Maurer School of Law: Indiana University

Digital Repository @ Maurer Law

Addison Harris Lecture Maurer Lectures

1985

Beyond Carolene Products

Bruce A. Ackerman
Columbia University

Follow this and additional works at: https://www.repository.law.indiana.edu/harris

Part of the Constitutional Law Commons, and the Courts Commons

Recommended Citation
Ackerman, Bruce A., "Beyond Carolene Products" (1985). Addison Harris Lecture. 1.
https://www.repository.law.indiana.edu/harris/1

This Article is brought to you for free and open access

by the Maurer Lectures at Digital Repository @ Maurer
Law. It has been accepted for inclusion in Addison
Harris Lecture by an authorized administrator of
Digital Repository @ Maurer Law. For more
information, please contact kdcogswe@indiana.edu.
VOLUME 98 FEBRUARY 1985 NUMBER 4

HARVARD LAW REVIEWI

BEYOND CAROLENE PRODUCTSt
Bruce A. Ackennan*
I. THE PROMISE OF CAROLENE PRODUCTS
'IP]rejudice against discrete and insular minorities may be a special condi-
tion . . . curtaiking] the operation of those political processes ordinarily to
be relied upon to protect minorities, and [so] may callfor a correspondingly
more searchingjudicial inquiry."I

T HESE famous words, appearing in the otherwise unimportant

Carolene Products case, came at a moment of extraordinary vul-
nerability for the Supreme Court. They were written in 1938. The
t @ 1985 by Bruce A. Ackerman. An earlier version of this essay was delivered in April
1983 at the University of Indiana, Bloomington, as the Addison C. Harris Lectures.
* Beekman Professor of Law and Philosophy, Columbia University. I am grateful to the
faculty and students of Indiana Law School for their thoughtful comments. As always, I am
indebted to my friends, at Columbia and elsewhere, who have greatly assisted me by their
critical readings of earlier drafts. My work on this essay was supported by a research grant
from the Columbia Law School's Rubin Program for the Advancement of Liberty and Equality
through Law.
I United States v. Carolene Prods. Co., 3o4 U.S. 144, 152 n.4 (1938). Although my central
concern here will be this passage from the third and final paragraph of Carolene's famous
footnote four, it will be useful to reproduce the footnote's entire text:
There may be narrower scope for operation of the presumption of constitutionality
when legislation appears on its face to be within a specific prohibition of the Constitution,
such as those of the first ten amendments, which are deemed equally specific when held
to be embraced within the Fourteenth. See Stromberg v. California, 283 U.S. 359, 369-
370; Lovell v. Griffin, 303 U.S. 444, 452.
It is unnecessary to consider now whether legislation which restricts those political
processes which can ordinarily be expected to bring about repeal of undesirable legislation,
is to be subjected to more exacting judicial scrutiny under the general prohibitions of the
Fourteenth Amendment than are most other types of legislation. On restrictions upon
the right to vote, see Nixon v. Herndon, 273 U.S. 536; Nixon v. Condon, 286 U.S. 73;
on restraints upon the dissemination of information, see Near v. Minnesota ex rel. Olson,
283 U.S. 697, 713-714, 718-720, 722; Grosjean v. American Press Co., 297 U.S. 233;
Lovell v. Griffin, supra; on interferences with political organizations, see Stromberg v.
California, supra, 369; Fiske v. Kansas, 274 U.S. 380; Whitney v. California, 274 U.S.
357, 373-378; Herndon v. Lowry, 301 U.S. 242; and see Holmes, J., in Gitlow v. New
York, 268 U.S. 652, 673; as to prohibition of peaceable assembly, see De Jonge v. Oregon,
299 U.S. 353, 365.
Nor need we enquire whether similar considerations enter into the review of statutes
directed at particular religious, Pierce v. Society of Sisters, 268 U.S. 5Io, or national,
Meyer v. Nebraska, 262 U.S. 39o; Bartels v. Iowa, 262 U.S. 404; Farringtonv. Tokushige,
273 U.S. 484, or racial minorities, Nixon v. Herndon, supra; Nixon v. Condon, supra:
whether prejudice against discrete and insular minorities may be a special condition,
which tends seriously to curtail the operation of those political processes ordinarily to be
relied upon to protect minorities, and which may call for a correspondingly more searching
judicial inquiry. Compare McCulloch v. Maryland, 4 Wheat. 316, 428; South Carolina
v. Barnwell Bros., 303 U.S. 277, 284, n.2, and cases cited.
Id.
 HARVARD LAW REVIEW [VOL. 98:713

