Louvain, Public Sphere—1
Democracy and a Practical Public Sphere Beyond the State?1
Joshua Cohen and Charles Sabel
1. The changing context of global politics
People used to say that beyond domestic politics lay the world of international
politics: thus the title of Kenneth Waltz’s book, which celebrated the bipolar
world, just as dusk was settling on it.2 “International politics” was the label for a
mix of relations between and among states and for the institutions (always of
contested significance) built by those states. Students of international politics
disagreed deeply about the right way to understand that world—with realists
emphasizing the underlying distribution of national power and interest; Bullians
arguing that relations among states were always set within a constraining “social
element” of common norms, rules and values; second image theorists holding
that the internal forms of regimes matter to conduct of international politics; and
regime theorists arguing that rule-governed regimes foster cooperation among
states. But there was considerable agreement about the main elements of the
object of inquiry.
1
Respectively, Leon and Anne Goldberg Professor of the Humanities and Professor of
Philosophy and Political Science, Massachusetts Institute of Technology; Maurice T. Moore
Professor of Law, Columbia University. This paper draws on material from three joint essays: our
Wade Lecture on “Democracy Beyond the State?”, Washington University, St. Louis, April 2005;
“Extra Rempublicam Nulla Justitia?,” Philosophy and Public Affairs (forthcoming, fall 2005); and
“Administrative Law and Global Politics,” Journal of International Law and Policy (forthcoming).
2 Kenneth Waltz, The Theory of International Politics (Boston: McGraw-Hill, 1979);
Louvain, Public Sphere—2
Something very different now seems to be happening: call it the condition
of “global politics,” marked by proliferation of political settings beyond domestic
boundaries that are not reducible to states and their voluntary interactions (at
least if we understand “voluntary” as a persisting trait of a relationship, and not
simply an observation about historical origins): a proliferation that expands the
range of relevant political actors, while shifting our understanding of political units
(human rights as limits on westphalian sovereignty was a first step, but not the
last) and of relations among them.
What are some of the elements of global politics? They include—with no
pretence either to completeness or precision—the following:
(1) Economic integration—as measured by communication and
transportation costs, trade and trade dependence, and movements of
capital—has made the global economy a substantial presence in the
economic lives of virtually all states.
(2) Cultures, economic circumstances, and political institutions and
traditions vary widely, and much more widely between states than within
them.
(3) While states remain essential players, to a considerable and growing
extent, rule making, as well as rule elaboration and application—
especially in the arena of economic regulation, but also in areas of
security, labor standards, environment, rights, food safety standards,
product standards among others—is taking place in global settings that,
even if established by states (and many regulatory functions are
Louvain, Public Sphere—3
provided by private or public-private bodies), conduct their activities of
making, elaborating, and applying rules activities with some de facto
decision-making independence from their creators.
(4) The rules made in those settings are consequential for the conduct and
welfare of individuals, firms, and states, in part because they provide
standards for coordinated action and in part because national rulemaking itself proceeds subject to rules, standards, and principles
established beyond the national level.
(5) Whatever the origins of these rule-making bodies, they are expected—
by states, firms, individuals, and organizations—to continue to exist and
to make consequential decisions, so that agents (including states, firms,
and non-governmental organizations) and movements need to take them
into account in making decisions and pursuing goals.
(6) Even when rule-making and applying bodies lack their own independent
power to impose sanctions through coercion, they have the capacity to
encourage conduct by providing incentives and permitting the imposition
of sanctions; moreover withdrawing from them may be costly to
members (if only because of the sometimes-considerable loss of
benefits).
(7) Those settings of global rule-making are the focus of a transnational
politics of movements and organizations—and not only an
intergovernmental politics between states—that contest and aim to
reshape the activities of supra-national rule-making bodies. Those efforts
Louvain, Public Sphere—4
work in part through protest, in part by representing interests to those
bodies, and in part by advancing norms, values, and standards of
reasonableness—that is, by suggesting potential elements of a global
public reason that might serve as a common ground of argument in
assessing the practices and performances in global politics.3
Global politics is thus not an occasional matter of sparse agreements;
while the ground is changing quickly, it seems to be enduring and institutionally
dense,4 featuring coordination and relatively autonomous administration (rulemaking). Putting (3)-(6) together, Kingsbury, Krisch, and Stewart have argued
that, even in the absence of a global state exercising authority over a territory,
there is now a relatively autonomous space of global administration: “the
classical dichotomy between an administrative space in national polities
and inter-state coordination in global governance.” In their view: “The rise
of regulatory programs at the global level and their penetration of domestic
counterparts means that the decisions of domestic administrators are
increasingly constrained by substantive and procedural norms established at the
global level; the formal need for domestic implementation then does no longer
provide for meaningful independence of the domestic from the international
realm. At the same time, the global administrative bodies making those decisions
3
Mary Kaldor, Global Civil Society: An Answer to War (Cambridge: Polity, 2003); John Keane,
Global Civil Society? (Cambridge: Cambridge University Press, 2003); John Ruggie, “Taking
Embedded Liberalism Global: The Corporate Connection,” in David Held and Mathias KoenigArchibugi, eds., Taming Globalization (Cambridge: Cambridge University Press, 2003).
4 Confining attention to intergovernmental organizations with permanent administrative staffs, the
world’s least integrated country is a formal member of 14 organizations, and virtually all other
countries are formal members of more than 100 organizations. In addition, there are agreements
that establish rights and obligations but do not create administrative capacity. See Shanks,
Jacobson, Kaplan, “Inertia and Change.”
Louvain, Public Sphere—5
enjoy too much de facto independence and discretion to be regarded as mere
agents of states.”5 In short, they describe an emerging regulatory state, with
potentials for de- as well as re-regulation—but without the state.
Moreover—now adding in (7)—that relatively autonomous space is the
focus of increasingly routine political activities and associated discourse about
how global politics and rule-making be conducted. So we have something new—
global politics without a global state.
