. 7
( 13)


There is an irony here, however. Pettit™s recent writings, and especially
his acknowledgment of the desirability of ˜a liberal or inclusive form of
republican theory™, have brought him closer than Skinner to the posi-
tion I favour. He may conceive of liberty more inclusively than Pettit, but
Skinner does not regard this inclusive conception as evidence that re-
publicanism and liberalism share a common foundation in autonomy. In
this respect, he resembles Maurizio Viroli, another neo-republican who
has his doubts about the relationship of republicanism to autonomy.

IIIb Republicanism vs. Autonomy?
Skinner™s remarks on republicanism and autonomy are con¬ned, so far
as I know, to a footnote in Liberty Before Liberalism. There he states that
one ˜might say that the neo-roman and classical liberal accounts of free-
dom embody rival understandings of autonomy. For the latter, the will
is autonomous provided it is not coerced; for the former, the will can
only be described as autonomous if it is independent of the danger of be-
ing coerced™.49 On this account, republicanism and liberalism both have
foundations in autonomy, but not a common or shared foundation. What
Richard Dagger

Skinner™s footnote does, in effect, is push the purported rivalry between
liberal and republican conceptions of liberty up, down, or back a level to
a rivalry between liberal and republican conceptions of autonomy.
Is this move justi¬ed? In the absence of a richer account of autonomy
than Skinner provides, it is hard to see how it is. Of course, if we already
believe that republicanism and liberalism are sharply distinct and incom-
patible, then we would expect either that one of the two theories lacks a
conception of autonomy altogether or that their conceptions are quite
different from each other. But that is to assume precisely what is in ques-
tion here; and Skinner offers no evidence to show that ˜neo-romans™ and
liberals really do differ as he says with regard to autonomy. Moreover, the
˜rival understandings of autonomy™ Skinner identi¬es both rest, like his
neo-roman and liberal conceptions of liberty, on a common element “ in
this case, coercion. The ˜liberal™ view is that autonomy is the absence of
coercion; the ˜neo-roman™ is that autonomy is the absence ˜of the danger
of being coerced™. Assuming that Skinner means to include the absence
of coercion itself in the neo-roman/republican view, and not merely the
danger of it, the result is an inclusive conception of autonomy. In this
respect, neo-roman autonomy as the absence of coercion and of the dan-
ger of being coerced is like his neo-roman conception of liberty as the
absence of interference and of dependence. But that is to say that in both
cases, the neo-roman/republican position absorbs and extends the sup-
posedly liberal position, not that it rejects it. If this is rivalry, then it is
rivalry of a friendly and intramural nature.
As with Pettit, in sum, so with Skinner. Both have made valuable con-
tributions to our understandings of republicanism and of freedom, but
neither has shown that the republican conception of freedom is so dif-
ferent from or hostile to (what they take to be) the liberal conception
as to demonstrate that liberalism and republicanism are fundamentally
incompatible. But what of Viroli, who distinguishes republican liberty
not only from the liberal but also from the democratic ideal of liberty as
Viroli™s understanding of republican liberty is in line with Pettit™s:
˜The central point for classical republican theorists is that dependence
is a more painful violation of liberty than interference™.50 Viroli extends
Pettit™s analysis, however, when he associates democratic liberty with au-
tonomy. As he puts it:

The democratic ideal of political liberty, understood as a condition in which
citizens have autonomy and are governed by laws that re¬‚ect their will, is in
Domination and the Republican Challenge to Liberalism 195

fact a radical version of the republican ideal of political liberty as absence of
domination. If to be free means that one is not subject to the arbitrary will of a
man or group, as republican theorists claim, we enjoy complete political liberty
when we are dependent only on our own will “ that is, when we live in a self-
governing polity that permits us to approve or reject the rules governing the life
of the collectivity.51

As stated here, the democratic ideal of political liberty may not seem
to be a truly ˜radical™ departure from the republican ideal. As a citizen
of ˜a self-governing polity that permits us™, my fellow citizens and me,
˜to approve or reject the rules governing the life of the collectivity™, I
apparently enjoy ˜complete political liberty™ in both the republican and
democratic senses. But Viroli has something much stronger in mind when
he refers to approving or rejecting the rules governing the polity. To enjoy
democratic autonomy in his sense of the term, I must be able not only to
have a say or cast a vote, but to approve or reject each and every law of
the polity “ and so must every other citizen. The ˜radical™ nature of this
democratic ideal emerges in the following passage:

The republican conception of political liberty approaches the democratic idea
of liberty as autonomy of the will in that it, too, sees constraint as a violation of
liberty; yet it is not identical, because it holds that the will is autonomous not when
the laws or regulations that govern my actions correspond to my will, but when I am
protected from the constant danger of being subjected to constraint.52

By implication, then, I am not autonomous according to the democratic
ideal unless I have the power to veto any law or regulation that I disap-
prove. No wonder that this account of democratic liberty as autonomy of
the will appears in a chapter entitled ˜The New Utopia of Liberty™!
What are we to make of this conception of democratic autonomy?
Viroli presumably wants us to reject it in favour of the more sensible
republican ideal of liberty, but others may try to turn the tables on him
by using it to reject republicanism. Robert Paul Wolff, for one, relies
on much the same notion of autonomy “ ˜the refusal to be ruled™ “ yet
Wolff argues for ˜philosophical anarchism™ because unanimous direct
democracy is impossible to achieve, and anything less is incompatible
with autonomy.53 In any case, there is no reason to accept this radical
view as the democratic conception of autonomy. None of the chapters in
the present volume, for example, entails or even implies that a person
is autonomous only when she is able to approve or reject every rule or
law that applies to her; indeed, Rainer Forst™s and Bert van den Brink™s
separate discussions of ˜political autonomy™ resemble Viroli™s republican
Richard Dagger

ideal much more closely than his ˜democratic idea of liberty as autonomy
of the will . . .™ (see Chapters 10 and 11 in the present volume).
Nor does Viroli himself hold, in the end, that republicanism is thor-
oughly hostile to autonomy. As he says, the ˜republican conception of
political liberty approaches the democratic idea of liberty as autonomy of
the will . . .™ (emphasis added). It does this because freedom from domi-
nation or from dependence upon the arbitrary rule of others enables a
person to be self-governing in a meaningful sense of that term even when
that person must sometimes accept a rule or law that he or she did not
approve. To see how such a person can be autonomous, however, and
how autonomy underpins a republican-liberal political theory, requires,
¬nally, a closer look at the concept of autonomy itself.

IV Autonomy
As the chapters in this and other volumes testify, autonomy is a rich and
multi-faceted concept.54 In the space remaining, I cannot even pretend
to approach a comprehensive treatment of the subject, but I can offer
remarks on four points that are especially pertinent to the republican
challenge to liberalism.
The ¬rst point begins with the basic observation that autonomy is a
matter of self-government. This observation may seem to be singularly un-
helpful, as it leads to dif¬cult questions about the nature of the self and
how it may be said to govern “ questions such as the nature of the relation-
ship between personal and moral autonomy that Gerald Gaus and Jeremy
Waldron explore in Chapters 12 and 13, respectively, in the present vol-
ume. Nevertheless, this basic observation provides a useful starting point,
as it indicates that autonomy is something available only to people who
have both a reasonably secure sense of self and the ability to govern their
conduct. Someone who suffers from multiple-personality disorder can-
not be autonomous; nor, as the ¬lm Memento illustrates, can someone
who cannot remember whom he has just met, where he has just gone,
what he has just said, or how any of these ¬t into his plans or purposes.
Less dramatically, people who are unable to resist any impulse that strikes
them also lack autonomy, for they are incapable of self-government.
It is equally important to notice that external forces can prevent some-
one who is quite capable of self-government from exercising this capacity.
This may happen, for instance, when a person who could be autonomous
is subject to constant interference or coercion; it may also happen when
such a person is dominated by or utterly dependent upon others. This
Domination and the Republican Challenge to Liberalism 197

is why autonomy is a concern of liberals and republicans alike “ and of
those who believe that republican liberalism is an especially powerful po-
litical theory. Autonomy is the capacity to lead a self-governed life, but
this capacity, like others, will atrophy if it is not exercised. Liberals, re-
publicans, and republican liberals will all have an interest in protecting
people, or enabling them to protect themselves, against interference or
domination that threatens their ability to govern themselves. There will
be disagreements and differences of emphasis among them, to be sure,
but their fundamental concern for self-government demonstrates that re-
publicans and liberals share a common foundation in their commitment
to autonomy.
This claim leads to my second point about autonomy: it is not the
peculiarly liberal concept that critics of liberalism sometimes take it to
be. These critics are doubly mistaken, in my view, as they misconceive
both liberalism and autonomy. Mark Tushnet provides a colorful case in

