In this article I argue that theorizing about justice at the level of ideal theory is inherently flawed and thus
has impoverished liberal egalitarianism. Ideal theorists (falsely) assume that a political philosopher can
easily determine (or has privileged access to) what constitutes the ‘best foreseeable conditions’. Further-
more, by assuming full compliance, ideal theorists violate the constraints of a realistic utopia. More
specifically I argue that liberal egalitarians who function at the level of ideal theory adopt a cost-blind
approach to rights and a narrow view of possible human misfortune. The former issue leads liberal
egalitarians to give priority to a serially ordered principle of equal basic liberties or to treat rights as
‘trumps’; and the latter to a stringent prioritarian principle (Rawls’ difference principle) or luck
egalitarianism.Taken together, the cost-blind approach to rights, coupled with the narrow view of human
misfortune, mean the liberal egalitarian theories of justice cannot address the issue of trade-offs thatinevitably arises in real non-ideal societies that face the fact of scarcity.This makes liberal egalitarianism
an ineffective theory of social justice.
Political philosophers have recently begun to take seriously methodological
questions concerning what a theoretical examination of political ideals (e.g.
justice) is supposed to accomplish and how effective theorizing in ideal theory isin securing those aims. Andrew Mason (2004) and G. A. Cohen (2003), for
example, believe that the fundamental principles of justice are logically indepen-
dent of issues of feasibility and questions about human nature. Their position
contrasts sharply with political theorists like John Dunn (1990) and Joseph Carens
(2000) who believe that normative theorizing must be integrated with an appre-
ciation of the empirical realities of one’s society. Rather than bracket questions of
feasibility and human nature, empirically oriented political theorists believe that
real, non-ideal considerations (like our historical circumstances, problems of
institutional design, etc.) must be taken seriously when deriving normative
theories of justice.1 And some justice theorists, most notably John Rawls (1971;
1996), attempt to occupy a middle position that acknowledges some moderate
feasibility constraints (e.g. pluralism) but also employs a number of idealizing
assumptions (e.g. society is closed, full compliance, etc.) when deriving the
principles of justice.
The disagreement between those political philosophers who feel inclined to
invoke highly abstract hypotheticals when deriving the principles of justice, and
those political theorists who take seriously real, non-ideal considerations, is a
disagreement over how fact-sensitive a theory of distributive justice ought to be.Mason raises a challenge for the more empirically grounded political theorists
doi: 10.1111/j.1467-9248.2007.00656.x
P O L I T I C A L S T U D I E S : 2 0 0 7 VO L 5 5 , 8 4 4 – 8 6 4
when he asks:‘what reason do we have for thinking that any adequate analysis of
an ideal such as justice must be conducted in the light of an investigation of what
is feasible?’ (Mason, 2004, p. 255). In this article I hope to provide a compelling
response to Mason’s question. I believe there is some conceptual incoherence
involved in saying ‘This is what justice involves, but there is no way it could beimplemented’ (Mason, 2004, p. 255).This incoherence stems from the fact that a
theory of social justice, and the principles of justice it endorses, must function as
an adequate guide for our collective action. A theory of social justice that yields
impotent or misguided practical prescriptions is a deficient theory of justice. If the
collective aspiration to implement the conclusions of a theory would not result in
any noticeable increase in the justness of one’s society, then it fails as a normative
theory.
In this article I argue that theorising about justice at the level of ideal theory is
inherently flawed and thus has impoverished liberal egalitarianism. I believe thatmoderate ideal theorists, such as Rawls and Ronald Dworkin, are actually much
closer to the idealizing end of the spectrum and thus their theories are not
adequately fact-sensitive to be considered realistically utopian.2 Ideal theorists
(falsely) assume that a political philosopher can easily determine (or has privileged
access to) what constitutes the ‘best foreseeable conditions’. Furthermore, by
assuming full compliance, ideal theorists violate the constraints of a realistic utopia.
Determining what is feasible in partially compliant societies that exist in the
modern area of rapid globalisation is perhaps one of the major sources of political
disagreement in contemporary democratic societies. Rather than side-steppingsuch disagreement, political philosophers should advance theories of justice that
adopt a critically reflective attitude towards their own background empirical
assumptions concerning what is realistically possible.The moderate ideal theories
of Rawls and Dworkin fail to internalize such a reflective attitude. By illustrating
the shortcomings of Rawls and Dworkin I hope to convince political philoso-
phers that they need to take the empirical realities of real societies more seriously.
Rather than moving in the direction advocated by Cohen and Mason (i.e.
towards a more extreme idealized position) political philosophers should take
more seriously non-ideal theory. This will help equip them with a theory of
justice that can provide some normative guidance for real, non-ideal societies.
More specifically, I argue that liberal egalitarians who function at the level of ideal
theory adopt a cost-blind approach to rights and a narrow view of possible human
misfortune.The former issue leads liberal egalitarians to give priority to a serially
ordered principle of equal basic liberties (Rawls, 1971; 1996) or to treat rights as
‘trumps’ (Dworkin, 1978); and the latter to a stringent prioritarian principle
(Rawls’ difference principle) or luck egalitarianism. Taken together, the cost-
blind approach to rights, coupled with the narrow view of human misfortune,
mean the liberal egalitarian theories of justice cannot address the issue of trade-offs that inevitably arises in real non-ideal societies that face the fact of scarcity.
This makes liberal egalitarianism an ineffective theory of social justice.
J U S T I C E I N I D E A L T H E O R Y : A R E F U T A T I O N 845
Liberal egalitarian theories of justice are theories that typically function at the
level of ideal theory.The distinction between ideal and non-ideal theory is not
given rigorous classification in the existing literature. As Mason (2004, p. 265)notes, this distinction is employed by Rawls in The Law of Peoples.An account of
justice in ideal theory must recognize ‘some moderately strong feasibility con-
straints which require it to be realistic in the best of foreseeable conditions’
(Mason, 2004, p. 265). Rawls describes ideal theory as being realistically utopian.
Political philosophy is realistically utopian ‘when it extends what are ordinarily
thought of as limits of practical political possibility’ (Rawls, 1999a, p. 6). This
contrasts with non-ideal theory, which is concerned with problems of non-
compliance or unfavourable (historical, social or economic) conditions.
