When people think of democracy nowadays, they tend to think of it primarily in terms of ‘one person one vote’ – and perhaps (if particularly sophisticated) ‘one vote one value’. They tend to think in terms of enfranchising all affected interests. Call that the ‘Reform-Act model of democracy’, in honor of nineteenth century British legislation expanding the franchise.
That Reform-Act model of democracy is, at root, an electorally oriented, vote-centric vision of what democracy is all about. Academic political theorists point to all sorts of reasons for thinking that merely voting is not all that democracy is about, but for present purposes I eschew all those subtleties. Here, my focus will be on practical politics. In popular political discourse, the Reform-Act model is, for better or worse, the thumbnail version of democracy that is most readily available both for popular consumption and for political implementation.
To appreciate the pervasiveness of that electorally oriented, vote-centric model of democracy just note these examples:
• That is the model of democracy that is most celebrated by liberal internationalists. In awarding the Nobel Peace Prize to Jimmy Carter for ‘his decades of untiring efforts to… advance democracy’, the Norwegian Nobel Committee (2002) pointed particularly to his service ‘as an observer at countless elections all over the world’.Footnote 1 Note similarly that ‘encourag[ing] popular participation and support for free and fair elections’ comes second on Secretary-General Ban Ki-Moon’s (Reference Ban2009: 5) list of priorities for UN democracy assistance.
• That is the model of democracy that neo-conservatives would impose on the rest of the world, by force of arms if necessary. The first thing said under the heading of ‘Democratic Governance’ in the National Endowment for Democracy’s (2007) ‘Statement of Principles and Objectives’ is: ‘Democracy requires a system of representative government in which leaders are chosen in freely contested fair and periodic elections…’.
• That is the model that lies at the heart of the emerging ‘right to democratic governance’ at international law. There, ‘a preoccupation with elections is a striking feature’ – to such an extent that ‘to raise the question of democracy in international law [just] is largely to raise the question [of] whether international law requires states to hold periodic and genuine elections’ (Crawford and Marks, Reference Crawford and Marks1998: 80; cf. Franck, Reference Franck1992).
The italicization in all those cases is mine. The emphases, however, are very much those of the sources cited. Wherever we look, the electorally oriented, vote-centric model really does seem to dominate practical political discourse on democracy.
In discussions of how to democratize international institutions themselves, that same emphasis on fully enfranchised electoral democracy is also much in evidence. That has long been the preferred model of strong internationalists.Footnote 2 When discussing what shape the new United Nations should take, British Foreign Secretary Ernest Bevin (Reference Bevin1945: 785) urged ‘there should be … study of a house directly elected by the people of the world…’. A half century later, the pre-eminent theorist of cosmopolitan democracy still sees, as ‘an unavoidable institutional requirement’ of that, ‘the establishment of an independent [UN] assembly of democratic peoples’, directly elected by them and accountable to them’ (Held, Reference Held1995: 273).Footnote 3
Any Second Assembly, apportioned according to population as a counterweight to the General Assembly’s ‘one country one vote’, may initially be elected indirectly, consisting of representatives elected by existing national legislatures. Initially, it may be merely a toothless ‘talking shop’. Eventually, however, the aspiration is very much for that Second Assembly to become a directly elected People’s Assembly with real power.
That proposal was first elaborated in its current form in 1982, at a UN Special Session on Disarmament of all places (Segall, Reference Segall1982, Reference Segall1991). Since then, it has attracted the support from a bevy of Non-Governmental Organizations (NGOs), from the United Nations Development Program (1999: 111), and from a range of public intellectuals.Footnote 4 Conspicuous among the latter is Richard Falk, who with Andrew Strauss wrote an influential (Reference Falk and Strauss2001) Foreign Affairs article in support of the proposal. There they make the case for a ‘popularly elected global assembly’ – one that ‘would represent individuals and society instead of states’. They envisage that as involving:
• the establishment of electoral districts throughout the world…
• global voter rolls…
• a system of campaign finance and other election rules
• [safeguards against] attempts to manipulate or undermine elections… (Falk and Strauss, Reference Falk and Strauss2000: 217, 219).
Clearly, what they propose is nothing short of a ‘Global Reform Act’.
As an ultimate goal, that is one with which I have considerable sympathy. Philosophically, enfranchising all affected interests surely is the right way of constituting the demos from a democratic point of view (cf. Gould, Reference Gould2004: Ch. 7; Goodin, Reference Goodin2007). In addition, in today’s world, just about everyone really is affected (or at least potentially affected)Footnote 5 by just about everyone’s decisions and choices, in ways that ought to entitle them to a say in those decisions. Therefore, a Global Reform Act is precisely what democratic principles should prescribe.
In practice, however, such suggestions are hopelessly visionary – or so self-styled realists insist. They see them as, at best, very long-term aspirations that are not remotely realistic in any near future.Footnote 6 As Joseph Nye (Reference Nye2001: 4) says, in starkly realist mode: ‘Alfred, Lord Tennyson’s “Parliament of Man” made for great Victorian poetry, but it does not stand up to contemporary political analysis’.Footnote 7
The pronouncements of self-styled realists ought always be taken with a grain of salt. They often tend to be too realistic. Realists often fold their highest moral aspirations far too quickly in the face of obstacles that might well have been eminently surmountable, had only they tried (Goodin, Reference Goodin1992). Speaking as president of the UN General Assembly, Edvard Hambro wisely remarked, ‘Politics should be the art to make possible tomorrow what seems impossible today’ (quoted in Kuper, Reference Kuper2004: 45). Long before, Max Weber had said the same at the end of his famous lecture on ‘Politics as a Vocation’.Footnote 8
Generally skeptical, though I thus ordinarily am of realists’ excessive realism, on the present point I fear they are almost certainly correct. It is simply too early to be thinking of democratizing the global order along Reform-Act lines.Footnote 9 But remember – and this is the key point of this article – the Reform Acts came very late in the process of democratizing the domestic polity. Those were the very last steps, not the first steps, in that long process.Footnote 10
When it comes to the global polity, we are still very much in the early days – both of developing a global polity, and still more of democratizing it. What we should be looking for in that context are ‘first steps’, not final steps. It should be no cause for despair that we are not (yet) in a position to undertake the sorts of reforms that would mark the culmination of democratization on a global scale.Footnote 11
When looking for a template for democratizing our emerging global institutions, we should not be looking at how democracy has developed ‘at the end of the day’ (Fukayama, Reference Fukayama2006). Instead, we should be looking at how democracy developed ‘in the beginning’. Modeling like-on-like (Hui, Reference Hui2001), we should be looking at domestic developments, not in the nineteenth century, but instead five or six centuries earlier.