Court was just beginning to dig itself out of the constitutional debris
left by its wholesale capitulation to the New Deal a year before. With
the decisive triumph of the activist welfare state over the Old Court,
an entire world of constitutional meanings, laboriously built up over
two generations, had come crashing down upon the Justices' heads.
Indeed, the Court had been so politically discredited by its constitu-
tional defense of laissez-faire capitalism that it was hardly obvious
whether any firm ground remained upon which to rebuild the insti-
tution of judicial review. How, then, to begin the work of reconstruc-
tion?
Only once before had the Court confronted a similar challenge.
Just as the triumphant New Deal Democrats had destroyed the laissez-
faire constitutionalism of Lochner v. New York, so too a triumphant
Republican Congress had destroyed the slavocratic constitutionalism
of Dred Scott v. Sanford after the Civil War. Just as many contem-
porary observers doubted the institutional independence - let alone
the constitutional importance - of the Supreme Court during Recon-
struction, 2 nobody could be confident about the future of judicial
review in the aftermath of the Court-packing crisis.
Only one thing was clear. If the Court were to reassert itself after
the Great Depression, it could not do so through the same constitu-
tional rhetoric with which it had rehabilitated itself after the Civil
War. During the long period between Reconstruction and New Deal,
the Court had risen to the heights of power by insisting upon the
fundamental right of free men to pursue their private aims in a free
market system. Yet it was precisely this ideological elixir, which had
given the judiciary new life after the Civil War, that proved nearly
fatal during the constitutional birth agony of the activist welfare state.
If the Court were to build a new foundation for judicial review, it
would need an entirely new constitutional rhetoric - one that self-
consciously recognized that the era of laissez-faire capitalism had
ended.
Against this historical background, we may glimpse the promise
of Carolene Products.3 Rather than look back longingly to a repu-
diated constitutional order, Carolene brilliantly endeavored to turn the
Old Court's recent defeat into a judicial victory. As far as Carolene
was concerned, lawyers could dispense with their traditional effort to
organize their concern for individual rights through a constitutional

2 A fine study of the Chase Court's retreat before the Reconstruction Congress in Georgia
v. Stanton, 73 U.S. (6 Wall.) 5o (1867), and Ex parte McCardle, 74 U.S. (7 Wall.) 5o6 (1869),
may be found in i C. FAIP.iAN, RECONSTRUCTION AND REUNION 1864-1868, at 366-514
(Oliver Wendell Holmes Devise History of the Supreme Court of the United States No. 6, 1971).
Another thoughtful treatment, emphasizing the continuing support for the Court during this
time of adversity, is S. KUTLER, JUDICIAL POWER AND RECONSTRUCTION POLITICS (1968).
3 For a deeper historical grounding, see Cover, The Origins of Judicial Activism in the
Protection of Minorities, 9i YALE L.J. 1287, 1287-1304 (1982).
1g85] CAROLENE PRODUCTS

rhetoric glorifying private property and free contract. Instead, Caro-

lene proposed to make the ideals of the victorious activist Democracy
serve as a primary foundation for constitutional rights in the United
States.
Fifty years onward, the basic idea is familiar, but it requires re-
statement if we are to examine it carefully. Carolene promises relief
from the problem of legitimacy raised whenever nine elderly lawyers
invalidate the legislative decisions of a majority of our elected repre-
sentatives. The Carolene solution is to seize the high ground of demo-
cratic theory and establish that the challenged legislation was pro-
duced by a profoundly defective process. By demonstrating that the
legislative decision itself resulted from an undemocratic procedure, a
Carolene court hopes to reverse the spin of the countermajoritarian
difficulty. For it now may seem that the original legislative decision,
not the judicial invalidation, suffers the greater legitimacy deficit.
Assume, for example, that the people of a state, after excluding
blacks from the polls, elect an all-white legislature that proceeds to
enact some classic Jim Crow legislation. Under the Carolene ap- Jim Crow legislation
proach, the court does not purport to challenge the substantive value -character
derogatory minstrel
- laws
judgments underlying the legislative decision; instead, it simply denies segregating
and whites -
blacks

