Icon-03.

qxd 11/12/02 11:06 AM Page 13

ARTICLE

The constitution, social rights, and

liberal political justification
Frank I. Michelman*

Proposals to put positive social and economic guarantees into constitutional law
typically meet an objection linked to judicial review (the “institutional objection”).
It seems that a related objection would hold even assuming away judicial
“enforcement.” The constitutionalized rights would, after all, be no less intended to
curb and constrain the choices of current majorities (the “majoritarian objection”).
Constitutionalization of social rights may, moreover, be thought to render the con-
stitution nontransparent in a way disallowed by leading liberal accounts of political
legitimacy (the “contractarian objection”).
On the table, then, are three possible objections to constitutionalization of social
rights—institutional, contractarian, majoritarian. The first, the author argues,
ought to be the least of our concerns; the second is manageable, or at any rate politi-
cal liberals cannot deny that it is while upholding in general the practice of constitu-
tionalism; and the third is grave only if we choose an ideal or normative conception
of democratic decisionmaking that is not the only one available to us, or the best one.

1. Introduction: the terms of debate

Whatever else it may also be, a country’s written constitutional bill of rights is
a high-ranking regulatory law, a “statute” fraught with direct legal conse-
quences. Granted, the constitution may not be “simply” that.1 No doubt it may
figure as something beyond positive law: “a ‘mirror reflecting the national
soul’,” perhaps; an expression of national ideals, aspirations, and values
expected, as such, to “preside and permeate the processes of judicial interpre-
tation and judicial discretion” throughout the length and breadth of the
national legal order.2 But had bills of rights not also and always registered as

* Frank Michelman is Robert Walmsley University Professor, Harvard University, USA. The author thanks
Norman Dorsen and Patrick Macklem for helpful comments.
1
S v. Acheson 1991 (2) SA 805 (Nm), 813A–B (1991 NR 1, 10A–B) (Mahomed AJ).
(“The Constitution of a nation is not simply a statute which mechanically defines the structures of
government and the relations between the government and the governed.”)
2
Id. Doubtless Justice Mahomed drew inspiration from the Lüth-jurisprudence of the German
Federal Constitutional Court. See the Lüth Case, 7 BVerfGE 198 (1958); VICKI C. JACKSON & MARK
TUSHNET, COMPARATIVE CONSTITUTIONAL LAW 1402–35 (Foundation Press, 1999). Cf Carmichele

© Oxford University Press and New York University School of Law 2003, 13
I.CON, Volume 1, Number 1, 2003, pp. 13–34
Icon-03.qxd 11/12/02 11:06 AM Page 14

14 F. I. Michelman

direct, regulatory legislation—as laws to be enforced like other laws—jurists

and scholars the world over would not have conducted their debates over the
constitutionalization of social rights in the terms that we have grown used to.
Constitutions, to be sure, are regulatory laws of a special kind, setting terms
and conditions for the making and execution of all other laws. Typically,
although not necessarily, some of the terms and conditions are cast in the form
of a bill of rights: a list of certain interests of persons, upon whom are conferred
what are considered to be legal rights, not just background moral claims,3 to
have these interests at least negatively respected,4 and maybe positively secured
and redeemed, by the state’s legislative and other actions yet to come.
Among the constitutional-legal rights thus conferred may be the rights to
satisfaction of certain material needs or wants, or of access to the means of
satisfaction. Take, for example, section 26 of the Constitution of South Africa:
1. Everyone has the right to have access to adequate housing.
2. The state must take reasonable legislative and other measures, within its
available resources, to achieve the progressive realisation of this right.5
Without a doubt, that declaration of a right that everyone has is seriously
meant. The surrounding constitutional text makes emphatically clear that sec-
tion 26 lays down a full-fledged constitutional right, no less a legal right than
any other declared by the Bill of Rights—no less intended to be obligatory
upon those to whom it is addressed than are legal rights in general and no
less intended to be subject to enforcement by some means from among those
available for the effectuation of legal rights in general.6
Opinion divides over whether it is a good idea thus to confer full constitutional-
legal status upon “social” rights guarantees of this kind. The division doubtless
stems in part from substantive disagreement. You will not support constitu-
tionalization of social rights guarantees unless you are convinced, at least pro-
visionally, that the claims for social support of individuals and families for
which such guarantees would speak are ones that no morally legitimate or
would-be successful political society can ignore. Obviously, not everyone
shares that view. Some do, though, and what I have to say here is addressed
mainly to them. I want to concentrate on possible nonsubstantive objections to
constitutionalization. Assume, then, for the sake of the argument, that one sees

v. Minister of Safety and Security 2001 (4) SA 938 (CC), ¶54 (Ackermann & Goldstone JJ) (“Our
Constitution is not merely a formal document regulating public power. It also embodies, like the
German Constitution, an objective, normative value system.”)
3
See RONALD DWORKIN, TAKING RIGHTS SERIOUSLY 101–02 (Harvard Univ. Press 1977) (distinguishing
between “institutional” rights and “background” norms).
4
See infra Part II.
5
Constitution of the Republic of South Africa, Act 108 of 1996, §26.
6
See id. §§ 2, 8(1), 38, 172(1).
Icon-03.qxd 11/12/02 11:06 AM Page 15

Constitution, social rights, and political justification 15

a persuasive prima facie moral and practical case for constitutionalization.7

Might one still hesitate, perhaps out of a regard for the proper workings of
constitutional-democratic political and legal institutions?
Those who say yes almost always cite concerns about over-extension of the
judiciary. Courts, they fear, will find themselves unable to make clear assessments
of the government’s compliance or noncompliance with social rights guarantees,
or to fashion apt and pointed remedial orders in case of a finding of noncompli-
ance, without getting themselves disastrously mixed up in matters beyond their
province and their ken as judges of the law. It may thus come to seem a bad idea
to invite the judiciary to tussle with the government over material resource allo-
cations and distributions. Of course, there is another side to the debate. Judges
who know their business, it is said, drawing lessons from administrative law, can
find both properly adjudicative standards for testing claims of social-rights viola-
tions and worthwhile, properly judicial remedies for violations when found.8
Whichever way one tilts, though, it is clear that the debate throughout has been
centered on a concern about the place and work of the judiciary in the demo-
cratic political order. We seem to think the problem with constitutionalizing social
rights comes down mainly, if not solely, to a matter of the separation of powers.
These terms of debate are inadequate. With a view to their modification—not
their total transformation—I shall suggest that there is more to the problem of
constitutionalized social rights than questions of judicial role and competence,
and also less.
There is less for two reasons, both of which have received attention in the past.
First, courts exercising constitutional review in entirely conventional, nonworri-
some ways almost certainly can play a useful role in the promotion of the distribu-
tive aims of social rights guarantees.9 Second, even were that not true and the
choice therefore had to be made to bar courts entirely from review of government
action for compliance with social rights guarantees, that would not be a good
argument against constitutionalization in the sight of anyone who believes that
a morally legitimate political regime must include a visible, effective commitment
to certain forms of positive social support for individuals and families.10
Assuming both those points are accepted, the case for constitutionalization
of social rights is not yet fully made. Two possible grounds for hesitation