Determining exactly how fact-sensitive a theory of justice ought to be goesbeyond the aspirations of this particular article.3 A central aim of this article is to
convince political philosophers that even moderate ideal theorists are too ideal-
ising and that this obstructs the value of their central prescriptions. It is useful to
conceptualize the issue of fact-sensitivity as occurring along a spectrum. Figure 1
captures the different positions political philosophers can take on the issue of how
stringent the requirement of fact-sensitivity ought to be, and the dangers inherent
with the extreme positions.At the extreme of fact-insensitivity (what we can call
extreme ideal theory), one runs the risk of invoking an account of justice that fails
to function as an adequate guide for our collective action in the real, non-idealworld. At the opposite end of the spectrum is the danger that all existing
constraints (even those imposed by an unjust social structure) are taken as
legitimate constraints and thus justice simply reaffirms the status quo. One danger
with this overtly fact-sensitive approach is that it is prone to what Mason (2004)
calls ‘adaptive preference formation’. This occurs when one believes that the
current situation is the best possible situation.
In this article I limit my critique to those who occupy the moderately ideal
position, which I believe is most effectively illustrated by the theories of John
Rawls and Ronald Dworkin.4 Defenders of this moderate idealizing position willargue that their position shares little with the extreme idealizing position advo-
cated by G. A. Cohen. Indeed, Cohen is critical of Rawls’ position precisely
because Rawls takes certain non-ideal considerations (e.g. human nature) seri-
ously. I agree that there are important differences between the extreme and
moderate idealized positions. Rawls and Dworkin are not completely blind to
non-ideal considerations. In fact, a good deal of their theories are motivated by
non-ideal considerations. Rawls, for example, defends a theory of justice that
seeks to take seriously what David Hume called the ‘circumstances of justice’.
These include the facts of moderate scarcity and limited altruism. Furthermore,in Political Liberalism Rawls revised his earlier theory of justice because he believed
it failed to take the fact of reasonable pluralism seriously enough and thus it could
not secure the stability of a just democratic regime. Dworkin (2000) sees his
defence of equality of resources as superior to Rawls’ theory of justice because
Dworkin takes inequalities in natural endowments and concern for personal
responsibility more seriously than Rawls. So a good deal of the debates that take
place within ideal theory are debates about which particular constraints a theoryof justice should function within.
Rawls is not alone in modifying his theory so as to make it better equipped to
take more seriously real, non-ideal considerations. Various supporters of the
Rawlsian project have also proposed some modifications. So some tinkering with
Rawls’ comments about the family gives you liberal feminism (à la Okin, 1989),
some tinkering with Rawls’ assumption about society being closed gives you
global justice (à la Beitz, 1979), some tinkering with the assumption that all
people fall within the normal range of functioning gives you an account of just
health care (à la Daniels, 1985) and some tinkering with Rawls’ assumption thatsociety is culturally homogeneous gives you a liberal account of multiculturalism
(à la Kymlicka, 1989).
Figure 1: The Spectrum of Fact-Sensitivity Ranging from the Extreme IdealPosition to the Extreme Non-ideal Position
But there comes a stage where those inclined to defend Rawls must realize that
the sheer volume of qualifications they make in defence of his theory itself
constitutes a reason for rethinking the viability of Rawlsian justice. What is
surprising about the Rawlsian canon is that, despite all the tinkerings with Rawls’
theory – to equip it to address patriarchy, global poverty, health care, multicul-turalism and reasonable pluralism – Rawls himself did not change (at least
substantially) the content of the theory. The two serially ordered principles of
justice remain the principles of justice. How can that be so? How can the
principles of justice that are (arguably) appropriate in a scenario where concerns
of patriarchy, global poverty, health care,multiculturalism and reasonable pluralism
do not arise be the same principles of justice that are appropriate in a society
where such concerns do arise? Surely this is not just a coincidence! Rather,
defenders of Rawls have perhaps been too vigilant to defend the theory in light
of overwhelming difficulties and I, for one, have finally thrown in the towel. In
fact, I am willing to abandon ideal theorizing more generally in search of a new
approach to distributive justice. In the conclusion of this article I hope to make
some progress on this front when I detail some general strategies for taking
non-ideal theory more seriously.
Perhaps the most pressing failure of Rawlsian moderate ideal theorizing is that it
fails to take scarcity seriously.This is a particularly troubling problem for Rawls’
theory because he serially orders his two principles of justice. Such ordering
yields impotent prescriptions for real societies that face conditions of scarcity.
Defenders of Rawls will no doubt reply that such a charge is misplaced as Rawlsacknowledges the so-called ‘circumstances of justice’. These include the fact of
moderate scarcity. Following Hume, Rawls argues that justice is only necessary and
possible when society falls between the extremes of severe scarcity, and abundance
of goods.While it is true that Rawls does mention moderate scarcity this fact does
not figure in the deliberation of the parties in the original position and Rawls
makes a number of idealizing assumptions (e.g. full compliance, society is closed
and filled with healthy people) that obscure how constraining scarcity will be on
fulfilling the demands of his two principles of justice.This becomes most evident
if we consider some of the comments Rawls makes concerning maximin in Justice
as Fairness:A Restatement .
Maximin and the Cost-Blind Approach to Rights
Rawls violates the assumption of moderate scarcity when he explains why the
contracting parties would choose his two principles of justice over the principle
of utility. Let me expand on this particular critique for it illustrates the problem
with functioning at the level of (even moderate) ideal theory. By assuming full
compliance Rawls’ idealizing assumptions collapse into what Onora O’Neill(1996, p. 41) calls ‘idealization’. Idealization involves making claims that are
actually false, in order to simplify an argument. Rawls engages in idealization
when he argues that the maximin strategy would lead the parties in the original
position to endorse his two serially ordered principles of justice over the principle
of utility.
I will assume that most of my readers are familiar with the basic details of Rawls’
original position so I shall not rehearse, in any great detail, those points here.The
parties in the original position are placed behind the ‘veil of ignorance’ and are
presented with a shortlist of traditional conceptions of justice.They must decide
which of these will secure them the largest share of social primary goods (rights
and liberties, powers and opportunities, income and wealth and self-respect).
Rawls believes that it is rational for the parties to choose his two principles of
justice over the rival principles. He defends this claim by arguing that the two
principles of justice are the maximin solution to the problem of social justice.‘The
maximin rule tells us to rank alternatives by their worst possible outcomes: we are
to adopt the alternative the worst outcome of which is superior to the worstoutcomes of the others’ (Rawls, 1999b, p. 133). Given that the parties do not
know what their social position will be, or what their level of natural assets will
be, it is rational, Rawls claims, for them to adopt the conservative attitude
expressed by this rule.The maximin rule is ‘a useful heuristic rule of thumb for
the parties to use to organize their deliberations’ (Rawls, 2001, p. 97, n. 19).
