Symposium on Huemer's Problem of Political Authority, Academic Philosophy

On Legitimacy

I’ve been meaning to post on state legitimacy for a while now. Our symposium on Michael Huemer’s fine book seems like as good a reason to do this as any. So here goes.

Many philosophers who talk about legitimacy understand legitimacy in a way that is implausible. Most understand legitimacy as consisting of either the right to be obeyed and the right to enforce the law, or (weaker) as only the right to enforce. [1]

In this simple form, neither conception will do. They lead to at least two problems. I will focus here on the weaker conception since what it is true of it will also be true of the stronger view.

Problem #1: A Dilemma

Suppose a legitimate really does have the permission to coercively impose all the laws on its subjects. This view leads to a dilemma. We face one horn of the dilemma if states at least somewhat like the ones we live among can be legitimate. For in that case everything such a (imaginable) state would do would be permissible. But such a state will inevitably be occupied and run by people that are fallible (not to mention corruptible, whatever that may mean). This means that such a state will end up doing things that are under ordinary circumstances unjust.

If the state is legitimate, then it will be permitted to do these otherwise unjust things. And if it is permitted to do them, then it cannot be wrong for it to do them. And if it is not wrong for the state to do these things, then its doing them cannot be unjust. It follows, then, that this conception of legitimacy makes it impossible for a legitimate state to commit an injustice, pass an unjust law, or otherwise do wrong. It also follows that subjects of a legitimate state cannot have rights that their state could violate. This is ridiculous.

Suppose, then, that we accept what is obviously true: all existing states occasionally do things that are morally wrong, unjust, and otherwise regrettable. This makes us face the other horn of the dilemma. For according to the present conception of legitimacy it follows from this basic observation that no actual state can be legitimate. After all, any such state will commit injustices. And if a legitimate state has the permission to use coercion enforcing all its laws, any such state will not be legitimate.

But surely a standard for legitimacy is too demanding if it renders states illegitimate merely for being a human creation, run by human beings. If this is what legitimacy means, then saying that states are illegitimate is not very interesting.

Problem #2: Narrowing the gap between justice and legitimacy

On either horn of the above dilemma, we end up equating a state’s legitimacy with the justification of its policies, laws, and actions. To be a legitimate state is, on that view, to be a state that is justified in undertaking its activities.

This brings us to the second problem. Such a view narrows the distance between judgments of a state’s legitimacy and justification. This is, as A. John Simmons has showed in perhaps the most important article published on this topic, a problem. These judgments track significant and different dimensions of moral evaluation. Justice concerns the content of individual laws. Legitimacy concerns the standing of the institution issuing them. Showing that everything a state does is justified does not suffice for showing that it is legitimate. And showing that a state is legitimate cannot be sufficient to show that all its activities are permissible.

By reducing the gap between justification and legitimacy, we risk losing sight of one of these dimensions. We might overlook the importance of evaluating laws on their specific merits (as many who defend an obligation to obey the law tend to do). Or we might overlook that a state’s right to rule is not proportional to the justice of its laws (as many public reason theorists who equate legitimacy with the justifiability of laws tend to do).

One thing I like about Huemer’s book is that it offers a nice improvement over this sorry state of affairs. Here is how Huemer defines legitimacy:

the right, on the part of a government, to make certain sorts of laws and enforce them by coercion against the members of its society – in short, the right to rule. (5)

For Huemer a state is legitimate if it is morally justified (permitted) to issue and enforce at least some of a state’s laws. This is better because it allows that a state could be legitimate and still not be justified to do everything it does. That is, Huemer allows that a legitimate state can have the right to rule and yet do things that are, quite simply, wrong. This is an important insight.

But while this conception of legitimacy is better than the one philosophers usually adopt, it also undercuts Huemer’s argument. Consider the following list Huemer offers (pp. 95-6). It includes nine types of laws that states typically undertake:

  1. Laws designed to protect citizens’ rights
  2. Policies designed to provide public goods
  3. Paternalistic laws
  4. Moralistic laws
  5. Policies designed to aid the poor
  6. Rent-seeking policies
  7. Laws designed to secure the state’s monopoly and promote its power and wealth
  8. Policies designed to promote other things that are regarded as good in general, including education, art, etc.
  9. Laws and policies that seem motivated by emotion, such as limits to immigration and bans on gay marriage

Huemer thinks that only laws (1) and (2) and possibly (7) are justified. I suppose many people would think that (5) should be included as well, although Huemer has interesting objections here. No matter – our concern is Huemer’s argument for anarchism. Here’s what Huemer writes shortly after presenting the list above:

If the range of coercive actions that the state is actually entitled to take is only a small fraction of what it is generally thought to be entitled to do and of what the state in fact does, then I think the state does not truly have legitimate authority. (p. 98)

But this simply does not follow. If state legitimacy entails only the permission to enforce some of the state’s laws, then we cannot conclude that states that undertake (1) through (9) are illegitimate. It might be that some of these classes of laws are more relevant to legitimacy than others. Perhaps a state that is really good at (1) and (2) can survive some unjustified laws in the other classes. Perhaps the injustices must not just be many but severe. Perhaps some other condition is to be met. We simply do not know.

