page 1 of 36
Defining the Demos
Until relatively recently, few democrats had much to say about the constitution
of the ‘demos’ that ought to rule. A number of recent writers have, however, argued
that all those whose interests are affected must be enfranchised if decision-making is
to be fully democratic. This paper criticizes this approach, arguing that it
misunderstands democracy. Democratic procedures are about the agency of the
people so only agents can be enfranchised, yet not all bearers of interests are also
agents. If we focus on agency, rather than who is affected, then this leads us to focus
on the permissible limits of action, rather than who makes the decision. Just as
individual sovereignty is ordinarily limited by the requirement not to injure others, we
may apply similar restrictions to what a demos can permissibly do. Thus, any
collection of individuals may choose to regulate their affairs collectively for mutual
benefit, but should be prohibited from imposing negative externalities on outsiders.
On this view, the constitution or definition of the demos is arbitrary, from the
democratic point of view. Moreover, democracy does not require the expansion of the
franchise, unless injuries cannot be avoided.
Keywords: all affected interests; boundary problem; democracy; externalities;
franchise; Goodin; voting.
Democracy means rule by the people. One problem with this concerns who ‘the
people’ in question are.1 Since who makes a given decision can effectively determine
its solution, this is crucially important. For instance, a referendum on the future of
Palestine would doubtless produce very different results depending on whether the
page 2 of 36
franchise consisted only of Palestinians (whether only those in the occupied territories
or also refugees overseas), of Israelis and Palestinians, the wider Arab world, or the
whole world. Until we know the relevant group of decision makers, it seems that there
is no prospect of a genuinely democratic decision.
This article criticizes the recently popular ‘all affected principle’ and,
particularly, applications of it that suggest the need to enfranchise almost everyone in
almost every decision. Such proposals, I argue, would be neither desirable nor even
recognizably democratic. Though it is possible that alternative principles could be
found, for instance focusing on those subject to coercion, I suggest that any such
principle is likely to be misguided, because it focuses on outputs rather than inputs,
whereas democracy is about agency. There is no need for a decision-making group to
include all who are affected by a decision. This does not mean that groups are entitled
to inflict injuries upon outsiders any more than individuals can. Agents, whether
groups or individuals, are constrained in what they can legitimately do to others. I
may not ordinarily injure you to further my own interests, and my society (acting as a
collective) cannot impose injuries on you to further our collective ends.2 The rights of
a given people to sovereign self-determination are restricted by the rights of others.
This paper makes two related main claims: firstly, that we should focus on rights
rather than mere interests, and secondly that rights are better respected by limiting the
power of groups to infringe them, rather than by requiring them to include or
enfranchise the rights-holders. Rights set limits on what we may permissibly do to
others without their consent. This does not mean that we must enfranchise others in
our decision-making so that we can impose costs upon them. Indeed, it is unclear that
mere enfranchisement is sufficient to justify the imposition of costs, since it clearly
falls short of individual consent. I shall argue that individuals may be understood as
page 3 of 36
waiving certain rights where they are part of a democratic group, thus explaining why
groups are able to affect their members in ways that they cannot legitimately affect
outsiders.
Democracy can be seen as a procedure that allows some of us to impose costs on
others for our own benefit, provided that all expect to benefit overall. We are prepared
to waive some of our rights provided that others do so too and all expect to gain. The
gains, however, are confined to members of a particular cooperative group. We are
entitled to impose losses on some, but this does not entitle us to impose losses on
those neither consent nor stand to gain.3 This explains why the existence of a
permanent minority has long been thought problematic for democracy. If members of
this group never actually win, then they never receive the benefits of democracy, so
the procedure is not justified to them. More importantly, for present purposes, this
suggests that we are not entitled to impose costs on those who are not part of the
cooperative scheme. Prohibition of or compensation for negative externalities
imposed on outsiders is the first-best ideal.
Section I introduces the boundary problem and the need for inclusion principles.
Sections II and III respectively outline the recently popular all affected interests
principle and Goodin’s particularly expansive interpretation of it. Section IV criticizes
these proposals as over-inclusive and section V develops one particular diagnosis of
this problem: that the focus on who is affected diverts our attention from decisionmaking agents to the patients affected. Section VI argues that democracy is essentially
about the agency of the people and section VII develops a liberal account of collective
agency. Agents, whether individual or collective, face various moral limits on their
permissible range of action. Though a full account of these cannot be given here,
section VIII argues that we are forbidden from violating others’ rights, but not from
page 4 of 36
affecting them in other ways. Section IX argues that the best way to respect others’
rights is to limit the decision-making power of demoi, as of individuals, rather than
simply to make them more inclusive and finally section X concludes.
I. The Need for Inclusion Principles
Historically, democracy was an inclusive ideal, contrasted to the rule of the few
(oligarchy) or one (monarchy), but the political unit in question was always taken as a
given.4 Athenian democracy may be criticized for excluding women and slaves, but
few would have called it less democratic for not enfranchising Spartans. Nonetheless,
the decisions of a particular people have always had knock-on effects on their
neighbours; damning or polluting a river, for instance, affects the water supply of
those downstream. In an increasingly global world these problems are more
widespread, since decisions taken within a single community almost invariably affect
outsiders.5 Consequently, many of those who have considered the so-called ‘boundary
problem’ have argued that there is no justification for restricting the franchise to
members of a particular state.6
Until relatively recently, democratic theorists had surprisingly little to say about
who constitutes the relevant demos.7 Some recent writers, however, suggest that the
reasons we have to favour democracy, since they are reasons to give ‘the people’ a
vote, also provide grounds for determining the relevant people who ought to be
enfranchised.8 If the justification of democracy lies, for example, in allowing people
to protect their interests, then it seems reasonable that all whose interests are at stake
in a decision should have the chance to participate in making it.9 It is frequently
claimed that i) it is unjustifiable for one group to impose costs on outsiders
unilaterally, so ii) those outsiders ought to be included in the decision-making
page 5 of 36
process. I will argue that, while the first claim is true, the second does not follow and
ought to be rejected.
II. The All Affected Principle
Amongst those who believe that democratic principles provide some answer to
the boundary problem, the most common solution has been that all of those affected
by a decision ought to be enfranchised. The appeal of this ‘all affected’ principle has
been augmented by the observation that it seems implicitly presupposed by many
other proposed principles for constituting the demos.10 Until comparatively recently it
was the case that those living in close proximity were most likely to affect each other
by their actions, so this principle provided derivative justification for territorial
political units.11 Even now this can be seen to justify practices such as federalism or
regional devolution, where certain decisions are devolved to the local constituents
most affected by them. Some writers have even suggested that people should have
unequal shares of decision-making power, proportional to the interests they have at
stake in a given decision, which would make being affected the basis both of
inclusion and exclusion.12
This widespread acceptance of the all affected interests principle, however, may
be a consequence of its lack of definite content, as can be seen when we try to specify
its precise meaning.13 An expansive notion of what it is to be ‘affected’ focuses
simply on whether one is causally affected by the decision to be made.14 This seems
too broad: the decision to tunnel through a mountain causally affects the rocks that are
to be tunnelled through, but it is unclear what it even means to say that the rocks
ought to be enfranchised. Since democracy involves some form of collective agency,
non-agents cannot meaningfully be included. Some have suggested narrower readings
page 6 of 36
of ‘affected,’ for instance that only those subject to the decisions made count as
relevantly affected.15 This proposal may well have substantive merits, but ‘subjected’
is not what is intuitively meant by ‘affected,’ so this should be regarded not as a
version of the all affected principle but as a different principle altogether.16 I shall
focus on the all affected principle, though I think similar objections could be raised
against the ‘all subjected’ principle too.