The maximin rule,argues Rawls, instructs the parties to choose the two principles
of justice for they guarantee the highest minimum pay-off. Rawls believes that
utilitarianism permits the interests of some to be sacrificed for the greater good
of others and thus he argues that the principle of utility would not be chosen as
it does not satisfy the requirements of maximin. Some have questioned whether
it is rational for the parties to adopt such a conservative stance. John Harsanyi
(1975), for example, argues that the maximin principle has some very irrational
conclusions. But let us put these issues aside as my point is to show that Rawls
falsely assumes that the rights entailed in the equal basic liberties principle are
costless rights. This assumption permits Rawls to make the case for serially
ordering the first principle of justice over the principle of fair equality of
opportunity and the difference principle.
Without knowing the probability of being talented or born into a favourable
social position, Rawls maintains that it is rational to assume you have the least
desirable genetic endowments5 and family position. A crucial assumption that
Rawls slips into his discussion of maximin is that the contracting parties are
deliberating about what their place would be within a society that exists in the
circumstances of justice under reasonably favourable conditions. Elaborating on what
the latter entails Rawls claims that they are the ‘conditions that, provided the
political will exists, make a constitutional regime possible’ (Rawls 2001, p. 101).
These are conditions such as sufficient economic and technological development,sufficient natural resources and an educated citizenry. But this assumption is a
blatant violation of one of the considerations blocked by the veil of ignorance,
J U S T I C E I N I D E A L T H E O R Y : A R E F U T A T I O N 849
namely, how rich or poor our society is. Now perhaps Rawls thinks that even
poor societies could, if only they had the political will, become liberal democ-
racies like America. Such an assumption is puzzling and really demonstrates how
localized and insular Rawls’ discussion of social justice is. How many of the
world’s non-democratic countries are non-democratic simply because they lackthe political will to become a constitutional democracy? It is more than just a
coincidence that the world’s most stable constitutional democracies are also those
that enjoy the highest levels of economic and technological development, are rich
in resources and have an educated citizenry.
Defenders of Rawls might respond that such an assumption is tenable because
Rawls is functioning at the level of ideal theory.Thus Rawls can assume that the
society in question is closed and one where there will be full compliance . It is
reasonable to assume, they might contend, that most societies in such circum-
stances will be in the reasonably favourable position Rawls describes. But if Rawls’ theory is supposed to yield principles of justice that can serve as a guide
for the collective action of citizens in open, partially compliant societies, then
Rawls’ ‘simplifying’ assumptions will prove problematic. It is one thing to say that
we should serially order the principle of equal basic liberties when no-one would
seek to transgress such rights (e.g. the state, individual citizens or foreigners). But
it is quite another thing to say that these rights should be given absolute priority
over other concerns (e.g. equal opportunity, the difference principle) when the
reality is that a significant portion of the government’s budget will need to be
invested in protecting and promoting the first principle of justice.The assumption that these favourable conditions will hold in our society violates
the requirements of the maximin rule.This rule requires us to rank alternatives by
their worst possible outcomes.The worst possible outcome would be to be the
least advantaged in the least advantaged society that satisfies the circumstances of
justice. Just as we cannot assume that we will be born with ‘reasonably favourable
genetic endowments’ or into a family with ‘reasonably favourable social advan-
tages’, nor can we assume that we will be born into a society that has the
reasonably favourable conditions stipulated by Rawls. One of the things the
parties in the original position do not know is what the particular circumstancesof their society (e.g. its economic or political situation) will be.We do not know
how rich or poor our society will be. So Rawls is not justified in claiming that
the parties can assume that whatever society they end up in it will be one in
which the reasonably favourable conditions that make a constitutional democracy
possible exist. Behind the veil of ignorance we do not know what the probability
of being born into an affluent country is, just as we do not know what the
probability of being born into an affluent family is.
We can thus make a distinction between what I call the Lax Maximin Rule and
the Stringent Maximin Rule .The former is adopted by Rawls and requires us toassume that we will be members of the least advantaged group in a society that
exists in reasonably favourable conditions.The Stringent Maximin Rule requires
Reflecting on the budget costs of protecting various rights is something which
the parties in the original position must contemplate as they decide whether
the principle of utility or Rawls’ two principles of justice will maximize the
minimum.Assuming, as stringent maximin requires them to, they will be the least
advantaged members of the least advantaged society, they will know that it is
inevitable that trade-offs in rights protection must occur. With only limited
revenues available the state will not be able to make rights absolute in the way
Rawls’ two principles presuppose.
Faced with these kinds of considerations, which are ones the contracting parties
must reflect upon, is it rational for them to opt for Rawls’ two principles of
justice? No. The equal basic liberties principle, for example,tells us that individualrights should take a priority over concerns of equal opportunity and socio-
economic inequalities. But Rawls assumes that protecting and promoting these
J U S T I C E I N I D E A L T H E O R Y : A R E F U T A T I O N 851
principle can guarantee the rights of minorities (or the rights of anyone).And this
is precisely why we should reject Rawls’ priority rules which give an absolute
priority to liberty. Such a strategy presumes rights are costless. But rights have
costs; vast amounts of revenue are needed to legally enforce these rights effec-
tively. Rawls’ idealizing assumptions (e.g. society is closed and there is fullcompliance) no doubt help explain why he believes that the equal basic liberties
principle can be guaranteed. If all citizens are healthy and fully compliant, and
society is closed, then legally enforcing these rights only necessitates that the state
refrain from interference.By ignoring the realities of non-compliance and scarcity
of resources, Rawls’ theory of ‘justice as fairness’ insulates itself from the most
pressing concerns that face every real society. The assumption that protecting
these ‘negative’ rights will have no costs might (arguably) be true if we lived in a
society with full compliance and abundance of resources. But such an assumption
goes beyond what ideal theory is supposed to presuppose – namely, that we be
realistic about what the best of foreseeable conditions are. The best foreseeable
conditions do not include the possibility of living in a society of saints or
transcending moderate scarcity. Such conditions would violate Hume’s circum-
stances of justice.We would not need justice if this kind of utopian society were
possible.