Democratization before the vote, domestically
Pulling us out of that nineteenth-century Reform-Act mindset, and casting our minds back to those earlier periods, what were the key developments that contributed most crucially to democratization in the very beginning?
Two steps stand outFootnote 12:
• The first was curbing the arbitrary exercise of power.Footnote 13
• The second was rendering power-holders accountable to someone or another.
Of course, the way in which we eventually came to curb the arbitrary exercise of power was to make power holders accountable. Therefore, nowadays those two steps have come to be conflated. But note well: that was a later conflation of two steps that, at the beginning, were clearly distinct. In the very first instance, what was essential to repudiate the principle that the sovereign’s word was law, and to establish instead that the sovereign ruled under law. That, roughly, was the accomplishment of the Magna Carta of 1215, as it was interpreted and reinterpreted over the years.Footnote 14
Once it was established as a matter of principle that the sovereign could not exercise power in any arbitrary way he (or occasionally she) pleased, discussion could then turn to questions of how best to embed that principle in practice. Making the sovereign accountable to someone or another – requiring the sovereign to give reasons for acting as he or she had done – was seen to be one way of doing that. Establishing that was, roughly, the accomplishment of the Bill of Rights of 1689, as that was interpreted and reinterpreted over the years.Footnote 15
Historically, those were two distinct steps separated in Britain by over four centuries. In between times, arbitrary exercises of sovereign power were prohibited, but it was beyond the power of any human agent to sanction them.Footnote 16 Not only were non-arbitrariness and accountability separated historically. They are logically distinct, as well. Requiring that people have reasons is clearly distinct from requiring them to give reasons. As a purely contingent matter, of course, requiring people to give reasons (through some accountability mechanism) might be the most effective practical way of ensuring that they actually have reasons, and are not just acting on some arbitrary whim. But that is a practical convenience, not a logical necessity.
Furthermore, there is a natural sequence to those two requirements.Footnote 17 Certain political rights presuppose – practically, if not strictly logically – certain civil rights. A right to vote pragmatically presupposes, for example, a right of habeas corpus. No election is free and fair if rulers can lock up opposition supporters to prevent them from voting. Giving people a right to demand that they be charged before a court of law protects against (among many other things) arbitrary arrests being used in this way to alter election outcomes.Footnote 18
More fundamentally, unless the sovereign is bound by the rule of law, there is no pragmatic point in anyone else’s having the right to vote to determine the content of that law.Footnote 19 Requiring an election without requiring that rulers be bound by its outcome would be a nonsense, or anyway a sham. In this most fundamental respect, the rule of law truly is a prerequisite, and not a merely accidental precursor, to democratic political rights of more recognizably Reform-Act sorts.
Whether there is some similarly inexorable logic dictating where it all must necessarily end up – whether it is somehow inevitable that curbs on the arbitrary exercise of power and institutions holding the powerful to account must necessarily end in democracy of the fullest Reform-Act sort – is a question to which I shall return below.
First steps toward global democracy
With these domestic analogies in place, let us return to the topic of global democracy. Similar things seem to be happening in today’s international order as happened centuries ago in the domestic sphere to curtail the arbitrary exercise of power and to make it accountable. Of course, there is no reason to suppose that these developments need necessarily follow exactly the same path everywhere, internationally as well as domestically. Interestingly enough, however, international developments do seem to be following uncannily similar paths to those domestic ones long ago.
Curbing arbitrary exercises of power
As UN Secretary General, Kofi Annan (Reference Annan2006) was fond of quoting a remark of President Truman’s (Reference Truman1945) to the closing session of the San Francisco conference that founded the United Nations. There Truman said, ‘We all have to recognize – no matter how great our strength – that we must deny ourselves the license to do always as we please’ (quoted in Annan, Reference Annan2006).
Doubtless the UN Secretary General was so fond of those words for a reason. US presidents – and power-holders more generally – clearly need to be reminded of that home truth at least once every generation. However often forgotten, the principle was nonetheless clearly established back then, and it is increasingly being enshrined in international law and international political practice today.
At the level of international practice, we can point to a plethora of treaty regimes with increasingly strong enforcement mechanisms. Emblematic of this is the shift from the voluntary dispute settlement mechanism found in the General Agreement on Trade and Tariffs to the mandatory one found in World Trade Organization.Footnote 20 One could equally well point to the International Criminal Court as acknowledgment (at least among the treaty’s signatories) of norms of international law that are binding even on notional sovereigns in the state system.
Treaty regimes, however, are just the formalized tip of the international legal iceberg. If our concern is with the emergence of norms curbing arbitrary exercises of power by state actors, an even better place to look might be to what international lawyers call jus cogens. Treaties are things that sovereign states negotiate among themselves, and they bind only those who sign them. Jus cogens, in contrast, does not require the consent of any particular state before it is legally binding on that state. Shifting from treaties to jus cogens as a basis of international law marks a ‘shift from consent to consensus as the basic source of international law’ (Bull, Reference Bull2002: 150).Footnote 21
Thus, in addition to international conventions and international custom, sources of international law under jus cogens also include ‘the general principles of law recognized by civilized nations’ (even if this particular state does not recognize that particular principle) and ‘judicial decisions and the teaching of the most highly qualified publicists of the various nations’ (even if states generally fail to recognize them).Footnote 22 Prohibitions on the use of force, on genocide, and on gross violations of human rights have all become part of binding jus cogens in just that way. They apply even to states that have not ratified treaties on those topics.