For political philosophy, this new context has placed three large, closely
related normative issues on the table: first, what are the appropriate forms of
governance for global politics, and what place might democratic norms have in
governance beyond the state? Second, how are we to understand the nature and
content of the rights engendered by new forms of global politics. And third, are
there norms of access to basic goods or of distributive justice that apply in the
world of global politics? Though much of the recent philosophical writing about
issues of global justice has focused on the third of these—in particular, on
whether egalitarian distributive justice has as much normative force globally as
domestically—the normative discourse that is already part of the world of global
politics focuses with equal intensity on the first and second as well. And we will
concentrate here on the first two (not for reasons of relative importance, but
simply to keep the presentation tractable), with a particular focus at the end on
the issues about democracy: about how the context of global politics might
Benedict Kingsbury, Nico Krisch, and Richard Stewart, “The Emergence of Global
Administrative Law,” Journal of International Law and Policy (forthcoming).
5
Louvain, Public Sphere—6
induce a new form of public sphere, demos, and democracy, but without a global
state.
2. Two conventional responses to these unconventional developments
Though global politics is a new historical phenomenon, clashing theoretical
responses—particularly in normative political thought—to these developments
remain wedded to conventional ideas.
One pole is defined by a political cosmopolitanism that transposes the
idea of unitary sovereignty from the state to the world. In this view emerging
global politics is breaking down the artificial barriers that divide individuals from
one another; individuals are all at last emerging as equal citizens of the world,
and the emerging global politics is providing the settings in which they are finally
treated as equal citizens of the world (perhaps to be perfected by a global state).
At the second pole, Statist/Nationist views argue the key actors in global
politics are still states because states, or other communities with shared political
culture, remain the key providers of the basic good of security or because states
remain the principal site of enforced obligations. Thus while the new world of
global politics is more complexly organized, the agents in the new world of global
politics are still bounded collectivities not individuals.
Taken together these views suggest, implausibly, that whether the world
governance is on the verge of radical change (cosmopolitanism) or very far from
it (statism), our familiar ideas of sovereignty will continue to provide the axial
principles of politics.
Louvain, Public Sphere—7

The roots of an alternative are suggested by the idea,
advanced by John Rawls, that “the correct regulative principle for a thing
depends on the nature of that thing.” Tom Nagel has called this the “political
view” of justice, drawing the terms from Rawls’ account of a political conception
of justice. The main idea is that (as against what has been called “monism”)
there is no set of basic moral principles from which we can derive both moral
requirements for individuals and principles for associations and institutions.
Instead, and in particular, norms of justice—beyond the humanitarianism that
binds us even in the absence of any organized cooperation—apply to us only
when we stand in certain kinds of relations, and the content of the principles vary
according to the nature of the relations.
Some variants of this view assume that various domains of life have
natural, correlative regulatory principles: different spheres, different principles.
But a more plausible version says that working out the right principles depends in
part at least on a view about the nature of the relations and of the problems to be
solved, rather than fixed features of this or that problem-solving domain, or of the
good that a domain distributes, or our moral or political nature, that set conditions
on the regulative principle. The idea that we need to fashion an appropriate
response to our problems with the materials presented by the situations
(including the relations) in which they arise—that no such appropriate response
is antecedently given and awaiting discovery—is one way to summarize the
general idea of constructivism, and the political view is a species of that general
conception.
Louvain, Public Sphere—8
In a recent essay on global justice, Nagel endorses this politicalconstructivist view in the generic form just stated, but then presents a truncated
application of it, which issues immediately in a reaffirmation of a strong from of
statism, according to which the existence of a state is necessary and sufficient to
trigger any norms beyond humanitarianism’s moral minimum. With the
compression required for current purposes the argument goes like this:

There could in theory be (infinitely) many constellations of normgenerative relations, with corresponding norms of justice, on the notional
continuum between humanitarianism and a world state that makes all
humanity subject to the same Hobbesian sovereign and equally entitled to
authorize its conduct.

But, according to Nagel, the only situation that actually triggers new norms
of justice is the one in which we are all subject to a common sovereign,
which has the power to coerce our conduct and whose conduct we are
seen as authorizing—only when we are subjects in law’s empire and in
citizens in law’s republic. Here, the involvement of will associated with the
combination of coercion and co-authorship issues in new, demanding
requirements of egalitarian justice.

And given the evident impossibility of achieving Hobbesian sovereignty on
a world scale, neither the increase in global inter-dependence, nor the
increasingly explicit efforts to co-ordinate response to the higher order
effects of this globalization, are justice-generative.
In short: extra rempublicam nulla justitia.
Louvain, Public Sphere—9
But there is a third possibility: that the kinds of interdependence,
cooperation, rule-making characteristic of global politics trigger—even in the
absence of a state—normative demands that are greater than humanitarianism
but may yet fall short of the full measure of egalitarian principles. As a general
matter, we think that this third possibility is right, and that a normative idea of
inclusion, both procedural and substantive, is central to the domain of global
justice. Conceptions of global justice offer ways to address the three questions
mentioned at the end of the previous section: they offer accounts of fair
governance, human rights, and fair distribution (including access to such basic
goods as health and education); competing conceptions can be understood as
advancing alternative accounts of what inclusion demands—of the kind of
respect and concern that is owed by the variety of agencies, organizations, and
institutions (including states) that operate on the terrain of global politics, and of
how a particular conception of rights, governance, and access to resources
provides the respect and concern that is due.
A common name for the normative requirements on governance
(processes of rule-making) arising from organization of inter-dependence and
cooperation is accountability (including transparency, reason giving, ex ante
hearings and ex post review, standing for those affected). And once there is a
rationale for norms more demanding than humanitarianism on the governance
side, there is also a case for norms more demanding than humanitarianism on
the outcome side: that is, for claims to basic rights and a fair distribution of
Louvain, Public Sphere—10
benefits and burdens—though as I mentioned earlier, we will be concentrating
here on governance (in section 4) and rights (in section 5).