Even the most affluent liberal democracies cannot escape a cost–benefit analysis
of protecting fundamental rights and freedoms.Take, for example, a fundamental
right like the right to vote.The right to vote is a basic liberty and a just society
should ensure that no adult citizen is denied the right to vote. But the difficulty
arises when decisions must be made concerning the allocation of the public funds
needed to run an election and ensure citizens can exercise their right to vote.7
Prohibiting citizens from voting is not the only way citizens can be disenfran-
chised.The distribution of polling stations within a geographical territory and the
hours of operation of a polling station, etc. will also have an impact on the
opportunity citizens have to exercise the right to vote. These provisions have
budgetary implications which can run into millions of dollars. In The Costs of
Rights Stephen Holmes and Cass Sunstein describe some of the costs involved in
American elections:
In Massachusetts, a state law passed prior to the 1996 presidential elections man-
dated longer hours for polling stations. Implementing this tiny amendment to the
law cost Massachusetts taxpayers $800,000. In California, where a study of electoral
expenses was commissioned by the state government, the cost of any statewide
election (whether presidential, senatorial, gubernatorial, etc.) runs around $45–50
million.This is also true for any referendum requiring a separate ballot. Printing and
mailing costs for voter guides alone, including those printed in Spanish as well as
English, can range from $3 million to $12 million. In California, the cost per voter
is estimated to run from $2 to $5, depending on each municipality’s voting system
(Holmes and Sunstein, 1999, p. 114).
All rights, even something as basic as the right to vote, cost money and this means
that giving an absolute priority to the basic liberties will severely constrain the
J U S T I C E I N I D E A L T H E O R Y : A R E F U T A T I O N 853
public funds available to promote other laudable aims. How much should we
spend to ensure that every citizen enjoys what Rawls calls the ‘central range of
application’ (Rawls, 1996, p. 297) for each of the basic liberties and exactly what
constitutes this range of application? The answers to these questions cannot, for
Rawls, be answered by appealing to considerations of the common good. Rawlsbelieves that justice denies that the loss of freedom for some is made right by a
greater good shared by others. Indeed, justice even denies that the loss of freedom
for some is made right by conferring greater benefits on those individuals whose
rights are being restricted (unless those benefits are greater liberty). In Lecture
VIII (s. 2) of Political Liberalism entitled ‘The Special Status of Basic Liberties’,
Rawls makes a distinction between restricting and regulating the basic liberties.
He claims that the first principle of justice permits regulations but not restrictions
of the basic liberties. However, Rawls limits his discussion of regulating the basic
liberties to cases where the liberties conflict with each other, and thus he limits
the scope of possible reasons for regulation to the basic liberties themselves.
Consider the following passage from Political Liberalism:
Since the various basic liberties are bound to conflict with one another, the
institutional rules which define these liberties must be adjusted so that they fit into
a coherent scheme of liberties.The priority of liberty implies in practice that a basic
liberty can be limited or denied solely for the sake of one or more other basic liberties, and
never, as I have said, for reasons of public good or perfectionist values.This restriction holds
even when those who benefit from the greater efficiency, or together share the
greater sum of advantages, are the same persons whose liberties are limited or
denied (Rawls, 1996, p. 295, emphasis added).
The only way one can make sense of Rawls’ absolutist stance on the basic liberties
is that he (falsely) assumes these are costless rights. But a basic right like the right
to vote, as Holmes and Sunstein illustrate, has costs. We could spend endless
amounts of public funds trying to ensure that the right to vote is adequately
regulated (as well as all the other fundamental rights and freedoms).The more we
spend on improving the enjoyment of this one right the less we will have for
other laudable aims that do not involve basic liberties, such as universal health
care, education, etc. But Rawls does not permit us to appeal to benefits to the
public good when making decisions about the regulation of basic liberties. Suchdecisions would violate the equal basic liberties principle which is serially ordered
above the principles of fair equality of opportunity and the difference principle.
So the equal basic liberties principle cannot be subjected to a cost–benefit analysis
that transcends the metric of liberty itself. But the reality is that every society, even
affluent societies like America, must (and do) make such decisions when they
invest scarce public funds into certain aims (e.g. national defence, domestic
security, etc.) rather than other aims. These constraints are ignored by Rawls
because he discusses a society that exists in an idealized scenario that is insulated
from many of these issues.
A number of other simplifying assumptions that Rawls makes in his account of
justice demonstrate that his abstraction collapses into idealization. Rawls (1999b,
p. 83) invokes the assumption that everyone has physical needs and psychological
capacities within the normal range, for example when he attempts to identify
who are the least advantaged.The difference principle requires the basic structure
of society to be organized so that social and economic inequalities are to the
greatest benefit of the least advantaged. But if all citizens are within the normalrange of functioning then we would not need to worry about health care
concerns that arise in a world where there are accidents and disease. Is such a fact
one that we can ignore because, in ideal theory, we are concerned with what is
required in the best possible scenario? No. No society (not even a realistic utopia)
can transcend the fact of human vulnerability. The idea that we can derive the
principles of justice by simply ignoring or postponing such facts is severely
misguided.
Rawls’ reason for making this assumption is not that issues of just health care fall
outside the scope of justice; rather they should be dealt with ‘at the legislativestage when the prevalence and kinds of these misfortunes are known and the costs
of treating them can be ascertained and balanced along with total government
expenditure’ (Rawls, 1996, p. 184). But once Rawls’ legislators have the infor-
mation about the prevalence of these misfortunes and the costs of treating them,
they will need some criteria by which to make a judgement about how much
should be spent on health care. If the fundamental principles of justice that are
supposed to help guide their deliberations are ones that are derived for ideal
societies that do not have people in need of medical treatment, then such a theory
will be of very limited use to them.The principles of justice will be ineffective inhelping them determine how much they should spend on health care.
Rawls’ assumption that society is a closed system also results in further idealiza-
tions. This is evident when one considers one of the central institutional pre-
scriptions he advocates in the paperback edition of Political Liberalism.This is the
prescription that society should take the role as employer of last resort through
general or local government, or other social and economic policies (Rawls, 1996,
p. lix).This prescription might sound like a good idea when we assume that our
society is the Rawlsian closed society (and a society consisting of fully compliant,
healthy citizens). But the case for such a prescription is much harder to make ina society whose economy is subject to the pressures of a competitive global
economy. Rawls assumes that 100 percent employment is something that can be
attained in the best foreseeable circumstances. But this is an enormous empirical
assumption. It is arguably an easy assumption to make if one assumes society is
closed but it is a very contentious empirical claim to make about real, open
societies. Ballooning deficits, the migration of skilled and unskilled (and illegal)
workers, coupled with the pressures for remaining competitive in a global
economy, all pose formidable obstacles to the aim of having 100 percent employ-
ment. Nothing in Rawls’ account of justice addresses these non-ideal considera-tions.And thus the institutional prescriptions he advocates are of little use when
applied to real societies.