that the Jim Crow statute would have emerged from a fair and open separate facilities for
blacks and whites
political process in which blacks were allowed to participate. In like schools,
essence, the court is trumping the statutory conclusions of the deeply transportation
public
and

flawed real-world legislature by appealing to the hypothetical judg- accommodations

ment of an ideally democratic legislature.
No wonder, then, that Carolene Products seemed so promising in
1938. Not only did it point the Supreme Court toward the path of
racial justice and minority rights, but it also explained why the new
road to minority rights was fundamentally different from the old road
to property rights that had so recently led the Court to the brink of
self-destruction. Whereas the Old Court had protected property own-
ers who enjoyed ample opportunity to safeguard their own interests
through the political process, the New Court would accord special
protection to those who had been deprived of their fair share of
political influence.
No less significantly, the Carolene Court sketched its new mission
in exceptionally broad strokes. It did not limit its prospective inter-
vention to the straightforward cases in which blacks or other unpop-
ular groups were excluded from the polls or denied other fundamen-
tal rights ot political expression. Carotene suggested an enduring role
for the judiciary, one that would continue even after every adult
American had secured his right to participate in politics. 4 To take

4 Indeed, one can best appreciate the scope of Carolene's concern by contrasting it to the
only other assertion of judicial authority that is remotely comparable - the Court's effort, a
generation later, to reapportion state legislatures, see Reynolds v. Sims, 377 U.S. 533 (1964),
 HARVARD LAW REVIEW [Vol. 98:713

the case that will serve as a paradigm in this essay: the Carolene
footnote suggests that, even in a world in which blacks voted no less
frequently than whites, and in which election districts strictly con-
formed to the Court's reapportionment decisions, blacks would still
possess, by virtue of their discreteness and insularity, a disproportion-
ately small share of influence on legislative policy - a disproportion
of such magnitude as to warrant the judicial conclusion that a fair
democratic process would have generated outcomes systematically
more favorable to minority interests. This suggestion, moreover, an-
6
imates countless modern discussions - judicial5 as well as academic

and the House of Representatives, see Wesberry v. Sanders, 376 U.S. x (1964). As in Carotene,
the reapportionment cases go beyond the situation in which the government deprives some
group's members of their individual rights of citizenship. They deal instead with the way the
political process aggregates citizens' views to form a collective judgment: some citizens have too
much influence; others too little. Like Carotene, the reapportionment cases try to identify
situations of disproportionate influence without making substantive judgments about the political
interests that the judges think the legislature ought to favor. Instead, the concept of illegitimate
influence is elucidated through a formula - "one person, one vote" - grounded on the theory
of democratic process.
At this point, however, Carotene breaks with the reapportionment cases. In implementing
its one-person, one-vote principle, the Court has contented itself with a strictly formal conception
of democracy. Although it has been very strict in insisting that each congressional district
contain equal numbers of inhabitants, see Karcher v. Daggett, 103 S. Ct. 2653 (1983), it has
been extremely hesitant about moving beyond formal equality to more substantive conceptions
of democratic power-sharing. But see id. at 2670-78 (Stevens, J., concurring) (advocating
stricter scrutiny of legislative gerrymandering). In contrast, Carotene Products takes a strikingly
substantive view of the problem. As the Carotene footnote makes plain, the Court's concern is
not merely with a group's formal right to vote, or its formal right to have its votes counted on
a one-person, one-vote basis, but with its capacity to effect substantive political outcomes.
5 For example, Sugarman v. Dougall, 413 U.S. 634, 642 (973), and Graham v. Richardson,
403 U.S. 365, 372 (I97I) (discussed further in note 27), explicitly invoked Carotene's concern
with the political powerlessness of "discrete and insular" minorities to justify special judicial
protection for resident aliens. As Justice Powell has noted, moreover, "the influence of Footnote
4 cannot be measured accurately by simple enumeration of cases in which it has been cited."
Powell, Carolene Products Revisited, 82 COLUM. L. REv. 1087, 1087 n.4 (1982). Thus, even
when it is not cited explicitly, the Carotene idea plays a role in standard judicial justifications
for strict judicial scrutiny of legislation burdening "suspect" classes. For example, in San Antonio
Indep. School Dist. v. Rodriguez, 411 U.S. 1 (1973), the Court, in discussing the treatment of
the "traditional indicia of suspectness," observed that the critical question is whether "the class
is . . . saddled with such disabilities, or subjected to such a history of purposeful unequal
treatment, or relegated to such a position of politicalpowerlessness as to command extraordinary
protection from the majoritarian political process." Id. at 20 (emphasis added).
6 Dean Ely's book, J. ELY, DEMOCRACY AND DISTRUST (1980), is the most important effort
to develop footnote four's larger implications for the practice of judicial review. The high
quality of the debate it has engendered attests to the book's value. See, e.g., Estreicher, Platonic
Guardians of Democracy: John Hart Ely's Role for the Supreme Court in the Constitution's
Open Texture, 56 N.Y.U. L. Rv. 547 (x98i); JudicialReview versus Democracy (Symposium),
42 OHIO ST. L.J. I (i981); Laycock, Taking Constitutions Seriously: A Theory of Judicial
Review (Book Review), 59 TEx. L. REv. 343 (r981); Tribe, The Puzzling Persistenceof Process-
Based Constitutional Theories, 89 YALE L.J. ro63 (198o). Although Ely's work currently
dominates the field, important insights may also be found in Cover, supra note 3, and L. LUSKY,
By WHAT RIGHT? (1975).
1985] CAROLENE PRODUCTS