7
See infra Part IV.
8
See Cass R. Sunstein, Social and Economic Rights? Lessons from South Africa, 11 CONSTITUTIONAL FORUM
123 (2001). For recent supporting evidence (received as this goes to press), see Minister of Health v.
Treatment Action Campaign (Unreported, South African Constitutional Court, July 5, 2002), available
at http://www.concourt.gov.za/cases/2002/tacsum.shtml. See also the exchange in South Africa
between Etienne Mureinik and Dennis Davis. Etienne Mureinik, Beyond a Charter of Luxuries: Economic
Rights in the Constitution (1992) 8 SAJHR 464; Dennis M. Davis, The Case Against the Inclusion of Socio-
economic Demands in a Bill of Rights Except as Directive Principles (1992) 8 SAJHR 475.
9
See infra Part 2.
10
See infra Parts 3 and 4.
Icon-03.qxd 11/12/02 11:06 AM Page 16

16 F. I. Michelman

remain, even for those who are persuaded morally of reasons to go ahead and
fear no resulting evil of judicial overreaching. These are, first, a democratic
objection (as I shall call it) to the effect that adding social rights to the constitu-
tion constricts democracy unduly, regardless of judicial involvement in the
enforcement of such rights; and, second, a contractarian objection to the effect
that adding social rights to the constitution defeats a crucial function of the
constitution-as-law, that of providing legitimacy to the coercive political and
legal orders. Having put those two possible further objections on the table,
I shall suggest that (1) their force may vary with how sweepingly—or,
conversely, specifically—the constitutional social rights guarantees are
couched; (2) the contractarian objection is manageable even with a maximally
sweeping constitutional guarantee, or at any rate political liberals cannot deny
that it is while upholding in general the practice of constitutionalism; and (3)
the democratic objection is grave only if we choose a normative conception of
democracy that is not the only one available to us, or the best one.

2. Social rights and conventional judicial review

As we have noticed, proposals to put positive social and economic guarantees
into constitutional law typically are met by an objection linked to the expecta-
tion of judicial review. By constitutionalizing social rights, the argument often
has run, you force the judiciary to a hapless choice between usurpation and
abdication, from which there is no escape without embarrassment or discredit.
One way, it is said, lies the judicial choice to issue positive enforcement orders
in a pretentious, inexpert, probably vain but nevertheless resented attempt to
reshuffle the most basic resource-management priorities of the public house-
hold against prevailing political will. The other way lies the judicial choice to
debase dangerously the entire currency of rights and the rule of law by openly
ceding to executive and parliamentary bodies an unreviewable privilege of
indefinite postponement of a declared constitutional right.
The objection quite obviously is overstated. Constitutionalization of social-
rights guarantees can provide both a prod and a hook for ho-hum forms of
judicial action in furtherance of the distributive aims these rights represent.
The fact that social rights make budgetary demands, or call for government
action and not just forbearance, does not in itself differentiate them radically,
from the standpoint of justiciability, from constitutionally protected rights to
property, to equality before the law, or to so-called negative liberties.11 At “the

11
See Ex Parte Chairperson of the Constitutional Assembly: In re Certification of the Constitution
of the Republic of South Africa, 1996 (4) SA 744 (CC), ¶78; STEPHEN HOLMES & CASS R. SUNSTEIN,
THE COST OF RIGHTS: WHY LIBERTY DEPENDS ON TAXES (W. W. Norton & Company 1999); Frank I.
Michelman, Welfare Rights in a Constitutional Democracy, 1979 WASH. U. L.Q. 659 [hereinafter
Welfare Rights]; Frank I. Michelman, Foreword: On Protecting the Poor through the Fourteenth
Amendment, 83 HARV. L. REV. 7, 17–18, 25–26 (1969).
Icon-03.qxd 11/12/02 11:06 AM Page 17

Constitution, social rights, and political justification 17

very minimum,” social rights can sometimes be “negatively protected” by

comfortably kosher forms of judicial intervention.12 For example, municipal
zoning and land-use laws, insofar as they constrict the local housing supply or
escalate its cost, quite plausibly lie open to challenge under the South African
Constitution’s guarantee to everyone (in section 26(1)) of a right to access to
adequate housing.13 It seems a court would be expected to test such a law
under the standard laid down by section 36 of the Constitution, which permits
legislative limitation of rights in the bill of rights, but only by lawmaking that
is “reasonable” as well as “justifiable in an open and democratic society based
on human dignity, equality, and freedom.” “Reasonableness” is not, in any
legal discourse known to me, a nonjusticiable standard (or are negligence law
and general clauses beyond the pale?), and the remedy for violation, if found,
would be a simple prohibitory injunction.
A recent South African case suggests how a court may act usefully in
furtherance of a constitutional social rights guarantee by the most conventional
of all forms of judicial action, namely, dismissal of a case (where relief would
have been forthcoming but for the guarantee). In Minister of Public Works v.
Kyalami Ridge Association,14 the government proposed to relocate homeless
and destitute flood victims to housing it would cause to be built on state-owned
land. Neighboring homeowners sued for an interdict against this plan on the
ground, among others, of ultra vires, pointing out that no act of parliament
authorized the government to build housing for the purpose in question. The
Constitutional Court allowed the neighbors standing to lodge the ultra vires
claim, agreed with them that the constitutional conception of the rule of law
requires positive legal authorization for any government activity that disturbs

12
See In re Certification of the Constitution, 1996 (4) SA at 801. See generally Welfare Rights, supra
note 11, at 660–64, 686–93.
13
See supra note 5; Government of Republic of South Africa v. Grootboom, 2001 (1) SA 46 (CC),
¶ 34 (“Although [section 26(1)] does not expressly say so, there is, at the very least, a negative
obligation placed upon the state and all other entities and persons to desist from preventing or
impairing the right of access to adequate housing.”).
In Minister of Health, supra note 8, the government had refused, for the time being, to supply an
antiretroviral drug, navirapine, to all of the country’s public hospitals and clinics where attending
physicians might dispense it in appropriate circumstances to women about to give birth and their
newborns, despite (1) the drug’s having been approved as safe through the normal South African
governmental channels, (2) its efficacy in reducing mother-to-child transmission of HIV/AIDS even
when administered without ancillary precautions by a simple, nonintrusive means, and (3) the sup-
plier’s standing offer of an unlimited supply to the government, free of charge. In holding this refusal
unreasonable, hence unconstitutional, the Court at some points seemed to regard it as tantamount
to a state-imposed prohibition—an active interference by the state with the freedom of physicians
and their patients to make use of navirapine—while at other points regarding it as a failure by the
government to take reasonable positive “measures within available resources,” as required by the
Constitution of the Republic of South Africa, supra note 5, § 27(2), to secure everyone’s right of
“access to health care” guaranteed by § 27(1)(a). See, e.g., Minister of Health, supra, at ¶¶ 46, 135.
14
2001 (3) SA 1151 (CC).
Icon-03.qxd 11/12/02 11:06 AM Page 18

18 F. I. Michelman

the interests of others, and agreed further that no act of parliament answered
the need in this instance. The Court, however, denied the plaintiffs the relief
they sought. In effect, it found authorization for the government’s action in
section 26 of the Constitution, as construed in Government of Republic of South
Africa v. Grootboom15 to place the government under a duty to make reasonable
provision within its housing plans for short-term aid to persons blamelessly in
conditions of housing crisis. “The government,” wrote the Court in Kyalami,
“contends that these obligations require it to come to the assistance of the
victims of the flooding throughout the country, . . . and that in doing so it
cannot be said to be acting contrary to the rule of law.”16
The Court apparently agreed. Like other owners of land, the Court reasoned,
the government has in general the right to erect buildings on land it owns. Alone,
that general property-owners’ right cannot satisfy the rule-of-law demand for
positive legal authorization for government action that trenches upon others’
interests, but the Court found that it can in combination with section 26 of the
Constitution as construed in Grootboom. If the government asserts its property-
owner’s building right “within the framework of the Constitution and the restric-
tions of any relevant legislation,” the Court concluded, “it acts lawfully.”17