J U S T I C E I N I D E A L T H E O R Y : A R E F U T A T I O N 855
Like Rawls’ assumption that no major obstacles face legally enforcing constitu-
tional rights (except having the political will), the idea that the only obstacle
facing 100 percent employment is the political will to make this a reality is naïve.
Contemporary political disagreements are largely disagreements over how best to
achieve such an aim, rather than over whether such an aim is itself desirable.Neither American President George W. Bush nor Democratic candidate John
Kerry would object to the idea that being employed is crucial for achieving
self-respect, but the former believes that the best way of promoting job growth in
America is through tax cuts for the wealthy while the latter believes that
preventing the outsourcing of jobs is the more appropriate tactic. So there is
fundamental disagreement about what is possible, though one might retort that
this disagreement is not one that concerns what is possible in the best conditions.
But how confident should political philosophers who engage in ideal theorizing
be in their judgement that they know what is feasible in the best possible
conditions? I think we should be very modest about such judgements as they will
rely on large, contentious empirical assumptions – assumptions that one does not
have to make explicit (let alone defend) if one can simply say ‘I am talking about
justice in ideal theory’. But such assumptions will receive the attention they
deserve if one explicitly functions at the level of non-ideal theory.
Dworkin and Equality of Resources
In the early 1980s Dworkin (1981a; 1981b) published a series of influentialarticles on equality. These two articles now form part of Dworkin’s book
entitled Sovereign Virtue . Dworkin’s theory is in many respects an attempt to
improve upon Rawls’ theory. His account of ‘equality of resources’ attempts to
take seriously the concern for mitigating natural inequalities (i.e. handicaps) and
holding people responsible for their choices. Dworkin’s defence of equality has
given rise to the position Elizabeth Anderson (1999) calls ‘luck egalitarianism’.8
Luck egalitarians construct their theories around what is called the choice/chance
(or choice/circumstances) distinction. ‘People’s fates are determined by their
choices and their circumstances’ (Dworkin, 2000, p. 322) and this must remain,argue luck egalitarians, a fundamental insight when considering what constitutes
a just distribution. Luck egalitarians disagree on exactly what should be equalised
(e.g. resources, opportunity for welfare, etc.) but they believe that inequalities in
the advantages that people enjoy are just if they derive from the choices people
have voluntarily made, but that inequalities deriving from unchosen features of
people’s circumstances are unjust. Luck egalitarians also function at the level of
ideal theory and thus are guilty of idealization. And this results in impotent or
misguided normative prescriptions.
Dworkin’s egalitarianism is premised on two fundamental principles of ethicalindividualism – the principle of equal importance and the principle of special
responsibility.These principles maintain the following:
Principle of Equal Importance : it is important, from an objective point of view, that
human lives be successful rather than wasted, and this is equally important, from an
objective point of view, for each human life.
Principle of Special Responsibility: though we must all recognize the equal objective
importance of the success of a human life, one person has a special and finalresponsibility for that success – the person whose life it is (Dworkin, 2000, p. 5).
Dworkin’s attempt to merge the two fundamental principles of ethical individu-
alism are most stark in his hypothetical tale of shipwrecked survivors who are
washed up on a desert island that has abundant resources. I will assume that most
of my readers are familiar with the basic details of Dworkin’s tale. My criticism
focuses on the normative prescriptions that Dworkin believes can be derived
from his hypothetical insurance scheme .
After telling the story about the hypothetical auction, the challenge for Dworkin
is translating how this applies to real, non-ideal societies. In the real world, he
argues, there is a need for taxation and redistribution. Income tax is a device
society can use to neutralise the effects of handicaps and differential talents.A tax
system, admits Dworkin, can only roughly approximate the results of the insur-
ance scheme and will not achieve a truly ambition-sensitive/endowment-
insensitive distribution. But the problems facing Dworkin’s theory are more
formidable than the concern that it is only a second-best solution. Dworkin
assumes that the appropriate level of public funding for health care and welfare
can be determined by asking how much people would insure against these
misfortunes in the hypothetical auction (where people do not know what their risk is of faring poorly in natural endowments). But this makes a number of
idealizations that obscure our liberal and prioritarian commitments. Firstly, we
must ask why Dworkin’s insurance scheme is limited to only these two forms of
misfortune? Citizens of even affluent liberal democracies are vulnerable to
numerous forms of misfortune beyond handicaps and differential talents.We face
risks of misfortune from crime, food preparation and sanitation, terrorist attack,
natural disasters, injury at work, pollution, etc. So if Dworkin is to utilize the
insurance scheme as the model for determining the appropriate level of state
provisions then we must add protection against all forms of brute luck disadvan-tage to the list of things we could insure against. Like Rawls, Dworkin assumes
that protecting the equal basic liberties of all will be costless; hence he is only
concerned with the costs of health care and welfare. But this assumption is
mistaken.Rights cost money. Legally enforcing any right (e.g.protecting property,
security of the person, education, etc.) will mean that numerous strains will be
placed on the funds available for promoting the aims of health care and welfare.
So the question ‘How many of your 100 clam shells would you spend insuring
against disease, illness, (brute luck) accidents and unemployment? must be
extended to something like the following:
Non-ideal Insurance Scheme (or Real Politics!): How much would you be willing to
spend on insurance to mitigate the misfortunes of disease, illness, (brute luck)
J U S T I C E I N I D E A L T H E O R Y : A R E F U T A T I O N 857
famine, etc.), injury at work and every other conceivable misfortune that citizens
could face?
Dworkin is not justified in limiting his insurance scheme to the two brute luck
considerations he limits himself to (i.e. handicaps and differential talents). Theproblems with Dworkin’s limited strategy are evident if we consider how Justine
Burley utilises Dworkin’s argument to justify the state provision of fertility
treatment. In ‘The Price of Eggs: Who Should Bear the Cost of Fertility
Treatments?, Burley argues, via recourse to Dworkin’s theory of equality, ‘that
justice demands that [infertile] individuals be compensated for all or part of the
costs of the assisted conception techniques that they undergo’ (Burley, 2000,
p. 129). Burley believes the Dworkinian approach to justice is attractive because
it is an alternative to needs-based approaches to health care which get drawn into
a debate about what our ‘fundamental needs’ are. The needs-based approach istypically viewed as being opposed to state funding of assisted conception tech-
niques because infertile individuals are not at risk of physical harm if they do not
undergo such treatments.The Dworkinian model only requires us to determine
if infertility counts as a deficit in personal resource holding (i.e. a handicap).