How, then, should we understand the ‘all affected’ principle? I believe it is most
charitably interpreted as the claim that all those whose interests are affected ought to
be enfranchised. Since this requires that those objects affected have interests, it is
narrower than the ‘causally affected’ interpretation, but broader than ‘all subjected,’
since one can be affected by a law even if not legally bound by it. Obviously, this still
needs to be fleshed out by an account of interests; I simply suppose that such a task is
possible and that what I have to say will be compatible with any plausible theory of
interests. Even so, there are still many different formulations of the principle.17 We
may distinguish, for example, between enfranchising ‘all actually affected interests’,
‘all probably affected interests’, ‘all possibly affected interests’, ‘all and only affected
interests’, and so on. As Goodin argues, some of these are unappealing, while others
may even be incoherent – for instance, we cannot enfranchise only those actually
affected, because who is affected depends on what the decisions is, which depends on
who makes it.
Goodin argues for the most expansive possible reading of the all affected
principle, suggesting not only that we should enfranchise all whose interests are
possibly affected but also that over-inclusiveness is less problematic than underinclusiveness. Hence, we should ensure sufficient inclusiveness by enfranchising
everyone.18 I return to the principle in its more general form below, but begin with
page 7 of 36
Goodin’s interpretation since, by taking the principle to its logical extreme, his
argument clearly demonstrates its main defects.
III. Goodin’s Expansive Interpretation
Goodin argues that all affected interests must be enfranchised, but that we need
not restrict the franchise to only affected interests.19 Consequently, adequate inclusion
can be ensured by enfranchising everyone on every decision. This is a radical
conclusion, but we should not reject it too hastily.
Goodin does not make the, surely exaggerated, claim that any genuine
democracy must be global.20 This would imply either that there has never yet been
anything approximating true democracy in the world or, at least, that increasing
international interdependency means that democracy on a limited scale is no longer
possible. It is perfectly possible that, for some local decisions, all possibly affected
interests are included in a narrower franchise. Thus, Goodin does not insist that we
must always enfranchise everyone, only that doing so is the most practical way to
ensure that all possibly affected interests are actually included. While he may allow
smaller units, such as nation-states, to operate within a larger system of federal
government or regulation, it is for him second- or third-best to a wider franchise that
would effectively eradicate these distinct collectives.21
Goodin takes being affected as sufficient but not necessary for inclusion in the
demos, explicitly rejecting the ‘all and only affected interests’ principle.22 There seem
to be good pragmatic reasons to extend the franchise in this way. Firstly, it avoids
many of the problems associated with finding an adequate account of interests and
determining who is or may be affected by a given decision. Secondly, it yields a
relatively stable demos – i.e. everybody – rather than a different constituency for
page 8 of 36
every decision. This simple solution only works, however, because Goodin does not
consider over-inclusiveness a problem. Not all advocates of the all affected principle
agree; some also take it as a principle of exclusion, saying that those unaffected
should not be included. The next section argues that over-inclusiveness is a problem;
later I argue that even more restrictive versions of the all affected principle are
necessarily over-inclusive, because they include non-agents.
IV. Over-Inclusion
The tendency towards a global demos results from Goodin’s swift dismissal of
any concerns about the over-inclusiveness of the demos. His argument is simply that
people required to vote on some issue that does not affect their interests will do so
randomly and thus be equally distributed across all options, leaving the outcome
unchanged.23 This argument can be challenged at a number of points.24
Firstly, we may question whether unaffected voters really will vote entirely
randomly. Voters may be influenced by ‘external preferences’, such as racism, or
more innocuous factors, such as a tendency to check the first name on the ballot
paper. Enough people voting in this way may affect the outcome, especially if voters
are not independent, resulting in a ‘cascade’.25 Secondly, even if voting is random,
there is still the possibility that it may affect outcomes. Goodin offers an example
where there are 1,000 affected voters, all of whom support a proposal, and 1,000,000
unaffected individuals who also vote on it.26 While we might expect those unaffected
to be divided about 50/50, a division such as 499,499 in favour and 500,501 against
would be plausible and enough to defeat the motion, even if all affected favoured it
(500,499 for the proposal and 500,501 against).27 Thus, we cannot be sure that
outcomes will be unchanged when we permit large amounts of random ‘noise’.28
page 9 of 36
Of course, Goodin may have good responses to these objections. He might
suggest, for example, that democracy favours a global franchise, but should
sometimes be restricted for the sake of other values, much as some regard liberal
rights as imposing external checks on democratic majorities. He might also reply that
people may always vote on objectionable grounds, such as external preferences, but
this does not mean that this is what they should do. It is possible to condemn the way
that people exercise their votes, while maintaining that they ought to have the vote
and that they ought to use it for the common good (or vote randomly).
The logic of Goodin’s argument suggests that we ought always to err on the side
of over-inclusiveness. This rule of thumb can be applied to other cases, such as that of
minors. It is widely accepted that those below a certain age ought not to be
enfranchised, in part on grounds of competence29 Goodin’s arguments, however,
suggest that we ought to enfranchise all children, even newborns, to ensure that all
those who should get the vote do so. While this would mean enfranchising many who
should not get the vote, this is unproblematic if we assume either that they will not
vote or that their votes will cancel out. I find these implications counter-intuitive, but
of course it is a live question how much weight we should attach to pre-theoretical
intuitions in such cases. I raise these concerns not to refute Goodin’s application of
the all affected interests principle, but merely to underline how radical it is.
Goodin could mitigate these difficulties by distinguishing between the distinct
questions ‘Who should be entitled to vote?’ and ‘Who should exercise that vote?’ We
often think that people should have certain rights, even if they should not use them in
particular ways or at particular times. I have the right to say something that you might
find offensive, for example, but it may be wrong of me to do so. Similarly, there
might be good reasons to grant someone the right to vote – for instance because it
page 10 of 36
shows that they are recognized as an equal – even though it would be wrong of them
to exercise it on a given matter.30 Goodin fails to make this distinction and speaks of
people being required to vote on some issue that does not affect their interests. It is
not clear why anyone should be required to vote, though some democracies (including
Australia) practice compulsory voting, which has recently received support
elsewhere.31 It is possible that individuals may sometimes be under a duty not to vote,
for instance when they are uninformed about the issue or unaffected by the decision.32
In these cases, abstention may be a democratic virtue. Thus, one could hold that
everyone should be enfranchised, but that people should only exercise their votes if
they have interests at stake, and otherwise abstain.
A universal franchise would be less objectionable if those unaffected by a
decision could be trusted to abstain. Even this proposal, however, faces problems,
since we cannot be sure that people will actually abstain when they ought to.
Moreover, while it may be an improvement to know that our decisions are not
actually determined by outsiders, it might still be objectionable that they could
influence our decisions if they wished to. Our self-government is not secure where
others have the power to interfere arbitrarily if they wish.
As with any foundational normative premise, it is difficult to prove that overinclusion is problematic to anyone who does not already accept that it is. Nonetheless,
we might draw an analogy here to the individual case. We ordinarily think it better
that you can choose what to do for yourself than be directed by someone else, even if
what she directs you to do is what you would have done anyway. The same, it seems,
is true if the ‘someone else’ were a group, rather than an individual, and, plausibly,
even if you are a member of that group.33 A central feature of liberalism is that
individuals have rights to direct their own lives, whatever the majority thinks they
page 11 of 36
should do. Republicans express a similar concern in terms of non-domination.34 It is
true that the over-inclusive demos differs from ordinary cases, where one individual is
dominated by another, since it includes the individual(s) who should be decisive. This
may be less objectionable, but it still dilutes the power of the locals to govern their
own lives. Even where the wishes of insiders are not swamped by outsiders, the
insiders will find their freedom to govern themselves compromised if they get their
way only at the pleasure of others.