So here’s the point. Legitimacy is the right to rule. And like many rights, that right will give the state (if it has it) a right to do wrong. That is, a state’s right to rule is consistent with (some of) its activities being unjustified. The most important question therefore is precisely what activities are relevant to judgments of legitimacy. A book arguing for the illegitimacy of all states must answer this question. Huemer here does better than most. But he still does not do good enough.


[1] Here’s one representative example. In his Philosophy and Public Affairs article “Liberalism, Samaritanism, and Political Legitimacy” Christopher Wellman writes: “An account of political legitimacy explains why [state] coercion is permissible. In doing so, it explains why the state has a right to coerce its citizens and, correlatively, why its citizens have no right to be free from this coercion.” (p. 211-2)

Published on:
Author: Bas van der Vossen
  • Jason Brennan

    “So here’s the point. Legitimacy is the right to rule. And like many rights, that right will give the state (if it has it) a right to do wrong”

    I think this is a good analysis, if we understand “right” here as a permission right and not a claim right. I’d want to make the issue of a claim right separate, and use the term “authority” for that.

    Legitimacy gives states permission to stand and to create and enforce rules (within some scope over some people over some area).

    Authority grounds a duty to obey.

    I think it’s plausible that some states have legitimacy, but I don’t accept that any have authority.

    • Murali

      You don’t agree with Renzo’s account of authority?

    • Basvandervossen

      I think I agree. With a small caveat: legitimacy (the right to rule) is most likely at the same time also a claim-right against outsiders that they not interfere, etc.

  • John Phillips

    As I understand it legitimacy (of the Simmons variety) is separate from political authority, which is what I understand as the right to create content independent duties of obedience and perhaps enforce them, but that’s terminology for you. One thing I though was odd about your post Bas:

    “And if it is permitted to do them, then it cannot be wrong for it to do them.” This moves too fast. I have a right to do all kinds of things it is morally wrong for me to do. Perhaps states work this way as well.

    To say that the state has a right to do it does not mean that it is right. If the state does have the right to pass a law, then this imposes constraints on what I may do and these bind my actions. I am morally required to perform/refrain from the prescribed behavior (assuming the injustice isn’t too great). However, they need not bind in conscience. I don’t have to agree or think that this is the right law. I can rant and rave, agitate for legal reform, etc, I have not conceded that the law is right, merely that I must obey it because it (or its issuer) has the requisite characteristics.

    I will say this. The problem with discussions of justice and legitimacy is that a lot of people (Plato for example) take Justice to mean something like “overall rightness” in a society whereas others take it to mean something much more limited. If it means “overall rightness” then the state can’t have a right to be unjust, but if justice means something nearer to “the rightness of the content of the law” irrespective of who passed it or how it was passed, then the problems you outline seem less urgent to me.

    • Basvandervossen

      I’m quite confused about what your thoughts are. But insofar as I understand this, we may be arguing for similar points. So let me clarify: I think that states have a right to rule that includes the right to do wrong. Those wrong things are things a state is not permitted to do (by definition). What does a state’s right to do wrong come to? A few things come to mind: a claim-right that others not interfere with the state even though it’s acting wrongfully (see the Law & Phil. paper I linked to in the post), perhaps a claim-right that disobedience to unjust laws be done in ways that does not harm the legitimate state’s justified actions, etc.

  • Christopher Morris

    Interesting argument, Bas. I don’t like the idea of identifying legitimacy with a right; rather, legitimacy is a particular status possession of which grants the holder certain liberties, claim-rights, powers. That’s a detail. It seems as if what’s important for your argument is that common idea that a claim-right allows one permissibly to do something that is wrong. Now suppose that there are only two kinds of acts that states can perform, and only the first are acts that legitimate state may perform (illegitimate states may do nothing). The legitimate state’s right to rule permits it to perform acts of the second (impermissible) form only if its agents believe they are performing acts of the first kind. So, for instance, if paternalistic acts are not something a legitimate state has the right to perform, then if that states enacts a paternalistic law understanding it as paternalistic, then it does something it may not do; it exceeds its authority. So I should think that if we find a state performing lots and lots of the impermissible acts and we concluded that it is not confused, thinking that these acts are instances of the permissible kind, then we may rightly conclude that this state has lost its right to rule.