3. Laws and remedies

Suppose we have a constitutional norm, N. Maybe, for some reason, we do not
expect or wish our judiciary to get too mixed up with enforcing compliance
with N. But still we do want to say that N is meant to be fully binding, obliga-
tory, on those state officials to whom it is addressed. Indeed we want to say
N is binding in just the ways—whatever we think they are—that laws in gen-
eral are understood to be binding even at moments when they are not being
externally enforced.18 We want to say not only that the addressees are not
meant to have a free choice about heeding N, but further that an addressee
who consciously disregards N, without special excuse or justification, is blam-
able for contempt of the rule of law in the same way that anyone who flouts
the law is blamable. How can we say these things, if we cannot call N a law?
And yet we may not do so, if the judicial-involvement-based objection
to constitutionalization of social rights is sound. The objection’s unstated

15
2002 (1) SA 46 (CC)
16
2001 (3) SA at ¶ 39.
17
Id. at ¶ 40.
18
Lawrence Sager, most prominently among American constitutional legal scholars, has
explained at length the reasons for taking this combination of positions. See Lawrence Sager,
The Domain of Constitutional Justice, in CONSTITUTIONALISM: PHILOSOPHICAL FOUNDATIONS 235 (Larry
Alexander eds., Cambridge Univ. Press 1998), Lawrence Sager, Justice in Plain Clothes: Reflections on
the Thinness of Constitutional Law, 88 NW. U. L. REV. 410 (1993).
Icon-03.qxd 11/12/02 11:06 AM Page 19

Constitution, social rights, and political justification 19

premise is that to classify an obligation as one of constitutional law is ipso facto

to make the obligation one that the judiciary is responsible to enforce to the hilt
against a recalcitrant government. It seems that by naming something a con-
stitutional right you also name it a legal right and so willy-nilly make the judi-
ciary responsible for its effectuation. The objection thus strictly conditions a
norm’s dignity as law on its availability for judicial enforcement. Is that what
we believe?19
A recent Massachusetts case20 puts our belief neatly to the test. The
Massachusetts Constitution, article 48 of the amendments, provides for law-
making by popular initiative. The elected state legislature may choose to repeal
a law thus “approved by the people” in a statewide vote, but insofar as the le-
gislature does not do so it stands under article 48’s express command to “raise
by taxation or otherwise” and to “appropriate” whatever funds may be needed
to carry the law into effect.21
In 1998, acting pursuant to article 48, a large majority of Massachusetts
voters approved a “Clean Elections” law providing for sizeable payments of
state funds to candidates for state elective office who undertake to limit their
fundraising from private sources and who demonstrate fulfillment of certain
other qualifying tests.22 The law makes these payments collectible from a
designated state official, the Director of the Office of Campaign and Political
Finance (“director”)—adding, however, that the director’s obligation to pay is
“subject to appropriation.”23 “Appropriation” evidently refers to an act of the
legislature, on recommendation of the Governor, releasing state-owned funds
for expenditure by one or another state office or official on specified objects
during a specified fiscal period.24

19
It plainly is not what we have always believed. See Christine A. Desan, Contesting the Character of
the Political Economy in the Early Republic: Rights and Remedies in Chisolm v. Georgia, in THE HOUSE
AND SENATE IN THE 1790S: PETITIONING, LOBBYING, AND INSTITUTIONAL DEVELOPMENT 178 (Kenneth R.
Bowling & Donald R. Kennon eds., Ohio Univ. Press 2002) (describing the strictly legislative remedy
for contract claims against the state in the early Republic).
20
Bates v. Dir. of Campaign & Political Fin., 763 N.E.2d 6 (Mass. 2002).
21
MASS. CONST. amend. art. XLVIII, § 2.
22
The Massachusetts Clean Elections Law, Mass. Gen. Laws Ann. ch. 55A (West Supp. 2001).
23
Id. §§ 1, 7.
24
See MASS. CONST. AMEND. ART. LXIII (setting procedures for the annual budget and appropriations).
The Massachusetts Constitution expressly prohibits any “issuance” of “money” out of the state’s
“treasury” except under warrant from the executive branch drawn “agreeably to the acts and
resolves of the general court [i.e., the state legislature].” MASS. CONST. PT. 2, C. II, § 1, ART. XI. Final
resolution of the Bates case, which has not yet occurred as this goes to press, appears likely to carry
some lessons regarding the scope of this prohibition. See Memorandum of Decision and Order on
Plaintiffs’ Motion for Relief (March 12, 2002); Memorandum of Decision and Order on Plaintiffs’
Emergency Motion for a Levy on Property of the Commonwealth (April 5, 2002), both available at
http://www.nvri.org/library/index.shtml#Massclean. See also infra note 33, and accompanying text.
Icon-03.qxd 11/12/02 11:06 AM Page 20

20 F. I. Michelman

Despite an obvious strong dislike of the Clean Elections Law on the part of
most of its members and leaders, the legislature had trouble mustering the
super-majority of open votes against it to sustain a repeal over a promised
gubernatorial veto. The statute thus remained on the books as Massachusetts
politicians worked their way deep into the state election cycle of 2001–02. The
legislature also, however, refused (or at any rate failed) to appropriate funds to
meet the payment obligations that were beginning to accrue according to the
terms of the law.
Various plaintiffs sued for a mandatory injunction to the director to make
the distributions the law provides for—leaving open, however, the question of
how the director might manage to comply in the absence of any covering
appropriation.25 The plaintiffs included Warren Tolman, a candidate for the
office of Governor who had applied for and received the director’s certification
of compliance with the law’s prerequisites for an initial collection of money.
The case quickly reached the Supreme Judicial Court of Massachusetts.
Speaking through Chief Justice Margaret Marshall, the Court’s majority
had little difficulty concluding that the legislature, having not repealed the
Clean Elections Law, stood in breach of a clear constitutional and legal man-
date to enact statutes making sufficient appropriations from the state treasury
to fund the payments that the director is required by the law to make to quali-
fied candidates.26 But appropriation there was none, and that lack precluded
any injunctive order to the director to pay: “Because the director currently has
no clean elections funds to distribute and no authority on his own to reach
those funds, he cannot be ordered to distribute them.”27 The Court refrained
from any suggestion that a mandatory judicial remedy might lie against the
legislature to make the requisite appropriation. Whether the Court thus for-
bore because the plaintiffs had not dared to request such a remedy, or because
it read the Constitution expressly to preclude it,28 or because it believed such a
remedy would abrogate an unwritten rule of separation of powers, one cannot
definitely say.
The case might have ended just there, had the plaintiffs had naught but
public law—the direct commands of article 48—with which to back their
claims of entitlement to be paid out of Commonwealth assets. Having per-
formed its “ ‘imperative duty’ to say what the Constitution requires,”29 and
having accordingly affirmed the legislature’s constitutional-legal obligation to

25
If you were the state’s ministerial officer to whom the director issued his warrant to pay, would
you write the check? If the check were written on a state-owned account in your bank, would you
honor it?
26
763 N.E. 2d at 23–24.
27
Id. at 28.
28
See supra note 24.
29
763 N.E. 2d at 24.
Icon-03.qxd 11/12/02 11:06 AM Page 21