These problems need not be fatal. Indeed, I do not think that being affected
should be a necessary condition for inclusion. My aim is merely to show that Goodin
is too swift in dismissing the idea that over-inclusiveness is a problem. Moreover, the
all affected interests principle is actually far more over-inclusive than so far shown. It
is possible for any one or thing with interests to be affected, but not all of these can be
included because not all creatures with interests are also agents.
V. Agency and Patients
The all affected principle, in its purest form, says that it is sufficient for
inclusion in decision-making that one has interests that may be affected by that
decision. One difficulty is that not all entities with interests capable of being affected
are capable of participation in a decision-making process.
It is common to distinguish between moral agents and moral patients. Moral
agents are actors, capable of deliberate, reasoned choice, and bound by moral
requirements. Moral patients, conversely, are simply those beings deserving of moral
consideration. On most views, all moral agents are also moral patients, but the reverse
does not hold. We may think that all sentient life is deserving of moral consideration,
in virtue of having interests that can be advanced or setback by our actions. This is
page 12 of 36
what distinguishes, say, a dog from a stone, which we can treat in any way that we
like. Not all bearers of interests are also agents. We may consider animals, very young
children and the severely mentally handicapped to be moral patients, though not
agents. This means that, though they are not moral decision-makers, their interests
must be taken into account by others.
Since moral patients have interests, their interests can be affected by a decision.
Thus, the all affected interests principle requires that even non-agent moral patients
are enfranchised if affected, even though incapable of voting. One may reply that this
is unproblematic in practice, since either these non-agents will not be able to vote at
all (as in the case of animals or future generations) or we can assume that their votes
will cancel out (as in the case of young children or the mentally incompetent).
Nonetheless, it seems at least odd, if not problematic, that the theory tells us that
certain groups ought to be enfranchised though they cannot meaningfully be so.35 It is
bizarre to suggest, even in principle, that animals or unborn children ought to be
granted the vote, though they cannot possibly exercise it.
Goodin has considered the problem of mute interests, and suggested that the
solution lies in ‘democratic deliberation within’ – that is, the voting members of the
community should imagine themselves justifying their choices to those whose
interests are affected but who cannot vote for themselves.36 This may be appealing,
but one should not confuse the distinct questions of who should have a vote and how
those who have a vote should exercise that vote.37 One may say that members of a
demos should consider the interests of outsiders in voting, without thinking that this is
merely a second-best and that those outsiders should, ideally, have been enfranchised
themselves.
page 13 of 36
The fact that the all affected principle can be construed as recommending the
inclusion of non-agents within the democratic process at all suggests a faulty
conception of democracy. Rule in the interests of the people (or affected patients) is
not democracy, but guardianship. Democracy is essentially a matter of rule by the
people, that is, a matter of their agency.38 It does not suffice for a democratic process
that my interests are taken into account. Rather, it is necessary that my agency
influences the process, since what I vote for may not be my own interests. This
explains why we should heed people’s actual, expressed wishes, rather than simply
bypassing their opinions whenever we know their interests or preferences.39
Moreover, it explains why we do not think that democracy consists in, or even
requires, that all people be equally satisfied. Some will get what they want while
others will lose out. This is acceptable, provided that the procedure is a fair one,
because it respects these people’s agency – that is their votes – and not because it
results in equal satisfaction of their interests.
VI Democracy as Agency
Democracy is not simply about the equal satisfaction of different parties’
interests, but about giving people equal (and positive) inputs into what the decisions
are. This is why choosing a course of action by lottery, though it shows equal respect
for all people’s interests, is not democratic.40
There may be disagreement about what it is to give people equal inputs into
decision-making. Suppose, for example, we have a society divided into two factions,
which oppose each other over a number of similar issues, such as trade-offs between
economic interests and aesthetic values.41 If two-thirds of society consistently favour
economic efficiency at the cost of aesthetic values, while the other third would prefer
page 14 of 36
to bear financial burdens in order to enjoy a more attractive environment, then we
may question whether majority rule really treats them all fairly. The majority
principle would result in the economizers getting their way on all decisions, and the
aesthetes being effectively excluded. Perhaps it would be more democratic either to
compromise on each decision, or to give the economizers their way on two-thirds of
the decisions and aesthetes their way on the rest, respecting proportionality.42 But no
one thinks it would be democratic to give each group their way on half of the
decisions because, while this would satisfy all people equally, it would ignore their
preferences (it would make no difference whether society was divided 2:1 or 1:5).43
It seems that democracy is, in some sense, a matter of collective agency. If this
is so, then it can only include agents, since others are simply incapable of
participation. James Hyland, in a discussion hitherto largely ignored, proposes
restricting the democratic constituency to agents.44 He starts by observing that almost
all systems exclude the young, the insane and non-nationals,45 but claims that it is
ludicrous to consider a state less democratic simply because a baby boom reduces the
portion of the population enfranchised. He argues that we can avoid this conclusion if
we restrict the ‘domain of relevance’ of democracy. Just as how a society treats
animals is irrelevant to our judgement of its racial justice, so the exclusion of animals
or very young children is irrelevant to an assessment of a society’s democratic
character. They are outside the concept’s domain of relevance because they are not
agents and cannot participate in collective decision-making. Hyland’s positive
proposal is that the quantitative dimension of democracy depends on the extent to
which a state enfranchises all of those who are subject to the decision and capable of
participating in such decision making.46 These two conditions are each necessary, and
jointly sufficient, for inclusion. Consequently, there is nothing undemocratic about
page 15 of 36
excluding non-resident foreigners (who are not subject to the decision) or very young
children (who are incapable of participation) – though Hyland accepts that excluding
twelve-year-olds who satisfy both conditions is undemocratic, even if justifiable.47
Could such a domain restriction rescue the all affected interests principle? It
seems not. While it has the merit that it does not imply anything undemocratic about
the exclusion of non-agents from the franchise, it suggests that being affected is not
sufficient for inclusion. The all affected interests principle is then shown to be
importantly incomplete and, at best, only part of a principle of inclusion: it is also
necessary that democratic participants be agents. Maybe advocates of the all affected
principle will be inclined to go down this avenue, developing a multi-part criterion of
inclusion, but this comes at the cost of the simplicity that is part of the appeal of the
all affected interests principle. I cannot show that no such modified principle could
work, but I think that we should look elsewhere. Once we regard the restriction of
inclusion to agents as not merely a practical constraint, but the essence of what
democracy is, we have the potential to reach very different conclusions about who
should be enfranchised. The following sections describe an agency-centred
conception of democracy and argue that it implies limits on what may be done to
outsiders, rather than a need to enfranchise them.
VII. Democratic Agency
Democracy is based on respecting people as agents, rather than merely as
patients. This is why we should heed people’s expressed wishes, rather than merely
acting in their interests, and why outcomes should be sensitive to these inputs, rather
than serving all interests equally. I cannot, here, offer a full theory of democracy, but I
page 16 of 36
shall sketch an agent-centred conception of democracy to show why we may value it
and the limits on what democratic groups may permissibly do.
If democracy is simply a form of pseudo-collective agency, we may wonder
why anyone would accept such an arrangement, given that it seems to entail a loss of
individual autonomy because they will not always get their way.48 There are several
reasons why autonomous individuals may accept some process of collective decisionmaking. One is that there is a trade-off between having complete power over only
one’s own life and having a share of power over larger decisions.49 Added to this is
the fact that there are certain decisions that an individual cannot take, e.g. individuals
cannot have policies on matters such as national defence.