More generally, there is movement toward a ‘rule of law’ within the international sphere.Footnote 23 That does not necessarily presuppose any centralized lawgiver. Instead, the rule of law emerges out of customary international law in the same dispersed way the common law emerged domestically. It is inevitably ‘soft law’, at least in its early stages, based on vague norms rather than precise enforceable edicts with authorized interpreters (Abbott and Snidal, Reference Abbott and Snidal2000; Abbott et al., Reference Abbott, Keohane, Moravcsik, Slaughter and Snidal2000). In the very first instance it is far from being universally acknowledged. But the more states that are party to those arrangements, the more advantageous it often is for others to join in those arrangements and the more disadvantageous it is for others to remain outside them.Footnote 24 Therefore, the cooperative network naturally expands, eventually encompassing virtually all countries in ways I shall further elaborate below.
Of course, when it comes to imposing rules of international law on powerful states that insistently do not want any part of them, questions of enforceability inevitably arise. Paraphrasing Andrew Jackson’s riposte to the US Supreme Court, they can always just adopt the stance: ‘The World Court has made its decision; now let’s see the World Court enforce it’ (Warren, Reference Warren1926: vol. 1, 759). The US famously took just the same stance toward the International Court of Justice’s (1986) decision in Nicaragua vs. United States. Therefore, if a global superpower is hell-bent on throwing its weight around in completely arbitrary and illegitimate ways, it can still do so even in the face of all this emerging jus cogens and international rule of law.
But before despairing too deeply about that, remember that we were in broadly the same position domestically for a couple of centuries after the Magna Carta. The arbitrary exercise of power had been repudiated. It had been accepted that the sovereign’s word was not itself law; it had been accepted that, instead, the sovereign ruled under law. It simply took a little more time before any practical way was found for mortal men to (as was said back then) ‘put a bridle on him’ (McIlwain, Reference McIlwain1947: 69). Still, the point having been established, the path was set.
So, too at the international level, ‘almost all nations observe almost all principles of international law and almost all their obligations almost all of the time’, even if there is no one who can strictly make them do so. There is every reason to hope that may continue to be true even as the scope of international law expands from the strictly contractual to the more broadly consensual (Henkin, Reference Henkin1968: 42; Koh, Reference Koh1997). It is to those reasons that I turn, next.
Mutual accountability via networked governance
In today’s international society, there is also a flowering of accountability mechanisms analogous to those in the earlier stages of domestic democratization (Grant and Keohane, Reference Grant and Keohane2005: 36). Virtually none of them is itself directly democratic, in the first instance. But they are nonetheless valuable as accountability mechanisms, for that fact. In addition, as I shall argue below, there is a natural dynamic by which, once such mechanisms are put in place, the circle of those to whom the powerful are accountable almost inevitably expands over time.
The early thought, domestically, was that the sovereign must consult ‘his curia…, and the earls and barons who are his associates there’. In addition, as one early slogan put it, ‘one who has an associate has a master’Footnote 25 – or anyway he has an equal to whom he is accountable.
The basic idea there was one of mutual accountability among associates joined in cooperative networks to pursue shared (or anyway parallel) purposes. That same sort of mutual accountability within networks increasingly characterizes relations among states, International Organizations, Non-Governmental Organizations and Inter-Governmental Organizations in today’s increasingly globalized world.Footnote 26
Policy networks, epistemic communities and professional associations all ‘create and maintain transnational norms to which NGOs, Inter-Governmental Organizations (IGOs) and government officials can be judged accountable’ (Keohane and Nye, Reference Keohane and Nye2002: 239–240). These networks involve everything from states and International Organizations through non-governmental actors (ranging from Amnesty International, Transparency International, and Greenpeace International) all the way to private actors (such as Moody’s and Standard and Poor’s). Serving in effect as ‘private overseers of national governments’ (Scott, Reference Scott2002: 60), these networks have notched up some very real accomplishments as itemized in a recent UN Report: ‘Global policy networks have significantly influenced policy, shaped public opinion and helped to resolve disputes on such issues as debt, landmines, small arms, conflict diamonds, big dams, and crimes against humanity…’ (Cardoso, Reference Cardoso2004: 33, para. 51).
In the limiting case, a pure network has no hierarchy. No one stands in authority over anyone else. Therefore, strictly speaking, no one would have to answer to anyone. Such networks are purely voluntary. They operate purely on the basis of trust and the reputation for trustworthiness (Kreps, Reference Kreps1990; see further Sørensen and Torfing, Reference Sørensen and Torfing2007). Of course real-world networks are not pure in this respect – but insofar as they have more structure than that, cooperation within them becomes only easier to orchestrate.
In the limiting case, a pure network would also lack any formal mechanisms for sanctioning members. A pure network would rely instead on more informal processes of ‘naming and shaming’ and ‘shunning’, insofar as enforcement is required at all.Footnote 27 Nevertheless, insofar as participants value the esteem of others in the network and want to continue doing business with them, those seemingly weak sanctions might actually suffice to facilitate productive collaboration. All of that is of course all the more true in the more mixed cases of networks found in the real political world that have far greater power to sanction members through consumer boycotts, international sanctions, withholding resources, and such like (Braithwaite and Drahos, Reference Braithwaite and Drahos2000; Castells, Reference Castells2000).
Having more structure and sanctions, those real-world accountability networks are even easier to mobilize. But even in the limiting case, pure networks could make a very real difference by linking people and organizations that share common concerns and values. Mutuality would be their hallmark. Each participant would internalize the perspective of the others. Naturally, differences of opinion would still exist within any given network. But confident that they all share some important purposes, participants in such networks would be more likely to respect one another’s opinions, to explore the bases of their differences and to make good-faith efforts to find common ground (March and Olsen, Reference March and Olsen1995: Ch. 2).
To a hard-bitten realist, all that may seem pretty airy-fairy. Such skepticism notwithstanding, that is precisely how networks do often work in the real world. One good example is Concert of Vienna multilateral conference diplomacy and the conventions that grew up around it (Mitzen, Reference Mitzen2005). Another example is found in the informal interactions among delegates that go on in the corners of formal meetings of UN General Assembly committees and that do so much to shape outcomes there (Alger, Reference Alger1966).