    • Basvandervossen

      I agree on the detail. Which, albeit a detail, matters. So thanks.

      One point of disagreement. I would not say that a “legitimate state’s right to rule permits it to perform acts of the second (impermissible) form”. It cannot be permissible to do things that are impermissible. But legitimate states have the right to do wrong in at least two senses: (a) the right to rule can survive doing things that are wrong; and (b) the right to rule can include a claim-right that protects the state’s status (as you helpfully put it) even when it is doing wrong. (See my reply to John Philips above for some details if you’re interested.)

      I think that allows us to steer clear of the swamp of objective/subjective justification that your invocation of beliefs invites. I’m not sure I accept the distinctions you there draw. (Then again, I’m not sure I reject it either.)

  • Murali

    Bas, what about a purely positivist account of legitimacy where a government is legitimate iff there is a rule of recognition acknowledged by the people designating that body as the government?

    • Kevin

      “Consent of the governed” is a good start for a moral standard. Of course, there are a bunch of details related to the granularity of consent, like: must it be acknowledged by all the people or just a super-majority or majority? And is the designated body legitimate regardless of whether they act within the acknowledged (e.g. constitutional) constraints on their governing?

      But there are two standards there: (1) a constitutional standard and (2) individual agreement. A government may be illegitimate relative to its own constitution and yet individuals may lend legitimacy by not leaving for lack of a better alternative.

      The basic problem is that “legitimacy” is always relative to some standard, and Bas doesn’t seem to pick one. It seems like he’s going for a moral standard, but he still has to define it.

    • Basvandervossen

      Unless the rule of recognition somehow has morally relevant features capable of justifying the state’s right to rule, I would not accept that positivist account. (So yes, as Kevin put it, I’m interested in the moral issue of legitimacy here.)

      Also, the mere acknowledgment of a rule by subjects is not sufficient for legitimacy.

  • Chris Bertram

    Isn’t one option here to adopt a threshold conception of legitimacy such that serious injustices undermine the state’s legitimacy but wrongs below the threshold do not? So, a state that engages in serious and systematic human rights abuses cannot be legitimate, but one that fails to live up to the best-supported standard of distributive justice (ymmv) nevertheless can be. Obviously, there would be a lot of detail to argue about here …. On such a standard at least some existing (liberal democratic) states would count as legitimate (Iceland?) but arguably others such as the US (drone strikes, penal policy, mass surveillance) would not.

    • Basvandervossen

      That’s more or less the view I think is right, Chris. But notice that if you accept that view and you wish to make judgments on the legitimacy of existing states, we need to know what the threshold is.

  • Chris Bertram

    I also think you should worry about this inference: “And if it is permitted to do them, then it cannot be wrong for it to do them.” I accept that the inference seems a natural one, but I’d suggest that – justice being the complex thing that it is – we might want to look not just at permissions but at whether an agent exercises those permissions as a whole in ways consistent with the duty of justice. Here’s an example (though it may not be the best one): I have a permission to decide which shops I give my business to, but if I use that permission to boycott all and only Jewish shops I am acting wrongly.

    • Basvandervossen

      Well, in that case it seems natural to me to say you are not permitted to boycott all and only Jewish shops.

      • Chris Bertram

        Well indeed. But I’m presuming that the permissions are set by rules and that it will always turn out to be possible to act in ways that are consistent with the rules but that we nevertheless judge to be unjust.

  • Chris Bertram

    And on Huemer’s list of laws (which includes categories of which he disapproves), a missing category is laws designed to maintain equality of status, or the fair value of political liberties (an important category for relational egalitarians such as Elizabeth Anderson). I suppose you might class them under 1, 2 or 5, maybe 8, but they seem sui generis to me.

  • shemsky

    To those of us who believe that individuals are sovereign, and that individuals must each consent to be governed, and that no individual may legitimately be forced to give up their sovereignty for the sake of others, these arguments are old and tired nonsense. Regarding Wellman’s article, I reject the idea that A has any obligation to associate with or submit to B and C because it would benefit B and C. The only obligation I recognize that A has to B and C is to refrain from harming them. I also reject the idea that government must be territorial. Governments deriving their authority by individual consent may tend towards territorial govenments, but I believe it is a mistake to make it an absolute requirement.