3. New Directions
To exploit fully the constructivist insight associated with the political conception,
then, we need to suppress the Hobbesian reflex—that coordination and
consequential rule-making require a state of a familiar kind—state the distinctive
coordination problems posed by global politics, and inquire whether there are
regulative principles—including principles of governance—suited to those forms.
Take first the distinctive problem. We just saw that global politics reflects
profound heterogeneity in the normative and factual terrain: and even in its
current, limited form, it addresses (however imperfectly), through more or less
institutionalized settings, urgent problems across a wide span of domains
(trades, security, environment, rights, health, education…), each of which is itself
extremely diversified (by North/South, gender, culture, and so on), and,
moreover, builds expectations (both defacto and normative) of continued such
address, through coordinated activity.
Solutions to each of these problems require public goods that are best
provided when they are tailored to the circumstances of the individuals and
groups to whom they are addressed. Determining what these public goods are,
and how to provide them in particular cases, requires constant learning, which in
turn requires organizational innovation. (And, to forestall a concern by
anticipating a later point: the learning requires global comparison, even if the
Louvain, Public Sphere—11
tailoring focuses on local circumstance.) Contrast such new, elusive public goods
with traditional generic ones, conceived within states, whose nature and utility is
taken to be self evident to all potential beneficiaries—so self evident, indeed, that
each is tempted to free ride on the willingness of the others to finance them
The provision of such particularized and changing goods is in turn
dependent on cooperation among (specialized subgroups) of providers and
recipients, and expectations about providing goods build up around forms of
cooperation. Such cooperation, in turn, invariably gives rise to political contest:
disagreements about best solutions, and about proper ways of determining
acceptable solutions, as well as disagreements about the distribution of benefits
(though we should resist the familiar temptation to think that politics begins and
ends with distributive issues, as if everything else about the provision of public
goods was somehow technical, and a matter for extra-political expertise). Hence,
given defacto interdependence among a wide range of differing people and
groups, Hobbesian, one size for all sovereignty is unworkable and undesirable,
but so too is pre-political cooperation understood as mutual adjustment in light of
strategies of others, rather than coordinated decision-making.
So what is normatively desirable is a form of governance or decision
making, with corresponding powers of coordination, enforcement authority and
surrounding expectations, which builds on existing forms of global politics, makes
problems of mutual adjustment explicitly accountable to the diverse, concerned
parties, and is capable of producing outcomes responsive to their various and
changing needs. We say that the form of governance needs to be accountable to
Louvain, Public Sphere—12
concerned parties and responsive to them because these (abstractly stated)
norms express the idea of inclusion, and seem appropriate to existing global
politics, with its mix of interdependence, coordinated decision-making in the
provision of basic public goods, and expectations of continued such provision.
Even if more exacting demands of egalitarian justice are not called for, the
regulative principles appropriate to existing forms of global politics surely are
more demanding than the (already-demanding and endlessly breached)
requirements of humanitarianism.
We call the form of governance that fully embodies these principles, while
promising the mix of local tailoring and global learning suited to the provision of
the relevant public goods “deliberative polyarchy.”6
4. Deliberative Polyarchy
Deliberative polyarchy is shaped by mutually re-enforcing moral and practical
concerns. Its appeal within global politics (and national politics as well) is
precisely to create the prospect of a legitimate—accountable and responsive,
thus more inclusive—form of problem solving under conditions of diversity so
extreme as to seem to preclude either broad effectiveness or general legitimacy.
We first set out the structuring ideals of deliberative polyarchy and then show
how this mode of governance can well survive in a world that often operates on
other principles.
See Joshua Cohen and Charles Sabel, “Directly-Deliberative Polyarchy,” European Law Journal
(1997), and [[our EU paper]].
6
Louvain, Public Sphere—13
Deliberative polyarchy is, obviously and to begin with, deliberative, and that
means that decision-making works through mutual reason-giving. Essentially, the
point of deliberation is to subject the exercise of collective power to reason’s
discipline, to what Habermas famously described as “the force of the better
argument,” not the advantage of the better situated.7 Questions are decided by
argument about best ways to address problems, not simply exertions of power, or
expressions of interest, or bargaining from power positions on the basis of interests.
More particularly, the aim is to find solutions that others can reasonably be expected
to support as well—or at least to acknowledge the relevance and importance of the
supporting reasons. The point of deliberative democracy is not to do away with
power, whatever that might mean; it is not simply to subject power to the discipline
of talking, because talking is not the same as reasoning (consider verbal assaults,
threats, insults, racial slurs, lies, blowing smoke, cheering, exchanging pleasantries,
exploring common experiences); nor is it simply to reason together, because
reasoning together may have no effect on the exercise of collective power.
In deliberative polyarchy, deliberation is situated: the actors are presumed to
know a good deal about their current dilemmas. But they also know relatively little
about the implications, for themselves or others, of the solutions from which they
choosing. Aware of these bounds on their knowledge, they aim to find measures
that address their immediate problems while suggesting next steps and allowing
evaluation of choices so far. Their uncertainty, moreover, and the consequent need
for mutual learning, makes it costly to game the information they provide; and limits
on the strategic use of information make it more difficult to pursue self-serving
7
See Juergen Habermas [ref].
Louvain, Public Sphere—14
solutions. Because of this prudent forbearance deliberation’s normative
requirement—of finding solutions that others can reasonably be expected to
embrace as well—is easier to meet. The normative and the practical may be happily
congruent.
Though the choices in this situated deliberation are informed by the facts as
known and addressed to solving specific problems step by step, deliberative
polyarchy is neither technocracy nor a familiar type of incremental reformism.