Constitution, social rights, and political justification 21

raise and appropriate clean-election funds,30 it seems that the Court—but for its
detection of the private-law claim soon to be described—might have left the rest
to the legislature acting under the gaze of the voters, adding its own vote of con-
fidence in the legislature’s law-abidingness.31 A court’s thus leaving to politicians
and voters the implementation, or “enforcement,” of a legal obligation would
have been unusual in our times, but so is the case unusual. On the Supreme
Judicial Court’s own advice, ubi ius, ibi remedium is not an exceptionless maxim;32
conversely, an obligation laid down by positive legislation would not cease to be a
legal one just because breaches are found to be judicially nonremediable.
As matters finally turned out, one of the plaintiffs—Tolman—was awarded
a judgment against the director, satisfiable by execution levy against tangible
assets of the Commonwealth, but so far not including treasury cash.33 That
judgment was not, however, imposed in respect of any breach—although the
Court found a grave one—of a public-law obligation imposed directly by
article 48. Rather, it was based on a private-law claim of, roughly, unilateral
bargain-contract or promissory estoppel, arising out of certain “affirmative
actions” taken by the director, by which he induced changes in position by
candidates, detrimental to them and beneficial to the Commonwealth.34
The clean-elections controversy in Massachusetts may thus in fact directly
instance the class of positive enactments found by a court to raise obligations
of law, which nevertheless defy judicial enforcement. That is exactly how the
Supreme Judicial Court treated the plaintiffs’ claims of obligations in their
favor imposed directly by public law (article 48)—those being primarily
the legislature’s obligation to raise and appropriate funds and secondarily the
director’s obligation, hinging on the legislature’s, to pay specified sums of
money to certified candidates out of funds appropriated for the purpose.

30
Id. at 29.
31
Id. But compare the very different view of the legislature’s posture in Memorandum of Decision
and Order on Plaintiffs’ Emergency Motion for a Levy on Property of the Commonwealth, supra
note 24.
32
See id. at 24. (“Not every violation of a legal right gives rise to a judicial remedy.”)
33
See Memorandum of Decision and Order on Plaintiffs’ Motion for Further Relief, supra note 24.
The plaintiffs have appealed to the full Court from this order of a single justice denying access to
treasury cash. Their brief is available at http://www.nvri.org/library/index.shtml#Massclean.
34
763 N.E. 2d at 11. See id. at 24–26, 30–31. The Court’s private-law theory of candidates’ enti-
tlement is somewhat contentious because the state’s offer to pay qualifying candidates—assuming
that an “offer” to potential qualifiers is what it was, see, e.g., RESTATEMENT (SECOND) OF CONTRACTS
§ 24 (1981) (characterizing an offer as “a manifestation of willingness to enter into a bargain, so
made as to justify another person in understanding that his assent to that bargain is invited and
will conclude it”)—was expressly made “subject to appropriation.” This fact would seem to bear on
both (i) whether all of the offer’s stated preconditions to liability had in any case been met and (ii)
the reasonableness of any candidate’s reliance during a time when, notoriously, there had as yet
been no appropriation. See 763 N.E.2d at 37–38 (Spina J., dissenting).
Icon-03.qxd 11/12/02 11:06 AM Page 22

22 F. I. Michelman

4. The constitutional-contractarian case for

social rights
I have been positing what may be found an extreme position: that a normative
proposition can be law despite total, flat-out exclusion of the judiciary from
enforcing it. I do so for the sake of the argument. For anyone who thus severs
the conceptual bond tying constitutionalization of a social-rights guarantee to
a demand for direct judicial enforcement,35 judicial-role concerns no longer
can support opposition to constitutionalization. My aim is to consider whether
either the democratic or the contractarian objection may still support it. But in
order to appraise objections to a practical proposal, you must first have in mind at
least the broad outlines of some general case in its favor. Here the proposal is to
confer the status of a constitutional legal right on claims of individuals to certain
forms of support and provisioning by the state. I shall now sketch out a general
affirmative case for this proposal. This will be a Rawlsian sort of a case, a political
liberal’s sort of a case, indeed a contractarian liberal’s sort of a case—no doubt not
the only sort that might be fashioned, but one that you won’t find outlandish and
one that sets up most clearly both the majoritarian and the contractarian objec-
tions to giving social rights a place in the constitution.
The general case I shall present for constitutional social rights is quite
specifically limited to the question at hand. It is not a case for a general moral
obligation on the part of every separate person having the means to do so to
come to the aid of sundry others who find themselves in need. It is only and
strictly a case for including social rights in a country’s constitutional bill of
rights. I suggest there exists at least one appealing argument to the conclusion
that a constitution ought, as a moral matter, to affirm the claims of individuals
to be assured of satisfaction of certain material needs, by the state if necessary,
on reasonable conditions of effort and cooperation. I suggest the argument
holds regardless of whether we believe there exists any other sort of moral
right to succor from one’s neighbor. I suggest, in other words, that there is
something about the particular moral point or purpose of constitutions, from
which it follows that a constitution, in particular, is morally defective—it fails
of its moral purpose—by reason of its lack of social-rights guarantees; and
that this is so regardless of what any of us individually, outside the context of
politically governed society, may or may not owe morally to others in the way
of aid or support.36
The argument I have in mind stands on a certain foundational commitment
to positive legal ordering, presupposed by constitutionalism and hence by the
question of constitutionalizing social rights. In any land where constitutional-
ism prevails, people wake up each day to find in place effectively compulsory
regulations of social life—we call them “laws”—with which the publicly

35
On the possibility of indirect modes of judicial vindication, see supra Part 2.
36
Cf CHARLES FRIED, RIGHT AND WRONG Ch. 5 (Harvard Univ. Press 1979).
Icon-03.qxd 11/12/02 11:06 AM Page 23

Constitution, social rights, and political justification 23

supported authorities in the land predictably will demand everyone’s compliance,

and, in doing so, predictably will have most everyone’s support. No one who
thus is subject to the laws of a country has chosen these laws for himself. In a
democratic country, the laws normally will have been decided in voting proce-
dures by which majorities rule over dissenters. Whatever may be the precise
history of how a democratically governed country’s laws came to be what they
are, it will not be that any law effectively was chosen by the actions of any
single one of the individuals who are called upon to abide by it.37
There arises thus the question of political justification or legitimacy. The
challenge is to supply a moral warrant for the application of collective force
in support of laws produced by nonconsensual means, against individual mem-
bers of a population of presumptively free and equal persons. For countries
under democratic rule, this means, as John Rawls has expressed it, to explain
how “citizens [may] by their vote properly exercise their coercive political power
over one another”—to explain how your or my exercises of our shares of polit-
ical power may be rendered “justifiable to others as free and equal.”38 Rawls has
offered in response what he calls the “liberal principle of legitimacy”:
Our . . . political power is . . . justifiable [to others as free and equal] . . .
when it is exercised in accordance with a constitution the essentials of
which all citizens may be expected to endorse in the light of principles
and ideals acceptable to them as reasonable and rational.39
That exemplifies perfectly what we may call a constitutional-contractarian
approach to the question of political justification. At its core stands an uncom-
promising moral concern with the freedom and equality of each and every
individual. From that concern flows the demand that potentially coercive polit-
ical acts be acceptable from the standpoints of each (not “all,” in some collec-
tivized sense of “all”) of countless persons among whom rational conflicts of
interest and vision abound. Acceptable, that is, in principle. Acceptable in all
reason; acceptable in the sight of whoever applies the test of acceptability.
In Rawls’s formulation, political coercion is justified when it is exercised in
support of laws issuing from a constitutional regime, on condition that the
regime is one that all may be expected to endorse, assuming everyone to be not
only rationally self-interested but also “reasonable.” “Reasonable” here means
three things. First, a reasonable person accepts the inevitability of positive
legal ordering. She doesn’t pretend we somehow are going to get along with-
out lawmakers making laws that have to bind everyone regardless of who in
particular likes each law and who does not. Second, a reasonable person