Burley (2000, p. 141) argues that two conditions must be satisfied in order for
fertility to count as a handicap: (1) the infertile person must envy the capacity of
others to bear genetically related offspring and (2) the deficit in personal resources
at issue must not be traceable to their tastes or choices. After stating the case for
why these two conditions are satisfied in the case of infertility, Burley remarks:
Recall that only those handicaps individuals in the aggregate would have insured
against will be compensated. In the hypothetical insurance scheme people would
have knowledge of the actual incidence of this handicap.They would know that 20
percent of all couples experience infertility, and that up to 10 percent of this
number will not conceive other than through artificial means ... I contend that in
this situation the average individual would deem having genetically related off-
spring a constitutive element of leading a good life ... It is therefore plausible to
insist that individuals in the aggregate would stipulate infertility as one handicap
that they were particularly concerned to receive compensation for (Burley, 2000,
p. 142).
The problem with Burley’s argument is that she does not consider how enormous
the list of possible misfortunes that we could insure against is. Focusing on one
handicap in isolation will of course garner support for the position that this
handicap is one we should insure against. But when placed against the back-
ground of all possible handicaps (and other possible misfortunes) where does the
interest of having genetically related children figure? That is the question that
Burley needs to address and it is not one that political philosophers can answer by
simply engaging in armchair theorizing.The danger of invoking the Dworkinian
model is that we will tend to treat particular misfortunes in isolation, as Burleydoes with infertility, and then make the leap to the real-world prescription that
compensation for this particular disadvantage is required. Given scarcity of
resources, this conclusion could jeopardize state funding for other, more pressing ,
treatments. By attempting to forgo a debate about what our ‘fundamental needs’
are, Burley’s position does not give adequate weight to the other forms of possible
misfortune that we are vulnerable to. Engaging in the ideal theorizing of the
Dworkinian model obstructs the importance of the empirical information (e.g.how pervasive and severe different forms of disadvantage are, the costs of miti-
gating these disadvantages, the likelihood that medical intervention will prove
successful, etc.) one would need to have in order to make informed decisions
about the allocation of scarce public funds.
Conclusion: Where the Action (Really) Is – On the Siteof Non-Ideal Theory
I have argued that armchair theorizing about justice in ideal theory severely limitsthe practical insights of liberal egalitarianism as such theorizing often brackets the
complex and contentious issues that make the struggle for justice in real societies
difficult. In this concluding section I seek to make a positive contribution to
helping political philosophers shift from ideal to non-ideal theory. My argument
here is primarily methodological and I do not attempt, in this article, to take on the
more ambitious project of actually defending a conception of liberal justice in
non-ideal theory.9 But drawing from the lessons to be learned from the short-
comings of the theories of Rawls and Dworkin, liberal egalitarians should take
seriously, at a minimum, the issue of trade-offs and the value of democracy.
Since the publication of Rawls’ A Theory of Justice in 1971, political philosophers
have invested a vast amount of their energies into defending what Amy Gutmann
and Dennis Thompson (2004) call ‘first-order’ social theories. First-order theories
are theories ‘that seek to resolve moral disagreement by demonstrating that
alternative theories and principles should be rejected’ (Gutmann and Thompson,
2004, p. 13). So Rawls attempts to show that his two serially ordered principles
of justice are preferable to the principle of utility. Robert Nozick (1974) attempts
to show that principles of entitlement are superior to any ‘patterned’ principles of
justice. And a range of contemporary egalitarian political philosophers posit rivalaccounts of what should be equalized (e.g. equal opportunity for welfare, equality
of resources, equality of capabilities, etc.).
I believe a shift from ideal to non-ideal theory will also bring about a shift from
first-order to ‘second-order’ theorizing. Second-order theories are theories that
are primarily concerned with dealing with the claims of conflicting first-order
theories (Gutmann and Thompson, 2004, p. 13). The most prominent second-
order theory that has recently begun to rise to the fore of debates in democratic
theory is deliberative democracy.There are of course a variety of different positions
advanced by deliberative democrats. One version that I believe is particularly wellsuited to tackling non-ideal theory is the version advanced by Gutmann and
Thompson (1996; 2004).10 Their version of deliberative democracy does not
J U S T I C E I N I D E A L T H E O R Y : A R E F U T A T I O N 859
champion the priority of one particular substantive (e.g. liberty, equality, etc.) or
procedural value (e.g. democracy); rather it acknowledges a plurality of such
values. Instead of trying to win a philosophical argument concerning the viability
of first-order principles, deliberative democrats are more concerned with deter-
mining what would constitute a reasonable balance between conflicting fundamen-tal values.
Deliberative democracy is well suited for tracking fact-sensitivity because it is a
theory that takes what Gutmann and Thompson call ‘provisionality’ seriously. ‘A
theory is morally provisional if its principles invite revision in response to new
moral insights or empirical discoveries’ (Gutmann and Thompson, 2004, p. 57).
Rather than starting off with the assumption that the political philosopher has
privileged access to what constitutes the best possible conditions, deliberative
democrats recognise that these are contested assumptions that are subject to
debate and will evolve over time as circumstances (both domestic and global)change.
By emphasising the importance of provisionality and deliberation, political phi-
losophers can take a purposeful (rather than cost-blind) approach to rights in a way
that guards against the criticisms typically levied against cost–benefit analysis
(CBA). Henry Richardson (2001), for example, argues that CBA is a ‘stupid’
model for public choice because its underlying normative standard of choice
makes no room for intelligent choice concerning how best to use our resources.11
By linking the purposeful approach to rights protection with democracy, non-
ideal theorists can defend a cost–benefit analysis that is normatively justified,
flexible and intelligent.12 ‘Cost–benefit analysis is a rich area of interdisciplinary
work’ (Adler and Posner, 2001, p. 1) and by shifting to non-ideal theory political
philosophers will be better positioned to make a substantive contribution to this
field and will also be better placed to relate their theory to real politics and the
challenges that real societies face.
In order to enhance the transformative potential of liberal egalitarianism political
philosophers must engage in a self-conscious dialogue, not only with other phi-
losophers (e.g. libertarians, conservatives, etc.), but also with executives, legisla-
tures, courts and other public administrators. Doing this ensures that non-ideal
considerations, like budget constraints, disagreement, limited knowledge and
indeterminacy, will figure prominently in our philosophical analysis.This does not
mean political philosophers must take all existing constraints as a given. But they
must provide compelling reasons why some constraints are valid while others are
not.This way they will avoid engaging in idealization.