On many issues we are only able to get what we want if we co-ordinate our
activities with others. This is because we must all live together but different uses of
our property may cause conflict – for example, I will find it hard to use my room as a
quiet study if you use your adjacent room for drum practice.50 Sometimes the actual
solution to these problems may be a matter of indifference, provided that all coordinate on the same solution, for example which side of the road we drive on. In
other cases, it may be that parties have competing preferences or interests, but the
need for co-ordination figures heavily, either because one policy has to be adopted by
all or simply because the parties in question have a higher-order preference for coordination. To take an individual example, suppose you and I want to go for a meal
together. We may have different preferences over where to go – say, I prefer Indian
food and you prefer Chinese food – but the most important thing for each of us is to
go together, so it is no solution for each of us to go our own way. In this case, there
are a variety of possible solutions to our problem. It may be possible to find some
compromise, such as going to a Thai restaurant. If our arrangement is a regular one,
page 17 of 36
we may take it in turns to choose restaurants. In a one-off case, we may simply toss a
coin. Any of these solutions allow us to achieve co-ordination while treating our
competing preferences fairly.
Most democratic decisions involve more than two people, but can be seen as
essentially larger versions of the same problems. We may have two or more
competing groups who need co-ordination on certain matters. Suppose, to use our
earlier economizers and aesthetes example, the latter group favour an expensive urban
regeneration programme. Again, we may be able to find some middle ground
compromise, such as a less expensive scheme, or, if the situation must be ‘winner
takes all’, then we use some fair process, such as a lottery or a vote, that gives each
group a chance of getting what they want. All might subscribe to the democratic
process because it gives weight to their preferences. They may lose on any given
decision, but at least they have a chance of winning next time. (Or, in one-off cases,
they may accept it because it gives them a chance of winning.) Winners are allowed to
impose certain costs on losers, but in the long-run all expect to benefit from these
arrangements compared to a world without co-ordination.51 Thus, all have reason to
comply with democratic outcomes, even when defeated, because, provided the
process is fair, all stand to benefit from the on-going arrangement.52
The upshot is that democratic procedures allow a group to act together, as if they
were a single agent, with some members getting their way and therefore imposing
certain costs on other members.53 The use of a fair collective decision procedure can
be seen as weakening the ‘separateness of persons’, allowing some to impose costs on
others for their own benefit, provided that all expect to benefit in the long-run.
Democratic arrangements are one form of social interaction that imposes particular
costs on particular agents on particular occasions, but can be justified because in
page 18 of 36
general beneficial to all. Other examples include traffic regulations, including
provisions for example that allow emergency vehicles to ignore the usual rules such
as red lights and speed limits.54
VIII. Limits on Agency
Agents normally face limits on what they can permissibly do. I cannot simply
push past you because I am in a rush, for example. Similar limits apply also to
collective agents. We cannot permissibly violate your rights simply because it is in
our interest to do so.55 This applies whether the ‘we’ in question are a sub-state group,
such as a firm, or indeed a state.
Whatever the group, there is an important difference between how it may treat
insiders and how it may treat outsiders. Participation in democratic procedures renders
everyone liable to the costs of defeat on some occasions in return for the expectation
of net benefits in the long-term. Some members of the demos are allowed to impose
costs on others as a consequence of their joint decision-making. This rationale does
not, however, extend to outsiders. Those who are not part of the demos cannot expect
to receive compensating benefits, so should not be liable to bear costs either.56 Thus,
the decision-making power of the demos ought to be limited to distributing benefits
and burdens internally and not to affecting outsiders (negatively, at least).
Goodin considers and rejects such a proposal.57 He argues that, given his very
inclusive reading of the all affected principle, there will be few significant decisions
that can be left to the demos without potentially affecting outsiders. If this is so, it
suggests that a bounded democracy will be unable legitimately to do anything
significant. Our only hope then is the global cosmopolitan democracy discussed
above, which, by enfranchising all, allows us to affect anyone. Since this is
page 19 of 36
impractical, Goodin suggests two more realistic approximations of this ideal: a federal
world government and lateral compensation for negative spillovers.58 Though he is
unspecific about exactly how a federal government would work, the very fact that he
considers it suggests that there may be some day-to-day decisions that can safely be
devolved to a localized demos, without needing universal enfranchisement. The upper
tier of world government, he imagines, would serve only as a higher court of appeal,
for cases where decisions taken within one sub-unit have spillover effects on
outsiders.59 His other proposal – which he describes as third-best to the ideal of the
global cosmopolitan democracy and second-best of global federalism – is a system of
international law that internalizes negative externalities. Localized demoi are thus
prohibited from imposing costs on outsiders, or required to compensate those
outsiders for any harms. Of course, compensation is a difficult issue.60 Nonetheless, I
shall argue that limiting the power of the demos in this way is more attractive than
Goodin allows and does not leave demoi as devoid of power as he suggests.
In his earlier discussion, Goodin argued that anyone who could possibly have
his or her interests affected ought to be enfranchised. No doubt, if the demos could
only affect those who were enfranchised, this would significantly constrain its
decision-making power. The discussion of compensation, however, focuses not on all
possible affects, but only negative externalities. It is far from obvious that we need to
enfranchise others before performing actions that bestow positive benefits on them
(assuming perhaps that these benefits are recognized as such by the recipients). The
decisional power of a group need not, therefore, be limited to what does not affect
outsiders, but only what does not harm outsiders. This already allows much greater
scope for action.
page 20 of 36
Moreover, not all negative affects are illegitimate. Non-consequentialists
typically distinguish between harming and not benefiting others. A full account of
where this boundary should be drawn is beyond the scope of this paper, but we need
not enfranchise outsiders in a decision simply because we could benefit them. If this
were so, then any developed country considering devoting significant sums of money
to some domestic project, such as a new sports stadium, would have to enfranchise the
whole of the developing world, since that money could instead have been given as
aid.61 If the money is legitimately ours – as I shall assume that it is, setting aside
disputes about global justice – then we should be broadly free to spend it as we see fit.
We may be constrained by general moral principles not to actively harm others, but it
does not follow that it is wrong for us to spend the money on domestic projects simply
because we could instead have benefited outsiders with it.
Thus far, we are permitted to impose or withhold benefits on others without
their consent or enfranchisement. There are also cases where one person is morally
permitted to act in ways positively harmful to others. I may, for example, compete
with you for business, even if it leaves you worse off. The legitimacy of my action
does not depend on any version of the acts/omissions doctrine, but rather on the
assumption that there are some things that I have the right to do, even if others are
adversely affected by my decision. I have the right, for example, to open my own
restaurant, even if other local restaurateurs will be harmed by the increased
competition.
If each of us has this (Hohfeldian) liberty, then it will ordinarily be the case that
the two of us could together open a restaurant.62 Of course, I cannot open a restaurant
with you without including you in the decision, and vice versa. Nonetheless, we do
not need to include other affected parties, such as other local restaurateurs, patrons, or
page 21 of 36
our immediate families, in the decision. Though these people may have their interests
affected, positively or negatively, they have no right to inclusion. Agents acting
within their rights do not need the permission of others, so do not need to consult or
include others in their decision-making. This is true whether the agent in question is
an individual or a collective. Thus, the citizens of one country may act in ways they
have a right to, without including others affected by the exercise of that right.
It might be said that, while this action is permissible, it is still undemocratic. But
that is so only in a sense of ‘democracy’ that is of little, if any, normative interest. The
motivation for asking what democracy requires is presumably the belief that we have
at least pro tanto reason to do what democracy requires. A purely descriptive notion
of democracy is, therefore, of little interest. We only care about the requirements of
democracy to the extent that these are normative requirements period. Thus, it seems
that our understanding of democracy should be sensitive to rights.63
A further objection is that I have not only refrained from giving a list of rights,
but even from specifying a procedure by which these rights might be determined. If
there is disagreement between different groups as to what rights people have, we need
some method of adjudication (democratic or otherwise) to establish the limits of
permissible action. My aim, however, is to specify the limits to what groups may
permissibly do, not to say how groups can know these limits in practice. Hence, the
limits that I refer to are those set by a correct account of rights. Of course, in practice
we will need mechanisms to determine what rights people have. These will
presumably be undemocratic, because we cannot have a democratic process without a
prior determination of who should be included in that process, which merely opens a
regress. Nonetheless, I do not think that my proposal is on more shaky ground here
than the all affected interests principle, since that may also be afflicted by
page 22 of 36
disagreement between different groups over what interests people have and whether
particular individuals have their interests affected by a given decision. The advocate
of the all affected interests principle must either stipulate that what she means is all
whose interests are affected on the correct account of interests (a move analogous to
my appeal to the correct account of rights) or perhaps evade the problem by taking
Goodin’s maximally inclusive line, which has been criticized above (and which might
itself be disputed, if some maintain that those unaffected should be excluded).