37
See FRANK I. MICHELMAN, BRENNAN AND DEMOCRACY 14–16, 31–33 (Princeton Univ. Press 1999).
38
JOHN RAWLS, POLITICAL LIBERALISM 217 (Columbia Univ. Press 1996) [hereinafter POLITICAL
LIBERALISM].
39
Id.
Icon-03.qxd 11/12/02 11:06 AM Page 24

24 F. I. Michelman

accepts the fact of deep and enduring conflicts of interests and ethical visions
within her society—what Rawls calls the fact of reasonable pluralism.40 Third,
she is imbued with the liberal spirit of reciprocal recognition by persons of
each other as individually free and equal. As a result, a reasonable person
stands ready to accept the laws as long as (a) she sees everyone else generally
supporting and complying with these laws, and (b) she sees how these laws are
ones that merit mutual acceptance by a competently reasoning group of per-
sons, all of whom desire, and suppose each other to desire, to devise and to
abide by laws reflecting fair terms of social cooperation in conditions of deep
and enduring but reasonable disagreement over questions of the good.41
But wait a minute. No matter how reasonable we ask each other to be,
surely none of us really expects that every discrete act of lawmaking could
pass a test of rational acceptability to every supposedly reasonable inhabitant
of a modern, plural society. Realistically, our hope is more modest, and
more procedural. It is that an aptly designed general system or regime for
lawmaking—or call it a constitution—might be able to pass such a test. Maybe
we can imagine some such regime, about which we’d be prepared to say that it
ought to be found acceptable, as a regime, by every rational person who is also
reasonable. If so, then we might further maintain that the rational acceptabil-
ity to you, as reasonable, of the constitutional regime commits you to accept-
ance of whatever specific laws may issue from the regime.42 That, after all, is
the apparent point of Rawls’s claim, in his “liberal principle of legitimacy,”
that exercises of political coercion are justifiable insofar as they issue from
“a constitution, the essentials of which all citizens may be expected to endorse.”
In effect, we have arrived at the idea of a sufficient, legitimating constitutional
agreement.43 Four terms compose this idea, as follows.
First, what is supposed to be “legitimated” (in the sense of justified morally)
by this agreement is some specific practice of positive legal ordering—of the
coercive exercise of collective power, through lawmaking, by and among
citizens considered as individually free and equal.

40
See id. at 36–37.
41
See id. at xliv, xlvi, 226–27; See also John Rawls, The Idea of Public Reason Revisited, in COLLECTED
PAPERS 573, 576–79, 581, 605–06 (Samuel Freeman ed., Harvard Univ. Press 1999).
42
See Samuel Freeman, Original Meaning, Democratic Interpretation, and the Constitution, 21 PHIL.
AND PUB. AFF. 3, 26, 36, cited with approval in POLITICAL LIBERALISM, supra note 38 at 234 n. 20;
Cf. JOHN RAWLS, A THEORY OF JUSTICE 195–201 (Harvard Univ. Press 1993) (on the “four-stage
sequence”). This means that your finding particular ordinary laws unjust gives you no ground for
resort to unlawful force, not that it gives you no ground for denunciation, civil disobedience, or
conscientious refusal.
43
I develop and examine this idea at greater length in Frank I. Michelman, The Problem of
Constitutional Interpretive Disagreement: Can “Discourses of Application” Help? in HABERMAS AND
PRAGMATISM 113 (Mitchell Aboulafia, Myra Bookman & Catherine Kemp eds., Routledge Press
2002).
Icon-03.qxd 11/12/02 11:06 AM Page 25

Constitution, social rights, and political justification 25

Second, what is supposed to have the desired legitimating effect is agreement

by each person affected. Not, however, actual agreement but hypothetical
(what some would call “counterfactual”) agreement—the “acceptability” of the
political practice among persons affected by it, envisioning those persons not
only as rational but also as reasonable.
Third, the legitimating hypothetical agreement is a constitutional agree-
ment. We don’t apply the universal-reasonable acceptability test to each and
every specific law that crops up in a country’s politics. We rather apply it to the
country’s system for lawmaking.
Lastly, then, “sufficiency.” In order to meet the test of rational acceptability
to every reasonable person, a lawmaking system has to include a principle or
guarantee affecting every topic for which a rational person, responding
reasonably, would demand a guarantee as a condition of willing support for the
system as a whole. (For present purposes—and here I depart a little from Rawls’s
usage of the term—we may say that the set of “constitutional essentials” is
equivalent to the set of minimally required principles and guarantees.)44 The set
must be extensive enough to compose a system for political decisionmaking
about which every affected, supposedly reasonable person rationally can say:
“A system measuring up to these principles and terms—all of them—is sufficiently
regardful of my and everyone’s interests and status as free and equal persons that
I ought in all reason to support it and its legislative products, provided everyone
else does.”
The makings of a general contractarian case for constitutionalized social
rights are now before us. How can we reasonably call on everyone, as reason-
able but also as rational, to submit their fates to a democratic-majoritarian
lawmaking system, without also committing our society, from the start, to run
itself in ways designed to constitute and sustain every person as a competent
and respected contributor to political exchange and contestation and further-
more to social and economic life at large?45 If we cannot do so, then no
“constitutional agreement” is a “sufficient” one if it lacks all trace or token of
such a commitment. It thus seems that social rights guarantees of some kind
would have to appear in a legitimate liberal-democratic constitution.

5. The democratic objection

We may leave the affirmative case for constitutionalization of social rights
resting just there for the time being, while we now take up for consideration

44
For Rawls’s discussion of “constitutional essentials,” see POLITICAL LIBERALISM, supra note 38, at
227–30.
45
The argument can be cast as well in parallel terms of membership, commitment, and identity.
See William E. Forbath, Constitutional Welfare Rights: A History Critique and Reconstruction, 69
FORDHAM L. REV. 1821, 1875–76 (2001) [hereinafter Constitutional Welfare Rights].
Icon-03.qxd 11/12/02 11:06 AM Page 26

26 F. I. Michelman

the democratic objection. It is raised in dramatic form by a proposal from

William Forbath.46 Forbath asks us to conceive of constitutionalized social
rights not as what he calls “welfare rights” but as what he calls “social citizen-
ship” rights. Roughly, his idea is this. We assume the general affirmative case
for constitutionalized social rights to be something like the one I have sug-
gested, in which the motivating moral ideal is that of a society committed to
run itself in ways designed to constitute and sustain every person (at least
every person who so chooses on fair terms) as a competent and respected con-
tributor to political, social, and economic life. Accordingly, argues Forbath, the
key universal right of persons—at any rate the key universal interest of
persons—would be assurance that one can make a respectable living through
forms of social participation that are themselves a source and support of
satisfaction, energy, pride, and social respect.47
This would be, then, a conception in which a central concern would be
interests related to work: the availability and terms of work, the character of
work, the organization and governance of work.48 Corresponding to such
a conception of a universal interest would be what Forbath calls a “social
citizenship” conception of constitutionally guaranteed social rights,49 to be
contrasted with a “welfare right” conception focusing on guarantees of money
income or of access to specifically listed, basic material necessities regardless of
work.50 Section 26 of the South African bill of rights exemplifies a somewhat
amphibian constitutional social right that Forbath probably would classify as
predominantly a welfare right, not a social-citizenship right.51
The two forms of imaginable constitutional guarantees differ drastically in
the apparent breadths of their respective, potential applications to the policy
choices of legislatures. Consider again the South African Constitution’s man-
date to the government (laid down by section 26(2)) to take reasonable meas-
ures, within available resources, to achieve progressive realization of the right
of every South African (declared by section 26(1)) to have access to adequate
housing. In Grootboom, the Constitutional Court found that the state’s housing