Liberalism™s psychology posits a world of autonomous individuals, each guided
by his or her own idiosyncratic values and goals, none of which can be adjudged
more or less legitimate than those held by others. In such a world, people exist
as isolated islands of individuality who choose to enter into relations that can
metaphorically be characterized as foreign affairs.55

Setting aside the caricature of liberalism here, the pertinent question is
whether autonomous individuals really are ˜isolated islands of individu-
ality™. The answer, quite clearly, is no. Autonomous individuals must be
able to make choices, certainly, including the choice to enter into and
break off various relations with others. But that is hardly to say that one
is autonomous only if he or she takes part in nothing but self-chosen
relationships.56 We are born, most of us, with the capacity to lead self-
governed lives, but we cannot develop or exercise this capacity without
the assistance of other people, and it would be silly to think of our rela-
tions with all of them, even metaphorically, as ˜foreign affairs™. Even as
mature and presumably independent adults, we ¬nd ourselves entangled
in relationships “ with relatives, neighbors, co-workers, compatriots, and
others “ that we have not fully chosen. Yet we may still be re¬‚ective per-
sons capable of judging the options available to us and making choices
in light of those judgments. In short, we may achieve autonomy despite
our inability to become ˜isolated islands of individuality™.
As these remarks suggest, autonomy is not a simple on/off concept “
something that one either does or does not have. On the contrary, one
Richard Dagger

autonomous person may have more or less autonomy than another; or
someone may be autonomous in one aspect of her life but not in an-
other. In most discussions of personal autonomy, however, we are talking
in global terms of whether this or that person or group of people should
be deemed, all things considered, to be autonomous. We can do this “
and this is my third point “ because autonomy, like freedom, is a thresh-
old concept. That is why someone who gives in to every impulse will not
be autonomous, ceteris paribus, but someone who occasionally acts impul-
sively may be. One need not be perfectly autonomous, in other words, in
order to be autonomous. It is only necessary to go beyond that vaguely
de¬ned threshold that distinguishes the autonomous from those who are
not (quite) autonomous.
This is an especially important point in the present context because it
helps to resolve those paradoxical situations that have worried Skinner
and bedeviled Pettit. As we have seen, Pettit resorts to a distinction be-
tween ˜unfree™ and ˜non-free™ in his attempt to explain how someone
who experiences non-arbitrary interference, such as the imposition of
a tax levy, may not suffer a loss of freedom. In making this move, I ar-
gued, Pettit implicitly trades on the sense in which freedom is a threshold
concept; and a more straightforward way to deal with the problem is to
say that the person subject to the levy remains a free person even though
she is not as free to spend as she was before the levy. An even better way
to resolve the problem is to employ the concept of autonomy. Doing so
certainly makes it easier to handle the case of the person whose impris-
onment seems, almost by de¬nition, to drop her below the threshold
that separates the free person from the unfree. It is easier, at least, if we
have reason to believe that the prisoner committed the crime of her own
volition in full knowledge of the illegality of her act and of its likely con-
sequences. To say that this prisoner remains a free person strains, at best,
the concept of freedom. Yet there is little strain, if any, in describing the
prisoner as autonomous but not free. That is because ˜autonomy™, unlike
˜freedom™ or ˜liberty™, is typically used to characterize persons in a global
sense. I may ask whether you are free this weekend, but only in frivolity
or in a philosophy seminar would I ask whether you are autonomous
this weekend. The threshold element is stronger in autonomy than in
freedom, in short, because autonomy is more of a global concept than
These considerations lead to my ¬nal point, which is that a commit-
ment to autonomy does not also commit one to the populist or plebisc-
itary forms of democracy that Pettit and Viroli deplore. If we want our
Domination and the Republican Challenge to Liberalism 199

political arrangements to respect individual autonomy, or to acknowledge
that people are ˜discursive subjects™ with voices and ears of their own,
those arrangements will have to be in some sense democratic. But auton-
omy does not require unanimous direct democracy, for a person does not
cease to be autonomous whenever a vote goes contrary to his wishes. Nor
does autonomy require unbridled majority rule. Indeed, majority rule is
both friend and foe of autonomy: friend because it is the only decision
procedure that gives equal weight to everyone™s vote, and foe because it
may allow those who constitute the majority to dominate those in the
minority. As Viroli remarks, a ˜law accepted voluntarily by members of
the most democratic assembly on earth may very well be an arbitrary law
that permits some part of the society to constrain the will of other parts,
thus depriving them of their autonomy™.57
Viroli™s remark is important both for what it says about the threat that
an excess of democracy poses to self-government and for what it implies
about republicanism and liberalism. As the words I have italicized indi-
cate, Viroli™s defense of republican liberty against unchecked majoritari-
anism is entirely consistent with a commitment to autonomy. But it is also
consistent with liberal fears that individual rights and liberties will fall vic-
tim to the tyranny of the majority. That is why the rule of law, separation
of powers, checks and balances, and other devices for constraining the
majority are neither peculiarly republican nor distinctively liberal. That
is also why republicans and liberals alike should be concerned with prob-
lems such as permanent or persistent majorities, which inevitably lead
the people who are on the losing side of almost every vote to ask whether
they are really self-governing or merely subject to the domination of the
majority. It is, in sum, the commitment to autonomy that unites repub-
licans and liberals in their quest for political arrangements that protect
and promote the individual™s ability to be self-governing. In this, as in
other respects, there is no reason to regard republicanism and liberalism
as hostile or even sharply divided political theories.

V The Republican Challenge to Liberalism
What, then, is the nature of the republican challenge to liberalism? It
is the challenge to take more seriously the commitment to individual
autonomy. Liberals too often seem to think that respecting autonomy is
simply a matter of leaving people alone to pursue their own conceptions
of the good, at least as long as they do not harm or violate the rights of
others. Many liberals are thus vulnerable to the two lines of attack that
Richard Dagger

neo-republicans have brought against them: ¬rst, as Sandel and others
have urged, that liberal societies give too little attention to the cultiva-
tion of the civic virtues necessary to sustain a self-governing polity; and
second, as Pettit, Skinner, and Viroli insist, that freeing people from in-
terference is not the same as enabling them to be free, self-governing
persons. Anyone who hopes to foster autonomy will do well to take these
criticisms seriously. For liberals, this means that they should correct their
course where necessary to respond to the republican challenge.
Can this be done? The examples of Constant, John Stuart Mill, and
other liberals who have displayed markedly republican tendencies indi-
cate that it can. For Constant, as we have seen, the challenge is to cherish
˜modern™ liberty while guarding against the danger that, ˜absorbed in the
enjoyment of our private independence, and in the pursuit of our partic-
ular interests, we should surrender our right to share in political power
too easily™.58 For his part, Mill gave classical expression to the ˜liberal™
view of freedom: ˜The only freedom which deserves the name, is that of
pursuing our own good in our own way, so long as we do not attempt to
deprive others of theirs, or impede their efforts to obtain it™.59 But he also
denied, in Considerations on Representative Government, that a truly benevo-
lent despotism is the ideal form of government, and he called attention
to the evil of domination in The Subjection of Women. That is not to say, of
course, that Mill and Constant were always right, or that their writings ex-
haust the possibilities of republican liberalism or liberal republicanism.
What these writings do show, however, in line with the arguments set out
in this chapter, is that a republican liberalism is not only possible but, for
anyone committed to the promotion of autonomy, remarkably attractive
as a theory of politics.60

See, for example, Zera Fink, The Classical Republicans: An Essay in the Recovery
of a Pattern of Thought in Seventeenth Century England (Evanston, IL: Northwest-
ern University Press, 1945); Caroline Robbins, The Eighteenth-Century Com-
monwealthman (Cambridge, MA: Harvard University Press, 1959/1985); and
Gordon Wood, The Creation of the American Republic, 1776“1787 (Chapel Hill,
NC: University of North Carolina Press, 1969).
Michael Sandel, Democracy™s Discontent: America in Search of a Public Philosophy
(Cambridge, MA: Harvard University Press, 1996); Philip Pettit, Republican-
ism: A Theory of Freedom and Government (Oxford: Clarendon Press, 1997).
For liberals who reject republicanism, see, for example, Steven G. Gey, ˜The
Unfortunate Revival of Civic Republicanism™, University of Pennsylvania Law
Review, 141 ( January 1993): 801“98; and, more temperately, Gerald F. Gaus,
Domination and the Republican Challenge to Liberalism 201

