Social Choice and the Grammar of Rights and Freedoms
| Published date | 01 March 2004 |
| Author | Keith Dowding |
| Date | 01 March 2004 |
| DOI | 10.1111/j.1467-9248.2004.00469.x |
| Subject Matter | Article |
Social Choice and the Grammar of
Rights and Freedoms
Keith Dowding
London School of Economics
The techniques of social choice and game theory are increasingly being used to analyse concepts
in political theory. Although these techniques may prove invaluable for teasing out contradictory
formulations, puzzles and problems with traditional concepts, formal writers often begin their
analysis with simplistic intuitive accounts rather than building on earlier traditions in analytic
political theory. This is no more obvious than with the social-choice and game-theory analysis of
rights and freedoms. This paper reviews these approaches and demonstrates that by ignoring the
grammar of rights and freedoms, social-choice and game-theory analysis goes wrong from the very
beginning. Formal writers need to take more account of the history of their subject, as developed
in the analytic theory tradition.
The tools of social choice and game theory are being marshalled in order to get a
new handle on old concepts in political philosophy. There is a large literature on
the nature of equality as a value and what we should expect to be equalized (see,
for example, Roemer, 1996, 1998). There is a growing literature on the nature of
liberty, on its potential measurability1and on the concept and form of rights.2Much
of this literature on rights was inspired by Sen’s ‘paradox’ of the impossibility of
the Paretian liberal; Sen has been a key writer in the literature on the concept and
measurement of freedom, and his ‘Equality of What?’ article (1982a) inspired
much of the work on the nature of equality and its maximand. The justification
for using the tools of social choice and game theory is that it leads us to re-examine
our intuitions and understand the true or valid implications to be drawn from
approaching these concepts in one way rather than another (Carter, 1999,
pp. 17–18; van Hees and Wissenburg, 1999, pp. 80–1). Where intuitions are found
to be contradictory or wanting in other regards, they need to be altered or given
up; though equally, if the results of analytic research prove to be too far adrift from
our intuitions, we may wish to consider the appropriateness of the particular ana-
lytic framework adopted. This is not quite the same as the ‘reflective equilibrium’
of Rawls (1972), which is concerned with the way in which we develop grand the-
ories of ethics and politics from our intuitions, reflecting back on the intuitions
from the implications drawn from the theory. The interrogation of theory by intu-
itions, and intuitions by theory, leads to reflective equilibrium (1972, pp. 48–51).3
Rather, we are discussing here the language in which we can represent both our
intuitions and our grand theories. The analytic language adopted must be rich
enough and utilized well enough to translate the intuitions and grand theorizing
of ordinary language and traditional political philosophy if it is to fulfil the role of
challenging the implications that may be drawn from them.
POLITICAL STUDIES: 2004 VOL 52, 144–161
© Political Studies Association, 2004.
Published by Blackwell Publishing Ltd, 9600 Garsington Road, Oxford OX4 2DQ, UK and 350 Main Street, Malden, MA 02148, USA
SOCIAL CHOICE AND THE GRAMMAR OF RIGHTS AND FREEDOMS 145
Does social choice and game theory handle rights adequately? I will argue in this
paper that the grammar of rights and freedoms developed by traditional legal and
political philosophy has all too often been ignored by writers in the social-choice
tradition, much to the latter’s detriment. I do not suggest that formal approaches
cannot appropriate the correct grammar; my aim in that regard is much more
modest – merely to point out their failure to do so.4
Minimal Liberalism (ML)
Recent work on rights in social-choice literature has grown from a critique of Sen’s
characterization of rights in terms of his Condition L* or minimal liberalism (ML)
introduced in Collective Choice and Social Welfare (Sen, 1970; see also Sen, 1982b).
The ‘impossibility’ result has produced an enormous literature discussing whether
or not it is possible to produce Pareto-efficient outcomes in a liberal society.5We
need not be distracted by that debate here. We are only concerned with the char-
acterization of rights contained in condition ML to see if they are adequate and to
consider the nature of the inadequacy if they are not.
Condition ML states that there are at least two individuals such that for each of
them there is at least one pair of alternatives over which she is decisive. An indi-
vidual is decisive if there is a pair (x, y) such that it is always the case that if she
prefers x(respectively to y) to y(respectively to x), then society should prefer x
(respectively to y) to y(respectively to x). The intuitive problem with Sen’s char-
acterization is that the domain of the rights-holding – or that which an individual
is said to be decisive over – is broader than that over which we normally think of
people as having rights. This can be illustrated with his own example of the rights
of two individuals (iand j) to read Lady Chatterley’s Lover.
In Sen’s example, person i(the lewd) would like to read the book; person j(the
prude) does not want to read the book. However, the lewd would prefer the situ-
ation where the prude reads the book, over the situation where he reads it himself.
Similarly, the prude would prefer to read the book himself than have lewd reading
it. Sen’s impossibility theorem was designed to show that each player choosing in
terms of his rights to read or not read the book would mean that the outcome both
preferred (the Paretian or welfarist outcome) would not be attained.6This is the
paradox of the Paretian liberal. This can be represented as a standard normal form
game which looks like a prisoners’ dilemma (PD):7
Prude (j)
R~R
~Ryw
Lewd (i)Rxz
where R means read the book, and ~R means not read the book. Thus, outcome
y(which from now on I will usually call (~R, R)) is what the Pareto principle sug-
gests should be chosen and is the second choice of both iand j. Since R is a dom-
inant strategy for iand ~R is the dominant strategy for j, game theory says z
(hereafter (R, ~R)) will be the outcome. The ‘paradox’ of the single-play PD has
To continue reading
Request your trial