Our IO article “Democracy-Enhancing Multilateralism”Footnote 1 reframes the ongoing debate over the consequences of multilateralism for democracy by challenging conventional claims about the “democratic deficit” problem in global governance. First, we argue that democracy is an internally complex political ideal that should not be identified simply with mass electoral accountability. We defend instead a constitutional conception of democracy, in which properly constituted democratic political institutions promote informed deliberation, faction control, and minority rights protection, as well as political accountability. Second, we argue that the impact of various multilateral organizations (MLOs) on constitutional democracy depends on how particular MLOs interact with domestic democratic institutions within specific issue-areas. These two arguments lead us to question the conventional wisdom that MLOs are necessarily democracy-degrading because they are not directly participatory. We maintain that assessing whether a multilateral commitment degrades or enhances domestic democracy requires careful empirical inquiry.
We appreciate the thoughtful response by Gartzke and Naoi (GN). They agree with us that MLOs are significant enough to make a real difference for domestic democracy. Like us, they reject the crude realist claim that MLOs do not matter. They agree that MLOs should be viewed as thoroughly political: MLOs are not technocratic but are governed by states and serve complex sets of private and public interests. Most importantly, we appreciate GN's effort to advance the debate in a constructive direction by emphasizing the need for further empirical research.
Yet on four important issues we diverge. GN misstate our basic position on how multilateralism affects democracy; fail to engage with our multidimensional conception of democracy; overlook the distinction, central to Madisonian constitutional theory, between interest groups that support the general interest and those that do not; and overgeneralize from existing research. To their credit, however, GN point to some areas in which we could further elaborate our argument, and in closing we suggest some possibilities for doing so.
Multilateralism and Democracy: The Need for Empirical Nuance
GN attribute to us the claim that multilateralism has uniformly positive consequences for domestic democracy—an argument they then seek to “disprove” with scattered counterexamples. This is a misreading. We do say that “multilateral institutions may (and frequently do) enhance the workings of domestic democracy in established democracies.”Footnote 2 But we stress that this effect is not uniform. Our conclusion is very clear: “We emphatically do not claim that multilateralism always enhances domestic democracy.”Footnote 3 For example, we describe antiterrorism measures taken by the UN Security Council as examples of a multilateral institution promoting “antidemocratic norms.”Footnote 4 Indeed, we recommend empirical analysis precisely to explain variation in democratic consequences of multilateralism, and present our approach explicitly as a critical policy tool to facilitate the assessment and reform of existing institutions. Hence, when GN offer examples of multilateralism undermining domestic democracy, they in no way challenge our central argument.
Similarly, GN are incorrect to claim that we “[advance] MLOs as a blanket remedy for domestic tyranny.”Footnote 5 We do not argue that the primary purpose or effect of most multilateral institutions is democracy promotion. Governments rarely construct multilateral institutions with the intent to improve domestic democracy. Expanding international cooperation in areas such as trade, the environment, and human rights is a response to functional incentives of globalization. The resulting institutional form of this policy coordination is highly contextual, depending heavily on distributions of interests and power in the world. Therefore, many of GN's criticisms are irrelevant. They take us to task for failing to demonstrate that multilateralism is superior as a method of democracy improvement as compared with other means, such as domestic reform or unilateral or bilateral foreign policy. Such criticism is misdirected, because we make no such claims.
The Constitutional Conception: Why Democracy Is Multidimensional
Our constitutional conception of democracy holds that human rights, control over factions, and well-informed deliberation, like electoral accountability, are constitutive elements of democracy because they help realize the ideal of collective self-government among political equals, based on the interests of all. Publics delegate various powers to institutions insulated from direct electoral accountability for the sake of improving democratic performance. On this understanding, periodic mass elections and opportunities for public participation are indeed crucial features of democracy, but they are far from sufficient. The fact that majoritarianism is balanced by other values helps explain why people regard democracy as an ideal worth striving for.
GN assert instead that protection of minority rights and interests, well-informed debate, and control over special interest factions (along with rule of law) are not essential elements of democracy, but “social benefits” that democrats can take or leave. They further claim that representation is not merely one among a number of democratic values, but the sole “defining and necessary attribute of democracy.”Footnote 6 Although these claims are central to their critique, GN misstate our position and fail to clarify their own.
GN's central error is to conflate popular participation and representation, as shown by their misstatement of our argument. “Representation,” they write, “is not, as KMM argue, ‘only one among a number of political values to be balanced in a well-ordered constitutional democracy… .’”Footnote 7 What we actually wrote was that “popular participation is only one among a number of political values to be balanced in a well-ordered constitutional democracy.”Footnote 8 Limiting popular participation, as we explain in the article, can often help make democratic systems more broadly representative: by blocking the power of special interest factions, by representing minority interests that might otherwise be dominated by majorities, and by improving a political system's capacity for well-informed deliberation.
This misunderstanding on GN's part is of critical conceptual importance. If representation is the cardinal goal of democracy, as they claim, then it makes little sense to speak, as they do throughout their critique, of representation as an alternative to suppressing special interests, protecting minority rights, or ensuring better deliberation. They ought, rather, to consider suppressing special interests, protecting minority rights, and improving deliberation as means to more inclusive and effective representation. This conceptual confusion infects GN's empirical analysis as well. GN wrongly attribute to us the view that “MLO intervention makes democracy less representative.”Footnote 9 This again is misleading. What they appear to mean is that multilateral regimes often limit opportunities for direct popular participation in policymaking. Still, by suppressing factions or helping to protect the basic rights of minorities, multilateralism may render policy outcomes more representative—as they admit is the case in trade policy. The net result may be to enhance or degrade democracy, but any such judgment requires both empirical investigation and normative analysis.
