply won’t be met. And whenever this is the case, the libertarian does indeed get to play dictator—not by virtue of his sensitivity towards coercion, but because of his denial that anything could improve upon pure laissez faire in vast areas of human life.

Second, Gaus concedes that there may exist people who reject minimal standards as worse than no standards at all. If such people exist, then they might restrict the set of laws acceptable to them so as to include only more expansive laws. This might even apply to the basic rights of life, liberty, and property. Such a person might hold that “a classical liberal state that enforces property rights with a modest provision for the poor imposes coercion costs on the poor that exceed the benefits, and so such a state is not justified in his view” (p. 507). Gaus is, however, very dismissive towards this hypothetical possibility. “Given the basic place of the rights of the moderns in the order of justification, because the state is necessary for protecting the rights of agency, privacy, and freedom of association, even modest-sized states that perform the job of adequately articulating and developing these core rights typically will have justified authority” (p. 507). In other words, reasonable people will not reject the minimal state if the only alternative is no state at all. And if they have similar preferences about less basic issues (e.g., public education), they are simply out of luck. If they think that no public education system at all is terrible, and a minimal educational system is even worse, this provides no case for a maximal educational system. For there will surely be libertarians who regard any public education system as worse than none at all, and none at all is the default position that wins in cases like these.

Gaus ends the chapter by acknowledging that his analysis of the political order has a definite “classical tilt” to it. For

all reasonable persons devoted to the basic rights of body, the person, speech, and so on accept that the benefits of political authority exceed the costs across a wide range of systems. Given the order of justification, the basic liberties of the person and civil rights are powerful justifiers of small states—although larger states certainly reasonably can be thought to be optimal. Consequently, while some may reasonably believe that a larger (more coercive) state would better satisfy their evaluative standards, critics of classical liberalism will not generally be in the position to reasonably claim that smaller states do more harm than good (p. 508).

If anything, Gaus understates the strength of the “classical tilt;” given his acceptance of the issue-by-issue treatment of political issues, I am not at all sure why he thinks a libertarian would not simply win on virtually every issue, simply because 1) when the issue concerns libertarian rights, everyone will accept the libertarian position, and 2) when the issue concerns something else, the libertarian can effectively veto anything other than the libertarian position (i.e., the state does nothing at all).

The only way anything other than the minimal state could come into being according to this argument is at the level of the basic political issues. If one considers basic issues regarding life, liberty, and property, there will be a number of important questions to be settled. And here it’s not as clear that libertarians get a veto. When the issue is freedom of speech or freedom of religion, for example, libertarians cannot readily endorse the status quo, which in this case would mean no protection of freedom of speech. The libertarian has reason not to want the maximal state, but he also has reason to want the nightwatchman state. But there may well be many alternative political principles governing freedom of speech that the libertarian would regard as superior to no protections at all. The same will presumably hold for freedom of religion, protection of private property, etc. The result may be a substantive set—maybe even a large set—of alternatives that everyone (including libertarians) can endorse as superior to no political solution at all. And if one of those alternatives is incorporated into a political order—through an admittedly arbitrary, Hayekian process—libertarians cannot reasonably object to the result.

But Gaus still isn’t done finding obstacles to throw in the way of anything larger than the minimal state. For Gaus is also concerned about coercion. Very concerned. “All coercive legislation, even that which improves our system of morality, must overcome a presumption against employing force against persons” (p. 459). This concern creates a presumption against the state tinkering with existing moral equilibria.

The political process is a deeply imperfect way to arrive at equilibrium, as it can so easily miss the mark, and move us from a current equilibrium to a moral disequilibrium that, because it is backed by coercive punishment, can be a stable social equilibrium…Because the political process is such an imperfect way to bring about moral equilibrium, a current bona fide moral equilibrium must be respected by the political process” (p. 460).

Each coercive law must be justified on an issue-by-issue basis. “A free social and political order cannot simply coerce its citizens to improve current morality through a specific law without ensuring that the coercion employed in this specific law can be justified to all” (p. 489).