Technocratic reasoning focuses exclusively on selecting the best means to achieve
well-defined goals. Deliberative polyarchy is founded in part on the idea the actors’
goals are too broadly and diversely defined to allow for technocratic means-end
calculations. Thus situated deliberation acknowledges, indeed underscores, that all
complex practical problems—from trade and security to organizing schools and
transportation, providing clean water and public safety, allocating health care and
ensuring fair compensation—are political in the sense that they implicate a range of
distinct values, that reasonable people disagree about the precise content of and
weights to be assigned to those values, that some form of collective decision is
needed despite these disagreements, and that a normatively desirable way to make
such collective decisions is by a process in which participants offer reasons that
others can be expected to acknowledge, even as they disagree about how the
variety of relevant reasons add up.
By incrementalism, we mean reform through adjustment at the margin, and
only there: the idea is that the best you can do is to make the best of the current
situation, taking the background situation itself as unalterable. Though deliberative
Louvain, Public Sphere—15
polyarchy proceeds step by step, the steps it takes may be cumulatively
transforming—differing choices of means to purse variously interpreted ends leads
to an improved understanding of ends, and reinterpretation of these through
subsequent specification of means, and so on. Proceeding by way of their (variously
interpreted) principled problem solving, the actors can thus transform the constraints
on their actions, and their own views of what constitutes a legitimate constraint.
What makes deliberative polyarchy polyarchic is its use of
situated deliberation within decision-making units and deliberative comparisons
across those units to enable them to engage in a mutually disciplined and
responsive exploration of their particular variant of common the problems. It
resembles federalism insofar as it authorizes yet limits diversity, and anarchy
insofar as it does away with an ultimately authoritative center; but it is a novel
form of each precisely because of its resemblance to the other. Polyarchy
addresses a dual problem: i) comparable problems, arising in different settings,
need solutions appropriately tailored to those settings, and; (ii) solutions in each
setting need to be subjected to the pressure of deliberative comparison with
solutions adopted in the others so that all reflect understanding of what was done
elsewhere, and why.
Put another way, while differences in the circumstances in which problems
arise suggest a need for differences in solution, the commonality of problems
indicates a need to discipline local solutions against those adopted elsewhere:
the aim is not to achieve uniformity, but to pool information, identify best
practices, and compare solutions across locations. Thus the basic architecture of
Louvain, Public Sphere—16
deliberative polyarchy is to have situated deliberation within and among distinct
decision-making units.
For the individual decision-making units, diversity implies that decisionmaking in each needs to be friendly to local experimentation in the policy area in
question, drawing on local knowledge and values. As each unit is distinct, none
does best by simply copying solutions adopted by others, though they may do well
to treat those solutions as baselines from which to move, or at least as providing
some information about possibilities. Local problem solving is well-suited to bringing
the relevant local knowledge and values to bear in making decisions. Deliberative
participation helps because it encourages the expression of differences in outlook,
and the provision of information more generally.
But the same concern for a form of decision-making that it is attentive to
unexplored possibilities and unintended consequences requires
institutionalization of links among local units—in particular, the institutionalization
of links that require separate deliberative units to consider their own proposals
against alternatives provided by other units. A natural place to look for promising
alternatives—including alternatives previously unimagined in the local setting—is
in the experience of units facing analogous problems. So we need deliberative
coordination: deliberation among units of decision-making directed both to
learning jointly from their several experiences, and improving the institutional
possibilities for such learning—a system with continuous discussion across
separate units about current best practice and better ways of ascertaining it. The
idea of deliberative polyarchy is thus to have a mix of flexibility for adjustment to
Louvain, Public Sphere—17
distinct conditions along with a discipline of comparison/learning that respects a
norm of accountability.
In describing deliberative polyarchy we have proceeded so far as though
situated deliberation was the modal form of political decision-making, and largely
ignored the mix of power and interest that is commonly considered the essence
of politics. Deliberation is about addressing problems through mutual reasongiving, and people are not always prepared to listen to reason. So here we want
to indicate some ways in which deliberative polyarchy may well be less
vulnerable to disruption than the widespread idea of politics as power and
interest suggest.
Take first the apparently-attractive flexibility (not uniformity) of polyarchic
solutions. The open-ended standards associated with flexibility may be seen as
invitation to letting powerful decide the rules. An important reason to favor rigid,
uniform rules is to cabin the exercise of power; by sacrificing such rules for
flexibility, the deliberative view may be seen as revealing an intellectually
touching but practically dangerous naïvete about power. But however warranted
this fear may be in some settings, as a general proposition it simply reproduces
the reflex, Hobbesian assertion that anarchy, now in the form of an invitation of
the more powerful to master the less powerful, is the sole alternative to
uniformity. Yet deliberative polyarchy creates, through peer review of local
exercises of discretion, a means for constraining power without recurring to rigid
rules.
Louvain, Public Sphere—18
But safeguarding ongoing deliberation is only one of the tests, and
perhaps not the greatest, that deliberative polyarchy must meet. Deliberation can
fail to get off the ground at all if one party, for example, thinks it can get by its
power a result it could never obtain by reasonable argument. Or deliberation,
once begun, can fail because the parties, having moved far from their initial
divisions, discover a new divide that cannot be bridged, in the relevant time, by
argument. These are of course real possibilities—all too real. A commitment to
deliberation cannot not require commitment to the belief that collective decisionmaking through mutual reason-giving is always possible.
It may be necessary to resort to destabilization, threats, and open conflict,
then, as answers to actors who won’t reason in good faith. But saying “if you
don’t listen to reason, you will pay a high price” is not a subtle joke—“I will give
you a reason you can’t refuse”—nor does the resort to power defeat the idea of
deliberation: people may need to be forced to listen to reason. As a
consequence, there may need to be an acceptable balance of power in the
background of deliberative polyarchy, if deliberation is to be a real possibility (a
point we return to). But it is equally true that once people do listen to reason, the
results need not simply reflect the balance of power that led them to be willing to
listen and defeated their previous imperviousness. So paying attention to power
and threats to exercise it doesn’t reduce deliberation to bargaining. To suppose
otherwise is like thinking that if you need to trust your math teacher in order to
learn how to do a proof then there is nothing more to proof than trust.