˜Backwards into the Future: Neorepublicanism as a Postsocialist Critique of
Market Society™, Social Philosophy and Policy, 20 (Winter 2003): 59“91.
Sunstein, ˜Beyond the Republican Revival™, Yale Law Journal, 97 ( July 1988),
p. 1590.
I develop this position in Civic Virtues: Rights, Citizenship, and Republican Lib-
eralism (Oxford: Oxford University Press, 1997). Others who make a case
for liberal republicanism or republican (or civic) liberalism include Shelley
Burtt, ˜The Politics of Virtue Today: A Critique and a Proposal™, American
Political Science Review, 87 ( June 1993): 360“68; Stephen Holmes, Passions
and Constraint: On the Theory of Liberal Democracy (Chicago: University of
Chicago Press, 1985), especially pp. 5 and 28; Jeffrey Isaac, ˜Republicanism
vs. Liberalism? A Reconsideration™, History of Political Thought, 9 (Summer
1988): 349“77; Alan Patten, ˜The Republican Critique of Liberalism™, British
Journal of Political Science, 26 (1996): 25“44; Thomas A. Spragens, Jr., Civic
Liberalism: Re¬‚ections on Our Democratic Ideals (Lanham, MD: Rowman and
Little¬eld, 1999); and Ronald Terchek, Republican Paradoxes and Liberal Anx-
ieties (Lanham, MD: Rowman and Little¬eld, 1997).
For book-length responses, see Bill Brugger, Republican Theory in Political
Thought: Virtuous or Virtual? (London: Macmillan, 1999); Iseault Honohan,
Civic Republicanism (London and New York: Routledge, 2002); and M. N. S.
Sellers, The Sacred Fire of Liberty: Republicanism, Liberalism, and the Law (Lon-
don: Macmillan, 1999).
Shelley Burtt, ˜The Good Citizen™s Psyche: On the Psychology of Civic Virtue™,
Polity, 23 (Fall 1990): p. 24. Cf. M. M. Goldsmith, ˜Republican Liberty Con-
sidered™, History of Political Thought, 21 (Autumn 2000), p. 555: ˜The good
citizen will develop the values of identi¬cation with the society, loyalty to it
and public spirit “ the willingness to put the good of the society above one™s
own good™.
Aristotle, The Politics, 1283b42“1284a3 (Book III, Chapter 13).
Rousseau, Du Contrat Social : note that ends Book I.
Sandel, Democracy™s Discontent, p. 6.
Ibid., p. 4.
Ibid., p. 323.
Richard Dagger, ˜The Sandelian Republic and the Encumbered Self ™, The
Review of Politics, 61 (Spring 1999): 181“208. See also Sandel™s response and
my rejoinder in ibid., pp. 209“17.
Pettit, Republicanism, esp. chapters 1“4; Skinner, Liberty Before Liberalism (Cam-
bridge: Cambridge University Press, 1998), and ˜A Third Concept of Liberty™,
London Review of Books, 16 (4 April 2002): 16“18; and Viroli, Republicanism,
trans. Antony Shugaar (New York: Hill & Wang, 2002).
Viroli, Republicanism, p. 61: ˜From a theoretical point of view, liberalism can
be considered an impoverished or incoherent republicanism, but not an
alternative to republicanism™.
Benjamin Constant, Political Writings, trans. and ed. Biancamaria Fontana
(Cambridge: Cambridge University Press, 1988), p. 317.
Ibid., p. 321.
Ibid., p. 326.
Richard Dagger

18. Ibid., p. 327.
19. Ibid., p. 328; emphasis added.
20. ˜Two Concepts of Liberty™, in Berlin, Liberty, ed. Henry Hardy (London:
Oxford University Press, 1969/2002). For the distinction™s troubles, see, for
example, David Miller™s introduction to Miller, ed., Liberty (Oxford: Oxford
University Press, 1993), esp. pp. 9“15; Holmes, Passions and Constraints, p.
28f.; and Rainer Forst™s Chapter 10 in the present volume.
21. Indeed, Green™s chief defense of positive freedom is found in his “Liberal
Legislation and Freedom of Contract,” which he delivered as a lecture to the
Leicester [England] Liberal Association in January 1881.
22. “Liberal Legislation and Freedom of Contract,” in Miller, ed., Liberty, p. 21.
23. Skinner, ˜The Paradoxes of Political Liberty™, in Miller, ed., Liberty, p. 202. In
his Introduction to Liberty, David Miller argues (p. 6) that Skinner goes too
far when he describes Machiavelli™s view of freedom as purely negative, ˜since
that overlooks the fact that a person™s freedom consists [for Machiavelli] also
in his membership in a self-governing state™. We need not settle this point
here, however, for on neither Miller™s nor Skinner™s reading does republican
freedom equate to Berlin™s positive liberty.
24. Skinner, ˜A Third Concept of Liberty™, p. 18; Pettit, Republicanism, p. 80.
See also Skinner, Liberty Before Liberalism, p. 84: ˜What the neo-roman writers
repudiate avant la lettre is the key assumption of classical liberalism to the
effect that force or the coercive threat of it constitute the only forms of
constraint that interfere with individual liberty. The neo-roman writers insist,
by contrast, that to live in a condition of dependence is in itself a source and
a form of constraint™.
25. The following discussion draws on my ˜Republicanism Refashioned: Com-
ments on Pettit™s Theory of Freedom and Government™, The Good Society,
9: 3 (2002): 50“53.
26. Ibid., p. 41.
27. Ibid., p. 64.
28. Ibid., p. 91.
29. Berlin, ˜Two Concepts of Liberty™, §III (pp. 181“87 in Berlin, Liberty [2002]).
30. Pettit, Republicanism: A Theory of Freedom and Government (Oxford: Clarendon
Press, 1999), p. 291; emphasis added.
31. Ibid., p. 76; emphasis added.
32. Ibid., p. 77.
33. See also ˜Republican Freedom and Contestatory Democratization™ (in Democ-
racy™s Value, eds. Ian Shapiro and Casiano Hacker-Cord´ [Cambridge: Cam-
bridge University Press, 1999]), where Pettit refers to ˜“old liberals” such as
John Locke™ (p. 166, with no explanation of the quotation marks), then sub-
sequently designates him ˜the hero of the later commonwealthman tradition,
John Locke™ (pp. 170“71).
34. Ibid., paperback edition, p. 301, n. 2.
35. Ibid., p. 10.
36. For the quoted passage, see Pettit, A Theory of Freedom: From the Psychology to
the Politics of Agency (Oxford: Oxford University Press, 2002), p. 152. But also
note the reference to ˜the characteristically liberal ideal of non-interference™
on p. 151.
Domination and the Republican Challenge to Liberalism 203

Ibid., p. 140.
Pettit, ˜Republican Freedom and Democratic Contestation™, p. 165.
Skinner, ˜A Third Concept of Liberty™, p. 18.
˜Republican Freedom and Contestatory Democratization™, p. 165.
Pettit, ˜Keeping Republican Freedom Simple: On a Difference with Quentin
Skinner™, Political Theory, 30 ( June 2002), p. 342; emphasis in original.
Skinner, Liberty Before Liberalism, p. 83, n. 54.
Pettit, ˜Keeping Republican Freedom Simple™, p. 347. The internal quotation
is from Pettit, Republicanism, p. 56.
Ibid., p. 344.
Pettit, A Theory of Freedom, p. 140.
Pettit, ˜Keeping Republican Freedom Simple™, p. 351.
Ibid., p. 351.
Ibid., p. 352.
Skinner, Liberty Before Liberalism, p. 84, n. 57.
Viroli, Republicanism, p. 10. See also pp. 47“52, where he argues against
Skinner™s conception of republican liberty as absence of dependence and of
Ibid., p. 10; see also p. 41.
Ibid., p. 42; emphasis added to ˜not when . . . my will ™, but not to ˜constant
danger™. See also ibid., p. 41: ˜The republican conception of liberty differs
from the democratic idea that liberty consists of the “power to establish
norms for oneself and to obey no other norms than those given to oneself ”.
This is liberty in the sense of autonomy™. Viroli does not give the source of the
internal quotation, but it appears from the context to be Norberto Bobbio,
Politica e cultura (Turin: Einaudi, 1974).
Wolff, In Defense of Anarchism, 3rd ed. (Berkeley, CA: University of California
Press, 1998), p. 18 et passim. I criticize Wolff ™s argument in Civic Virtues,
pp. 62“68.
See, for example, John Christman, ed., The Inner Citadel: Essays on Individual
Autonomy (New York: Oxford University Press, 1989), and Social Philosophy
and Policy, 20 (Summer 2003), which is devoted to autonomy.
Tushnet, ˜Following the Rules Laid Down: A Critique of Interpretivism and
Neutral Principles™, Harvard Law Review, 96 (February 1983): p. 783.
It is worth noting that for all their differences, not a single essay in the present
volume conceives of autonomous persons as ˜isolated islands of individuality™.
Viroli, Republicanism, p. 43; emphasis added.
Constant, Political Writings, p. 326.
Mill, On Liberty and Other Writings, ed. Stefan Collini (Cambridge: Cambridge
University Press, 1989), p. 16.
For very helpful comments on an earlier draft of this essay, I am grateful to
C´ cile Fabre and David Miller.