This aversion to coercion plays an important role in Gaus’ argument. Gaus occasionally tries to minimize how important it is. He states at one point that all he wants is for people to have “claims not to be coerced,” and emphasizes “just how weak this right is” (p. 485, n. 62). His claim is that “we have a claim not to be forced to do things, or threatened with force, and that a justified law must show that challenge of this claim can be met” (p. 483). This does sound rather  minimal; if you want to coerce someone, you need to have a good reason to do it. But Gaus interpretation of what a “good reason” is suggests that it is anything but minimal. He embraces the Nozickean idea that this right is a side constraint on action, and it is not consequentialist in nature. These side constraints are not absolute, contra Nozick, but they are nonetheless extremely strong. Thus, the mere fact that a state could, by coercing people a little bit, prevent private citizens from engaging in more coercion—maybe even a lot more—somehow does not count as sufficient reason to justify the law. He offers the following scenario:

Suppose we are living in a society in which a number of people defect from the social rules of morality and employ unjustified force against others. Suppose that Draconian laws would do the trick—they would reduce overall coercion, but at a cost of executing some people for trivial defenses. The Rights View [i.e., Gaus’ view] holds that such coercion against persons is still unjustified, and so we are prevented from moving from a more to a less coercive world. In the current social world where there is great private coercion, the right not to be coerced (which is honored by the state, let us say, but not by highway robbers) need not minimize the total amount of coercion employed against people (pp. 486-487).

This claim seems far from obvious to me. If the only political orders on the table are Somalia and Singapore, it’s not exactly clear that Somalia is the preferred option, even among people who value freedom highly. But what’s important here is that according to Gaus, one can believe that one state of the world is better than another and still reject the move from the latter to the former because it would require coercion to get there.

This is relevant to the justification of the minimal state—a state that, when it comes to basic rights, ensures the libertarian panoply of rights and nothing else. Surely, says the anti-libertarian, when it comes to property, you must agree that an extensive welfare state is better than no property rights regime at all? If so, you cannot reasonably reject it if we decide to set up such a welfare state. Not so, says the libertarian who has read Gaus. I can agree that an extensive welfare state is better than no property rights regime at all. But the welfare state has much more coercion than the minimal state. And so I can reject the move from the no-property-rights regime to the strong welfare state purely because I am justified in rejecting an increase in state coercion, even if it brings about other benefits. You, on the other hand, cannot reject the move from no property rights regime to the minimal state, for there is no comparable objection to the minimal state. This means that the minimal state wins.

This might be a little strong. Gaus does seem to think that there may be some non-minimal options that, even factoring in considerations surrounding coercion, would still be acceptable to the libertarian. Still, it is clear that “the limits of the eligible set will tend to be determined by those reasonable Members of the Public whose views about the nature of coercion are such that they see it as a steeply increasing cost and are most skeptical about its benefits” (p. 504). The libertarian dictator rides again; the only thing that has changed is the color of his horse.

The question, then, is whether coercion can do all the work Gaus wants it to do. There is at least one powerful objection to this deployment of the concept of coercion. Coercion, on some accounts, is an inherently moralized concept. It involves trespassing on rights that you posses. But we cannot speak about such trespassing coherently until we have defined what your rights are. Thus, you cannot object to someone taxing you as coercive until we have established that you have the right to the property sans tax.

Gaus recognizes the need to head off this objection. Earlier in the book, he argues for a non-moralized understanding of coercion. “Although it is correct to say that as one’s moral boundaries expand so does the range of possible coercion, it does not follow that all coercion involves a threat to violate the rights of another, and hence coercion is a thoroughly moralized concept” (p. 352). At the most basic level—before any moral principles have been accepted at all—“Members of the Public understand the right not to be coerced as a right not to have threats made against one’s natural person” (p. 352).

One easily could deny that this is a nonmoralized understanding of coercion at all; in effect, the Members of the Public have unanimously accepted some moral principles regarding the physical integrity of one’s natural person. But this move is not necessary to defeat Gaus’ argument. For Gaus pretty much equates increases in coercion with increases in state functions. As noted before, on page 508 he uses the phrase “a larger (more coercive) state;” for Gaus, the larger state is by definition the more coercive state. But there’s simply no way to make this move legitimately without a heavily moralized conception of coercion.