The lesson here is that proper comparative institutional evaluation requires not just subtle empirical investigation, but philosophical clarity about constitutional and democratic theory. We adopt the constitutional conception of democracy because it seems to us familiar and normatively attractive. If GN want to propose an alternative theory of democracy, they are welcome to do so. Until then, our conception seems to us a more philosophically coherent and pragmatically useful starting point for comparative analysis.
Madison, Special Interests, and Democracy
GN criticize our discussion of “special interests” or “factions” in domestic democracy, but they misstate our central Madisonian premise. Not all organized interests are factions, in Madison's sense, but only those with interests opposed to “rights” or “to the permanent and aggregate interests of the community.” Sound constitutional engineering, following Madison, rests on the ability to manage special interests in a way that reduces the power of factions and promotes the interests of the whole. Consider the example of trade, which GN cite as evidence that international institutions simply empower one set of special interests against another. Of course proliberal trade interests are “special interests” in one sense: these firms seek to increase their sales and profits. But in the Madisonian sense they need not be: in those cases where international cooperation helps offset a preexisting domestic bias toward the representation of protectionist interests, free trade interests are not opposed to those of the whole. Much sound pragmatic constitutional engineering in a democracy, according to Madison, lies in crafting institutions that are likely to have such effects.
In one respect, however, we agree with GN and think that they point in a productive research direction. GN argue correctly that MLOs, in addition to producing public goods, frequently also have distributional impacts, making them likely subjects of special interest pressure or influence.Footnote 10 MLOs might be corrupted or “bought off” outright, but more likely they may simply respond unduly to pressures from strong special interests, exercised through governments. In our view, this is the most important point that GN make, which suggests an important research direction (including for students looking for dissertation topics): under what conditions are MLOs themselves, or governments acting through MLOs, more or less subject to pressure from special interests (“factions”) than well-established democratic governments acting alone?
The Need for Contextual Empirical Research
We have consistently maintained that the varied implications of multilateralism for democracy can only be assessed through careful empirical inquiry, guided by a philosophically sound conception of democracy. We argue that scholars should examine the impact of multilateralism on domestic processes and outcomes, and that it is critical to determine when MLOs are more prone to capture by special interests, tyrannical majorities, or uninformed consensus than national governments, and when, conversely, involvement with MLOs can help domestic governments resist such capture. Such an inquiry requires theory-guided empirical analysis comparing, within specific issue-areas, the real-world functioning of autonomous domestic democratic institutions with the functioning of those domestic institutions as influenced by multilateral rules and regimes.
Such contextual analysis must be conducted with empirical and theoretical sensitivity. Valid claims about the relationship between public preferences and political outcomes are likely to be specific to particular institutional, issue, and international settings. It seems highly unlikely that scholars will arrive at sound blanket generalizations that all MLOs systematically undermine or enhance democracy. The question of whether multilateralism as such is democracy-degrading or democracy-enhancing is thus underspecified; MLOs, domestic societies, and issues are simply too diverse to support blanket claims. Instead, we call for detailed policy-relevant empirical research on the effects of the interaction between specific domestic political systems and particular multilateral forms within issue-areas—research that we and, we hope, others are now conducting.
Conclusion
GN call for a further elaboration of our theory, method, and empirics. We hope to do just that in the future, though not within the limited space allocated for this response. Three broad observations must suffice.
First, we now believe that our normative categories should be refined. In particular, we need to think more about the conceptual relationship among participation, representation, and accountability. These values are distinct, and we may need to rethink the relationship between participation, on the one hand, and representation and accountability, on the other. Representation and accountability are crucial to modern constitutional democracy and also to the assessment of MLOs, but the role of participation is less clear.
Direct popular participation can make public officials more responsive to broad public views, but the differential cost of participation may also amplify the voices and influence of special interests or zealots. Similarly, participation can enhance or degrade the quality of deliberation. Majoritarian participation often exists in similar tension with minority rights. For the sake of assessing MLOs' impact on democracy, participation might thus best be conceived of as an institutional mechanism or feature of social practices, rather than as a distinct and basic democratic value. We do not believe MLOs necessarily degrade any of these values, but clear measurement of any such effects requires more clearly defined concepts and theories.
Second, in thinking about the impact of MLOs on democracy, it now seems to us that we should not view specific institutions simply at one point in time or in isolation. It is often impossible to assess the effects of particular multilateral regimes without looking at their “life cycles”: as in domestic constitutional systems, the democracy-enhancing or democracy-degrading effects of multilateralism may be usefully distinguished and studied across successive phases.
Finally, multilateral institutions interact in complex ways not only with domestic politics but with one another. Particular MLOs might tend to counteract or magnify the effects on democracy (whether positive or negative) of other MLOs. Discerning the effects not only of particular MLOs but of the larger global institutional architecture will require careful empirical study, in which we would welcome the further engagement of Professors Gartzke and Naoi.