Perhaps the best description from the inside of how all this works is Gunnar Myrdal’s account of his time as Executive Secretary of the United Nation’s Economic Commission for Europe. That organization eventually gave rise to increasingly strong organizations – the European Economic Community, and thence the European Union itself. But in its early days, when Myrdal was its Executive Secretary, it was only a very loose intergovernmental organization, almost like the pure networks I have just been describing. Here is how Myrdal (Reference Myrdal1955: 8, 20) describes that organization as functioning:
[Once the] organization … settles down to a tradition of work, … the same state officials come together at regular intervals…. Certain substitutes for real political sanctions can then gradually be built up. They are all informal and frail. They assume a commonly shared appreciation of the general usefulness of earlier results reached, the similarly shared pride of, and solidarity towards, the ‘club’ of participants at the meetings…. Not upholding an agreement is something like a breach of etiquette in a club.
That is accountability of a sort. It worked in the medieval king’s curia, after a fashion. It worked in the Concert of Vienna and still does in UN committees, after much the same fashion. It is working after much the same fashion internationally today, through networks within and between states, International Organizations (IOs), NGOs, and IGOs.Footnote 28
The accountability is imperfect (Esmark, Reference Esmark2007). IOs, NGOs, and IGOs depend upon the support of states in myriad ways, making them less independent loci of accountability than ideally we would wish. But remember, so too were medieval barons dependent upon the king for all sorts of things, at the same time as he depended on them for others.
Worse, from the perspective of democratizing the global community, the groups to whom today’s international actors are accountable are themselves internally not particularly democratic (as yet). But remember, the unreformed seventeenth-century Parliament to which the Bill of Rights made William and Mary accountable was not all that democratic, either.
The first step is to get the accountability regime in place. The next steps are to strengthen and to democratize it. There are steps afoot to do just that, internationally, insofar as both the United Nations and European Union now decline to enter in formal consultative relationships with groups that do not themselves have democratic internal governance structures.Footnote 29
Dynamics of democratization, domestically
Thus it seems that democratization of the global community is up to roughly where it was domestically in Britain around the seventeenth century. So far, so reassuring. But what grounds are there for thinking the rest of the process will play itself out internationally in the same democratic way as it did domestically? Why think that we might end up with anything remotely resembling a Global Reform Act with one-person-one-vote worldwide?
I cannot pretend to predict particular paths toward global democratization. Even domestically, we have only sketchy outlines of possible ‘transition paths’ toward democratization (Linz and Stepan, Reference Linz and Stepan1996: Ch. 4). Likewise, highly stylized accounts are the best that can be expected internationally. Therefore, I shall once again be operating a couple degrees of abstraction above any particular historical experiences. In sketching paths toward global democratization, my aim will merely be to identify some generic features of the democratization process, leaving open how exactly they might be instantiated on any particular occasion.
Still less can I pretend to predict the pace of global democratization. Perhaps the fact that democratic ideas are already well established in one jurisdiction or at some other level of governance might enable democratization to proceed more quickly elsewhere or later. Nationally, late democratizers often accomplish in half a dozen years what took early democratizers like Britain as many centuries. Perhaps that same might prove true internationally. Or perhaps not,Footnote 30 I simply offer no speculations on the likely pace of global democratization.
As regard the direction of future developments, however, I shall be more committal. Here I shall sketch a process that, if I am correct, inexorably pushes the political system in a more democratically inclusive direction.
The basic model I shall be advancing is one of a ‘slippery slope in one direction’. There are two salient features of the process, as I shall characterize it:
(1) From time to time there emerges some crisis, to which expanding the range of people to whom power-holders are accountable is sometimes a solution.Footnote 31
(2) Accountability mostly only expands, it almost never contracts.
Taken together, those two factors guarantee the eventual predominance of democracy worldwide, both domestically (as I shall argue in this section) and internationally as well (as I shall subsequently go on to argue).
An example: extending the franchise
Think of the history of expansions of the franchise. At first, the monarch was answerable to only a handful of barons. Over time, more and more people were drawn into that circle, and eventually universal adult suffrage was achieved. It was not achieved because of any great good will on the part of those already within the charmed circle – Mill was quite wrong to expect that.Footnote 32 There were just crises, from time to time, to which extending the vote to additional groups was a solution.Footnote 33 In addition – here is the crucial point – once expanded, the franchise virtually never contracts.Footnote 34
Of course we can all think of exceptions.Footnote 35 In the United States, the most glaring is the way blacks in the South won voting rights during Reconstruction, only to lose them again once the federal forces departed. That is very much the exception that proves the rule, however. An eminent historian of that episode is emphatic on that point:
No major social group in Western history, other than African Americans, ever entered the electorate of an established democracy and then was extruded by nominally democratic means…, forcing that group to start all over again. In Europe, Latin America and elsewhere, liberal democracies never sponsored disenfranchisement. Once previously excluded social groups came into any established system, they stayed in (see further: Therborn, Reference Therborn1977; Freeman and Snidal, Reference Freeman and Snidal1982; Mackie and Rose, Reference Mackie and Rose1991; Valelly, Reference Valelly2004: 1–2).
That may be overstating the point a little – but only a little. Thus, for example, the French revolution enfranchised many people who were subsequently disenfranchised (although given the character of Napoleon’s regime, that was not by even ‘nominally democratic means’). The Reich Citizenship Law of 15 September 1935 disenfranchised Jews and anyone else who was not ‘of German or related blood’ (but again, while it was duly elected, one could certainly question whether the regime enacting that law was genuinely democratic). There are other smaller-scale instances. For example, convicted felons in the United States are often disenfranchised, sometimes for life, with racially differentiated consequences (Western, Reference Western2006). Or for another example, Commonwealth citizens resident in the UK enjoyed voting rights there until they were withdrawn by the Immigration Act of 1972. There are all those exceptions, to be sure. Still, those are exceptions to what proves to be an astonishingly strong general rule.
