My Theories of International Relations course spent this week discussing Rousseau, a theorist whose relevance to international relations is a little unclear at first glance. Hobbes and Locke have been — if badly — imported into the canon of IR theory, largely through the use of their definitions of the state of nature as accounts of the international system. Individuals in Hobbes’ state of nature, famously, lead lives that are “solitary, poore, nasty, brutish, and short,” as they are perpetually on guard against someone else’s killing them; IR realists often use this as a description of the relations between sovereign states, notwithstanding Hobbes’ own contrary thoughts on the matter of relations between states. Locke’s state of nature, by contrast, shows up in IR liberals’ account of the international system as primarily characterized by a global commitment problem; individuals for Locke, and states for IR liberals, are rational enough entities to adhere to their contractual agreements unprompted as long as they benefit from those agreements, and problems only arise when benefits are unclear or the terms of the contract itself has to be adjudicated.
I could go on elucidating the parallel, but the point is that Hobbes and Locke have some claim to being included on an IR theory syllabus because of this (mis)appropriation of their thought by contemporary IR scholars. Rousseau is another case, since as far as I know no one uses Rousseau’s account of the state of nature to describe the international system; even Alex Wendt’s tripartite updating of views of the international system uses Kant, not Rousseau, as the alternative to Hobbes and Locke. So what are we to make of poor Rousseau, with his concerns about popular sovereignty and the problem of how to preserve the natural liberty of individuals under conditions of modern social life?
Imagine my surprise, then, when this month’s chosen article for our faculty-and-PhD-student IR theory reading group here on campus — the lead article in the 2009 issue of International Organization, co-authored by none other than chart-topping influential scholar of IR Robert O. Keohane — turned out to contain precisely the kind of reflection that would have been strengthened by a dose of Rousseau. I say “would have been” because, sadly, Rousseau is nowhere in evidence in the piece. Instead, we are treated to a somewhat stilted conceptual discussion about aspects of democracy, a discussion which then abruptly turns into a set of testable hypotheses about the correlation between the public’s attentiveness to an issue and the extent to which the issue is governed by a multilateral international organization. The problem here is that these two tasks — philosophical reflection on the character of democracy and the testing of hypothetical claims about how an issue-area is governed globally — have basically nothing to do with one another. This makes it doubly odd that Rousseau doesn’t show up, since Rousseau is very clear on the difference between an exercise in philosophical legitimation and a concrete, empirical study of some specific issue or society. Keohane to the contrary, whether some institution is democratic or not is not an empirical question, and no amount of empirical research will even in principle put an end to the philosophical question of whether some institution is democratic or not. Rousseau knew this; it’s too bad that Keohane, and most of the rest of the IR field, has forgotten it.
The central puzzle in the IO article concerns what is sometimes called the “democratic deficit” displayed by international organizations. Unlike a state government, the traditional argument runs, which is directly accountable to their public and which can be directly influenced by the public’s actions, international organizations are distant from the public and for the most part insulated from popular agitation. The people can’t vote on what the IMF or the WTO or various organs of the UN do, which makes those institutions look “undemocratic” if by democratic we mean repsponsive to the people’s moment-to-moment express wishes. Keohane and his co-authors argue that participation is actually only one component of democracy, and that participation is not even the most important component; combating special interests, protection minority rights, and encouraging collective deliberation are, if anything, more important components of democratic practice. They call this “constitutional democracy,” and suggest that the basic idea is that popular rule can be enhanced by “complex procedural requirements” (p. 9). They are obviously not the first to suggest this, and James Madison shows up fairly often in the piece, along with more modern constitutional liberals like Robert Dahl or E. E. Schattschneider. The novelty here is extending the argument beyond the boundaries of the sovereign territorial state, and suggesting that multilateral international organizations, although relatively immunized from direct popular participation, can be likewise constitutionally democratic.
Here’s the first place where Rousseau might have been helpful. On p. 15, the authors make the following rather convoluted series of claims:
While constitutional democracy in our conception emphatically does not imply that the government should act as the majority prefers at any given time (that is, it is not government by poll or plebiscite) the essence of democracy is that in the long run, after due deliberation, the people rule. It would therefore be undemocratic for an elite multilateral institution, cosmopolitan and working in what its members considered the good of all, to override repeated demonstrations of informed, rights-regarding, fairly represented popular will. This would be benign technocracy, perhaps, but not democracy.
What is convoluted here is that the authors seem to lack a solid grounding for the argument that something insulated from direct public participation can nonetheless represent rule by the people; as a result, they have to blur the boundaries by suggesting that an institution can prove its democratic character by being responsive, at least in the long run, to what the people claim to want. But this, in turn, means that the only difference between a constitutional institution and a regular one is that the constitutional institution is somewhat slower to respond — and the qualitative distinction between constitutional democracy and participatory democracy collapses. One might easily imagine any given popular movement calling for greater “democracy” when facing a multilateral international organization, being told that the organization is looking after long-run interests, and replying by simply insisting that the timeline be accelerated and the institution conform to the public expression of its will in the moment, because there is no significant or fundamental difference between responding to the people’s declared wishes now or in a few months/years. And Keohane and his coauthors explicitly reject the argument that it is sufficient for an organization to be acting in the people’s actual interests even if the people don’t know what those interests are; democracy, it appears, can mean nothing but doing what the people say that they want.