46
Id. at 1821–91; William E. Forbath, Caste, Class, and Equal Citizenship, 98 MICH. L. REV. 1 (1999).
47
See, e.g., Constitutional Welfare Rights, supra note 45, at 1875–77.
48
Id. at 1824, 1833–35.
49
Id. at 1826, 1876.
50
Id. at 1854, 1871–72.
51
Following my cue, Forbath has called it a welfare right pure and simple. See id. at 1880.
However, we both may have spoken too soon. A right “to have access to adequate housing” might
plausibly be deemed fulfilled for anyone to whom work is available on fair terms, including pay
sufficient for procurement of adequate housing that is reasonably available and suitable.
Cf Grootboom, 2001 (1) SA 46 (CC), ¶¶ 36–37. (“For those who can afford to pay for adequate
housing, the state’s primary obligation lies in unlocking the system, providing access to housing
stock and a legislative framework to facilitate self-built houses through planning laws and access
to finance.”)
Icon-03.qxd 11/12/02 11:06 AM Page 27

Constitution, social rights, and political justification 27

measures failed to be “reasonable,” as required by section 26(2), specifically

because of their virtually total inattention to the matter of emergency relief of
persons blamelessly placed in conditions of housing crisis. The Court
demanded submission of revised plans to judicial scrutiny for their reason-
ableness, while making clear that no court would try to dictate solutions. This
outcome may pose stern tests to judicial skill and wisdom in future cases, but
it does not as it stands seem shockingly pre-emptive of legislative and executive
policy choice.52 Certainly it is no more so than the American Constitution’s
guarantee of freedom of expression, which has been construed to restrict quite
sharply a remarkable—in a world-wide view an amazing—range of legislative
policy choices respecting such varied and weighty matters as fomentation of
group hatred,53 civil rights legislation aimed at public civic equality,54 the flow
of money in politics,55 and the legal protection of reputation and persona.56
Matters seemingly would be different under a Forbath-style constitutional
guarantee of social citizenship. To see why, one need only take in Professor
Forbath’s approving summary of the lawmaking topics a social-citizenship
right was thought to cover by its late nineteenth-century proponents in the
United States:
. . . freeing [labor] from “the iron rule of the Money power” through
public credit and support for cooperative enterprise; . . . nationalizing the
railways; . . . ensuring for industrial workers the “right to a remunerative
job” through public works and countercyclical spending and, through
an end to the repressive common law constraints on workers’ collective
action . . . , encouraging robust unions and industrial cooperation; and
through these agencies . . . enabling workers to exercise the rights and
responsibilities of control over collective property.57
To which we in our own times may add: tax laws and policies; publicly
guaranteed education and training for all, of adequate quality; infant, child, and
elderly care; workplace health and safety, fair employment, wage and hour laws;
global trade issues, the World Trade Organization and so forth; macroeconomic

52
Accord, Minister of Health, supra note 8, at ¶ 38. See SUNSTEIN, supra note 8; Mureinik, supra note
8 (anticipating the Court’s focus on “reason” as a justiciable standard of review). See also Frank I.
Michelman, The Constitution, Social Rights and Reason: A Tribute to Etienne Mureinik, (1998) 8
SAJHR 499, 500–1.
53
See R.A.V. v. City of St. Paul, 505 U.S. 377 (1992); American Booksellers v. Hudnut, 771 F.2d
233 (7th Cir. 1985), aff ’d, 475 U.S. 1001 (1986).
54
See Boy Scouts of America v. Dale, 530 U.S. 640 (2000).
55
See, e.g., Buckley v. Valeo, 424 U.S. 1 (1976).
56
See, e.g., Hustler Magazine v. Falwell, 485 U.S. 46 (1988). See generally Paul D. Carrington, Our
Imperial First Amendment, 34 U. RICH. L. REV. 1167 (2001).
57
Forbath, supra note 46, at 49.
Icon-03.qxd 11/12/02 11:06 AM Page 28

28 F. I. Michelman

policy and controls; public oversight of industrial organization, including

antitrust and other legal counters to restraints of trade; anti-plutocratic polit-
ical institutions and practices including campaign finance regulation; and
I’m sure I’ve left a lot out.
In sum, it looks as though a constitutional social-citizenship right has
tentacles reaching in a hundred directions, into the deepest redoubts of the
common law and the most basic choices of political economy a modern soci-
ety can make. Abortion aside—if it is aside, which it very arguably is not58—I
can think of no leading issue on the current American political calendar that
a constitutional right of social citizenship would leave untouched.
Now suppose you think that a constitution’s list of guaranteed rights is also
its demarcation of the respective zones of supremacy of the judicial and other
branches of government. In other words, you think that constitutional rights
are for courts and always for courts to apply or put into action. Legislatures,
you think, are supposed to make their policy choices more or less oblivious of
constitutional law, then courts come along and police their actions for consti-
tutional compliance. Every proclamation of a constitutional right thus invites
the judiciary to add some further sphere or spheres of public decisionmaking
to the ones in which it already feels licensed to take a sometimes heavy hand.
If that is how you see matters, you very well may think that constitutionaliz-
ing a right of social citizenship, as Forbath conceives it, is a way of turning
over to an unelected judiciary a share of control over policymaking that is far
too extensive to be tolerable in a democracy.
We do not have to accept the view that a norm can’t be constitutional law,
can’t count as constitutional law, without its being turned over to judges for all-
out enforcement. Many of us indeed will be primed to deny that constitutional
law enforced by judges has to be all the constitutional law there is or that mat-
ters.59 We maintain that constitutional law outside the courts can figure import-
antly in the conduct of public affairs. We insist that contention outside the courts
over constitutional-legal meanings and obligations very possibly can be a politi-
cally cogent activity, a site for democracy in action.60 Is there any reason why we
who take this view should hesitate to embrace a social-citizenship conception of
constitutional social rights, in preference to a welfare-right conception, assuming

58
See, e.g., Planned Parenthood v. Casey, 505 U.S. 833, 927–28 (1992) (Blackmun, J., concurring
in the judgment in part, and dissenting in part) (“The decision to terminate or continue a preg-
nancy has no less an impact on a woman’s life than decisions about contraception or marriage.
Because motherhood has a dramatic impact on a woman’s educational prospects, employment
opportunities, and self-determination, restrictive abortion laws deprive her of basic control over
her life. For these reasons, ‘the decision whether or not to beget or bear a child’ lies at ‘the very
heart of this cluster of constitutionally protected choices’ ”) (citation omitted).
59
See supra Part 3.
60
See generally MARK TUSHNET, TAKING THE CONSTITUTION AWAY FROM THE COURTS (Princeton Univ. Press
1999).
Icon-03.qxd 11/12/02 11:06 AM Page 29