Liberal Autonomy and Consumer Sovereignty

Joseph Heath

There is obviously a very close af¬nity between the concept of citizen
autonomy, which is identi¬ed with a liberal organization of the political
sphere, and the concept of consumer sovereignty, which is used to justify cer-
tain sorts of free market arrangements in the economy. Depending upon
whose stock is thought to be higher at any given time, politicians will ap-
peal to economic metaphors in order to justify liberal political principles,
just as business leaders will appeal to democratic imagery in order to jus-
tify economic freedom. Thus, citizens are often described as “consumers”
or “clients,” browsing through the “marketplace of ideas,” deciding how
to “spend” their votes. Similarly, consumers are said to “vote with their
wallets,” and that the success of ¬rms re¬‚ects the “popular will.” The very
term “consumer sovereignty” borrows a concept from the political sphere
in order to legitimate a certain sort of economic arrangement.
These parallels are not illegitimate. The doctrines of liberal autonomy
and consumer sovereignty are usually grounded through appeal to very
similar normative considerations. More speci¬cally, the arguments that
are given for the claim that the state should respect the autonomy of citi-
zens are often identical to those made in favour of consumer sovereignty.
If individuals are, by and large, the best judges of their own interests, then
the society that leaves them free to pursue their own chosen plan of life
will be superior to one that does not. Autonomy in the political realm gives
individuals the legal freedom to pursue their goals, whereas sovereignty
in the marketplace allows them access to the resources needed to achieve
However, whereas autonomy is a very popular political ideal among
those of a broadly liberal persuasion, consumer sovereignty has not
Liberal Autonomy and Consumer Sovereignty 205

received a uniformly warm reception. The doctrine is often viewed with
suspicion, if not outright hostility. This hostility runs highest among those
who want to diagnose “consumerism” as one of the major ¬‚aws of capi-
talism. (In fact, the critique of consumerism is in many ways equivalent
to a rejection of the doctrine of consumer sovereignty.) According to this
sort of view, the doctrine of consumer sovereignty constitutes nothing
more than a crude misappropriation of democratic vocabulary in order
to justify callow corporate interests.
This is where the big question emerges. Is it possible to reject con-
sumer sovereignty while endorsing liberal autonomy? Economists have
an extremely strong tendency to simply equate the two doctrines “ to
assume that a democratic society must respect the wishes of consumers,
and therefore that a laissez-faire economic system is a natural expres-
sion of the political ideals underlying western liberalism. This close tie
between consumer sovereignty and liberal political ideals has been in-
advertently reinforced by many critics of consumerism, who have had a
lamentable tendency to develop their critique in a way that ¬‚agrantly
disregards the principles of liberal autonomy. In particular, these critics
have not been nearly careful enough to ensure that their critique of con-
sumer sovereignty does not imply an unacceptably paternalistic attitude
toward consumers. This has made it much easier to dismiss the critique
of consumerism as fundamentally illiberal.
It is my conviction that one can consistently criticize consumerism with-
out abandoning the liberal commitment to respecting the autonomy of
consumers, but that doing so requires greater care in the formulation
of the critique than is normally exercised. The key difference between
the two doctrines, as traditionally conceived, is that a commitment to lib-
eral autonomy does not preclude the possibility of legal remedy in cases
where state intervention is able to resolve a collective-action problem.
The doctrine of consumer sovereignty, on the other hand, has tradition-
ally involved a commitment to market freedom under all circumstances,
even in cases where consumers are clearly caught in a suboptimal equi-
librium. Thus it is possible to violate consumer sovereignty in order to
eliminate such equilibria, without thereby contradicting one™s commit-
ment to respect autonomy.

I Autonomy and Sovereignty
Many theorists regard the de¬ning characteristic of a liberal political
order to be an of¬cial commitment on the part of the state to remain
Joseph Heath

neutral with respect to the different goals and projects pursued by its
citizens. This strain of liberal thinking dates back at least to Hobbes, who
argued that because each person™s conception of the good is a function
of that person™s ever-changing desires, any attempt to establish a political
order dedicated to the pursuit of one particular conception of the good
would necessarily generate con¬‚ict and disorder. The solution was to cre-
ate a state dedicated to providing only the enabling conditions needed
for the pursuit of any conception of the good: personal security, enforce-
ment of contracts, and so on. Because these sorts of services are of value to
citizens regardless of the conception of the good that each one happens
to endorse, such a state should be able to enjoy stable, universal support
amongst the citizenry. Most importantly, the powers exercised by such a
state should be ones that all citizens could agree to accept ex ante.
It is not dif¬cult to derive from this commitment to neutrality the view
that a liberal state should respect the autonomy of its citizens. One need
only add the premise that the individual is, in general, best positioned to
ascertain his or her own good, and also the most likely to carry out the
pursuit of such good in a conscientious manner. From this, it is easy to
infer a general non-interference principle. Not only should the state re-
frain from imposing some conception of the good upon an individual,
the state should not even impose upon the individual when doing so
is thought to promote that individual™s own good. This is naturally sub-
ject to certain conditions: the action in question must not harm anyone
else, the individual making the decision must be mature, well-informed
and of sound mind, and so on. When these conditions are satis¬ed, it is
then claimed that the best overall outcome is to be achieved by granting
individuals whatever freedoms are needed for them to formulate and
pursue their own particular conceptions of the good life. It is precisely
in granting this freedom that the liberal polity is said to respect the au-
tonomy of its citizens.1
This commitment to respect autonomy has a number of different po-
litical consequences. One of the most important is that it acts as a prima
facie constraint on the enactment of paternalistic laws. If individuals are
the best judges of what is good for them, then the state is not justi¬ed in
making interventions that have as their primary intent thwarting a self-
regarding, voluntary choice made by those individuals, even when it is
thought that this is being done for their own good. The classic argument
for such an anti-paternalism constraint “ from John Stuart Mill “ is that for
every one or two times somebody is stopped from doing something that
he or she truly would have regretted, there will be dozens of interventions
Liberal Autonomy and Consumer Sovereignty 207

that simply represent other people™s imposing their own conceptions of
how best to live. Thus the net effect of a policy of paternalistic interven-
tion will be an overall reduction in social welfare.
Apart from this ef¬ciency argument, it is also often argued on contrac-
tualist grounds that the power to make paternalistic interventions could
never be grounded in consent (since the person whose decision is be-
ing overridden would always object), and so the state could not have the
legitimate authority to engage in such interventions. Any interventions
that do take place must therefore represent the unjust exercise of power
by one individual or group over another.
The case for consumer sovereignty rests upon substantially similar
grounds. Put crudely, the doctrine of consumer sovereignty states that
individuals should be able to spend their money however they want. More
generally, where “market transactions” are understood as contracts involv-
ing the exchange of rights over property, consumer sovereignty stipulates
that individuals should be free to dispose of any property they have rights
over through any market transaction they choose. They should not be
forced to buy anything or hold anything they do not want.
The term “consumer sovereignty” is of relatively recent vintage. Its
formal introduction is credited to William Harold Hutt, who in his 1936
book Economists and the Public, de¬ned it as follows: “the consumer is
sovereign when he has not delegated to political institutions for author-
itarian use the power which he can exercise socially through his power
to demand (or refrain from demanding).”2 Of course, a commitment
to consumer sovereignty is implicit in a lot of earlier work. Perhaps the
reason it was not articulated more explicitly is that the set of liberties in
question can easily be seen as simply a subset of those liberties accorded
to all citizens within a liberal political regime. If market transactions
are just contracts, and the state is simply in the business of enforcing
contracts “ not deciding which ones are wise or unwise “ then it follows
pretty closely that consumers can enter into whatever market transactions
they like.
It is unsurprising, then, that the primary arguments for consumer
sovereignty are substantially equivalent to the ones for liberal autonomy.
First, in terms of the ef¬ciency effects, consumer sovereignty gives con-
sumers (indirect) control over the major decisions pertaining to the al-
location of scarce resources within the society. This makes it far more
likely that these resources will be employed in ways that promote human
welfare than if the decisions are made by some other agency, such as the
state. The details of how the price system accomplishes this are too well
Joseph Heath

known to require repetition here. The second argument appeals to more
principled concerns. If individuals are not choosing for themselves how
to exercise their demand, then they are in effect being bound by contrac-
tual obligations that they have not freely chosen. The fact that they have
not consented, again, suggests that the relevant exercise of power is not
one that citizens could have unanimously agreed to delegate to the state.
It therefore represents the unjust exercise of power of one group within
society over another.