The most systematic empirical investigation on this subject has been undertaken by Adam Przeworski (Reference Przeworski2009). He provides an exhaustive survey of changes to election laws in 187 countries, from the time they first established national electoral institutions to the present.
Przeworski’s findings are striking. He finds that movement is in the ‘increasingly democratic’ direction – from more restrictive to less restrictive franchise provisions – in nearly ninety percent of cases.Footnote 36 On this, the best evidence we have, it thus seems that the franchise takes nine steps forward, for every one it takes backward.
Democracy as an absorbing Markov chain
Here is an analogy that might help in explicating the basic structure of the model I have in mind. Imagine a deep hole, such that once you fall into it you can never get out. People are not particularly drawn into the hole. They just pass by it from time to time, and occasionally they fall in. But since everyone passes the hole from time to time, and on some occasion or another eventually falls in – and since no one can get out once they have fallen in – everyone eventually ends up in the hole.Footnote 37
Two things make that model work. The first is the risk of falling in; the second is the impossibility of getting out. Both are obviously necessary for the working of the model. The way I have set up this model, however, it is the second rather than the first that does most of the work.
I emphasize that point because models of democratization usually take the opposite tack. Their emphasis is more ordinarily upon factors that push institutions in democratic directions: having a strong middle class or a strong union movement or whatever (Lipset, Reference Lipset1960; Jackman, Reference Jackman1972; Geddes, Reference Geddes1999). Or their emphasis falls teleologically on pull factors: in public-choice mode, on the way in which majority rule can (given sufficient uncertainty) maximize expected benefits for all concerned (Rae, Reference Rae1969); or on the way in which strong institutions (democratic or otherwise) can be sources of stability (Huntington, Reference Huntington1968).
I do not want to deny any of that. I want merely to ask, ‘Just how strong do those push-pull forces need to be, in order for democratization eventually to be fully accomplished?’ On the ordinary account, they need to be pretty strong and pretty persistent. On the model I am here suggesting, they can be pretty weak and pretty intermittent. All my model requires is that power-holders at least occasionally get themselves into jams, and that expanding the circle of accountability is one (perhaps among many) ways of getting themselves out of those jams.Footnote 38 Those occasional crises are what drive us in a more democratic direction, to be sure. They are important parts of the model, for that reason. But basically they can be treated just as random shocks, in my model.
My model can allow the pressures toward democratization to be just that weak, because on my model the ‘stickiness’ of democracy is strong. If democratization is basically a one-way ratcheting-up process, and the circle of accountability only expands and almost never contracts, then institutions will over time become increasingly democratic even with only very weak pressures in that direction.
What my hole analogy describes is, of course, just a folksy example of an ‘absorbing Markov chain’.Footnote 39 If there is some positive probability of transition from any other state into some ‘absorbing’ state, and zero probability of transition out of it, then everything will end up in that absorbing state eventually. How quickly depends just on how frequently transitions occur, and on the probability that any given transition will lead us into the absorbing state.
Completely absorbing states are of course rare. More commonly, there are merely highly differential probabilities of transitions into and out of the various states. The expanding-franchise story is one of an ‘almost-absorbing’ Markov chain, with movement into the expanded-franchise state being roughly nine times as likely as movement out of it. The point nonetheless remains: even with a merely almost-absorbing Markov chain, just about everyone will eventually fall into in the almost-absorbing state, and just about everyone will then stay there.
To the categories of completely absorbing and almost-absorbing states should be added the further category of ‘increasingly-absorbing’ states. Consider for example network externalities. There, the more people who are in a certain state, the more you want to be in that state yourself. The more people who are connected to the telephone line or to the internet, for example, the more beneficial it is for you to be connected yourself (Katz and Shapiro, Reference Katz and Shapiro1985; Grewal, Reference Grewal2008: ff. 25). Or for another different kind of example, consider the case of increasing returns to scale. The more you are already doing of something, the higher the payoff to you per unit cost of doing yet more of it (Pierson, Reference Pierson2000). In both cases, the allure of some state of the world is an increasing function of your proximity to it.Footnote 40
Mechanisms
Why should any of that be so, when it comes to transitions toward democracy? Here I point to three mechanisms – one cultural, one psychological, and one sociological. Together, they work hand-in-glove to produce that result in the domestic case. I shall focus in this section on the domestic case, reserving discussion of the international case for later.
The first mechanism that makes moves toward democratic inclusiveness so ‘sticky’ in the domestic case is the cultural shift that comes from acknowledging each other as equals. That having been once done, it is virtually impossible to go back. Sans-culottes can be forced back into their place, of course, but force is then required to do what had previously been accomplished by acculturation alone (Sonenscher, Reference Sonenscher2008). After a culture of egalitarianism had once been created, recreating the old culture of a status-differentiated society of ‘my station and its duties’ is well nigh impossible (Bradley, Reference Bradley1876; Shklar, Reference Shklar1991). That, as much as any ratchet effect of wartime taxing and spending (cf. Peacock and Wiseman, Reference Peacock and Wiseman1961), is what explains postwar expansions of the welfare state (Dryzek and Goodin, Reference Dryzek and Goodin1986).
The second source of ‘stickiness’ of moves toward democratic inclusiveness in the domestic case is a psychological mechanism, ‘loss aversion’. People are invariably much more sensitive to the prospect of losing something they already have than they are to the prospect of gaining the same thing if they do not yet have it. There is evidence aplenty of this tendency from social psychology and experimental economics experiments (Kahneman and Tversky, Reference Kahneman and Tversky1979; Tversky and Kahneman, Reference Tversky and Kahneman1991). There is evidence aplenty in the frequency of ‘hold harmless’ clauses and the need to ‘grandfather in’ existing uses and existing beneficiaries when changing public policy on everything from regulatory policy through land use controls to social security entitlements.
The third source of that ‘stickiness’ of moves toward democratic inclusiveness is sociological in character. Once people have been mobilized, it is hard to demobilize them (Tilly, Reference Tilly1997; McAdam et al., Reference McAdam, Tarrow and Tilly2001). Standing down armies after a war is perhaps the most literal case in point: decommissioning weapons, hard though it may be, is the easy part; harder to overcome is the fact that, among the former brothers in arms, the ‘strength of weak ties’ remains (Granovetter, Reference Granovetter1973). It is always easier to get people back together again, after they have once been together, than it was to get them together in the first place.