Many political philosophers will no doubt resist the push towards non-ideal
theory because such a transformation will alter the way they conceive of the
discipline itself. It will mean that there is less room for armchair theorizing andthat the primary focus will not be on winning a philosophical debate among
first-order theories of justice. Political philosophers can learn how to take non-
ideal theory seriously by considering how other normative disciplines have
attempted to grapple with similar issues. In particular, it is useful to consider
recent trends in the field of bioethics, such as the utilization of quality-adjusted
life year (QALY) in health economic evaluation studies.13 Concerns of scarcity
and pervasive disadvantage are arguably most stark when addressing the issue of distributing limited medical provisions. Many contemporary debates about justice
function almost exclusively on the distribution of what Rawls calls ‘social primary
goods’ (e.g. rights and wealth); while accounts of just health care have tended to
focus almost exclusively on access to health care.The former ignore the issue of
just health care and the latter tend to ignore the more general issue of societal
fairness. This disconnect between the disciplines of political philosophy and
bioethics is no doubt a major reason why ideal theorizing has been so rampant in
each discipline.
In ‘Broadening the Bioethics Agenda’ Daniel Brock addresses the shortcomings of accounts of just health care. He highlights two shortcomings of the bioethical
discussions of justice or equity: firstly, that ‘the focus on establishing a right to
health care has left bioethicists largely silent on the important and complex moral
issues faced when limited health care resources must be prioritized’ (Brock, 2000,
p. 22); and secondly, that the focus on inequalities in access to health care has
ignored the great impact social determinants play on our life prospects. I believe
parallel shortcomings exist in contemporary debates about distributive justice
more generally. As I argued above, many ‘rights-based’ theories of justice (like
Rawls’ theory) adopt a cost-blind approach to rights protection and a limited viewof human disadvantage. And thus they are ill equipped to address the issue of
trade-offs in primary goods which inevitably arises given the fact of scarcity. So
like Brock’s appeal to broaden the bioethics agenda, this article is an analogous
appeal to broaden the distributive justice agenda.
Søren Holm, in an editorial in a recent issue of the Journal of Medical Ethics, notes
the differences between the first-order approaches to ethics and deliberative
democracy. His insights are useful for they can be applied to political philosophy
and debates about distributive justice. Holm claims:
The standard [bioethics] model, and to an even greater extent the rhetoric of
ethics, often imply that there is only one right solution to each policy making
decision, and that it is possible to find this solution, and to know that it is the right
decision. ... In contrast, theorising about deliberative democracy proceeds from the
assumption that whereas there might be a right solution, it is often impossible to
find this solution or to know that it is the right solution in any absolute sense.The
best we can do is to outline the area of acceptable policies, and then choose a policy
within this area through a deliberative, democratic process (Holm, 2005, p. 1).
Rather than striving to win a philosophical argument among rival first-order theories, contemporary justice theorists who take non-ideal theory seriously will
seek to enhance public deliberation about legitimate public policy in a morally
J U S T I C E I N I D E A L T H E O R Y : A R E F U T A T I O N 861
pluralistic liberal democracy.This will make the discipline of political philosophy
more interdisciplinary and the conclusions theories of justice reach will be
somewhat indeterminate. But this fact should not surprise us given that the world
we inhabit is itself indeterminate.14 In order to function as a guide for our
collective action, an account of justice must give due consideration to the diversecomplexities that arise in the real world. This will better prepare us for taking
seriously the distinct demands of social justice.
(Accepted : 19 April 2006)
About the Author
Colin Farrelly, Department of Political Science, University of Waterloo, 200 University Ave. W.,
Earlier versions of this article were presented to the Canadian Political Science Association and Wilfrid Laurier University. I am grateful for the useful feedback I received on those occasions, especially to Michael MacMillan, LorenKing and Jonathan Quong. I am also grateful to three anonymous referees and Matthew Festenstein for their insightfulsuggestions and criticisms.
1 The concern for taking non-ideal theory seriously is gaining momentum in recent philosophical literature. See, for example, John Gray (1995), George Sher (1997), Jonathan Wolff (1998), Elizabeth Anderson (1999), ElizabethAshford (2000; 2003), David Miller (2001), Liam Murphy (2000), Samuel Scheffler (2001), Stuart White (2003) andBarbra Fried (2004).
2 The phrase ‘realistically utopian’ is one Rawls (1999a) employs. I explain what he means by this phrase in the nextsection.
3 The particular topic of justice (e.g. domestic or global) one is addressing will influence what the requirements of fact-sensitivity are. So, for example, a theory of domestic justice will grapple with a range of concerns that aredistinct from those that arise in the context of an analysis of global justice. For a recent discussion of the latter seeLeif Wenar (2003). The argument advanced in this article is primarily concerned with the issue of domesticjustice, in particular, how the theories of justice advanced by Rawls and Dworkin are to be applied to the issue of domestic justice.
4 I have chosen the moderate idealized position as my target, rather than the extreme idealized position (e.g. Cohen),as I believe that the case against the moderate position is much harder to make. Furthermore, if one can show thateven moderate ideal theorizing is inherently flawed then one has also shown that the extreme idealized position isflawed.
5 This is qualified, however, by Rawls’ assumption that all persons fall within the normal range of functioning. So weshould assume that the least endowed possess a decent minimum level of natural primary goods (e.g. health andvigour).
6 These figures are taken from the Budget of the United States: Enacted in the Fiscal Year 2006 . Available at: http://www.whitehouse.gov/omb/budget/fy2007/tables.html.
7 Having the right to vote is pointless if you cannot reasonably exercise the right.
8 Those who endorse this position include, among others, Richard Arneson (1989), Ronald Dworkin (2000), G. A.Cohen (1989) and Philippe Van Parijs (1995).
9 I defend a liberal conception of justice in non-ideal theory in my (forthcoming) book entitled Justice, Democracy and Reasonable Agreement .
10 For an examination of how practical other recent accounts of deliberative democracy are see Farrelly (2005).
11 Elizabeth Anderson (1993) makes a similar critique of CBA in Value in Ethics and Economics. Anderson argues thatCBA fails to consider the proper roles the protection of human life and environmental quality occupy in public life.This is so because, according to Anderson, CBA regards such goods as ‘mere commodities’ (p. 190). By prescribing
that we combine a purposeful approach to rights protection with concerns of democratic accountability, I believeone can guard against such concerns. In order to examine these issues political philosophers must abandon the‘rights-based’ approach to justice (and its prescription of judicial supremacy) that currently dominates a good dealof contemporary political philosophy.
12 This can be accomplished by endorsing, for example, a ‘dialogical model’ of judicial review. See my paper (2006)‘Civic Liberalism and the “Dialogical Model” of Judicial Review’ in Law and Philosophy vol. 25(5), 489–532 for adiscussion of this model of judicial review.