- in which the political weakness of "discrete and insular minorities"

is a crucial, if unexamined, premise in the elaboration of intricate
constitutional doctrines. 7 I shall argue, however, that the Carolene
formula cannot withstand close scrutiny.
Given the unsettling character of this thesis, it is best to begin by
emphasizing its limits. My critique applies exclusively to the paradigm
case I have just described, and not to the cruder case in which blacks
- or other racial or religious minorities - are excluded from the
voting booth or deterred from exercising their fundamental rights to
speak freely and organize politically. 8 During Carolene's first half-
century, it was these brutal efforts at political exclusion that rightly
were of central constitutional concern. Nothing I say is intended to
deny the obvious unconstitutionality involved in excluding minorities
from the nation's political life.
My concern here, however, is with the future, not -the past. Al-
though America has by no means worked itself clear of past practices
of political exclusion, it is not visionary to hope that we will indeed
put this grim aspect of history behind us and that, during the next
generation, we will inhabit a world that increasingly resembles my
paradigm case: a world in which, despite the existence of pervasive
social prejudice, minorities can and do participate in large numbers
within the normal political process. 9 In light of this prospect, a
reappraisal of Carolene is a pressing necessity: its approach to minority
rights is profoundly shaped by the old politics of exclusion and yields
systematically misleading cues within the new participatory paradigm.
Indeed, if we fail to rethink Carolene's dictum about discrete and
insular minorities, we will succeed only in doing two different kinds
of damage. On the one hand, we will fail to do justice to the very
racial and religious groups that Carolene has done so much to protect
in the past half-century. By tying their rights to an increasingly
unrealistic model of politics, we will place them on the weakest pos-

Not that there haven't been expressions of anxiety by some of our most thoughtful com-
mentators. See, e.g., J. ELY, supra note 6, at 152-53; Bennett, Abortion and JudicialReview:
Of Burdens and Benefits, Hard Cases and Some Bad Law, 75 Nw. U.L. REV. 978, 995-98
(ig8i); Goodman, De Facto School Segregation: A Constitutional and Empirical Analysis, 6o
CALIF. L. REV. 275, 312-14 (1972). These writers, however, have not attempted a systematic
analysis of the reasons that led them to doubt Carolene's equation of political powerlessness
with "discreteness and insularity." As a consequence, their expressions of doubt have not much
affected the conventional wisdom. Indeed, I believe that Dean Ely himself has not appreciated
the extent to which his passing doubts undermine the central thrust of his analysis. See infra
PP. 734-37-
8 In other words, my principal concern here is the footnote's third and final paragraph
(quoted in note I). I am not challenging the footnote's second paragraph, which suggests the
propriety of special judicial supervision of restrictions upon "those political processes which can
ordinarily be expected to bring about repeal of undesirable legislation." On the role of the
footnote's first paragraph, which envisions judicial intervention on behalf of "specific prohibi-
tion[s] of the Constitution," see pp. 743-44.
9 See infra notes 35 & 57.
 HARVARD LAW REVIEW [Vol. 98:713