Constitution, social rights, and political justification 29

we find the former to be morally the more appealing conception?61 If we don’t

think constitutional law enforced by judges is or need be all the constitutional
law that there is or that matters, then why should fears for democracy dictate
acceptance of what we deem a morally inferior, constitutional formulation of
social rights? We can both constitutionalize a social-citizenship right and tell (or
expect) courts to be discreet about any efforts to enforce it judicially. We can fol-
low that course at no cost in suppression of democracy. Is all that not so?
My answer for the moment is: not obviously. Such a copiously expansive
constitutional right as a right of social citizenship conceivably could call forth
a democracy-protective objection even if we assume that courts will abstain
totally from trying to enforce it.
Because suppose they do abstain. Then there are two possibilities. Over
time, the legislative branch either will or will not prove conscientious, in this
case, about compliance with constitutional law. Either it will or will not feel
constrained to comply, sooner or (not too much) later, even in the absence of
judicial enforcement. If it will not, then what would be the point of writing the
social citizenship right into the constitution? What would be the point of
naming something a constitutional right that we don’t mean and expect to
be taken seriously, by public officials presumed conscientious? But if we do sup-
pose that a law-abiding legislative branch really would feel itself constrained to
heed a constitutional conferral of social citizenship rights on everyone, would
that not subject our legislatures to serious curbs on their policymaking discre-
tion? How can we honestly name social citizenship a constitutional right with-
out intending a far-flung constraint on policy choice by majority rule, even
if we assume the courts will make no attempt to enforce the right thus named?
A different example may help make the point clear. Imagine that some
country adds to its constitution what we may call the Libertarian Amendment.
It reads:
The regulatory state is hereby abolished. Parliament shall make no law
attaching liabilities, penalties, or burdens of any kind to conduct that
would not be actionable at common law.62
Imagine also that in this country there is no practice of judicial constitutional
review. Constitutional law is left to be made effective on its direct, official
addressees through self-discipline aided by political pressures. The Libertarian
Amendment would nevertheless be seen by everyone, supporters as well as

61
Constitutional Welfare Rights, supra note 45, makes a stirring and persuasive case for the moral
superiority of the social-citizenship conception.
62
See Richard A. Epstein, Takings, Exclusivity and Speech: The Legacy of PruneYard v Robins, 64 U.
CHI. L. REV. 21, 21–28 (1997); Cf. Lucas v. South Carolina Coastal Comm’n, 505 U.S. 1003,
1029–30 (1992) (Scalia, J.) (making actionability under prior “background law,” of the uses of
property prohibited by a new regulatory enactment, the test of the state’s duty to pay compensa-
tion as if for a “taking” of property).
Icon-03.qxd 11/12/02 11:06 AM Page 30

30 F. I. Michelman

detractors, as profoundly counter-democratic. Putting shackles on democracy

would be exactly the point and aim of the amendment.
We see that a democratic objection to constitutionalization of one or another
class of rights can be formidable, even assuming “enforcement” of such a right
would occur wholly outside the courts, through political and moral pressure.
What about the specific case of a social-citizenship right? Does it give rise to a
formidable democratic objection? I wish to leave that question hanging for a bit,
while I now take us through what I’ve called the contractarian objection to
constitutionalized social rights.

6. The contractarian objection

Constitutional contractarians, concerned about political legitimacy within a
given country, will ask whether the country’s basic or constitutional laws—
the ones that shape, organize, direct, and limit the country’s political and legal
practice—compose in toto a sufficient, legitimating constitutional agreement.
In order to do so, those laws must include a guarantee affecting every topic for
which a rational person, responding reasonably, would demand a guarantee as
a condition of willing support for the lawmaking system as a whole.63
Suppose this constitutional set right now consists of twenty-five clauses,
A through Y, the wording of which is canonical and undisputed. The bill of
rights comprises clauses P through Y. P through Y say things like “no search or
seizure shall be conducted without a warrant, and warrants shall not issue with-
out probable cause,” “death is hereby prohibited as a punishment for crime,”
“freedom of the press is hereby guaranteed.” There is currently no guarantee of
social rights. Prompted by the constitutional-contractarian aim of filling out a
complete, legitimating constitutional agreement, the country just now is debat-
ing addition of a twenty-sixth clause, Z. Z would provide, and I quote, “Everyone
has the right of social citizenship as described in the collected works of William
Forbath. Within available resources, the state must direct and conform its
legislative and other measures to the progressive realization of this right.”
Suppose it is widely agreed that the goal of Z—effective social citizenship on
fair terms for all who seek it—indeed is one to which the country’s govern-
mental operations must visibly be committed in practice, in order that the total
governance system may be one that meets the constitutional-contractarian
standard of universal reasonable acceptability. Such a moral fact—so to call
it—might seem to pose a fatal difficulty for constitutional contractarian polit-
ical legitimacy. The apparent difficulty lies in the further fact that it will almost
always be impossible for anyone to say decisively whether Z is or is not being
pursued in earnest. Let’s say Parliament this year has done all of the following:
replaced welfare with workfare, increased by one half the budget allocation for

63
See supra Part 4.
Icon-03.qxd 11/12/02 11:06 AM Page 31

Constitution, social rights, and political justification 31

job training, reduced the minimum wage by one-third, extended the collective
bargaining laws to cover employers of as few as ten workers, abolished rent
control, budgeted an annual sum of 30 billion crowns for housing allowances
and job training, increased income tax rates by five percent, reduced the prime
lending rate by two percentage points, doubled the size of the employment
discrimination mediation corps, and approved a new tariff schedule somewhat
less protective than its predecessor, in exchange for reciprocal concessions
from abroad. Are they complying with clause Z?
Raging indeterminacy of this sort seems to disqualify a clause like Z from
figuring as a required component in a complete and legitimating constitu-
tional agreement. Remember how the constitutional-contractarian argument
goes. I can freely accept the daily run of coercive acts from a constituted polit-
ical regime, including acts I judge to be pernicious or unjust, because and only
because (1) I regard this regime qua regime as rationally acceptable by every-
one who is being reasonable, and (2) I see my fellow citizens and their government
abiding by this regime. In order for the regime to merit my willing compliance,
one of the conditions is that I can know at all times that the commitments that
make it universally reasonably and rationally acceptable are actual, not fake.
I have to be able to observe my fellow citizens and their government really
complying with the principles. And how can I, if Z is one of the principles? Just
because I can’t, it seems that Z cannot be deemed an indispensable part of any
constitution meant to do the political-justificatory work that constitutions
morally have to do, in a constitutional contractarian view.
And there, then, you have the contractarian objection to the idea that social
rights belong in a constitutional bill of rights. Because social rights lack the
trait of transparency, as we may call it—the trait of more-or-less detectably
being realized (or not) at any given moment—we seem barred from regarding
them as required parts of the essential constitution, lacking which the consti-
tution would fail to meet the political-liberal standard for legitimacy; lacking
which, in other words, the constitution would fail to provide an acceptable
basis for political rule, in the sight of every rational person responding reason-
ably.64 (And then what would be the affirmative case for constitutionalization
of social rights guarantees?)

7. Constitutional contractarianism in a bind:

public reason
At this point, we must question seriously the cogency and coherence of the
constitutional contractarian’s answer—his deeply, normatively individualist

64
A consideration of this kind may have entered into John Rawls’ own conclusion that social-
rights guarantees are not among the “constitutional essentials.” See POLITICAL LIBERALISM, supra
note 38, at 227–30; Frank I. Michelman, Rawls on Constitutionalism and Constitutional Law, in THE
CAMBRIDGE COMPANION TO JOHN RAWLS (Samuel Freeman ed.) (forthcoming 2002).
Icon-03.qxd 11/12/02 11:06 AM Page 32