Imagine a socialist state that supports a large and extensive welfare rights regime. In this state, everyone works for a wage issued directly by the state, which is electronically deposited in their bank accounts at the end every pay period. The heavy taxes needs to support this welfare rights regime are automatically deducted from these wages before the deposits are made. Now I contend that this entire tax system cannot count as coercive on any nonmoralized understanding of the term. No one’s body is ever touched. The government never lays a finger on anyone. It doesn’t break into anyone’s home to take anything. Indeed, people never even see the money before it is taken. There might be many objections a libertarian might raise to such a system. But the only way he could possibly claim it is coercing people is by claiming that it is taking money that belongs rightfully to them. In other words, by embracing a moralized understanding of coercion. (Indeed, one could go further. Suppose that the only punishment ever issued by this state consisted of monetary fines, which are simply deducted from the electronic bank accounts. If such a regime could be maintained, the state would literally be committing zero coercion, in the “nonmoralized” sense of threatening people’s natural persons.)

Like most libertarians, Gaus moves rather effortlessly from speaking of coercion-as-threats-to-my-person to coercion-as-threats-to-my-property. Consider the following passage:

So, as we might do any day when we open the newspaper, we ask, “is this new legislative proposal justified?”…Given that these more basic rights and claims are taken as settled, we can say that our law threatens coercion against noncompliers. What makes the proposal—even if it turns out to be justified—morally problematic is that it employs force, or threatens to use force, against the persons of its citizens are residents. Legal restrictions, requirements, and other rules require compliance, and if compliance is not forthcoming the citizen is typically threatened with the use of force against her person; she is threatened with the loss of freedom, or loss of her property (Gaus’ emphasis; pp. 480-481).

Here Gaus simply assumes that for a citizen, both “loss of freedom” and “loss of her property” count as instances of “force against her person.” Now admittedly, this might be a legitimate thing to do if, as Gaus assumes, the “basic rights and claims are taken as settled.” If we know what is yours and what is mine, then threatening to take away what is yours does look like coercion. And so one could imagine Gaus’ stricture against coercion looking legitimate whenever any non-basic issue is on the table. My point, however, is that whenever the basic rights themselves are on the table, it is incoherent to speak of one political order coercing more than another. For until the political regime, with its accompanying rights and liberties, is accepted, there is no sense in which anyone morally has anything for others to threaten.

In order for Gaus’ argument to yield a legitimate defense of the libertarian understanding of basic rights, he would have to show that some kind of private property regime, virtually unencumbered by taxes, regulations or other restrictions, is a system that one could not reasonably reject, and that no other property regime has this property. Gaus makes a show of doing so in §18.3, but his argument is very brief and, frankly, rather flip. Given the controversial nature of the topic, I simply cannot believe that Gaus would think any non-libertarian would find his defense of strong private property rights convincing. (I gather Gaus discusses property rights further in the following section. My brief glance at this section did not reveal anything that affected by argument here, but if there is something then I’m sure the next commentator will set me straight.) But without an argument to this effect, one is left with people who reasonably disagree strongly about property regimes, all of which everyone presumably agrees are better than no regime at all. And so any regime in this set could constitute part of moral order that no citizen can reasonably reject. Moreover, if citizens decided to alter the regime using the democratic process, it would be hard to see a reason why they could not do so, so long as they stay within the acceptable set. The classical tilt, if it still tilts at all, now has a rather gentle slope to it. The libertarian in Gaus recoils at this conclusion, and so he attacks it both by raising suspicions about the validity of state action (section 22), and by imposing philosophical obstacles on any effort to exceed the minimal state (section 23). I have suggested that Gaus’ attempt to make an analysis of the nature of social morality yield libertarian conclusions is less successful than he thinks, but even if one disagrees, one would be hard-pressed to deny that (Gaus’ protestations notwithstanding) this is what he is in fact trying to do.

About Peter Stone

I work in various contemporary areas of political philosophy, including democratic theory, theories of justice, rational choice theory, and the philosophy of social science. My present research project deals with the virtues and limitations of random selection.
This entry was posted in Books, Posts, Reading Group and tagged , , . Bookmark the permalink.