Louvain, Public Sphere—19
To be sure, discussion, even when it is founded on reasons, may not—
and often does not—issue in unanimous agreement. So even in a deliberative
democracy, collective decisions must often made be made through voting, under
some form of majority rule. Indeed, deliberation may work better when
participants do not (as in a jury setting) feel the pressure to adjust their views for
the sake of consensus—as if attention to reasons ensured convergence of
judgment, disagreement revealed bias or incapacity, and reasonable people
cannot disagree on complex practical issues.
It may be argued, however, that the prospect of majority rule reintroduces
power as a basis of decision: because collective decision-making concludes with
a vote, participants—anticipating that final stage of resolution—will not have any
incentive to deliberate in earlier stages, and instead will strategically seek allies
for their position, and focus on counting heads rather than weighing reasons.
But the decision rule may not drive process in so direct a way. Even if all
parties know that, at the end of the day, heads may be counted, they are unlikely
to be able to predict which decisions will end in head counting. And the
predictability of outcomes aside they may still may accept the importance of
finding considerations that others acknowledge as reasons: they only need
accept that something other than a resolution that advances their antecedent
interests matters to them. They may, for example, believe that reason-giving is
an important expression of respect, or the right way to acknowledge the
collective or public nature of the decision. If they do, they will be willing to
deliberate in the stages leading up to the vote even when they know a vote is
Louvain, Public Sphere—20
coming. In short, the objection supposes that, once voting is in prospect,
interaction must turn strategic. But this view is no more than plausible than the
obverse claim that moral advantages and possible mutual gains from deliberation
eliminate all strategic maneuvering.
In arguing that global politics in the form of deliberative polyarchy may be
resilient in the face of disparities of power, however, we may be ignoring a much
greater issue about power that defeats the entire idea of a functionallydifferentiated, deliberative-polyarchic, global-political architecture, with no
superintending state in the background: military aggression by actors indifferent
to an international order of any kind, or, more generally, the problems of security
that are commonly said to dominate global politics and determine their traditional,
statist shape. After all, the standard argument against the binding power of moral
or legal constraints in international relations, and for the view of world politics as
an anarchical, self-help system dominated by concerns about security—is
precisely the threat of military force. Facing this threat, or fearing they will,
nations must defend themselves by accumulating power. But with no clear
standard of when power is sufficient, and no bright line between defensive and
offensive forces, secure defenses themselves can look like a project of
aggression; so universal prudence—a product either of the nature of states or of
the competitive pressures they are subjected to—keeps the international order
on the brink of conflict, and undermines normative constraints on state action.
Moreover, because resources can translate into power, there is always suspicion
Louvain, Public Sphere—21
about cooperation because of concerns about the distribution of the gains from
cooperation.
Worse yet, as each state prepares to defend itself, it concentrates power
internally, creating, where one did not already exist, a centralized, unitary
sovereign to act with the necessary decisiveness in the battles to come (war
makes the state as much as the state makes war). On this view, absent a global
state with the power to enforce its will against all possible challengers, the
international order will remain an anarchy, and the anarchy will create actors who
will resist (with reason) normative constraint and cooperation abroad and
regularly reassert a more consolidated form of sovereignty at home.
But this structural realist account is misleading, in two ways. First, it
seems mistaken about the substance of currently dominant security issues.
There is much warfare, and horrific killing by regular and irregular forces. But the
protagonists of military action today are much more likely to be non-state
actors—armed ethnic groups, often operating across national border;
transnational leagues of religious fundamentalists—than nation states. And it is a
bad strategic error to assimilate this dispersed, networked enemy to the familiar
image of the enemy Leviathan. More fundamentally, even if non-state enemies
could be ultimately shown to be under the controlling influence of one or more
states, the decentralization of their financing, logistics, strategizing and tactical
operations mean that effective defense against them requires not centralization,
but networked decentralization on the lines of deliberative polyarchy: the way to
detect and defeat the plots of a dispersed enemy is to constantly pool and
Louvain, Public Sphere—22
evaluate, within and across the efforts of many, independent efforts to identify
and neutralize threats.
More broadly, we are skeptical—for familiar reasons—about the structural
realist’s account of the iron logic of global politics in a world with states. That
account does not survive the relatively-robust finding that regime type does
matter (the democratic peace theory) or the observation that rational conduct
depends on beliefs that agents have about one another or the general fact that
equilibria are generically multiple. Together, these considerations undercut the
case for the conclusion that there is a single form of global politics, dominated by
security issues that swamp all norms and possibilities for mutually beneficial,
functionally specific cooperation. Security matters, but it need not produce the
preoccupation with power, absence of normative constraint, anxiety about the
distribution of the benefits of cooperation, or internal consolidation associated
with the realist argument.
Suppose, then, we accept provisionally both that eliminating power
disparities is not a precondition to deliberation and that security concerns will not
turn global politics into another occasion for the expression of national power and
interest. Still, the reduction of imbalances of power surely favors deliberative
polyarchy, if only by reducing the occasions for recourse to open threats of the
sort, deliberate or else, and the risk of violence attached to them. Hence the
deliberative-polyarchic project for global politics must be attentive to the
background balance of power, within which deliberation is set. Actual
arrangements must provide some basis for confidence that joint reasoning will
Louvain, Public Sphere—23
actually prevail in shaping the exercise of collective power. So discussion that
expresses the deliberative ideal must, for example, operate against a
background of free expression and association, thus providing minimal conditions
for the availability of relevant information. Equally, if parties are not somehow
constrained to accept the consequences of deliberation, if “exit options” are not
foreclosed, it seems implausible that they will accept the discipline of joint
reasoning.