13 See Schwappach (2002) for a review of the QALY debate in the context of health economics.
14 Russell Hardin (2003, p. 103) captures this insight when he claims: ‘In a world that is inherently indeterminate, asuitable theory of distributive justice must perhaps itself be indeterminate, and its indeterminacies must accom-modate those of the world where relevant’.
References
Adler, M. and Posner, E. (eds) (2001) Cost–Benefit Analysis: Legal, Economic, and Philosophical Perspectives. ChicagoIL: University of Chicago Press.
Anderson, E. (1993) Value in Ethics and Economics. Cambridge MA: Harvard University Press.
Anderson, E. (1999) ‘What Is the Point of Equality?’, Ethics, 109 (2), 287–337.
Arneson, R. (1989) ‘Equality and Equal Opportunity for Welfare’, Philosophical Studies, 56 (1), 77–93.
Ashford, E. (2000) ‘Utilitarianism, Integrity and Partiality’, The Journal of Philosophy, XCVII (8), 421–39.
Ashford, E. (2003) ‘The Demandingness of Scanlon’s Contractualism’, Ethics, 113 (2), 273–302.
Beitz, C. (1979) Political Theory and International Relations. Princeton NJ: Princeton University Press.
Brock, D. (2000) ‘Broadening the Bioethics Agenda’, Kennedy Institute of Ethics Journal , 10 (1), 21–38.
Burley,J. (2000)‘The Pr ice of Eggs:Who Should Bear the Cost of FertilityTreatments?,in J. Harris and S. Holm(eds), The Future of Human Reproduction, paperback edition. Oxford: Oxford University Press, pp. 127–49.
Carens, J. (2000) Culture, Citizenship, and Community. Oxford: Oxford University Press.
Cohen, G.A. (1989) ‘On the Currency of Egalitarian Justice’, Ethics, 99 (4), 906–44.
Cohen, G. A. (2003) ‘Facts and Principles’, Philosophy and Public Affairs, 31 (3), 211–45.
Daniels, N. (1985) Just Health Care . Cambridge: Cambridge University Press.
Dunn, J. (1990) ‘Reconceiving the Content and Character of Modern Political Community’, in J. Dunn (ed.),Interpreting Political Responsibility. Oxford: Polity, pp. 193–215.
Dworkin, R. (1978) Taking Rights Seriously. London: Duckworth.
Dworkin, R. (1981a)‘What is Equality? Part 1:Equality ofWelfare’, Philosophy and PublicAffairs, 10 (3), 185–246.
Dworkin, R. (1981b) ‘What is Equality? Part 2: Equality of Resources’, Philosophy and Public Affairs, 10 (4),283–345.
Dworkin, R. (2000) Sovereign Virtue . Cambridge MA: Harvard University Press.
Farrelly, C. (2005) ‘Making Deliberative Democracy a More Practical Political Ideal’,European Journal of Political Theory, 4 (2), 200–8.
Fried, B. (2004) ‘Left-Libertarianism: A Review Essay’, Philosophy and Public Affairs, 32 (1), 66–92.
Gray, J. (1995) Enlightenment’s Wake: Politics and Culture at the Case of the Modern Age . London: Routledge.
Gutmann, A. and Thompson, D. (1996) Democracy and Disagreement . Cambridge MA: Harvard University Press.
Gutmann, A. and Thompson, D. (2004) Why Deliberative Democracy? Princeton NJ: Princeton University Press.
Hardin, R. (2003) Indeterminacy and Society. Princeton NJ: Princeton University Press.
Harsanyi, J. (1975) ‘Can the Maximin Principle Serve as a Basis for Morality? A Critique of John Rawls’sTheory’, American Political Science Review , 69 (2), 594–606.
Holm, S. (2005) ‘Bioethics Down Under – Medical Ethics Engages with Political Philosophy’, Journal of Medical Ethics, 31 (1), 1.
Holmes S. and Sunstein, C. (1999) The Costs of Rights:Why Liberty Depends on Taxes. New York: W.W. Norton& Company.
Kymlicka, W. (1989) Liberalism, Community and Culture . Oxford: Oxford University Press.Mason, A. (2004) ‘Just Constraints’, British Journal of Political Science , 34 (2), 251–68.
Miller, D. (2001) Principles of Social Justice . Cambridge MA: Harvard University Press.
J U S T I C E I N I D E A L T H E O R Y : A R E F U T A T I O N 863
Murphy, L. (2000) Moral Demands in Nonideal Theory. Oxford: Oxford University Press.
Nozick, R. (1974) Anarchy, State and Utopia. New York: Basic Books.
Okin, S. (1989) Justice, Gender and the Family. New York: Basic Books.
O’Neill, O. (1996) Towards Justice and Virtue:A Constructive Account of Practical Reasoning . Cambridge: CambridgeUniversity Press.
Rawls, J. (1971) A Theory of Justice . Cambridge MA: Harvard University Press.
Rawls, J. (1996) Political Liberalism, paperback edition. New York: Columbia University Press.
Rawls, J. (1999a) The Law of Peoples. Cambridge MA: Harvard University Press.
Rawls, J. (1999b) A Theory of Justice , second edition. Oxford: Oxford University Press.
Rawls, J. (2001) Justice as Fairness: A Restatement . Cambridge MA: Harvard University Press.
Richardson, H. (2001) ‘The Stupidity of the Cost–Benefit Standard’, in M. Adler and E. Posner (eds),Cost–Benefit Analysis: Legal, Economic, and Philosophical Perspectives. Chicago IL: University of Chicago Press,pp. 135–68.
Scheffler, S. (2001) Boundaries and Allegiances. Oxford: Oxford University Press.
Schwappach, D. (2002) ‘Resource Allocation, Social Values and the QALY: A Review of the Debate andEmpirical Evidence’, Health Expectations, 5 (3), 210–22.
Sher, G. (1997) Approximate Justice: Studies in Non-ideal Theory. Lanham MD: Rowman & Littlefield Publishers,Inc.
Van Parijs, P. (1995) Real Freedom for All . Oxford: Oxford University Press.
Wenar, L. (2003) ‘What We Owe to Distant Others’, Politics Philosophy & Economics, 2 (3), 283–304.
White, S. (2003) The Civic Minimum. Oxford: Oxford University Press.
Wolff, J. (1998) ‘Fairness, Respect, and the Egalitarian Ethos’, Philosophy and Public Affairs, 27 (2), 97–122.