The rights that we have limit what others can permissibly do to us. I have no
complaint against someone simply because their action affects me. It is only if they
violate my rights, without my permission, that they wrong me. This applies whether
the actor is an individual or a collective. Individuals should be free to act as they wish,
provided they do not thereby infringe others’ rights, and groups should be free to act
as their members wish, provided they do not infringe outsiders’ rights. The content of
these rights cannot be specified here, but this should be the focus of our attention.
This can be seen if we consider ordinary cases of group decision-making.
Consider a group of friends deciding where to go for a meal. Plausibly, a member of
the group not attending the meal should have no say, and one attending only part of
the evening a lesser say than one attending the whole meal. We cannot, however,
conclude that people ought to have power proportionate to the degree to which their
interests are affected. Surely the group of decision-makers ought to be the diners in
question, even if their decision affects the owners, staff and patrons of nearby
restaurants. It does not seem that members of these other groups ought to be
enfranchised; merely that the group should not impose injuries (that is wrongful or
rights-infringing harms) on them.
page 23 of 36
It might be objected that I have merely introduced a new version of the all
affected principle. It might be acknowledged that it is implausible to enfranchise all
whose interests are affected, yet maintained that my proposal amounts simply to
enfranchising all whose rights are affected. This avoids over-inclusiveness, since even
on the Interest Theory of rights not every interest is protected by a right.64 The ‘all
affected’ formula is sufficiently contentless that I have no quarrel with those who
choose to read my proposal as an interpretation of it, provided it is acknowledged that
we have already come some way in moving from affected interests to rights. In the
remaining sections of this article, however, I argue that enfranchising others does not
allow us to override their rights, so it is better to limit the scope of democracy than to
expand the franchise.
IX. Inclusion or Compensation?
Our usual intuition is not that all whose interests are affected by any given
decision have a right to be included in its making. Rather, agents – whether
individuals or groups – have rights to make certain decisions, even where those
impose costs on others.65 The immediate implication for the boundary problem, then,
is that groups do not need to include outsiders when making decisions that they have a
right to. If we legitimately own a piece of land, for instance, then we are entitled to
exclude others from that land, without including them in the decision.66
We are not, however, entitled to violate others’ rights. If our action infringes on
the domain of others’ rights then we need to obtain their consent. Note that this need
not take the form of including them in our democratic procedure. The members of my
bridge club may need to borrow your van for a proposed trip, so we need your
permission to use the van, but it does not follow that you must be included in the
page 24 of 36
decision-making in the same way as the rest of us. Where we choose to go is up to us
alone, but whether we can use your van is up to you alone.
Simply including someone in a democratic process is not itself sufficient to
allow the overriding of their rights. Someone included in the democratic process may
waive some of their rights, provided that they expect to benefit from the process as a
whole (perhaps due to others also waiving their rights), but if they do not then those
rights may still constrain what the rest of us may do. I leave open whether we call this
constraining democracy or prefer to think that what democracy really requires is
respect for rights, rather than unconstrained majoritarianism. Including someone in
the process of decision-making does not necessarily entitle us to override their rights.
Even if all whose rights are affected by a decision have a further right to be included
in the making of that decision, respecting this right to inclusion is no substitute for
respecting the original right; it is merely a condition that might lead them to waive
that original right.
The obvious implication is that group agency – like that of individuals – should
be constrained by others’ rights. There are some things that we cannot permissibly do,
without others’ consent. If we go ahead and violate some other person’s rights, the
usual remedy is not to say that that person should be (or should have been) included in
the decision-making, but that they ought to be compensated for the rights violation.
Thus, it seems i) that groups have the right to act as their members see fit, provided
that they do not violate the rights of any outsiders, regardless of whose interests may
be affected, and ii) that groups ought to be prohibited from violating rights without
permission. One way this permission can be secured is through inclusion in the
democratic process, but this is not to say we must include others because we violate
their rights, since even inclusion is not necessarily justification. The more obvious
page 25 of 36
solution is to cease the rights violation and, if necessary, pay compensation. A theory
of rights will provide us with an account of the limits of group action, but not an
account of who ought to be enfranchised in democratic decision-making.
X. Conclusion
The bulk of this paper has focused on criticizing the all affected principle, which
has become the most popular answer to the boundary problem. I have argued that we
ordinarily have some prior notion of the relevant group of decision-makers, and that
they should be able to make decisions as they wish, subject only to certain restrictions
on not injuring outsiders. I have said little about how such groups are to be
constituted. My main, negative, argument has been that any variant of the all affected
principle and related proposals (such as all subjected) go wrong in a similar fashion,
by focusing on the outputs of democratic decision-making and neglecting the
importance agency.
This negative conclusion, of course, invites questions about how people come to
form a demos, conceived of as in some sense a collective agent. One possibility is that
political communities may be formed by some contract of association, as supposed by
Locke and Rousseau.67 This is the case for some groups, such as voluntary
associations, but it is questionable whether the state can be regarded as voluntary.
Sometimes, we find ourselves simply thrown together by chance, for instance a group
of shipwreck survivors washed up on a desert island. In this case, we have little
choice as to who we cooperate with, particularly if cooperation is necessary for
survival, though it could be said that there is still a choice of sorts, even if not a fully
voluntary one.
page 26 of 36
Some groups may be fit for democratic self-determination in virtue of some
shared property, such as nationality.68 If democracy requires all members of the
community to be willing to bear short-term losses for the benefit of others, democratic
arrangements are likely to function more smoothly where the people are united by a
sense of solidarity. Members of a demos need not share anything in common other
than their membership of that political institution. What determines their inclusion is
not simply that they are affected by decisions, but that they are part of a pervasive
basic structure or that decisions are made in their name.69 Finally, it may be that none
of these methods of constituting the demos are alone adequate, so ultimately we face a
trade-off between competing considerations favouring inclusion and exclusion.70
The idea that the democratic community is arbitrary from a democratic point of
view may strike some as unsatisfactory, but it is not a unique situation. Democracies
require various constitutional essentials, such as the procedure for proposals and
amendments to be made and decisions on the voting rule, to be settled. While these
procedures, once they are in place, can be changed reflexively, their initial
establishment cannot itself be a matter of democratic process, since there is as yet no
determinate democratic procedure to follow. Initial decisions between these possible
forms of democracy cannot be made democratically because, until one has been
chosen, none is privileged over the others.71 It may be that, on some issues, such as
the voting rule, the considerations favouring democracy give us reason to choose one
arrangement rather than another, but it cannot be assumed that the value of democracy
will be fully determinate here. Different groups may realize democracy in different
ways and all of these can be legitimate. How a demos constitutes itself is a matter for
its members, provided it does not transgress the rights of others.72
page 27 of 36
Democracy is the process whereby a group – any group – of agents unite to
coordinate their collective action together. If the members of the demos have the right
to act in certain ways, despite possibly negative effects on outsiders, then ordinarily
so will the group. There is no need to enfranchise outsiders on matters where their
rights are not infringed. Whose interests are affected is irrelevant, since we often have
the right to affect others’ interests, even negatively. Further, when the group’s action
does run up against the limits imposed by others’ rights, the usual solution is not to
include the rights-holders in the franchise, but to limit the action of the group and,
where these limits are exceeded, to require them to compensate outsiders for the
infringement.