Enter Rousseau, who famously distinguishes between the sovereign and the government: the sovereign is the people assembled as a whole, whereas the government is what the sovereign establishes in order to handle day-to-day business. The sovereign, so to speak, only acts constitutionally, establishing the rules of the game and the parameters for governing; actual ruling is carried out by the government, which has to remain within the parameters established by the sovereign (which speaks with the General Will as opposed to any particular interest — indeed, as opposed to the “will of all,” i.e. what everyone says that they want at any given moment). The government derives its mandate and its authority from the act of collective, or general, will, and what makes it “democratic” is not whether it is at all responsive to the people at any given moment, but whether it is adhering to the constitutional mandate that it was given at the outset. If the people want to re-do that mandate, Rousseau suggests, all they have to do is to assemble as the sovereign, and the government automatically disbands because its jurisdiction ceases; then the sovereign can establish a new constitution and government, complete with “censorial tribunals” and other mechanisms designed to prevent the government from getting too far away from the constitution.
My point here is not that Rousseau is necessarily correct about any of this. In particular, there is a key ambiguity involving how one ascertains whether an expression of will is truly general and hence constitutional, as well as a particularly thorny problem involving the relationship of a general will to standards established by other groups of people or to claims about universally valid norms. Instead, my point is that introducing Rousseau into the discussion would help to clarify the issues involved — if the authority of a multilateral international organization can be traced to a constitutional document or expression of a general will, that puts a different spin on the whole debate. But no Rousseau in the article means no considerations of this sort, so we are left with a bit of a conceptual muddle.
The other thing that Rousseau does for the discussion is that he makes it clear that discussions about democratic legitimacy are philosophical discussions, not empirical ones. It is clearly not a realistic expectation that a government would disband simply because the people showed up as a unit and told it to disband; that said, Rousseau’s point is not that this is a feasible empirical scenario, but that the jurisdiction of the government ceases when the people assemble as the sovereign — if it remains in power, it does so by sheer force of arms, deprived of the legitimacy it enjoyed when it was operating under a popular constitution. Rousseau is not operating in the sphere of empirical facts, but in the sphere of moral principles, which is where a discussion about democratic legitimacy ought to be carried out. This is because when one boils it down, principles like “rights” and “authority” are something other than merely empirical objects. The validity of a claim to authority depends not on the simple claim itself (or, parenthetically, even on whether the claim is accepted; we can easily imagine a claim being accepted even though it is not, strictly speaking, morally correct — and it doesn’t matter which system of morality one uses to evaluate that correctness), but on whether the claim is defensible within some moral frame of reference. Whether a government is legitimate and whether a government behaves in some particular way are different kinds of issues, and Rousseau — like most political philosophers — troubles himself with questions of legitimacy, leaving questions of behavior for others.
Not so Keohane and his coauthors. After their conceptual discussion, which takes up most of the length of the article, they proceed to elucidate an empirical research agenda characterized by observable implications and testable hypotheses:
In areas of the highest priority to the public, where relevant publics are very highly organized and attentive, multilateralism will tend to be subject to more directly participatory democracy, whereas where publics are less organized and attentive, nonparticipatory mechanisms will be used.
Ignore for a moment that this formulation is basically tautological, unless there were some way to determine the public’s priorities without observing how they act in various issue-areas. And ignore the fact that this formulation shifts the focus away from whether an organization immunized from public participation is democratic to how particular issue-areas are governed by the public, and in so doing basically presumes away the entire animating question of the first two-thirds of the article (since “the public” is governing the issue-area in either case, by this definition). The most profound problem here is that this hypothetical proposition has nothing, diddly-squat, nada to do with the conceptual discussion that preceded it. The empirical proposition that publics act on their interests, and that those interests can be correlated with particular kinds of organizational outcomes and arrangements, is completely separate from the conceptual question of whether a nonparticipatory organization can be a democratic organization. One simply doesn’t matter to the other, because they operate in different conceptual spheres: whether something is democratic is a moral or philosophical (or political) question, while the effect of a certain kind and degree of public attentiveness on how an issue-area is governed is an empirical or causal (or scientific) question. Neither has any implications whatsoever for the other.
Now, it is of course always possible to claim that because democracy means being attentive to the (to steal a phrase from Madison) “permanent and aggregate interests of the community,” we can ascertain whether an institution is democratic by determining whether it upholds those interests. (The authors reject that option.) Or we could claim that democracy means responsiveness to the people’s will in the long term, and see whether an institution was democratic by determining whether it was in the long term repsonsive to the will of the people over whom it governs. But neither of these operations would settle the question of what democracy means, or whether an institution is democratic in some global or universal sense. Regardless of the results of any given empirical assessment of an institution, someone else could come in with a different definition of “democracy” and demonstrate that according to that definition, the institution either was or was not democratic. Empirical measures can’t resolve the debate unless we have prior agreement on the relevant conceptual standards; hence, empirical tests of hypotheses, or empirical traces of process, can’t tell us whether constitutional democracy is “actually” democracy — which is what the article seems to suggest. Rather, political philosophy inhabits its own sphere, separate from empirical controversies about how things factually hang together.
Just to be clear: what bothers me in the article is the fact that the authors appear to be trying to assimilate philosophical investigation/discussion to empirical research. It does not, however, follow that I think that philosophical discussion can actually resolve the question of what “democracy” is; I actually don’t think that it can, and I would rather characterize any discussion of “democracy” as a political discussion, and any resolution to that discussion as a contingent political settlement. But that’s a separate issue. My point for the moment is that I don’t think that Keohane and his coauthors can actually do what they are setting out to do, which is to resolve a philosophical controversy with empirical data. And when the lead article in the most important journal in our field, co-authored by the most influential person in our field, promulgates this kind of methodological confusion, I feel that it merits an extended response. In the end, you just can’t get there from here; the best way to get where they want to go is not to start where they start, and not to imagine that empirical social science can do things that it simply cannot, constitutively, do.