Those three mechanisms, separately and especially when acting in unison, are what make it so difficult to disenfranchise people after they have once won the franchise. Beyond the franchise, those factors make it similarly difficult to curtail accountability of any sort after it once has been expanded. Once power-holders have been made accountable to someone, electorally or otherwise, the same cultural, psychological, and sociological forces conspire to constrain any rollback of that accountability.
Ruptures
So far, I have been talking about dynamics within democratic systems of governance. The three mechanisms just discussed make it hard for a system that remains even ‘nominally democratic’ to reduce the range of people to whom power holders are democratically accountable.Footnote 41
But there can of course be a breakdown of the democratic system altogether (Linz and Stepan, Reference Linz and Stepan1978). Through revolution or counterrevolution, invasion or coup, democratic institutions can be suspended and replaced with others (or in the case of ‘failed states’, with none). Clearly in those cases, the cumulative progress toward expanding democratic accountability unravels.Footnote 42
I do not suppose there is any predicting what the future might hold for catastrophically failed states, or even for ones suffering very long bouts of foreign occupation or domination.Footnote 43 But consider democracies that have less dramatic interruptions to their democratic development, such as military coups.
Clearly, with a coup, democracy takes a step backwards. Elections are suspended and democratic rights are abrogated. Notice, however, this crucial fact. When democracy is eventually restored after the coup, it tends to pick up where it left off. The country does not ‘Return to Go’ and start all over again on the long and torturous route that had led it to the stage of democracy it enjoyed before the coup. Instead, the old constitution is typically restored pretty much holus bolus.Footnote 44 The same old rules apply once again. Everyone who used to have voting rights gets them back again. Democracy picks up where it left off.Footnote 45 In addition, the same might be true even with more dramatic interruptions to democratic development.
Students of comparative political history will doubtless think of a few counterexamples to that generalization.Footnote 46 As with expansions of the franchise, so too with restorations of democratic rule: the most we can confidently say is that the circle of accountability ‘almost never contracts’. Looks like a pretty reliable ‘stylized fact’, at least as regard domestic politics. There are counterexamples, to be sure, but the rarity of counterexamples constitutes confirmation of the rule.
Extending the model, internationally
How applicable is all of this at the international level? Well, casual empiricism offers plenty of confirmatory examples. Notice that the Organisation for Economic Co-operation and Development has only ever expanded, it has never contracted. The G-7 has become the G-8, which has become the G-20 (more of which below). The range of NGOs with official consultative status at the Economic and Social Council of the UN has mostly only expanded and virtually never contracted. The United Nations picked up pretty much where the League of Nations left off. And so on. All that makes the proposition that ‘accountability basically expands rather than contracts’ look pretty plausible as a stylized fact about politics in general, international as well as about domestic.
Still, generations of students of international relations have rightly been taught to be on guard against any too-easy appeal to ‘the domestic analogy’ (Suganami, Reference Suganami1989). It would be clearly mistaken to take it purely as axiomatic that politics between states is necessarily perfectly isomorphic with politics within states, any just as it would be clearly mistaken to take it purely as axiomatic that the way in which states behave in a situation of anarchy is necessarily perfectly isomorphic with the way individuals do.
Neither, however, is it necessarily the case they are completely different. Sometimes they are, sometimes not. We simply need to check whether the particular domestic analogy being drawn is or is not appropriate. The best way to do that, I suggest, is through a close inspection of the details of the model and the mechanisms at work within it (Tilly, Reference Tilly2001; Tilly and Goodin, Reference Tilly and Goodin2006; Elster, Reference Elster2007). Those are precisely the ‘points of analogy’ in the analogies being drawn between domestic politics and international politics. If similar mechanisms can be shown to be at work in both cases, then on that point the analogy will have been vindicated.
The model of domestic democratization offered above had two key components:
(1) From time to time there emerges some crisis, to which expanding the range of people to whom power-holders are accountable is sometimes a solution.
(2) Accountability, mostly only expands, it almost never contracts.
The first component is a very weak one (‘from time to time…’; ‘sometimes…’). Accordingly, all I need to do to vindicate the analogy on that point is to provide an existence proof – to show that some such thing also happens internationally, at least occasionally. I do that below via the case of the G-20. The second component is what does most of the work in my model. To vindicate the analogy on that point, I then go on to show that the three mechanisms identified above for the domestic case are also heavily at work internationally as well.
Crises expand accountability: the G-20
For a compelling example of how crises can arise in global affairs, to which expanding the circle of accountability can be a solution, consider how the G-20 supplanted the G-8 in response to the 2008 meltdown of the global economy. Although far from unique, that is a paradigm case in all sorts of respects: the crisis was truly a major one; the bodies in question had formal memberships, one larger the other smaller; and since those bodies work by unanimous consent, the newly included members gained real power when authority was transferred from the smaller to the larger group. Best of all, those facts can all be read right off the face of official communiqués issued at the end of each meeting.Footnote 47
The history in a nutshell is this. With the demise of the Bretton Woods exchange rate regime and in the wake of the 1973 oil crisis, there was a felt need for a small, focused forum to enable the world’s leading industrialized democracies to coordinate management of the global economy. The G-6 (US, UK, France, FRG, Italy, and Japan) was formed for this purpose at a 1975 heads of government summit convened by President Valéry Giscard d’Estaing in Ramoouillet.Footnote 48 That morphed into the G-7 (with the addition of Canada a year later) and thence the G-8 (with the addition of Russia a decade after that). Under that aegis, summits of heads of government were held annually and meetings of finance ministers quarterly.
Other entities were represented at G-8 meetings in more limited ways. Among them were the European Union, the World Bank and the International Monetary Fund. Some emerging economies – the Outreach Five (Brazil, China, India, Mexico, and South Africa) – were also invited to attend for some but not all items of business. Basically, however, the G-8 served as a ‘rich countries’ club’ that informally steered the world economy.