II Perfectionism
The most common objection to consumer sovereignty is that it leads to
a social pathology known as “consumerism.” This ailment is somewhat
ill-de¬ned. Obviously, prior to the twentieth century, the overwhelming
majority of human beings lived at, near, or below the subsistence level. It
is only in the past century, in the industrialized parts of the world, that any
signi¬cant segment of a population has acquired what might reasonably
be described as “disposable income.” This has meant that for the ¬rst
time, large numbers of people ¬nd themselves with actual consumption
choices “ that is, situations in which their choices are not directly dictated
by economic necessity. People in our society “ even many of the very
poorest “ now choose what they would like to eat for breakfast, choose
what clothes to wear, choose what forms of entertainment to consume,
and so on.
At ¬rst glance, this would seem to be a good thing. The fact that peo-
ple have such choices is a direct consequence of the fact that they are
no longer condemned to living out their lives in grinding poverty. Thus
there is a clear sense in which consumerism, insofar as it is a problem,
is something of a problem for happy people. Like obesity, it™s not some-
thing that people have had to worry about too much over the course of
human history. The concern, however, is that people may not be exercis-
ing this new-found freedom wisely. John Kenneth Galbraith put it best,
at the beginning of The Af¬‚uent Society, when he suggested that people
may not handle living under conditions of material af¬‚uence very well,
simply because the human race has no prior experience dealing with this
There are in fact some striking parallels between obesity and con-
sumerism. Both are understood as diseases of excess. However, unlike
the case of obesity, where the damage done to the person™s health is eas-
ily quanti¬able, the impact of consumerism is much harder to diagnose.
Liberal Autonomy and Consumer Sovereignty 209

There is the obvious fact that people often express dissatisfaction with
their lives, regardless of how much money they make, or how many mate-
rial possessions they have. But the underlying cause of this dissatisfaction
is very dif¬cult to determine, and tends to get expressed in a wide variety
of different ways. As a result, critics of consumerism tend to de¬ne the
pathology in a way that is quite intimately tied to what they happen to
think is wrong with consumer society.
The most common objection to consumer sovereignty is simply that,
according to the critic, consumers consistently make bad decisions. Of
course, this objection would have very little force if it were expressed
as simply a disagreement “ or worse, a difference in taste “ between
the critic and the consumer. The standard strategy is rather to argue
that consumption choices re¬‚ect a commitment to some broader set of
values, or conceptions of the good, and that these values can be organized
into some sort of evaluative hierarchy. It is then claimed that the choices
made by consumers consistently re¬‚ect values that are found towards the
bottom of this hierarchy
When formulated in terms of values, this critique of consumerism can
be referred to as perfectionist. The perfectionist critique boils down to the
claim that consumerism re¬‚ects a commitment to the wrong conception
of the good. This claim is often derived quite explicitly from a more
general perfectionist doctrine. Consider, for example, the reasoning of
Pope John Paul II:

To call for an existence which is qualitatively more satisfying is of itself legitimate,
but one cannot fail to draw attention to the new responsibilities and dangers con-
nected with this phase of history. The manner in which new needs arise and are
de¬ned is always marked by a more or less appropriate concept of man and of his
true good. A given culture reveals its overall understanding of life through the
choices it makes in production and consumption. It is here that the phenomenon
of consumerism arises. In singling out new needs and new means to meet them,
one must be guided by a comprehensive picture of man which respects all the
dimensions of his being and which subordinates his material and instinctive di-
mensions to his interior and spiritual ones. If, on the contrary, a direct appeal is
made to his instincts “ while ignoring in various ways the reality of the person as
intelligent and free “ then consumer attitudes and life-styles can be created which are
objectively improper and often damaging to his physical and spiritual health. Of
itself, an economic system does not possess criteria for correctly distinguishing
new and higher forms of satisfying human needs from arti¬cial new needs which
hinder the formation of a mature personality.4

According to John Paul II, it is precisely the neutrality of the capitalist
economy that generates the problem. Because it works to satisfy whatever
Joseph Heath

needs people happen to have, it has no mechanism to promote the sorts
of needs and desires that people should have. Unfortunately, people have
a tendency to let themselves be governed by their “instincts,” and so to
assign greater priority to their lower material needs than to their higher
spiritual ones. This tendency gets translated by the economic system into
a pattern of production and consumption that overemphasizes the value
of material goods.
Of course, this argument makes no pretence of respecting liberal po-
litical principles. It is self-consciously illiberal, insofar as it depends upon
the view that society should privilege one particular conception of the
good. The neutrality of economic institutions is what provides the pri-
mary grounds for the objection to the unrestricted exercise of consumer
choice. Its rejection of consumer sovereignty and its rejection of liberal
autonomy therefore go hand in hand.
There should be no doubt that the position articulated here is consis-
tent (regardless of whether it is politically judicious). However, it is clearly
not a view that is available to the liberal critic of consumerism, as least
as far as those terms are being used here. Much more problematic are
cases in which critics rely upon a tacitly perfectionist critique of con-
sumerism, while still espousing liberal principles in the political realm.
Consumerism as a social phenomenon is generally sustained through
individual behavior that we condemn on moral grounds. In particular,
consumerism is associated with character traits that we identify as vices:
vanity, greed, narcissism, venality, to name just a few. Critics often sup-
pose that if is wrong for individuals to engage in this sort of behavior,
then clearly it is also wrong for society to set up the economic system that
rewards this sort of behavior.
Much of this concern is clearly motivated by anxiety over the decline of
the traditional virtues associated with the “Protestant work ethic,” and the
growing shift in emphasis away from production and toward consump-
tion. For example, Daniel Bell™s The Cultural Contradictions of Capitalism is
motivated by the assumption that these virtues are somehow functionally
essential for the reproduction of our society, and that without them we
are condemned to civilizational decline.5 But despite decades of hand-
wringing, none of these dire predictions has come true. Capitalism ap-
pears to get along just ¬ne without the old package of Christian virtues.
Furthermore, it has scored some remarkable successes in parts of the
world where “materialist” values are arguably even more prominent “
various parts of Asia come to mind. So if there is any mileage to be had
from this anxiety about the work ethic, it must be that the critic regards
Liberal Autonomy and Consumer Sovereignty 211

these values as somehow intrinsically superior to the ones that prioritize
The problem with such arguments is that the hierarchy of values that
they hinge upon is extremely controversial. It is not like the case of
health and obesity. There is no question that the way we are wired up
gives us an impulse to gorge on fatty foods. This is an adaptation de-
signed to help us survive conditions of material scarcity, and it is clearly
a maladaptation in contexts of abundance. Thus there is very little ob-
jection when governments engage in public-health initiatives designed
to counter these effects (promoting healthy diet and lifestyle, requiring
nutritional information on food packages, and so on.) But the case of
consumerist values is quite different. We could certainly use a more ro-
bust discussion in our society about how it is best to live. We may also agree
that many people pursue forms of grati¬cation that are ultimately self-
defeating. But there is still room for a great deal of reasonable disagreement
about the merits of these different projects. Thus it is entirely inappro-
priate for the liberal state to intervene in order to discourage “improper”

III False Consciousness
While the perfectionist critique of consumerism rejects the liberal com-
mitment to autonomy either implicitly or explicitly, there are a number
of other views that endorse the basic principle yet deny that the condi-
tions needed for the exercise of autonomous choice by consumers are
satis¬ed. In order to count as fully autonomous, a choice must be made
by an agent who is suf¬ciently mature, well-informed, and rational. Many
critics argue that one or another of these conditions is usually absent, and
thus that the choices made by consumers will often not re¬‚ect their own
best interests. According to this view, the pattern of action that results
from such choices “ consumerism “ re¬‚ects a form of false consciousness.
Consumerism is therefore diagnosed as a type of ideology.
Early versions of this argument were direct descendents of Marx™s cri-
tique of “commodity fetishism.” The original critique of ideology was
intended to provide some explanation for why workers generally failed
to agitate for the revolutionary overthrown of capitalism, given that the
persistence of the capitalist order was supposedly so inimical to their
interests. Marx™s considered response was that the operations of the cap-
italist economy disguised what were essentially social relationships, mak-
ing them appear as though they were a relationship between things. This
Joseph Heath

gave the economy the appearance of being a natural order, when in fact
it was arti¬cial, and hence mutable.
This explanation became considerably less persuasive as the twentieth
century progressed, simply because the occurrence of communist rev-
olutions in Russia and elsewhere made it clear to almost everyone that
a capitalist organization of the economy was merely one option among
many. So then why did the workers not rebel? Part of the problem seemed
to be that workers were overly content with the material goods afforded
to them by the capitalist order. Thus “consumerism” was hit upon as a
fallback explanation for the fact that the communist revolution did not
occur. Consumer goods were identi¬ed as the new opiate of the people.
Workers had been seduced into thinking that they liked living in sub-
urban houses, watching television, and eating white bread. This caused
them to overlook their real interests, which involved building a classless
society, freeing themselves from oppression, and so forth.
Of course, this theory ascribes a pretty colossal self-misunderstanding
to the working classes. As a result, it takes on a substantial burden of
proof. The mere claim that the “objective interests” of the proletariat
could be deduced from a correct understanding of their world-historical
role is hardly suf¬cient to discharge this burden. What made the diag-
nosis more plausible was the development of advertising. The success of
Nazi propaganda during the Second World War showed that modern
techniques of manipulation, particularly when ampli¬ed through mass
media like print and radio, could have an extraordinarily powerful effect
upon individuals. Anyone who doubted that large numbers of people
could be quite systematically misled about where their interests lay had
only to look at the Nuremberg rallies in order to see quite dramatic coun-
It is therefore not surprising that after the war ended, many people
were quite disturbed to see exactly the same propaganda techniques ap-
plied with renewed vigor to the sale of consumer goods. It suggested the
following line of reasoning: If Germans could be systematically brain-
washed by the Nazis, why couldn™t consumers be similarly brainwashed
by corporations?
Thus the development of advertising lent a great deal of plausibility
to the diagnosis of consumerism as a form of false consciousness. Vance
Packard™s 1957 bestseller The Hidden Persuaders provides a canonical for-
mulation of this critique, and re¬‚ects quite clearly the alarm and anxiety
that the introduction of these new techniques of persuasion created. Of
Liberal Autonomy and Consumer Sovereignty 213