5. Deliberative Polyarchy and Human Rights
This discussion of deliberative polyarchy as an architecture of accountability in
global politics connects to the issue of human rights—the second issue of global
justice that we mentioned earlier—and suggests a distinctive understanding of
human rights as basic conditions of membership or inclusion in a social-political
arrangement. Human rights from this perspective are thus neither the natural
rights that individuals are sometimes said to have in a pre-institutional or prepolitical state of nature; nor are they equivalent to the full set of rights associated
with justice in a society.8
The core or baseline of these membership rights are simply those rights
that individuals have as constitutive rights in deliberative polyarchy. Thus in
establishing its own membership requirements (associated with its conditions of
reasoning, information pooling, comparison, and so on) deliberative polyarchy
8
For further discussion, see See Joshua Cohen, "Minimalism About Human Rights: The Most We
Can Hope For?," Journal of Political Philosophy 12:2 (2004): 190-213; and “A Human Right to
Democracy?”, in Christine Sypnowich, ed., The Egalitarian Conscience: Essays in Honour of G.A.
Cohen (Oxford: Oxford University Press, forthcoming).
Louvain, Public Sphere—24
defines rights that are often coextensive with human rights understood as
membership conditions in distinct political societies. These human rights of
deliberative polyarchy include the rights of association and expression (as well as
rights against torture, arbitrary imprisonment, etc.); they are both substantive
(e.g., the right to bodily security) and procedural (rights of expression and
association). All are abstractly described and differently applicable to different
conditions, and none is specified in detail as in a rule book.
Further specification of these rights is itself a responsibility of deliberative
polyarchy, which requires situated deliberation within and among decision
making units to devise a more determinate scheme of rights that is both sensitive
to the constitutive requirements of deliberative polyarchy and that elaborates the
abstract rights in ways that are suited to circumstances. So there is no fixed
content of human rights (any more than there is fixed content of basic
constitutional rights in contemporary democracies, which embrace, for example,
different understandings of the scope and limits of free expression).
The open character of membership rights invites a first, institutional
objection to their practicality. In highly developed legal systems the elaboration of
rights is chiefly the province of courts; and judicial elaboration produces detailed
rules, with the advantage of enforceability by traditional means, and the
disadvantage—especially burdensome in the deeply diverse circumstances of
global politics—of one-size-fits all solutions. The alternative to detailed rules is
said to be hortatory standards, too broad to be enforced and so not really rights
at all.
Louvain, Public Sphere—25
But just as deliberative polyarchy builds on the experience that there are
ways of learning from and disciplining the most promising results of diverse local
experience with regulation, so it builds on the same capacities to articulate rights
in way that avoids the rigidity of false precision without evaporating into
generalities. Courts or tribunals can play a role in moving other actors to organize
the necessary deliberation, and monitoring their results.
But even if we think that local deliberation, subject to disciplines of
deliberative polyarchy, suffices for defending and further articulating the content
of rights, there remains a second concern that rights themselves may not be
suited to some contexts of global politics. The worry here is that the reasons
rights express are particular to liberal political cultures, and thus that human
rights, including human rights as understood in deliberative polyarchy, is an
intolerant imposition. This concern is misguided. Rights may of course express
the liberal ideas of individuals as autonomous choosers and the obligation to
protect the autonomy of their choices. But that is not all they express. Other, very
different ideas of personhood and its correlative obligations and immunities may
also be interpreted as empower individuals to contest authority in defense of their
dignity; and the idea of rights, understood as the entitlements of membership,
can express these variants of individual empowerment as well.
Thus in Confucian views, for example, identity is relational, responsibilities
are defined by roles, and the good person is one who discharges all the
obligations associated with the relations defined by her position in life. The good
official in this system is one who uses authority to enable others to meet their
Louvain, Public Sphere—26
obligations; the objectionable official is one who affronts the dignity of all by using
authority selfishly, so endangering the fragile balance of duty and obligation. In
Islamic cultures, for another, individuals are God’s vicegerents, obligated to enact
the complex rules of divine law, and charged with executive responsibility for
ensuring that others do so as well. But interpretation of the divine law or its
application, even authoritative interpretation, is always in principle fallible and so
contestable. Otherwise the interpreting authority could be thought to be
blasphemously confusing a believer’s idea with the divine word. What is essential
is that different views endorse the idea of treating individuals as worthy members
of a grouping with ultimately contestable mutual obligations, and provide them
the legitimacy to claim the protections required for such treatment.
6. Democracy?
But what about democracy? Does the extension of deliberative polyarchy to
global politics amount to a form of democracy beyond the state? It is easy to feel
the pull of two different answers: that democracy has no application outside a
world of law and central authority, and the legitimacy of global rule-making must
be democratic. Resisting both pulls, we think that the answer is indeterminate in
two ways: because of conceptual indeterminacies associated with idea of
democracy, and because of real uncertainties surrounding the possible evolution
of global politics, even if it takes a deliberative polyarchic form.
Louvain, Public Sphere—27
Lets start on the conceptual side. Our understanding of democracy has a
range of elements, some of which are shared by deliberative polyarchy and some
of which are not, and it is not clear how to assess the balance.
Thus, deliberative polyarchy shares some features with and is in some
obvious ways different from the democratic practice with which we are familiar:
conventional democratic practice operates in the setting of state with a legal
system, and serves to authorize the state’s law and policy making. Democracy,
as conventionally understood and practiced, is about authorizing the state to
exercise its power. But deliberative polyarchy is not a form of rule-making that
operates in the shadow of a single state.
Moreover, democracy institutionalizes a kind of equality. Thus universal
political rights, political competition, and the connection between electoral victory
and control of government can be seen as providing a link between the interests
and opinions of equal citizens and the exercise of power. Deliberative polyarchy
does not have such organized linkages between equal persons and outcomes.
And furthermore, the existence of a central authority establishes a public,
institutional focus for the formation of a common identity as a people and for
debate—both informal and formal—that shapes public opinion about the proper
exercise of political power, which opinion is connected (more or less loosely) to
political power through formal politics. In conventional democracy, in short, we
have a demos—arguably formed by democracy itself, rather than given by history
and culture—whose “rule” gives democracy its name. But it is an open question
whether deliberative polyarchy is associated with a demos. It obviously does not
Louvain, Public Sphere—28
presume an antecedent demos, but neither does conventional democracy. Can
we expect it to form a demos through its own operations? We return to this later.