sible foundation. On the other hand, we will fail to do justice to

Carolene's basic insight into the problem posed by prejudice in a
pluralist democracy. The end of the politics of exclusion hardly im-
plies that pluralist democracy now functions fairly; it does mean,
however, that the groups most disadvantaged by pluralism in the
future will be different from those excluded under the old regime.
The victims of sexual discrimination or poverty, rather than racial or
religious minorities, will increasingly constitute the groups with the
greatest claim upon Carolene's concern with the fairness of pluralist
process.
To demonstrate the need for doctrinal reorientation, I shall ex-
amine separately each of Carolene'sfour operative terms: (i) prejudice,
(2) discrete, (3)insular, and (4)minorities. It is by means of these
four terms that Carolene hopes to identify groups that have been
unconstitutionally deprived of their fair share of democratic influence.
As America moves toward the participatory paradigm, however,
judges can no longer expect these familiar concepts to operate in a
way that will allow courts to solve the problem of countermajoritar-
ianism. To explain why, I have found it helpful to examine Carolene's
four basic terms in reverse order from the way they appear in the
standard litany. A concluding section glimpses the path of reconstruc-
tion that lies ahead.

II. DISCRETE AND INSULAR MINORITIES?

The Carolene formula limits its attention to the asserted political
weakness of minorities and fails to consider the analogous case of a
politically ineffective majority. In view of Carolene's larger ambition
to deflect the countermajoritarian difficulty, this is an especially odd
omission. Consider again the paradigm case: blacks are participating
no less frequently than whites in a political system that satisfies the
standards for electoral fairness elaborated in the Court's modern de-
cisions. Despite these formal safeguards, imagine that an all-white
legislature manages to get elected, and then enacts a series of laws
prohibiting interracial marriage and forbidding interracial adoptions.
Would we be less concerned about this outcome if we completed the
scenario by assuming that blacks amounted to 75 percent, rather than
12 percent, of the relevant electorate?
Of course not.' 0 Indeed, the existence of a commanding black
majority would encourage us to intensify our search for a set of
structural factors that somehow allowed whites to dominate the os-
tensibly democratic political process. Carolene casually disregards the

10Cf. Palmore v. Sidoti, 104 S. Ct. z879 (1984) (holding unconstitutional a state's modification
of a child custody decree based upon custodial parent's cohabitation with a person of different
race).
1985] CAROLENE PRODUCTS

easiest case for finding a substantive defect in a formally fair electoral

process: the case in which organizational difficulties have prevented a
commanding majority of the population from influencing the ongoing
flow of legislative decisions. After all, if democracy means anything,
it means a regime designed to further the majority's basic interests;
that is certainly not what is going on in the case we have hypothesized.

A. The Principle of Minority Acquiescence

If we begin with the easy case of an ineffective majority, we can
also begin to see how much harder it is to justify the Carolene concern
for ineffective minorities. To put the point simply, minorities are
supposed to lose in a democratic system - even when they want very
much to win and even when they think (as they often will) that the
majority is deeply wrong in ignoring their just complaints. This
principle - call it the principle of minority acquiescence - is abso-
lutely central to democratic theory. Of'course, a minority may not be
denied its right to participate within a democratic framework. Al-
though it must acquiesce in current legislative decisions, it is fully
entitled to use all its political resources to induce a future legislative
majority to accede to its demands. But Carolene promises minorities
more than formal rights: it asserts that they are sometimes entitled to
demand substantive victory now, not merely the chance of victory
later.
The problem this promise raises is all the more acute because
Carolene refuses to accept the solution that countless others have
embraced. It is easy to solve the problem of majority rule by positing
the existence of minority rights that are so fundamental as to trump
the value of democratic rule itself. Indeed, as the Carolene Court was
well aware, it is too easy to solve the problem in this way. Faced
with the political repudiation of Lochner's natural rights jurispru-
dence, the Court was determined to build another foundation for the
protection of minority rights: why not redefine the concept of democ-
racy itself in a way that would support the notion that minorities do
have a right to win some of the time?