32 F. I. Michelman

answer—to the question of political legitimacy. It may, after all, be true that a
constitutional system without a social-citizenship guarantee—by which
I mean a credible guarantee of constant, good faith pursuit by the powers that
be of assurance of the prerequisites to social citizenship to all who seek them
on fair terms—fails to provide every rational and reasonable person with
sufficient reason to accept whatever specific laws may issue out of the system
from time to time. Believing exactly that, political liberals are not free just to
shuck the belief if it happens to become theoretically inconvenient.65 How,
then, can a political regime possibly, in their sight, be legitimate?
A regime, they believe, is not legitimate if its basic law does include a social
rights guarantee, but it also is not legitimate if any of its basic-law guarantees,
required for legitimacy, are such that citizens cannot judge whether those guar-
antees in fact are being kept, or at least at all times being pursued in good faith.
If a social-citizenship guarantee fails the latter test, then, by constitutional con-
tractarian lights, it cannot form an indispensable part of a sufficient, legitimat-
ing constitutional agreement—and yet it does. That looks like a contradiction.
If so, the constitutional contractarian theory of political justification through a
complete, legitimating constitutional agreement must be a mistaken theory of
political justification. If it is mistaken, and if there is in sight no other liberally
acceptable theory of political justification for modern, plural, law-governed
societies, then, for liberals concerned with such societies, political justice appar-
ently lies beyond the possibility of coherent definition, let alone achievement.66
Rawlsian thought offers a way out of this bind. We see it in proposals to give
social rights a constitutional status of “directive principles” rather than
“rights,” as well as in John Rawls’ ideas about what he calls public reason and
“matters of basic justice.” The point for Rawls is this: A sufficient, legitimating
constitutional agreement has to provide fully firm, strict, and reliable substant-
ive guarantees of compliance with what he calls the central ranges of the basic
negative liberties—freedoms of conscience and expression, for example.
Regarding the rest of social citizenship, the requirement is a looser one. What
we need, and all we need, is assurance that, whenever political and legislative
choices bear upon the basic structural conditions of social citizenship, those
choices will be approached by all who take part in them under what Rawls calls
a constraint of public reason. Participants in such decisions must stand ready

65
John Rawls would classify such a guarantee as a “matter of basic justice,” covered by a “con-
straint of public reason,” but not as a “constitutional essential.” See POLITICAL LIBERALISM, supra
note 38, at 216–20, 223–27; Michelman, supra note 64.
66
The problem may well be graver than my text discloses. The bind I have described is not clearly
restricted to social-rights guarantees, but rather may extend to all the “basic liberties,” all the
members of the standard list of constitutional negative liberties. See Frank I. Michelman,
Postmodernism, Proceduralism, and Constitutional Justice: A Comment on van der Walt & Botha,
CONSTELLATIONS Vol. 9, Issue 2, 246, 256–59 (2002); Frank I. Michelman, Relative Constraint and
Public Reason: What Is “The Work We Expect of Law?” BROOK. L. REV. (forthcoming).
Icon-03.qxd 11/12/02 11:06 AM Page 33

Constitution, social rights, and political justification 33

to explain the consonance of their positions with some conception of a com-

plete, legitimating constitutional agreement that—they sincerely maintain—
deserves acceptance by every affected person who is rational and reasonable.67
The move to public reason eases the strain on constitutional contractarians
who honestly believe that a political commitment to the state’s constant, good
faith pursuit of social citizenship for all must be a term of any universally rea-
sonably acceptable constitution. The case in which every parliamentarian and
indeed every voter stands ready, in all sincerity, to explain and defend all their
votes, on matters affecting the structural conditions of social citizenship,
as expressions of their honest best judgments about which choice is most con-
ducive to assurance of social citizenship for all, is a case of what Rawls would
call fulfillment of the ideal of public reason. If citizens could have sufficient
confidence that public reason in that sense prevails in public decisionmaking
over matters affecting the structural conditions of social citizenship, then that
confidence (combined with formal, legal guarantees of everyone’s enjoyment
at all times of the core, basic negative liberties) might give every reasonable
person a sufficient basis for accepting the legislative outcomes, whatever they
turn out to be, of a democratic constitutional regime. And notice, then, the
converse: If the facts on the ground are such that citizens cannot reasonably
maintain confidence in the effective constraint of public reason on political
choices affecting the structural conditions of social citizenship, then the extant
system of positive legal ordering is unjust. It fails to measure up to the moral
demand for justice in politics, as political-liberal, constitutional contractarian
thought conceives of that demand.

8. Public reason and democracy

Let us now return to the majoritarian objection to a constitutionalized right of
social citizenship. As we left it, the objection was this: To name social citizen-
ship a constitutional right is to impose a far-flung drag on democracy, even
assuming courts are kept out of the picture. We are in position now to see how
this objection trades on a particular, contestable, and indeed poor conception
of democracy.
If a so-called right of social citizenship would be as loose—as indeterminate—as
we have said, if it would lack mechanical applicability to any hard or contested
question of public policy, then exactly how does anyone think it would hamper
democracy? The case, as we now can see, is very different from that
of the Libertarian Amendment. To remind you: the Libertarian Amendment
provides that
Parliament shall make no law attaching liabilities, penalties, or burdens
of any kind to conduct that would not be actionable at common law.

67
See supra text accompanying notes 41–44.
Icon-03.qxd 11/12/02 11:06 AM Page 34

34 F. I. Michelman

This provision, while it may leave some room for good-faith dispute around the
edges, over exactly which conduct is and is not “actionable at common law,”
really leaves—surely it is intended to leave—an honest parliamentarian with
very little to decide in the way of regulatory law. By contrast, a constitution-
ally declared right of everyone to the enjoyment of social citizenship would
leave just about every major issue of public policy still to be decided. Its maxi-
mum (but maybe not trivial) effect on democratic decisionmaking (the courts
being kept away) would be a certain pressure on the frame of mind in which
citizens and their elected representatives would approach the sundry questions
of public policy always waiting to be decided. In Rawlsian language, the point
of naming social citizenship a constitutional right would be to give a certain
inflection to political public reason. Across a very broad swathe of public
issues, such a naming would amount to a demand that those issues be
approached as occasions for exercises of judgment—which choice will be con-
ducive to the social citizenship of everyone, on fair terms?—rather than as
invitations to press and to vote one’s own naked interests and preferences.
Of course, to call these matters of judgment is to see that they are matters
on which opinions can and will differ markedly, reasonably, and sincerely, and
very probably not independently of people’s particular social situations and
related interests. (In today’s United States, factory workers doubtless will tend
on average to see some of them differently than bond traders will, young
mothers differently from senior corporate personnel managers, blacks from
whites, etc.) But surely no harm to democracy lies there. Why should not dis-
agreements over constitutional-interpretive judgment make as good a seedbed
for democracy, or better, than do raw conflicts of interest and preference?
“Democracy,” then, would name the practice by which citizens communicat-
ively form, test, exchange, revise, and pool their constitutional-interpretive
judgments, only counting them as required to obtain, from time to time, the
“institutional settlements” a country needs in order to get on with the tolera-
bly orderly conduct of life.68
Granted, that is a pretty idealistic view of what democracy is and how it
works. I’m sure it is a minority view, too, by comparison with the view that
democracy means, quite strictly, that a country’s people are free to treat their
political agenda as a series of free-for-all contests of normatively unregulated
preferences. On it, however, seems to depend the idea of liberal justice or liberal
legitimacy within any possible system of positive legal ordering.

68
See Amy Gutmann & Dennis Thompson, Deliberative Democracy Beyond Process, 10 J. POL. PHIL.
153, 165–69 (2002); Frank I. Michelman, Why Voting? 34 LOY. L.A. L. REV. 985, 1001–04 (2001).
On “the principle of institutional settlement,” see HENRY M. HART, JR. & ALBERT M. SACKS, THE LEGAL
PROCESS: BASIC PROBLEMS IN THE MAKING AND APPLICATION OF LAW 1–9 (William N. Eskridge, Jr. & Philip
P. Frickey eds., The Foundation Press, Inc. 1994).