The case that deliberative polyarchy is a form of democracy turns on three
features, associated with democratic values of accountability, participation, and
persuasion, all of which depend on the deliberativeness of deliberative polyarchy.
Because deliberation is a kind of persuasion, we will focus here on issues of
accountability and participation.
First, as we emphasized earlier, the architecture of deliberative polyarchy
establishes forms of responsiveness and accountability, through its requirements
of local deliberative problem solving and comparisons with practice in other
settings: part of the case for deliberative polyarchy is that it responds to norms of
accountability and responsiveness that are, on the constructivist argument,
suited to the context of global politics, and help to ensure inclusion. Each of
these is locally accountable and responsive, and the linked system of
deliberation makes each local body accountable and responsive to interests and
judgments elsewhere. The accountability is not electoral, but it does connect
problem solving to both local interests and opinions and those held elsewhere—
in the limit, to global interests and opinions. Deliberative polyarchy is, in short, a
way to systematically pool a wide range of ideas and experiences: it thus
creates, in effect, a dispersed public political-policy discussion across different
locations, and it disciplines solutions in each location by reference to that broader
discussion—so influence extends across locations not through a uniform solution
established by a regulatory center, after inputs from different locations, but by
Louvain, Public Sphere—29
establishing standards elsewhere that others need to consider, and with full
expectation that there will be such mutual consideration.
Consider the emerging regulatory arrangements of the international trade
system housed in the World Trade Organization (WTO), which closely
approximates some elements of the system of EU governance, although it is of
course less well developed. Architecturally the similarity is this: Both the EU and
the WTO anticipate that the freedom of (regional or international) trade they seek
to foment will frequently conflict with, and need to be modified to accommodate a
wide range of normative concerns embodied in the domestic laws and
regulations of member states trading in the relevant markets. Both, furthermore,
permit members states to make domestic rules that inhibit trade on condition that
the inhibiting rules adequately reflect the relevant regional or international
standards. Thus the WTO Agreement on Sanitary and Phytosanitary Measures
(SPS)—which applies to agricultural, health and safety regulation—and the WTO
Technical Barriers to Trade (TBT) Agreement—which has been interpreted to
apply to a broad range of domestic regulation not covered by the SPS—require
that the trade-inhibiting rules, animated for example by a concern to protect
public health or ensure product safety, have a “basis” in international standards.
To show that a basis exists, in the relevant, sense, states must either use those
standards or show through an acceptable rule-making process that the domestic
rules are a reasonable departure from those standards, motivated, for example,
by an assessment of health risks.9 Put another way, membership in the WTO is
The TBT (along with SPS) “represents as big a paradigm shift to international economic law as,
say, the prohibition on the use of force and the introduction o the Security Council with binding
9
Louvain, Public Sphere—30
not equivalent to an agreement to substitute the particular, national rules with the
general laws of efficient commerce. Rather, member states agree to remake their
rules, in domain after domain, in light of the efforts of all the others to reconcile
their distinctive regulations with general standards in whose determination they
participate and that are assumed to be attentive to the interests of others
elsewhere.
But this architectural similarity masks an important difference between the
EU and WTO regulatory regimes that bears on the likelihood of the latter to
produce a democratizing effect. The processes, fora, networked agencies and
other institutions that set standards in the EU routinely revise those them through
peer review informed by the experience of implementation. Civil society actors as
well as governments can increasingly rely on a body of EU administrative law to
protect certain rights to participate in, or at least be informed of, these
deliberations; and it is, again, these features of EU administrative governance
that produce a democratizing effect.
The practice of the international standard setting bodies that produce the
rules in which trade-relevant domestic regulation of WTO members is more
various. In some cases the arrangements are the same because the same
standard setting bodies are active at both the regional and global level. Thus
within the SPS-WTO framework the Codex Alimentarius Commission has global
responsibility for setting standards for commodities, codes of practice and
resolution and police powers represented within the classical world of international law.” These
agreements produce “an internationally determined normativity.” Henrik Horn and Joseph H. H.
Weiler, European Communities—Trade Description of Sardines: Textualism and its Discontent”
[[ref]], 2005. p. 250. See also, Howse, “New Device,” and Joanne Scott [[reference]].
Louvain, Public Sphere—31
maximum limits for additives, contaminants, pesticides, residues and veterinary
drugs. Both the EU and its individual member states are parties to the Codex, as
well as the European Food Safety Agency (EFSA) (which itself cooperates
closely with the Codex); and from the standpoint of the leading national food
safety administrations, such as the Dutch, the EFSA and the Codex form part of
the same regulatory network.
Other domains, such as forestry, lack officially recognized standard-setting
bodies. NGO and industry sponsored codes of good practice tend of compete in
such areas; and even though the evidence suggests that such competition
encourages higher standards, a self-interested group could in theory established
a code of its own liking and offer this as a “basis” for domestic rule-making to
complicit governments. The magnitude of the democratizing effect depends on
the balance between international standard setting bodies that are accountable
and not to peer review, and the pressures to move towards or away from such
accountability. Research on such questions is likely to become a focal concern of
the study of global regulation, where the debate about deregulation and reregulation in the EU setting is now starting to be replayed in discussion about the
WTO.