course, The Hidden Persuaders was written at a time in which the intellectual
credibility of Freudian psychoanalysis was still untarnished. This meant
that people were far more likely to take seriously the suggestion that
consumers might be subjected to total manipulation at a subconscious
level. Packard cited studies, for example, that purported to show that con-
sumers entered into a “hypnoidal trance” upon entering supermarkets.
These poor shoppers, he reports, “were so entranced as they wandered
about the store plucking things off shelves at random that they would
bump into boxes without seeing them.”7
But setting aside these more extreme claims, it must be admitted that
there are some grounds to be concerned about the effects of advertising.
If the widely-reported claim that we are exposed to approximately 3,000
commercial messages every day is correct, then this is bound to have a
distorting effect upon our consciousness. It doesn™t matter how “media-
savvy” consumers have become, the idea that anyone could emerge from
this sort of bombardment without at least some distorted perceptions is
psychologically fantastic. Advertising must be having some effects, and
there is no reason to think that these effects work to the interest of
consumers “ simply because advertising is not designed to promote the
consumers™ interest.
Thus, critics of consumerism have called for restrictions upon adver-
tising, or on forms of consumption heavily promoted by advertising. Of
course, their case is bolstered by the fact that we already have laws that pro-
hibit certain types of misleading advertising. Advertisements that make
false claims about a product are illegal, as is packaging that misrepresents
its contents. All of these restrictions are designed to prevent consumers
from being manipulated by advertisers. So why not just extend these laws
further? We have had a tendency to focus legal efforts on the prohibition
of deceptive advertising. (Of course, almost all advertising is deceptive to
some degree “ regulation is usually a matter of drawing the line some-
where.) We might consider extending these efforts in other directions as
a way of combating consumerism.
However, as soon as this suggestion is made, it becomes clear that such
a proposal runs into trouble with autonomy. Take ¬rst the case of ad-
vertising aimed it manipulating people™s desires. It is common, among
critics of consumerism, to suggest that there is a distinction between our
“natural” desires and “arti¬cial” desires. (Such a distinction is, of course,
one of the oldest chestnuts in the literature on autonomy.) Advertising
is often held responsible for inculcating the latter. Because the agent is
Joseph Heath

manipulated into acquiring this desire, the conditions of autonomous
choice were not satis¬ed. As a result, it is argued, a paternalistic interven-
tion aimed at blocking the ad, or preventing the consumer from acting
upon the desire, would be legitimate.
Among liberals, there is still considerable controversy over the ques-
tion of whether actions based on imprudently formed desires count as
autonomous. Some have argued that an intervention is only paternalistic
if it prevents the agents from satisfyings desire that she would have had,
under conditions of full information and complete rationality. Others
argue that society has no business asking whether or not the agent has
exercised due diligence in cultivating one or another of his desires “ thus
all of them should be taken at face value.
But without even getting into this issue, we can see that there are still
greater problems with the idea that the desires formed through adver-
tising can be discounted or overruled. The most basic problem is simply
that for every one agent who acquires a desire through some purport-
edly “illegitimate” mechanism like advertising, there will be two or three
more who acquire it in a way that cannot be so easily impugned. For ex-
ample, people are far more in¬‚uenced by word of mouth, along with the
consumption habits of those around them, than they are by advertising.8
How could anyone possibly separate out these cases? There is a clear
sense in which all of our desires are “arti¬cial,” produced through a very
complex interaction between individuals and their cultural environment.
The idea that any one set can be privileged as somehow more authen-
tic than some other is deeply implausible. What is more likely is that
it re¬‚ects a mere difference in values. But if this is the case, then the
critique of false consciousness turns out to be just a disguised form of
The major problem is that whatever diagnosis the critique eventually
settles on is still likely to be deeply controversial. In the case of deceptive
advertising, we at least have some widely accepted criteria that allow us
to distinguish between false and true belief. One doesn™t need to be a
non-cognitivist about evaluative judgment to recognize that such criteria
are lacking in the case of desires. Even if it makes sense to judge one
desire to be more authentic, or natural, or even true, than another, there
is no consensus about how this should be done. And as long as such
disagreements exist, it would be unacceptably paternalistic to control
advertising because of its ability to manipulate our desires. We may place
restrictions on advertising aimed at children, but we cannot do so with
respect to adults.9
Liberal Autonomy and Consumer Sovereignty 215

IV Cultural Homogenization
The argument for consumer sovereignty rests quite heavily on the as-
sumption that markets are a relatively transparent mechanism for trans-
mitting consumer demand to suppliers. If markets are transparent, then
the fact that a particular commodity is being produced, in certain quan-
tities, re¬‚ects the fact that there is approximately that much demand
for it. So when you drive through the suburbs, past outlet malls, big-
box retailers, and franchise fast-food joints, you can rest content in the
knowledge that all is well; these places exist only because they are what
the people themselves want most. And when McDonalds or Pizza Hut
opens up shop in some new country somewhere in the world, putting
a number of local restaurants out of business, one should rejoice be-
cause, after all, the people there obviously prefer to eat at McDonalds or
Pizza Hut.
This story is, for better or for worse, one that many people have a hard
time believing. Looking at the incredible diversity of cultures in the world,
it is dif¬cult to imagine that much of this was sustained only through
the relative isolation of societies from one another “ but that once they
start trading, people turn out to all harbor a secret desire for Big Macs
and Coca-Cola. But there does seem to be a strong prima facie connection
between increased globalization “ in particular, increased integration into
the global capitalist economy “ and increased cultural homogenization.
It is tempting, however, to think that this homogenization, rather than
re¬‚ecting an underlying, preexisting homogeneity of taste, is actually an
effect of the global economy. The market economy, in other words, far
from being neutral with respect to taste, is positively biased towards the
satisfaction of certain sorts of tastes. In particular, it is often suspected that
the market has a leveling tendency “ reducing everything to the lowest
common denominator.
If this is correct, then intervention in the market might not violate au-
tonomy, because the market itself is not neutral. If the market is positively
biased towards increased homogenization, selective interventions aimed
at mitigating these effects need not privilege one particular set of values.
In fact, such intervention would be necessary in order to ensure that the
preexisting range of values and tastes are treated equally. Unfortunately,
critics are often less than explicit about why markets have this homoge-
nizing character. They also usually fail to take seriously the possibility that
this homogeneity may re¬‚ect uniformity of taste (or more speci¬cally, a
uniform range of diverse tastes). Thus, many critics of consumerism do
Joseph Heath

very little to free themselves from the suspicion that this cultural homo-
geneity simply offends their taste.
The closest thing to a stand-alone argument for limiting consumer
sovereignty in order to stave off cultural homogenization is due to Tibor
Scitovsky. Scitovsky points out that markets are only transparent under
very restrictive conditions. One of these conditions is that there must be
no economies of scale. But such economies do exist in the real world, and
as a result, mass-produced goods are often cheaper to manufacture than
others. This is what produces homogenization. Here is the argument:

Economies of scale not only cheapen large-scale production but by raising wages
they also raise the cost and diminish the pro¬tability of small-scale production.
This in turn raises the minimum volume of sales necessary to render production
pro¬table and thus leads to an ever increasing narrowing of the range of variants
of products offered and neglect of minority needs and tastes in the nature and
design of goods produced and marketed. The increasing neglect of minority
preferences is a bad thing, because it is illiberal, makes for uniformity, and destroys
to some degree the principal merit of the market economy: its ability to cater
separately and simultaneously to different people™s differing needs and tastes.10