B. The PluralistSolution
While the courts speak vaguely of "those political processes ordi-
narily relied upon to protect minorities," generations of American
political scientists have filled in the picture of pluralist democracy
presupposed by Carolene's distinctive argument for minority rights."
11The path of pluralist thinking may be traced from A. BENTLEY, THE PROCESS OF
GOVERNMENT (I908), through D. TRUMAN, THE GOVERNMENTAL PROCESS (1951), to R. DAHL,
DILEMMAS OF PLURALIST DEMOCRACY (1982). Cruder versions of pluralist theory are now
common among political economists of the Chicago school. See, e.g., Becker, A Theory of
Competition Among Pressure Groups for Political Influence, 98 Q.J. ECON. 371 (i983).
 HARVARD LAW REVIEW [Vol. 98:713

According to this familiar view, it is a naive mistake to speak of

democracy as if it involved rule by a single, well-defined majority
over a coherent and constant minority. Instead, normal American
politics is pluralistic: myriad pressure groups, each typically repre-
senting a fraction of the population, bargain with one another for
mutual support.
Once this picture of pluralistic politics is accepted, the stage has
been set for the rehabilitation of Carolene's concern with ineffective
minorities. We may now find that there is something about certain
minority groups - call them Carolene or C-groups - that makes it
especially difficult for them to strike bargains with potential coalition
partners. As a consequence, C-groups will find themselves in politi-
cally ascendant coalitions much less often than will otherwise com-
parable groups. Over time, then, C-groups will achieve less than their
"fair share" of influence upon legislation. And it is for this reason,
the pluralist concludes, that Carolene rightly suggests that judicial
protection for C-groups can be defended in a manner responsive to
the countermajoritarian difficulty afflicting judicial review. By inter-
vening on behalf of C-groups, a Carolene court merely produces the
substantive outcomes that the C-group would have obtained through
politics if it had not been so systematically disadvantaged in the
ongoing process of pluralist bargaining.
It thus appears, at first glance, that a Carolene court can draw
upon a well-developed body of pluralist political science to support
its special protection of minority rights. When we move beyond in-
tuition to analysis, however, the pluralist argument is full of traps for
the unwary. First, it is by no means clear that our Constitution
wholeheartedly endorses the bargaining theory of democracy. 12 Sec-
ond, all that pluralist theory explains is why minority groups can
expect to influence legislative outcomes some of the time; it is some-
thing very different to explain why minorities may dress up these
expectations in the language of constitutional rights and demand ju-
dicial protection for them.13
I shall defer these fundamental points to the concluding section.
My focus here is on a central doctrinal difficulty that persists even if
Carolene's pluralist foundations are secure. This problem concerns
the indiscriminate standard according to which Carolene proposes to
regulate the judicial protection of C-groups. In the common legal
understanding, Carolene is generally taken to imply that the same
level of strict judicial scrutiny should apply to legislation affecting

12 See Ackerman, The Storrs Lectures: Discovering the Constitution, 93 YALE L.J. 1013
(1984) (arguing that pluralism is only one aspect of the American constitutional tradition); infra
pp. 741-43.
13 I shall take up this question later on pp. 742-43.
1985] CAROLENE PRODUCTS