Second, on the issue of participation: assume that arrangements of
deliberative polyarchy form in distinct areas of concern. In the case of labor
standards, for example, imagine a coordinated system of monitoring that would
certify monitors to conduct audits of firms (also subject to independent audit), and
also require that monitors report back their findings to the “super-monitor”
Louvain, Public Sphere—32
constituted by international organizations such as the World Bank and
International Labor Organization, together with international NGOs and
international trade union confederations. This umpire organization would monitor
the monitors, conduct inspections to verify their integrity, assure the
comparability of monitoring data and methods, and make results publicly
accessible. Firms that depend on consumer loyalty would be eager earn high
grades. They would pressure less visible suppliers, and suppliers of suppliers, to
follow suit. Activists, consumer groups, analysts, and journalists would use
information to define acceptable behavior, press for improvements, and push the
monitors to improve their methods and their standards. And the same will be true
in other areas as well. Assuming a norm of transparency, as well as protections
of rights to associate and speak—all very substantial assumptions, of course—
deliberative polyarchies, once constructed, will become the focus of debate and
pressure from organizations that already exist to influence policy in these areas,
as well as new organizations that see new opportunities for influence. That is, as
arrangements of deliberative polyarchy crystallize, they become the focus of
political activity.
There is of course a very large question—analogous to questions about
conventional democracies, with substantial inequalities of resources—about the
extent to which opportunities for influence on such arrangements will be equal,
even if we assume transparency, and protections of expressive and associative
liberties. But part of the story is participation by non-state organizations that bring
some competence to the discussion of specific problems. Of course participation
Louvain, Public Sphere—33
by groups with special competences naturally prompts concerns about the
representativeness of those organizations, and therefore about the
democraticness of the decisions they reach. How can the outcomes plausibly be
represented as the result of a process in which people participate as equals?
The force of this concern may be fueled by a conception of participating
organizations as representing the interests of a group. But a deliberative
polyarchy may change the way that groups organize and understand their role in
local problem-solving bodies that are parts of a deliberative network. Because
problem-solving is deliberative, groups organize not simply to defend the
interests of members, but because of distinctive views about how best to solve
problems, views that themselves emerge from local discussion. Assuming, once
more conditions of transparency and openness to pressures under conditions of
expressive and associative liberties, democraticness may be fostered by
ensuring that deliberation takes appropriate considerations into account.
Putting these points together, we think that there is no easy way to say
whether the facts of accountability/responsiveness, participation, and persuasion
make the scheme democratic, and outweigh the dimensions on which
deliberative polyarchy evidently does not conform to democratic principles.
But the democraticness of deliberative polyarchy is not simply
conceptually open, but also—more fundamentally—really open as well, and
maybe reality will help to resolve our conceptual troubles. Thus consider again
the idea of democracy as a rule by a people. However we understand the notion
of a people—whether as a body of citizens who are authors of and subject to a
Louvain, Public Sphere—34
set of common laws and institutions, or a body that shares a common societal or
political culture—democratic arrangements institutions are not simply a way for a
people to express its judgments: those institutions also help to form the identity of
a people and the opinion that is, at least in our idealizations of democratic
politics, expressed through elections and then legislation. The people who rule
need not, in short, preexist the institutions through which they rule.
The history of the nation-state is famously the history of the interaction or
co-evolution of the nation—the demos—and the state in each sovereign country.
Early state institutions channeled nationalism, and coursing nationalism
redirected the institutional embankments through which it ran. Persistent
differences in national systems of political representation (consociational or neocorporatist as against pluralist or majoritarian) and of provision of social welfare
(the Bismarkian welfare state linking benefits to occupation history contrasted
with the Nordic Folkehjem providing universal benefits to all citizens) remind us
that the nation state could and did become very different things because of the
particularities of these interactions.
Could there, then, be a comparable formation of a global public fostered
by persisting global politics—with a relatively autonomous administrative space,
associated forms of mobilization and political discourse—organized as
deliberative polyarchy? It depends.
Suppose there emerges, in extension of current forms of global rulemaking, a form of global governance that makes rules and vindicates rights—a
form of governance understood to have significant impact on the fate of
Louvain, Public Sphere—35
individuals—but that this emergence occurs without recourse to the hierarchical
apparatus of the classic nation-state. Instead it occurs through the mutual
accountability of deliberative polyarchy, and an associated global public sphere
that includes persons and peoples the world round in decisions that mattered to
them without making them citizens of a cosmopolitan state? Suppose in
particular a deepening of global administration in a wide range of areas of human
concern, including security, health, education, environment, and conditions of
work and compensation. Suppose, too, that such global rule-making is
increasingly accountable: preceded by hearings, shaped by participation of
affected parties, subject to review, defended by reference to what are commonly
cognized as reasons, while also attentive to the differences of local
circumstance. And suppose that accountable administration has a substantial,
constructive impact of human well-being, and that its rule-making activities
reduce the current global dispersion in living standards—in mortality, morbidity,
income, education, housing, access to clean air and water, transportation,
communication—by leveling up, not down, or, less ambitiously, that these
activities help to reduce the current levels of human destitution. Suppose, finally,
that that impact is widely perceived, and that that perception leads to support for
the continued operation of those arrangements—more participation and greater
willingness to comply.
Couldn’t this all be true? Isn’t this a possible evolution of global politics,
even without a global state?
Louvain, Public Sphere—36
And if it were true, isn’t it plausible that dispersed peoples might come to
share an identity as common members of an organized global people, and not
only in the humanitarian sense that all are human beings on the same planet, nor
only in that we are all interdependent. More than that, our fates—despite our
cultural and linguistic differences—would be deeply and self-consciously shaped
by accountable rule-making that depends both on local debate and on global
comparisons. We would not belong to a single central state, with uniform rules
and rights for all, and global politics would not be defined around a competitive
process for control of that center. But accountable processes of global
administration would play a large role in shaping our lives. And if such common
identity were to emerge—though not as offspring of a conventional political
authority and an organized contest for control of it—then we would have a global
demos: an unusual demos, to be sure, an imagined community requiring newly
capacious acts of imagination, but with sufficiently many of the indicia of a people
to make sense of talk of a global democracy without a global state.
Maybe.
So whether deliberative polyarchy is a form of democracy beyond the
state but suited to the world of global politics depends. And it depends not simply
on our analytical judgments about how best to extend current political concepts
to new and unanticipated cases, but on the deeply uncertain evolution of global
politics itself.