This paper has been some time in the making. I thank audiences at the University of Oxford Graduate
Political Theory Workshop (January 2008), the University of Warwick Graduate Conference in
Political Theory (February 2008), my ‘Issues in Contemporary Democratic Theory’ graduate seminar
(Oxford, January 2009), the University of Oxford Philosophy Faculty Doctoral Thesis Seminar (May
2009), the Society for Applied Philosophy annual conference (Leeds, June 2009), the Association for
Legal and Social Philosophy annual conference (Edinburgh, July 2009), and at the University of
Stirling (May 2010). I am grateful, in particular, to Gustaf Arrhenius, Daniel Butt, Daniel Elstein,
Sarah Fine, Simon Hope, Robert Jubb, David Miller, Jussi Suikkanen, and two anonymous referees for
helpful comments.
1
Even the question is ill-defined until it is clear what it means to be part of the demos. Ludvig
Beckman, The Frontiers of Democracy: The Right to Vote and its Limits (Basingstoke: Palgrave
Macmillan, 2009), pp. 10-4, distinguishes between the questions i) who is a member of the demos (in
reference to whom the degree to which the polity is democratic must be assessed) and ii) who ought to
have the vote. This enables him to claim that it would be more democratic if children had the vote but,
nonetheless, they ought not to (p. 119).
My concern is with who should be enfranchised, following Robert E. Goodin, ‘Enfranchising All
Affected Interests, and Its Alternatives’, Philosophy & Public Affairs 35 (2007): 40-68 (which also
appears in Robert E. Goodin, Innovating Democracy: Democratic Theory and Practice After the
page 28 of 36
Deliberative Turn (Oxford: Oxford University Press, 2008), pp. 127-54). This is not to reduce
democracy to voting, but recognition that voting is the ultimate stage in decision-making. Some may
claim only that certain people should have a say – i.e. be entitled to participate in deliberation, but
without voting rights. I would not consider such participants members of the demos; their situation is
more like that of a counsellor, who offers advice but is not a decision-maker. In this, I agree with David
Miller, ‘Democracy’s Domain’, Philosophy and Public Affairs 37 (2009): 201-28, p. 224.
Note also that though the phrase ‘boundary problem’ may suggest spatial boundaries, I am not
concerned with territorial boundaries, but rather with delineating a group of people who should be
enfranchised in decision making, even if they are situated within different geographic states. See Arash
Abizadeh, ‘Democratic Theory and Border Coercion: No Right to Unilaterally Control Your Own
Borders’, Political Theory 36 (2008): 37-65, who distinguishes civic and territorial boundaries at p. 38.
2
‘Injury’ here is distinct from ‘harm,’ since we are sometimes permitted to harm others (e.g. in self-
defence). Injuries are sometimes analysed as something like ‘wrongful harms’; see Seth Lazar, ‘The
Nature and Disvalue of Injury’, Res Publica 15 (2009): 289-304. In fact, a wrong or rights-violation
need not harm someone’s interests at all, unless we postulate an interest in not being wronged.
3
I am unsure whether it is consent or standing to gain that does the important normative work here.
Each may be appropriate in different contexts. We might require actual consent to voluntary groups but
since some groups, such as the state, seem unavoidably non-voluntary, we might then take standing to
gain to be sufficient. The two alternatives might not be so sharply distinct if, for instance, we assume
that people would hypothetically consent if and only if they stand to gain; see Michael Blake,
‘Distributive Justice, State Coercion, and Autonomy’, Philosophy and Public Affairs 30 (2002): 25796. I remain deliberately agnostic, so as not to commit myself on questions of political obligation.
4
Who should be enfranchised does not directly address questions of either citizenship or the right of
immigration into the territory. On citizenship, see Heather Lardy, ‘Citizenship and the Right to Vote’,
Oxford Journal of Legal Studies 17 (1997): 75-100 and Bruno S. Frey, ‘Flexible Citizenship for a
Global Society’, Politics, Philosophy & Economics 2 (2003): 93-114. On immigration and border
controls, see Arash Abizadeh, ‘Democratic Theory and Border Coercion’, and, for further controversy,
David Miller, ‘Why Immigration Controls Are Not Coercive: A Reply to Arash Abizadeh’, Political
Theory 38 (2010): 111-20, and Arash Abizadeh, ‘Democratic Legitimacy and State Coercion: A Reply
to David Miller’, Political Theory 38 (2010): 121-30.
page 29 of 36
5
David Held, ‘The Transformation of Political Community: Rethinking Democracy in the Context of
Globalization’, in Democracy’s Edges, edited by Ian Shapiro and Casiano Hacker-Cordón (Cambridge:
Cambridge University Press, 1999): 84-111.
6
See Gustaf Arrhenius, ‘The Boundary Problem in Democratic Theory’, in Democracy Unbound:
Basic Explorations I, edited by Folke Tersman (Stockholm: Filosofiska Institutionen, Stockholms
Universitet, 2005): 14-29 and ‘The Democratic Boundary Problem’ (unpublished manuscript, 2007);
Beckman, Foundations of Democracy; and Goodin, ‘Enfranchising’. Note that the issue of who is to be
included concerns both spatial outsiders and internal exclusions, such as children. While the identified
principle of inclusion might differentiate between these groups, there is no reason to assume a
difference a priori. The principle that we are searching for should therefore tell us about the rights to
participation of foreigners, children, criminals, etc.
7
Those who grappled with the issue include Carl Cohen, Democracy (Athens, GA: University of
Georgia Press, 1971), pp. 41-54; Jack Lively, Democracy (Oxford: Basil Blackwell, 1975), pp. 10-12;
Frederick G. Whelan, ‘Prologue: Democratic Theory and the Boundary Problem’, in NOMOS XXV:
Liberal Democracy, edited by J. Roland Pennock and John W. Chapman (New York: New York
University Press, 1983): 13-47; Giovanni Sartori, The Theory of Democracy Revisited (Chatham, NJ:
Chatham House, 1987), pp. 21-5; and Robert A. Dahl, Democracy and Its Critics (New Haven: Yale
University Press, 1989), pp. 119-31, and After the Revolution? Authority in a Good Society revised
edition (New Haven: Yale University Press, 1990), pp. 45-51. None provided a satisfactory answer and
some concluded that there was no answer.
8
This paragraph draws principally on Arrhenius, ‘The Democratic Boundary Problem’, and Goodin,
‘Enfranchising’. Abizadeh, ‘Border Coercion’, also thinks that democratic principles can resolve the
boundary problem, though his proposed solution is notably different.
9
This is the primary meaning of democracy according to W. Arthur Lewis, Politics in West Africa
(London: George Allen & Unwin Ltd, 1965) p. 64.
10
Goodin, ‘Enfranchising’, p. 49.
11
Ibid., p. 48
12
David Heyd and Uzi Segal, ‘Democratically Elected Aristocracies’, Social Choice Welfare 27
(2006): 103-27, p. 104, fn. 1; and Harry Brighouse and Marc Fleurbaey, ‘Democracy and
Proportionality’, The Journal of Political Philosophy 18 (2010): 137-55, p. 139.
page 30 of 36
13
Beckman, Frontiers of Democracy, pp. 36-8.
14
Ibid., pp. 41-7.
15
Ibid., pp. 47-50. This principle is invoked by Abizadeh, ‘Border Coercion’, p. 45. He is not
committed to it as the correct principle of inclusion; he adopts it because it is the weakest principle
sufficient for his argument that a bounded community has no unilateral right to close its borders.