3 Responses to OPR VIII.23: The Justification of Coercive Laws

  1. I want to add to Peter’s doubts about the argument in this section (hopefully eliciting a response from those more sympathetic to the view, as he says). Section 23 starts with a reference back to 17.3c’s right not be to coerced. But where did this right come from?

    Section 17 begins with a “presumption in favour of liberty”; it is wrong to interfere or otherwise thwart someone’s agency without justification (341). However, it is presumably also wrong to fail to interfere when interference is morally required. So what does the presumption amount to? At this stage, it seems to be simply a tie-breaker; if the cases for and against coercion are otherwise tied, we should not coerce. This presumption is indeed weak, as suggested by the note on p.485 to which Peter draws our attention. Of course one needs a positive justification to engage in coercion; of course in cases of ties, not-coercing wins; and of course this applies to the state as it does to everyone else.

    However, Gaus quickly moves from this very minimal presumption against interference on p.341 to a right not to be coerced, beginning on 349, and in particular a right not to be coerced without public justification, as he says on p.350. The problem is that the minimal tie-breaking presumption doesn’t imply the heightened standard of justification (the unanimous qualified acceptability standard). The claim that when the balance of reasons is tied we should not coerce does not imply that the case for coercion must be made beyond a reasonable doubt, i.e. to the satisfaction of all reasonable / qualified points of view (all Members of the Public, in Gaus’s technical sense).

    Even if this stronger right doesn’t follow from the minimal presumption against coercion (and perhaps it isn’t intended to), perhaps it is still plausible. I don’t think it is, however, because coercion is such a broad category for Gaus and the right Gaus defends is a right in a strong sense. Coercion includes all threats against the person. And Gaus insists that the right not to be coerced is a right and not a goal. The fact that by coercing some we can prevent a greater number of others from being coerced by third parties may be a good reason for having a law, he concedes. But unless the law meets the standard of unanimous qualified acceptability, it is not publicly justified. That’s what I (like Peter) can’t accept; the application of the unanimous qualified acceptability standard to coercion directly, with a default of not coercing.

    For basic liberties, it is plausible that a right to X means that the state may not violate the right to X for one person even to prevent someone else from violating the right to X of two or more people. The right not to be tortured is like that, I suppose. With the principle of public justification, however, we are not talking about coercion of particular kinds in particular contexts, but reasonably rejectable coercion in general. Why should we think that such coercion is asymmetrically bad, i.e. worse to do than to fail to prevent? I don’t think Gaus has satisfactorily answered this question. His example of Draconian laws that reduce the overall level of coercion (which Peter cites) shows only that there are certain kinds of coercion that are not justified even if they reduce the overall level of coercion; threats to boil convicted murderers in oil, or to sever the hands of convicted thieves. One cannot conclude from such examples that there is a right against reasonably rejectable coercion in general.

    Moreover, if we are really just talking about an “abstract right”, the precise contours of which will be the subject of reasonable disagreement (341, 479), how has Gaus managed to arrive at such a particular specification? An abstract right not to be coerced would be abstract in that it would leave unspecified exactly what constitutes coercion, in what context the right applies, and so on. But Gaus’s right is definite and specific, in that it applies to all agents and all instances of threats against the person.

  2. I agree that a core problem of Gaus’ approach is that he justifies a minimal non-coercion principle, but some of the actual work is done by disputable specifications of this principle. I tend to think that decision rules are also crucial in this regard. For what seems to drive much of Gaus’ analysis is his very understandable worry that in the real world people are simply much too quick to coerce others with the help of the state, even if they have no good justification. Non-neutral (supermajoritarian) decision rules are important in his analysis because they are the constraints that our idealized selves can put on our real selves in the real world. But the problem (as I have argued elsewhere) is that the minimal non-coercion principle cannot justify particular decision rules. I find one aspect of this particularly noteworthy here, as it is linked, in my view, to points made by Andrew and Peter: when it comes to decision rules, Gaus seems to be conservative more than libertarian. He consistently turns his non-coercion principle into a “no-change-in-coercion principle”.