As a sop to the emerging economies, a larger G-20 was formally established in 1999.Footnote 49 Its initial task was described in milquetoast terms – ‘to broaden the discussions on key economic and financial policy issues among systemically significant economies’ (G-20, 1999: para. 2). The final communiqués of those meetings (initially just of finance ministers and central bank governors) confirm that just such broad ‘deep-background’ discussions of fairly inconclusive or inconsequential sorts were indeed the order of those early days.
With the global financial meltdown in 2008, however, the smaller G-8 clearly could no longer manage the global economy on its own. There was an urgent need to involve representatives of a wider set of the world’s economies. In Washington in December 2008, national leaders of G-20 members met for the first time, saying that ‘the global crisis requires global solutions…. The G-20, with its broad representation of major systemically important economies, has a critical role to play in ensuring global financial and economic stability…’. (G-20, 2008a: paras 2, 4; see further G-20, 2008b).
Expanding on that proposition in the final communiqué from their April 2009 London summit, G-20 national leaders proclaimed:
We start from the belief that prosperity is indivisible; that growth, to be sustained, has to be shared; and that our global plan for recovery must have at its heart the needs and jobs of hard-working families, not just in developed countries but in emerging markets and the poorest countries of the world too; and must reflect the interests, not just of today’s population, but of future generations too. We believe that the only sure foundation for sustainable globalization and rising prosperity for all is an open world economy based on market principles, effective regulation, and strong global institutions.
….
Emerging markets and developing countries, which have been the engine of recent world growth, are also now facing challenges, which are adding to the current downturn in the global economy. It is imperative for global confidence and economic recovery that capital continues to flow to them (G-20, 2009a: paras 3, 17).
In their November 2009 meeting in Pittsburgh, the leaders of the G-20 countries went further still. There, they formally proclaimed that the G-20 would henceforth take the place of the G-8 as the entity charged with principal responsibility for managing the world economy. The closing paragraph of their communique bluntly spelt the death knell for the old G-8, stating: ‘Today, we designated the G-20 as the premier forum for our international economic cooperation’ (G-20, 2009b: para. 50)
A clearer case could not be found of the emergence of a global crisis, to which widening the circle of global accountability was offered as a response. It happens, at least occasionally, at the international level just as at the domestic. How often it happens I leave as an open question, just as I did in my discussion of the domestic case above.Footnote 50 For the purposes of my model of democratization, it is sufficient that it happens ‘from time to time’.
Mechanisms: sources of stickiness, internationally
The real workhorse in my model of domestic democratization is the proposition that ‘accountability mostly only expands, it almost never contracts’. Three mechanisms were identified to account for that in the domestic case. The task of this section is to assess the extent to which the same sorts of mechanisms might also be at work internationally, leading expansions of accountability there to be similarly hard to reverse.
The first thing to note is that, even at the level of domestic politics, one of the mechanisms I identify – the psychological phenomenon of ‘loss aversion’ – is essentially an extrapolation from an individual-level phenomenon. The thought, domestically, is that groups are more likely organize to resist losing what they already have because individuals resist losing what they already have.
Of course we know from Olson’s (Reference Olson1965) logic of collective action that it is a fallacy to suppose groups will necessarily organize to pursue interests that members of that putative group share. In more instrumental settings, individuals might rationally be inclined to be free riders, benefiting from others’ paying the costs of the group activity that would benefit them whether or not they played a part in that activity themselves. But that is a rational, instrumental calculation. Psychological mechanisms evoke a more visceral and less considered response, operating (to some extent or another) ‘behind the backs’ of intentional agents. Insofar as ‘loss aversion’ evokes that sort of a response, the prospects for collective action are less undermined by Olson’s (Reference Olson1965) ‘logic’. Shared psychological responses may then translate fairly directly into concerted group action.
Furthermore, there is every reason to suppose that the same psychological dynamics would lead people who stand to lose to organize collectively to lobby against such developments, be they foreign or domestic. The costs of transnational organization might be higher (although they are getting lower all the time, with new communication technology). But there is every reason to think that the psychological impetus – loss aversion – would lead people to make common cause with others who stand to lose in the same way from the same policies worldwide.
Sometimes everyone in the same country would stand to suffer identical losses from some development, which they therefore all oppose. More often, everyone in some subset of people in a country (or straddling several countries) stands to lose, and they resist on those grounds. As a sort of shorthand, we might explain those reactions by saying that country or group as a whole was manifesting ‘loss aversion’, although strictly speaking of course loss aversion is a psychological phenomenon and it is only individuals, not groups or countries, that are possessed of psyches. The point remains: the ‘loss aversion’ mechanism works the same way in both internationally and domestically to make people, individually and in groups, resist efforts to remove democratic accountability once they have been granted it.
The other two mechanisms I discussed in relation to the stickiness of domestic democratization are more genuinely group-level phenomena. The story I told about the sociological effects of mobilization pointed to the ‘strength of weak ties’. One reason that it is easier to renew collaborations has purely to do with information: you already know who to talk to about what, on the basis of previous experience; you have some sense of what they are interested in, of what position they are likely to take, of what resources they have access to, and so on. Overlaid on that is a ‘reciprocity’ dynamic: people with whom you have interacted in the past, and with whom you expect to interact again in the future, are people whose past actions you will reciprocate, even if only as an investment in your own future payoffs (Bowles and Gintis, Reference Bowles and Gintis2002). A reasonably strong ‘mobilization effect’ is thus easily generated.
Networks thus mobilized have been shown to be powerful forces in international relations just as in neighborhood organizing. The stories of ‘activists beyond borders’, for both human rights (Keck and Sikkink, Reference Keck and Sikkink1998; Risse et al., Reference Risse, Ropp and Sikkink1999) and environmental protection (Zürn, Reference Zürn1998), bear rich testimony to that fact. Already knowing how important such networks can be in global politics, we need only add the further fact that networks are hard to destroy completely once they have been mobilized, in order to appreciate another important source of ‘stickiness’ in global accountability mechanisms. As with armies so too with global networks, once mobilized they are hard to demobilize completely.