This argument moves a bit too quickly though. Suppose that, initially,
goods are custom-ordered from a small-scale supplier, who produces
items that are tailor-made to each individual customer. A large-scale pro-
ducer comes along who makes the same type of product, but only in, say,
three styles. By limiting the number of styles, this producer is able to sell
the goods at much lower cost. Scitovsky infers that the large producer
will drive the small producer out of business.
But this is not necessarily the case. One must assume, given the vari-
ety of products made by the small-scale producers, that the three mass-
produced variants will not exactly match the taste of at least some con-
sumers. This means that they will suffer a loss of welfare if they switch to
the mass-produced goods. So if they do switch, it must be because they
prefer the money they save to the inconvenience of purchasing goods
not perfectly suited to their tastes. The small-scale producer would not
go bankrupt if people with minority tastes were willing to pay more for the
goods than those with majority tastes. Homogenization arises only be-
cause people are unwilling to pay the full cost associated with satisfaction
of their preferences when low-cost alternatives that are “close enough”
become available. There is nothing coercive or illiberal about this.
In any case, it is not really the market that is causing the problem
here. People who have extremely mainstream tastes bene¬t from that
fact, simply because these tastes take less effort and fewer resources to
Liberal Autonomy and Consumer Sovereignty 217

satisfy when mass-production techniques are available. This has nothing
to do with the market; it has to do with the distribution of tastes in the
population and with the productive technology that permits manufacture
on a large scale. Consumer sovereignty is not to be blamed. Furthermore,
it is not really unfair that people with minority tastes should have to pay
more, since satisfaction of these tastes imposes greater costs upon society
under such conditions.
The only case where Scitovsky™s argument holds is when a special sort
of feedback loop develops between production and purchasing decisions.
For example, if some portion of the value of a good is determined by its
anticipated resale value, people may purchase goods that are perceived
to satisfy majority tastes, even if they themselves do not especially like
them. If large numbers of people do this, it may become a self-ful¬lling
prophecy “ most people will buy a good just because it seems like the
sort of thing that most people will buy. But this is clearly a marginal
phenomenon, and it is not clear that very much can or need be done
about it.
Finally, despite all the rhetoric, it is not at all obvious that mass-
production does generate homogenization. One hundred years ago, most
people wore clothes that were made speci¬cally for them by a tailor. But
the overall variety of clothing was much narrower than it is now, especially
in terms of the variety that one could see in public display. This is primar-
ily because basic clothing has become so cheap that people can afford
to spend more on various aspects of style that make clothes distinctive.
The same can be said for food (which has become available in far greater
variety even in the past thirty years), housing, music and entertainment,
and so on.11

V Market Failure
So far, we have seen very little to suggest that the critique of consumer
sovereignty is consistent with respect for the autonomy of citizens. How-
ever, of all the arguments canvassed so far, the most dif¬cult one to fault
in this respect is Scitovsky™s. Despite its ¬‚aws, the argument does con-
tain a valuable theoretical strategy. Scitovsky tacitly acknowledges that
under conditions of perfect competition, there would be very little to
be said against consumer sovereignty. But the world we live in is very
far from meeting those conditions. As a result, the market will not func-
tion as a “transparent” mechanism for imposing consumer demand on
¬rms. When the signals being sent by consumers are being distorted, it is
Joseph Heath

possible that changing the pattern of demand will generate a new equilib-
rium that comes closer to satisfying the actual preferences of consumers.
Thus, imposing certain constraints on the sovereignty of the consumer
may in fact lead to a greater overall level of satisfaction of consumer wel-
fare. Such an intervention, despite limiting the freedom of consumers,
need not violate their autonomy.
Scitovsky™s particular example “ economies of scale “ fails to motivate
a policy of government intervention, simply because he is unable to show
that the market outcome is unfair to anyone in particular. However, if the
market imperfection is one that generates a Pareto-inef¬cient outcome,
then there would be nothing wrong with a regulatory intervention aimed
at remedying this inef¬ciency. Thus, if consumerism is identi¬ed, not as a
system of values or as a type of false consciousness, but rather as a pattern
of behavior that is symptomatic of market failure, then there is no reason
that a liberal cannot seek legislative remedies that involve restrictions on
consumer choice.
In order to see why legal intervention in such a case would not be
inappropriate, it is important to realize that the liberal conception of
autonomy has never been understood as preventing citizens from using
state power to eliminate collective action problems (or “prisoner™s dilem-
mas”). In fact, the primary justi¬cation for freedom of contract, since
Hobbes, has been that it allows individuals to work out their own cooper-
ative solutions to collective action problems, then to call upon the state
to enforce them. Every economic exchange, for instance, creates a collec-
tive action problem, because both parties have an incentive to “free-ride”
by eluding payment. Contracts of sale allow parties to work out the terms
of a solution, then to bind themselves to these terms by licensing the state
to punish any free riders. This sort of enforcement does not compromise
anyone™s autonomy, because the contract is mutually advantageous and
subject to general consent.
John Kenneth Galbraith was the ¬rst to point out that the existence of
market failure, along with missing markets, creates a serious problem for
the doctrine of consumer sovereignty. Property rights are very effective
at eliminating collective action problems when it comes to the exchange
of medium-sized dry goods, like chairs and bags of wheat. They are also
good at regulating the use of land. They are not so good when it comes to
more intangible goods, like knowledge, or goods that must be delivered
on a large scale, like bridges. They are not good at regulating the use
of air or water either. In all of these cases, these limitations are due to
enforcement problems that make it nearly impossible to eliminate free
Liberal Autonomy and Consumer Sovereignty 219

As a result, markets will systematically overproduce certain types of
goods, especially material ones, and will underproduce others. Goods that
have signi¬cant negative externalities will be overproduced, while those
with signi¬cant positive externalities will be underproduced. (We are
talking here about production levels relative to what consumer demand
would be, at the price level that would obtain when all externalities were
internalized.) Furthermore, consumers will be unable to correct this bias
through voluntary action, because the relevant economic incentives all
reinforce the biased outcome. Thus the presence of externalities locks
consumers into a giant collective-action problem.
Galbraith observed that under such circumstances, it would be entirely
reasonable for consumers to choose, collectively, to impose restrictions
on their private purchasing behaviour. Consumer sovereignty, under con-
ditions of “real existing capitalism,” will generate overproduction of “pri-
vate goods” “ that is, goods of the sort that property rights are good at
protecting. Thus, “vacuum cleaners to insure clean houses are praisewor-
thy and essential in our standard of living. Street cleaners to insure clean
streets are an unfortunate expense. Partly as a result, our houses are gen-
erally clean and our streets generally ¬lthy.”12 This is an enormous boon
for those who happen to have a strong preference for the private goods,
but it is unfair to everyone else. Thus neutrality requires that some of
the resources that are being used to produce private goods be redirected
into the production of non-market goods.
The primary mechanism through which this is done is the income tax.
Taxes in a welfare-state society are best understood as a form of mandatory
spending.13 Consumers are prevented from spending some fraction of
their income on private goods. The state takes this income and uses it
to ¬nance the provision of a wide range of goods that markets fail to
provide, or that markets would not provide at the right prices. This must
involve a restriction of consumer sovereignty, because given a choice
consumers would choose to free-ride rather than pay for these goods.
They want national defense, pest control, public education, and so on,
but they simply have no incentive to contribute. Thus, as Cass Sunstein
puts it, there are cases where “the force of law is necessary in order to allow
people to obtain what they want.”14 The consumer is only able to achieve
satisfaction of his preferences through delegation “to political institutions
for authoritarian use the power which he can exercise socially through his
power to demand” “ except of course that this use is “authoritarian” only
in the sense that it is coercive. This does not preclude the possibility that
the state remain democratically accountable for way that this spending
power is exercised.
Joseph Heath

Galbraith™s analysis also provides a useful insight into why con-
sumerism is often diagnosed as a form of false consciousness. Imagine
that each agent has two different preference schedules, one for market
goods and the other for non-market goods. Supply of the latter is compro-
mised by a collective action problem that markets fail to resolve, and so
will either be delivered through a non-market mechanism, or will not be
delivered at all. If one assumes that all consumption is subject to dimin-
ishing marginal satisfaction, this will mean that additional increments in
the supply of either type of good will be used to satisfy increasingly unim-
portant needs. Thus consumers will ¬rst ensure that their basic needs are
met “ food, shelter, clothing, and so on “ before they go on to purchase
“frills” or luxury items. Now suppose that a society is doing a very bad job
of supplying non-market goods. As a result, goods that are very high on
everyone™s list, like education or clean air, may not be provided in suf¬-
cient quantities. However, consumers who receive additional increments
in income have no means of purchasing non-market goods, because of
the unresolved collective action problem. As a result, they will wind up
buying increasingly inessential market goods that produce miniscule in-
crements in satisfaction, even when other very basic needs go unmet.
From the outside, it looks as though consumers are investing their re-
sources quite irrationally. It is important to see, however, that while this
behavior is collectively self-defeating, it is not irrational. Consumers are
merely responding to the existing set of economic incentives.