    Consider one of his examples. A law L is proposed to replace law K, with the effect that a large group g1 is coerced much less than before whereas a smaller group g2 is coerced somewhat more. Gaus’ point is: “Even if L would make the overall system of laws less coercive than K, L might still not be justified because it would violate some citizens’ rights not to be coerced.” (489). Hence he prefers a minority veto against L. My question is: what about the right of group g1 not to be coerced? Does this cease to exist because law K was passed before law L? Why should the temporal sequence of laws matter here? Gaus’ deliberately analysis conflates the passing of legislation with its coercive enforcement (480). Since coercive enforcement of a law is a continuous pattern of acts and threats, the justification of laws is also a continuous need. Hence if Gaus is serious about the right for a minority of some size to veto legislation, must this not be a sort of “atemporal” minority veto? That is, must not a minority of the same size be allowed to repeal any legislation that already exists? If there is a veto right available to g2 for trying to block L, shouldn’t there also be an ex post veto right available for g1 with respect to K?

    I think that this type of atemporal minority veto might have libertarian consequences; its usage might sooner or later move us to the libertarian end of the eligible set. I’m not entirely sure, however, because this kind of veto would provide for a real-world test of the degree of justificatory interdependence of issues (495-6). For instance, if some minority blocks banking regulation in response to financial crisis, a different minority might see a connection to other legislation, cherished by banks, and press for its repeal. An atemporal minority veto might not only lead to a great deal of legislative instability (which might greatly increase political transaction costs), it might also enforce a rather “holistic” justification of state activities, which Gaus wants to avoid as much as possible (496). I think there are good reasons why an atemporal minority veto could not be publicly justified, but Gaus’ actual argument about decision rules strikes me as more conservative (status quo preserving) than “libertarian”.

  3. Andrew, I think you’re right to point out that Jerry lays the groundwork for Sec. 23 in Sec. 17. If you accept the presumption in favor of liberty, much of Sec. 23 follows, if not directly. I don’t think Jerry moves quickly from the presumption to a right not to be coerced without public justification, though. I take it that the case for public justification is grounded partly in Chapter IV such that Jerry’s move to the more robust right against coercion will rely on material from this chapter.

    Also, I think Jerry addresses your concerns about reasonable disagreement about the level of coercion on pp. 503-504 in 23.3(d). I’d guess that your concerns will not be wholly addressed by this subsection, but what do you think about it? You will also want to check out 24.3. A choice quote: “Once we get beyond paradigmatic instances of coercion, claims about what constitutes coercion are, notoriously, open to dispute” (525).

    Peter, I will be interested to see what you think of 24.3 in light of your comments. Also, I take it that Jerry’s remarks about duties of assistance on p. 367, where he thinks it is hard to see how an extensive right to assistance “can go undefeated under full justification among a public deliberating on a diverse set of evaluative standards” might push back against your worries about implicit libertarian bias in Jerry’s defense of public reason’s classical tilt.

    I should emphasize that Jerry thinks that a wide range of welfare state provisions can be publicly justified under a number of circumstances and that the classical tilt is hardly libertarian. The tilt includes many modest and some extensive welfare states, though it does take a right to private property (even in capital) as basic. In this way, Jerry’s view is indeed well to the right of most political philosophers, but hardly controversial amongst nearly all Western liberal democracies. I am aware of few Western European and American governments over the last few decades that have denied the fundamental right to private property, including property in some forms of capital goods. With respect to redistribution, Jerry seems mainstream with respect to redistributive proposals that can attain moral authority. As Jerry says: “Property institutions that include significant redistributive elements certainly may be justified …” (526).

    The libertarian dictator point is complex, but I think that we can say correctly and briefly that the success of the dictator in determining the eligible set will depend on which proposals he ranks as better than no proposal. But as Jerry notes in 24.3, modest welfare states of the sort that now exist are quite good at protecting a vast range of basic liberties. This suggests that many of these policies might arguably be thought to be better than no law at all on the topic given libertarian Members of the Public. In that case, such states can achieve moral authority even over libertarians. Further, in lieu of that, libertarians are a small enough group that they may be better handled by exemptions of various sorts than allowing their concerns to defeat entire welfare state programs.

Leave a Reply