each and every C-group. But the pluralist model cannot justify such
a uniform judicial approach.
Consider, for example, an American constituency that includes 12
percent blacks and .5 percent Jehovah's Witnesses among its popu-
lation. 14 Doubtless, both groups will be encouraged by the pluralist
vision of democracy, since it suggests that neither group will inexor-
ably be excluded from the pluralist bazaar. Nonetheless, it should be
plain that these two groups have absolutely no reason to find the
prospect of pluralist bargaining equally gratifying. To the contrary,
the fact that blacks greatly outnumber Witnesses is bound to play an
important role in any plausible bargaining theory. Thus, even if the
two groups could somehow be compensated for their Carolene disad-
vantages, the Witnesses could not reasonably expect to win substan-
tive victories nearly as often as the blacks.
To put the point more generally, a bargaining approach to Carolene
does not suggest that each C-group has a right to be treated identically
to all other C-groups in the legislative process. Instead, the decisive
thought-experiment should involve the comparison of a particular C-
group with a hypothetical minority that I shhll call an unencumbered
or U-group. In each comparison, the relevant U-group should be
supposed to contain the same proportion of the population as the C-
group that invokes the Court's protection; the U-group differs, how-
ever, in that it is unencumbered by the bargaining disadvantages that
unconstitutionally burden the C-group. Thus, the Carolene question
for blacks entails a comparative analysis of the bargaining expectations
of a 12 percent minority unencumbered by those structural impedi-
ments that unconstitutionally impair blacks' bargaining position in the
ongoing pluralist process, while the question for Jehovah's Witnesses
involves a comparison with a much smaller U-group.
Such thought-experiments will most naturally result in a sliding
scale of Carolene concern. On one end of the scale are groups con-
sisting of ineffective majorities or very large minorities that find them-
selves disadvantaged in the political process by some constitutionally
impermissible barrier to bargaining. In cases involving these "major
minorities," a court can be quite confident that a comparable U-group
would have a decisive impact on the terms of pluralistic legislation.
In the middle of the Carolene scale are "middling minorities" in the
1o to 20 percent range. Here there is less reason for a court to expect
that a U-group of comparable size would radically change the terms

14 According to the 198o census, 26.5 million black Americans accounted for i.7 percent
of the total American population of 226.5 million. See BUREAU OF THE CENSUS, U.S. DEP'T
OF COMMERCE, i98o CENSUS OF POPULATION: GENERAL POPULATION CHARACTERISTICS, U.S.
SUMMARY 1-20 (1983) (PC8o-I-BI). The Jehovah's Witnesses claimed 643,000 active members
in the United States in 1984. Telephone call to Kingdom Hall, Brooklyn, New York (Mar.
i984).
 HARVARD LAW REVIEW [Vol. 98:713

of political trade, though its influence would be very substantial in

many plausible contexts. 15 And finally, on the other end of the scale,
there are groups so small as to elicit little solicitude from courts
concerned with correcting the failures of democratic bargaining.
When faced with "minor minorities" of .oi percent, for example, a
judge might well be unmoved by the enumeration of Carolene factors
that would generate substantial concern in the case of middling mi-
norities, not to mention major minorities. For the fact is that a U-
group of .oi percent has little to expect from a democratic political
process, unless it is very lucky, or exceptionally adept, in the bar-
gaining process. This point is essential to the responsible elaboration
of Carolene Products - whose promise, be it recalled, is to permit
courts to evade the thrust of the countermajoritarian difficulty by
appealing over the heads of real-world legislatures to the hypothetical
outcomes of a purified democratic process.
There is, then, an inevitably uneasy relationship between Caro-
lene's pluralist approach to democracy and the judicial protection of
minority rights. The tension reaches the breaking point in the pro-
verbial case of a minority of one: when the solitary citizen, having
little to expect from pluralist bargaining, challenges the invasion of
his fundamental rights by the normal political process. 16
My aim here, though, is to work out the doctrinal implications of
the Carolene formula rather than to criticize its foundations. So let
us focus our attention upon those groups, ranging from middling
minorities to encumbered majorities, whose role in the bargaining
process might well have a significant impact on the ongoing stream
of legislative decisions. How does Carolene propose to determine
whether a group suffers from severe enough bargaining disadvantages
to merit special protection? In other words, how are we supposed to
distinguish a C-group from a U-group?