16
Ibid., p. 45, Goodin, ‘Enfranchising’, pp. 49-50; and Sofia Näsström, ‘The Challenge of the All-
Affected Principle’, Political Studies 59 (2011): 116-34.
17
Goodin, ‘Enfranchising’, pp. 51-62.
18
Ibid., p. 58.
19
Ibid., p. 58.
20
Such a claim is made by Raffaele Marchetti, Global Democracy: For and Against: Ethical Theory,
Institutional Design and Social Struggles (New York: Routledge, 2008), p. 1. Abizadeh, ‘Border
Coercion’, p. 48 argues that cosmopolitan democracy is necessary to settle questions of border control,
but this issue is special because it involves subjecting outsiders to coercion.
21
Goodin, ‘Enfranchising’, pp. 64-7.
22
Ibid., p. 58
23
If what matters is simply the outcome, one might wonder whether under-inclusiveness matters,
provided that outcomes are unchanged. Perhaps we need only enfranchise a random sample of the
population, provided that their decision is representative of what the whole would choose. This idea is
suggested in Sarah Birch, ‘The Case for Compulsory Voting’, Public Policy Research 16 (2009): 21-7,
p. 26, and Ben Saunders, ‘Estlund’s Flight from Fairness’, Representation 46 (2010): 5-17, p. 13. It is
developed in Claudio López-Guerra, ‘The Enfranchisement Lottery’, Politics, Philosophy, and
Economics (forthcoming). But being granted the vote may be important for other reasons, for instance
because it expresses respect for individuals as competent decision-makers or because exercising it
responsibly serves their moral development.
24
See my review of Innovating Democracy; Ben Saunders, ‘Democracy After Deliberation’, Res
Publica 15 (2009): 315-9.
25
Keith Jakee and Guang-Zhen Sun, ‘Is Compulsory Voting More Democratic?’, Public Choice 129
(2006): 61-75. Goodin, Innovating Democracy, pp. 101-3 and 111-22, draws attention to similar
problems in deliberation.
page 31 of 36
26
Goodin, ‘Enfranchising’, p. 58, fn. 37.
27
I assume unanimity among the 1,000 affected voters. Had they been split 700/300 themselves, then
only a majority of 400 would need to be over-turned, which would be even more likely.
28
Consider further the suggestion that we should toss 1,000,000 coins, adding an extra ‘yes’ vote for
each heads and an extra ‘no’ vote for each tails. These extra votes are certainly random, but I think we
have good reasons for not risking such distortion of the vote. Goodin seems committed to the view that
there is little reason against this, besides the waste of effort.
29
There are various ways in which they may be thought incompetent. Very young infants may be
incapable choosing at all, while older children may be incapable of choosing well. C.f. Francis Schrag,
‘Children and Democracy: Theory and Policy’, Politics, Philosophy & Economics 3 (2004) 365-79;
Dahl, Democracy and its Critics, pp. 123-7; and Beckman, Frontiers of Democracy, pp. 90-119.
30
I defend the value of the mere opportunity to vote in Ben Saunders, ‘The Democratic Turnout
‘Problem’’, Political Studies (forthcoming).
31
See, for example, Arend Lijphart, ‘Unequal Participation: Democracy’s Unresolved Dilemma’,
American Political Science Review 91 (1997): 1-14; and Birch, ‘The Case for Compulsory Voting’.
32
Jason Brennan, ‘Polluting the Polls: When Citizens Should Not Vote’, Australasian Journal of
Philosophy 87 (2009): 535-49.
33
Robert Nozick, Anarchy, State, and Utopia (Oxford: Basil Blackwell, 1974), pp. 290-2.
34
See, for example, Philip Pettit, Republicanism: A Theory of Freedom and Government (Oxford:
Oxford University Press, 1997), pp. 150-3. Goodin has elsewhere criticized republicanism; see Robert
E. Goodin, ‘Folie Républicaine’, Annual Review of Political Science 6 (2003): 55-76. Note, however,
that his main argument is that liberals can consistently share this concern, seeing domination not as
reducing freedom but making it less secure.
35
What follows focuses primarily on children capable of decision-making. The all affected principle
may require enfranchising future generations, which is problematic given the ‘non-identity problem’.
See Derek Parfit, Reasons and Persons (Oxford: Clarendon Press, 1987), pp. 351-80.
Perhaps more problematic is that it may also require us to enfranchise past generations, since (on some
accounts) we may have interests that persist and can be affected after our deaths. The deceased are no
longer agents, but we can often know what they would have wanted. ‘Enfranchising’ past generations
might mean a majority in favour of, for example, criminalizing homosexuality. Of course, one could
page 32 of 36
respond that the majority’s external preferences should be trumped by individual rights. I leave aside
such problems here.
36
Goodin, Reflective Democracy, pp. 169-225.
37
As Abizadeh, ‘Border Coercion’, p. 41, observes, democratic legitimacy seems to require actual
participation, rather than merely hypothetical justifiability.
38
The importance of agency is stressed by Fabienne Peter, Democratic Legitimacy (London:
Routledge, 2009), pp. 25-30; and Margaret Gilbert, A Theory of Political Obligation (Oxford: Oxford
University Press, 2006).
39
David Estlund, Democratic Authority: A Philosophical Framework (Princeton: Princeton University
Press, 2008), pp. 76-8, notes that it may be important to do as someone says, even if we know their
preferences.
40
See Robert E. Goodin, ‘Democracy, Justice and Impartiality’, in Justice and Democracy: Essays for
Brian Barry edited by Keith Dowding, Robert E. Goodin and Carole Pateman (Cambridge: Cambridge
University Press, 2004): 97-111, p. 99; and Estlund, Democratic Authority, pp. 65-84.
41
This example is inspired by Peter Jones, ‘Political Equality and Majority Rule’, in The Nature of
Political Theory, edited by David Miller and Larry Siedentop (Oxford: Clarendon, 1983): 155-82, pp.
167-8.
42
I have considered a proportional alternative to majority rule in Ben Saunders, ‘Democracy, Political
Equality, and Majority Rule’, Ethics 121 (2010): 148-77.
43
This can be expressed in terms of what could be justified to all. A permanent minority might reject
majority rule as not treating them fairly, while the majority may reject equal outcomes as not respecting
their greater numbers. On equality at the justificatory level, see Charles Beitz, Political Equality: An
Essay in Democratic Theory (Princeton: Princeton University Press, 1989), pp. 16-9.
44
James Hyland, Democratic Theory: The Philosophical Foundations (Manchester: University of
Manchester Press, 1995), pp. 76-84. Note that Hyland focuses on those subject to a decision, rather
than merely affected. I am simply exploring whether his domain restriction argument can be borrowed
by advocates of the all affected principle.
45
For evidence of how widespread these exclusions are, see Beckman, Frontiers of Democracy, pp. 62-
71 and 146-51. The last of these goes beyond removing non-agents, since foreigners are agents and
could be enfranchised. For now, my focus is only on excluding non-agents, in order to reduce the over-
page 33 of 36
inclusiveness of the all affected interests principle. In effect, I am considering the revised principle that
all agents whose interests are affected ought to be included.
46
Hyland, Democratic Theory, p. 81. The reference to ‘quantitative dimension’ is because Hyland
thinks that an overall assessment of a state’s democraticness depends both on the extent of the
franchise and how equal those enfranchised are (the qualitative dimension).
47
Hyland, Democratic Theory, p. 82. One consequence of this is that a baby boom does indeed make a
society less democratic, albeit only some time later.
48
I say pseudo-collective agency because my conception of democracy does not, like perhaps
Rousseau’s, posit some collective will over and above that of individuals. Rather, it is thoroughly
individualistic, but assumes that the decision procedure allows a diverse group of individuals to act in
some ways as if they were a single agent. I believe that this theory is liberal, yet can still resist calls to
enfranchise all affected. On the inclusionary pressure faced by ‘liberal’ democrats, see Miller,
‘Democracy’s Domain’, pp. 205-13.