VI Conspicuous Consumption
The important feature of Galbraith™s diagnosis of consumerism is that he
treats it as a type of collective action problem. This is why his proposed
remedy does not violate the autonomy of consumers. Upon closer analy-
sis, however, one can see that these sorts of collective action problems are
far more widespread than Galbraith imagined, and that they need not
involve any sort of obvious market failure. In particular, certain forms
of conspicuous consumption generate a negative-sum game, in which
consumers “overspend” on goods that ultimately fail to produce any sat-
isfaction. This is particularly common in cases where some signi¬cant
fraction of the satisfaction consumers derive from a good depends upon
a comparison with others.
Thorstein Veblen very clearly identi¬ed the problem more than a cen-
tury ago.15 Veblen observed that goods were sought for both their mate-
rial properties and for the status they conferred upon their owner. The
Liberal Autonomy and Consumer Sovereignty 221

problem with status consumption, according to Veblen, is that it generates
waste. It is not called “waste” simply because Veblen disapproves of the
relevant sort of desires. He claims that his critique involves “no depreca-
tion of the motives or of the ends sought by the consumer,” and thus is not
merely a value-judgement, in the perfectionist sense.16 The problem with
status goods, according to Veblen, is that they work by generating an “in-
vidious comparison” between individuals. In other words, they increase
one person™s status only by lowering someone else™s. Thus the consump-
tion of status goods does “not serve human life or human well-being on
the whole.”17 Nowadays, we would say that status is a zero-sum game, and
thus status competition, insofar as it consumes other resources, is a nega-
tive sum. Conspicuous consumption is, in short, a prisoner™s dilemma.
Contemporary neo-Veblenians have formulated this basic critique at
a greater level of generality, in order to distinguish it more clearly from
the moralizing perfectionist critique. Consider Robert Frank:

Much has been written about our failure to achieve better balance in our lives,
which authors almost invariably attribute to dark forces of one kind or another “
some rooted within us, such as greed, impatience, or stupidity, and others that
work on us from the outside, such as exploitation by powerful corporate interests.
Yet these forces could be swept aside entirely and the fundamental problem would
remain, for its primary source lies not in individual or corporate imperfection,
but in the cold, impersonal logic of competition.18

Frank is at pains to emphasize that it is the comparative nature of our
preferences that generates this competitiveness. Comparativeness is often
dif¬cult to detect, however, simply because people are in general highly
adaptive. We tend to judge things big or small, beautiful or ugly, dirty
or clean, relative to what we are used to. When standards of cleanliness
in our environment go up, we adapt quickly, which in turn affects our
future judgments. Things are no different in an economic environment.
Frank uses a very telling example, involving the desire to live in a spacious
home. What counts as spacious is very much dependent upon the size of
everyone else™s house. Extremely rich people in New York live quite hap-
pily in apartments that would seem impossibly cramped by the standards
of Palo Alto. These apartments actually feel quite spacious when one is
in New York, simply because they are quite large relative to what other
people have. But because of this comparison, the desire to live in a spa-
cious home generates a prisoner™s dilemma. The only way to satisfy such
a preference is to buy a home that is of above average size, but when ev-
eryone does this, the average size creeps upwards. Thus, more resources
Joseph Heath

are invested in home construction and maintenance, while the increase
in satisfaction associated with the feeling of spaciousness is quickly
One could describe the problem here as a sort of market failure. If the
value of a good is based on a comparison with what others are consuming,
then all of the consumption choices made in this domain generates exter-
nalities for other consumers. (And, of course, if one de¬nes market fail-
ure as any circumstance in which markets do not achieve Pareto-optimal
outcomes, then it becomes true by de¬nition that these are all cases of
market failure.) The important point is that these effects can occur even
in markets that are well-structured, according to the standard economic
criteria “ where there is robust price competition, property rights are
well-de¬ned, there are few information asymmetries, and so on. As a re-
sult, the critique of competitive consumption can license interferences in
consumer sovereignty even in domains where markets are doing a good
job of producing goods in ample quantity. This may seem counterintu-
itive, until one recalls that it is not the sheer quantity of goods produced
by the economy that counts, from a normative point of view. What mat-
ters is the satisfaction that these goods generate. When consumers get
locked into competitive consumption, huge amounts of resources get fun-
neled in to feeding a competition that ultimately generates no increase
in satisfaction.
Frank™s solution to this problem mirrors Galbraith™s. His proposal is
for a progressive consumption tax (to be levied as a tax on income minus
savings). This is based on the observation that while increased income is
initially quite strongly correlated with increased happiness, the correla-
tion tapers off quickly once a certain basic level of need has been met. The
reasons for this must certainly be complex, but clearly one contributing
factor must be that most of the high-priority items on people™s general
preference schedules over consumer goods are items valued primarily for
their intrinsic properties, but that as one moves along to less important
items, the comparative dimension begins to loom larger.
Of course, there would be no point to such a tax if there were nothing
else that the money could be spent on (that is, nothing else that society
could commit its resources to). But this is not the case. Apart from the
market goods that people spend their money on, there is the full range of
goods that are not delivered through the market. Even if the progressive
consumption tax does no more than discourage work effort, it might
still generate an ef¬ciency gain. This is because people value the non-
market good of leisure. Thus a tax, insofar as it simply dampens down
Liberal Autonomy and Consumer Sovereignty 223

consumption in the upper disposable income range, will increase the
welfare of everyone “ including those who pay the bulk of it. The most
exclusive clubs will still be just as exclusive, and belonging to them will
still confer just as much prestige. It is just that they will cost less to join
(because everyone will have less money). The difference, however, is that
people will have more free time. Thus the satisfaction that people derive
from their consumption of private goods will be unaffected, while their
consumption of non-market goods will increase.

VII Conclusion
Critics of consumerism are constantly subjected to charges of elitism,
Puritanism, paternalism, and so forth. The reasons for this are not hard
to ¬nd. After all, given the absence of overt coercion in the supermarkets
and malls, criticizing consumerism must involve criticizing the choices
made by consumers. But it is dif¬cult to see why the social critic should
be better positioned than the consumer herself to decide what she should
be buying “ hence the suspicion that the critic is simply trying to meddle in
the private affairs of others. And since important features of our political
institutions were put in place precisely to prevent this sort of meddling,
it is no surprise that the critique of consumerism so easily runs afoul of
liberal democratic principles.
At the same time, it is very hard to avoid noticing that there is a gen-
uinely perverse inversion of priorities at work in the way our society al-
locates its resources (more money for SUVs, less money for schools . . . ).
Hence the problem facing the critic of consumerism: How to articulate
the perversity underlying these consumption choices, without simply at-
tributing it to the ignorance or venality of consumers? And how to formu-
late remedies that do not constitute illegitimate interferences with their
The clearest solution lies in the observation that the lopsided in-
vestment of resources into market over non-market goods constitutes a
Pareto-inef¬cient consumption pattern, generated by an underlying col-
lective action problem. Using the force of law to correct such collective-
action problems does not require any infringement on the autonomy of
consumers. If one takes the term “consumer sovereignty” to refer to a
blanket prohibition on legislative measures that limit individuals™ ability
to spend their income however they like, then the elimination of such a
collective-action problem will be inconsistent with consumer sovereignty.
Thus liberal autonomy and consumer sovereignty will not coincide “ in
Joseph Heath

fact, one might reasonable expect citizens to exercise their autonomy in
order to enact legislative measures that limit their own freedom of choice
as consumers.
Such measures may not do justice to all of the complaints that have
been made about the consequences of consumer sovereignty. It will do
nothing to encourage traditional virtues, like moderation in the satisfac-
tion of one™s desires. Nor will it do anything to cure the masses of their
inveterate bad taste. And it is very unlikely to diminish the popularity
of Big Macs and Coca-Cola. But the reason this kind of anti-consumerist
program “ the kind that diagnoses consumerism as a collective action
problem “ leaves these features of our society intact is that any effort to
use the force of law to remedy these supposed problems would be in-
consistent with the respect for autonomy and pluralism that is such an
important feature of civic life in democratic societies.

This statement of principle is, of course, merely the point of departure for
re¬‚ections on the role of autonomy in a liberal polity. I am using the term
autonomy here to refer only to what Forst calls legal autonomy. The chapters
in the present volume provide a good sense of the complexities that develop,
and the number of further distinctions that are required, when one begins
to explore the rami¬cations of this commitment in more detail.
William Harold Hutt, Economists and the Public: A Study of Competition and Opin-
ion (London: Jonathan Cape, 1936), p. 257.
John Kenneth Galbraith, The Af¬‚uent Society (Boston: Houghton Mif¬‚in,
Pope John Paul, Encyclical Letter: Centesimus Annus (1991), p. 75.
Daniel Bell, The Cultural Contradictions of Capitalism (New York: Basic Books,
Of course, there are some cases that are quite borderline. For instance, people
have a well-documented tendency to discount the future quite sharply, and
thus to adopt savings policies that are inconsistent with their own considered


. 7
( 13)