The final mechanism making expansions of domestic democratic inclusiveness hard to undo had to do with a more diffuse notion of a ‘cultural shift’. Once we have come to regard one another as equals, it is hard to go back to thinking of one another in status-differentiated terms. Of course, insofar as that is a fact about human psychology, it will hold true across national borders as well as within them. But insofar as the culture in question is a national political culture, then the effect of a culture shift might well stop at the national borders. Citizens of post-revolutionary France could think of one another as equals, without for a moment of counting Britons or Spaniards in that category as well.
For there to be some analogous mechanism at work globally, therefore, we would have to posit a political culture that transcends national borders and to point to hard-to-reverse shifts occurring in that culture. The best evidence I can offer of that comes from recent work on the role of ‘norms’ within the international community (Finnemore and Sikkink, Reference Finnemore and Sikkink1998). There are many penetrating accounts of the importance of such norms, of the processes by which they get established, and of the ways in which they change. For a quick and easy demonstration, let me here focus on one norm in particular, the ‘decolonization’ norm. That is a good place to fix our focus for multiple reasons. For one thing, independent self-government, is an important democratic accomplishment in its own right. For another, the number of countries under foreign rule, just like the number of people denied the vote, is readily countable; and we can therefore easily assess just how decolonization has spread and just how (in)frequently it is reversed. Finally, while decolonization is of course only one example of a norm of trans-national political culture, there is no reason to suppose that it is at all atypical of the way other democratic norms would work in international society.
Graphs plotting the rate of decolonization over time show modest flurries in the late eighteenth and early nineteenth centuries, as colonies in the Americas won their freedom. After a long hiatus, the rate picks up dramatically after the First World War and more dramatically again after the Second (Strang, Reference Strang1991: 436; see also Goertz and Diehl, Reference Goertz and Diehl1992: 653). As with extensions of the franchise domestically, so too with decolonization internationally: there was occasionally a little backsliding.Footnote 51 But basically, the movement was almost all in the one direction. Furthermore, close scrutiny of debates surrounding various cases of decolonization suggest that a cultural shift was indeed one key driver of that. Norms of decolonization took hold precisely when, and because, people – both in the metropole and especially in the colonies themselves – came to be imbued with notions of democratic equality.
Thus, it seems that all three of the same mechanisms might be at work internationally as have historically been at work domestically to ensure that moves toward ever-broader circles of accountability will be hard to reverse.
First find a state
There is one last potential point of disanalogy between democratizing domestic institutions and international ones. That might be encapsulated in the phrase, ‘first find a state’.Footnote 52
The thought is that, in order to democratize a state, you first must have a state to democratize.Footnote 53 Before you can democratize central institutions of authority, you have to have some central institutions of authority. In the domestic case, we had them. In the international case, we do not – or anyway not yet.Footnote 54 If we define democratization in terms of ‘a transfer of sovereignty from the prince to the people’, then we need a prince possessed of sovereignty before he can transfer it to the people (cf. Goodhart, Reference Goodhart2007: 574). Or so the thought might go.Footnote 55
Of course, as historians of state formation will immediately interject at this point, the central institutions of authority in the state were themselves awfully thin in the early period that is most analogous to today’s international order. Maybe the king could claim a territory, but what the putative sovereign could do by way of ruling over it was in practice pretty circumscribed (to claim authority is one thing, to exercise it is quite another…).
It is perfectly true that no one even claims sovereignty over the whole globe, as yet. But that is not to deny that there are any institutions, international in scope, which could be subject to a push for democratization. Obviously, there are many. These institutions exercise global authority in piecemeal fashion, functionally, defined. But whoever said – whoever would think – that only central authorities with a perfectly general remit admit of being democratized? Notwithstanding its functionally delimited scope of authority, a school board can be made more democratically accountable (by, e.g. making its members be popularly elected). Therefore too the World Trade Organization: its authority is functionally delimited, but that constitutes no conceptual barrier to its exercising that authority in a more or less democratic fashion.Footnote 56 In addition, so on down the list of functionally delimited power-holders in the global arena.
Would democratizing the partial and piecemeal institutions of global governance amount to a contribution toward global democracy? Surely it would. The only reason for doubting that is that we sometimes fail to notice that two dimensions are in play here. First, political institutions can be strong or weak; second, they can be democratic or not. If we have some strong central authority, and we democratize it, we end up with a strong democratic institution. If we have only weak central authorities with limited remits, and democratize them, we end up with weak democratic institutions with limited powers. But the institutions are no less democratic for their being weak in that way (Dahl and Tufte, Reference Dahl and Tufte1973).
That leads to one final speculation. Perhaps democratizing institutions might pave the way for strengthening those institutions.Footnote 57 That was certainly the case with kings and parliaments: they became stronger as, and because, they became more democratic. Maybe it will prove to be so with global institutions as well. Perhaps what is fundamentally required to strengthen those institutions, too, is that they be democratized.Footnote 58
Acknowledgements
This article has been presented as the Edmund Burke Lecture at Trinity College, Dublin, at the Universities of Bergen, Harvard, Oxford, Queensland, and Turku, at the City University of New York, at the Czech Academy of the Sciences in Prague and at the 2008 Annual Meetings of the American Political Science Association. The author grateful for valuable feedback to all those audiences, most particularly from: Daniele Archibugi, Magdalena Balcerak, Derek Bok, Michael Biggs, Dave Chalmers, Jurgen De Wispelaere, Keith Dowding, John Dryzek, Katrin Flikschuh, Carol Gould, David Grewal, Gudmund Hernes, Andy Hurrell, Stein Kuhnle, Leon Leontyev, Terry MacDonald, Jenny Mansbridge, Raffaele Marchetti, Larry May, Cillian McBride, Sue Mendus, David Miller, Hannu Nurmi, Martha Nussbaum, Claus Offe, Frances Fox Piven, Eric Posner, Mathias Risse, David Rodin, Duncan Snidal, Torbjörn Tännsjö, Dennis Thompson, Mike Titelbaum, Laura Valentini, Lea Ypi, Michael Zürn, and the editors of and referees for this journal.