49
Dahl, After the Revolution, pp. 76-77.
50
For similar problems, see Barbara H. Fried, ‘Left-Libertarianism, Once More: A Rejoinder to
Vallentyne, Steiner, and Otsuka’, Philosophy & Public Affairs 33 (2005): 216-22, pp. 217-8.
51
This argument is rather like J. S. Mill’s for free trade. He thought that, though competition involves
various immediate harms, all should expect to benefit from prosperity in the long-run. See Richard
Vernon, ‘Is There a Global Harm Principle?’, Critical Review of International Social and Political
Philosophy 12 (2009): 1-18, pp. 4-5.
52
This may suggest a ‘fair play’ theory of political obligation; see Herbert Hart, ‘Are There Any
Natural Rights?’, Philosophical Review 64 (1955): 175-91. I am trying to be as neutral as possible
about what grounds this obligation. It matters that the procedure is a fair one, so all expect to get their
way sometimes, but then we may simply appeal to presumptive benefits (see George Klosko,
‘Presumptive Benefit, Fairness, and Political Obligation’, Philosophy & Public Affairs 16 (1987): 24159) or what people might hypothetically consent to (see Blake, ‘Distributive Justice, State Coercion,
and Autonomy’).
53
It is possible that the costs permissibly imposed may be limited. For the idea that a majority may not
impose ‘uncompensatable’ losses on minorities, see Richard Vernon, Political Morality: A Theory of
Liberal Democracy (London: Continuum, 2001), pp. 72-92.
page 34 of 36
54
For an approach to such problems that emphasizes dealing in utilities, see Donald G. Brown, ‘Mill’s
Moral Theory: Ongoing Revisionism’, Politics, Philosophy & Economics 9 (2010): 5-45, pp. 23-4.
55
Who is responsible when collective agents do wrong? This question is beyond the scope of the
present paper, though I am inclined to say that outsiders can only hold the collective as a whole
responsible and that the distribution of any burdens within that collective is simply another matter of its
own internal distribution of costs. For various more detailed treatments of collective responsibility,
however, see Joel Feinberg, ‘Collective Responsibility’, Journal of Philosophy, 65 (1968): 674-688;
Margaret Gilbert, Sociality and Responsibility: New Essays in Plural Subject Theory (Lanham, MD:
Rowman and Littlefield); Chandran Kukathas, ‘Responsibility for Past Injustice: How to Shift the
Burden’, Politics, Philosophy & Economics 2 (2003): 165–90; David Miller, National Responsibility
and Global Justice (Oxford: Oxford University Press, 2007); and Avia Pasternak, ‘Sanctioning Liberal
Democracies’, Political Studies 57 (2009): 54-74.
56
As Vernon, ‘Global Harm Principle’, p. 1, notes these negative duties are generally accepted. I think
it unhelpful to draw an analogy here to Mill’s harm principle, however. The principle in question here
prima facie forbids causing harm to others. Mill’s principle was that preventing harm to others was the
only (necessary, not sufficient) condition for social regulation.
57
Goodin, ‘Enfranchising’, pp. 62-3.
58
Ibid., pp. 64-7.
59
C.f. Dahl, After the Revolution, pp. 68-75. This possibility is explored in Eric Cavallero, ‘Federative
Global Democracy’, Metaphilosophy 40 (2009): 42-64. It also has affinities with the ‘contestatory’ or
editorial democracy of Philip Pettit, ‘Democracy, Electoral and Contestatory’, in Ian Shapiro and
Stephen Macedo (eds.), NOMOS XLII: Designing Democratic Institutions, (New York: New York
University Press, 2000).
60
For a small sample of the literature on compensation and rectification, see: Robert E. Goodin,
‘Theories of Compensation’, Oxford Journal of Legal Studies 9 (1989): 56-75; Jeremy Waldron,
‘Superseding Historic Injustice’, Ethics 103 (1992): 4–28; Claus Offer, ‘Disqualification, Retribution,
Restitution: Dilemmas of Justice in Post-Communist Transitions’, The Journal of Political Philosophy
1 (1993): 17-44; Kukathas, ‘Responsibility for Past Injustice’; Andrew I. Cohen, ‘Compensation for
Historic Injustices: Completing the Boxill and Sher Argument’, Philosophy & Public Affairs 37 (2009):
page 35 of 36
81-102; and Daniel Butt, Rectifying International Injustice: Principles of Compensation and
Restitution Between Nations (Oxford: Oxford University Press, 2008).
61
Even nationalists admit that such projects should not take priority over the urgent needs of outsiders.
See David Miller, Citizenship and National Identity (Cambridge: Polity Press, 2000), pp. 177-8.
62
This is not a logical entailment. Each of us may have the Hohfeldian liberty to open a restaurant, but
have promised someone that we will not go into business together; then neither of us has a liberty to
open a restaurant together. On the relation between groups and their constituent members, see David
Copp, ‘The Collective Moral Autonomy Thesis’, Journal of Social Philosophy 38 (2007): 369-88; and
Seumas Miller, ‘Against the Collective Moral Autonomy Thesis’, Journal of Social Philosophy 38
(2007): 389-409. I take no stance on this debate and simply assume that ordinarily groups will be
permitted to act in ways that their members are.
63
See Corey Brettschneider, ‘The Value Theory of Democracy’, Politics, Philosophy & Economics 5
(2006): 259-78.
64
Matthew H. Kramer, ‘Rights Without Trimmings’, in Matthew H. Kramer, N. E. Simmonds, and
Hillel Steiner, A Debate Over Rights: Philosophical Enquiries (Oxford: Clarendon Press, 1998), pp. 7111, p. 62.
65
This is not to say groups have a right to autonomy in the same way as individuals. The group’s right
to self-determination may be derivative from that of its members. See Andrew Altman and Christopher
H. Wellman, ‘A Deontological Defense of Democracy: An Argument from Group Rights’, Pacific
Philosophical Quarterly 89 (2008): 279-93.
66
The implications of this for immigration depend on the legitimacy of states’ territorial claims. On
this, see Tamar Meisels, ‘Liberal Nationalism and Territorial Rights’, Journal of Applied Philosophy
20 (2003): 31-43; Cara Nine, ‘A Lockean Theory of Territory’, Political Studies 56 (2008): 148-65;
and Hillel Steiner, ‘May Lockean Doughnuts have Holes? The Geometry of Territorial Jurisdiction: A
Response to Nine’, Political Studies 56 (2008): 949-56.
67
Dahl, Democracy and its Critics, p. 122.
68
David Miller, On Nationality (Oxford: Clarendon Press, 1995), pp. 81-118.
69
These alternatives are meant to capture, respectively, the positions of John Rawls, Political
Liberalism (New York: Columbia University Press, 1996), pp. 257-88, and Thomas Nagel, ‘The
Problem of Global Justice’, Philosophy & Public Affairs 33 (2005): 113-47.
page 36 of 36
70
Miller, ‘Democracy’s Domain’, pp. 226-8.
71
Of course, we may consider the process by which democracy comes about more or less democratic –
e.g. contrasting a popular uprising to a military coup – but this differs from evaluating what is brought
about as better realising democracy.
72
This is not to presuppose democracy, but simply a prior right to self-determination. The group may
elect to govern itself non-democratically. See Altman and Wellman, ‘A Deontological Defense of
Democracy’.
About the author:
Ben Saunders is Lecturer in Philosophy at the University of Stirling (UK). His primary interests lie in
democratic theory and John Stuart Mill. His work has previously appeared in Ethics, The Journal of
Applied Philosophy, and Utilitas.