III. DISCRETE AND INSULAR MINOlITIES?

As in the case of its failure to consider the rightful claims of
politically ineffective majorities, Carolene disdains the easy case in its
Is The best courts can hope for in making these judgments is a rough approximation of the
orders of magnitude involved. Legal scholars, however, should reflect seriously on recent efforts
by analytic political scientists to model the strategic implications of different legislative arrange-
ments and to undertake experimental research that might test their conclusions in simulated
negotiating settings. See, e.g., McKelvey & Ordeshook, An Experimental Study of the Effects
of ProceduralRules on Committee Behavior, 46 J. POL. 182 (1984); Shepsle & Weingast, When
Do Rules of Procedure Matter?, 46 J. POL. 2o6 (1984). It may well be possible in a decade or
two to use this research to give more structure to commonsense intuitions about bargaining
power and procedures in real-world legislatures.
16 Of course, Carolene's defense of judicial review is not limited to the pluralist rationale
developed in this section. In particular, Carolene raises the prospect of judicial intervention in
cases involving "a specific prohibition of the Constitution, such as those in the first ten amend-
ments." See supra note i. For a critique of this alternative approach to the protection of
individual rights, see pp. 743-44.
X985] CAROLENE PRODUCTS

eagerness to pronounce on harder ones. A dispassionate survey of the

relevant literature does not reveal a single-minded concern with the
political weakness of insular minorities. Instead, it expresses a per-
vasive anxiety over the way in which inequalities of wealth distort
the operation of a democratic process formally based upon egalitarian
principles. 17
I do not suggest that every thoughtful democrat believes that a
systematic effort to check the influence of wealth is an indispensable
condition of democratic legitimacy. There are respectable arguments
- though I am thoroughly unconvinced by them - to support the
claim that any effort to purge our regime of its plutocratic vices would
prove worse than the disease itself.' 8 For the present, however, I
need not evaluate these claims on their merits. It is enough that I
make my point explicitly conditional: if, as Carolene plainly supposes,
it is legitimate to move beyond a purely formal conception of demo-
cratic rule, we should begin building a substantive conception of
undue influence by considering the disproportionate impact wealth has
on American politics. It is here where the easy case for undemocratic
influence may be established.
Yet, as we all know, Carolene does not assert that prejudice against
"impoverished and uneducated minorities" may call for a more search-
ing judicial inquiry. Instead, it professes a concern for the status of
"discrete and insular minorities." This way of framing the issue has
strongly influenced the judicial understanding of American democracy
over the past half-century. Thus, although modern courts regularly
express concern about the bargaining position of discrete and insular
minorities, they often react skeptically to the very idea that legislatures
may constitutionally attempt to curb the influence of wealth over
formally democratic processes. 19
It is too kind to Carolene, however, to regard it as yet another
example of the American judiciary's eagerness to emphasize the symp-
tom while ignoring the disease. Even when considered as an exercise
in symptomatology, Carolene is utterly wrongheaded in its diagnosis.
Other things being equal, "discreteness and insularity" will normally

17 For a thoughtful review of the literature emphasizing this point, see Beitz, Political
Finance in the United States: A Survey of Research, 95 ETHICS 129 (1984).
IsSee, e.g., R. WINTER & J. BOLTON, CAMPAIGN FINANCING AND POLITICAL FREEDOM
(1973).
19 The courts, though, have not been very consistent skeptics. The very same opinion that
condemns in the strongest terms the effort to neutralize the impact of wealth, see Buckley v.
Valeo, 424 U.S. i, 49 (1976) (stating that such a policy is "wholly foreign to the First Amend-
ment"), also allows the government to pursue egalitarian objectives - as long as it pursues
these goals by offering candidates a public subsidy if they refuse to accept private contributions,
see id. at 9o-io8; see also Republican Nat'l Comm. v. Federal Election Comm'n, 487 F. Supp.
280, 286-87 (S.D.N.Y.) (three-judge court) (upholding prohibition of private contributions in
connection with public subsidy), aff'd mem., 445 U.S. 955 (i98o). For a pointed discussion of"
the subsidy problem, see Polsby, Buckley v. Valeo: The Special Nature of PoliticalSpeech, 1976
SuP. CT. REV. I, 26-41.
 HARVARD LAW REVIEW [VOL. 98:713

be a source of enormous bargaining advantage, not disadvantage, for

a group engaged in pluralist American politics. Except for special
cases, the concerns that underlie Carolene should lead judges to pro-
tect groups that possess the opposite characteristics from the ones
Carolene emphasizes - groups that are "anonymous and diffuse"
rather than "discrete and insular." It is these groups that both political
science and American history indicate are systematically disadvan-
taged in a pluralist democracy.

A. The Free-Rider Problem

To see my point, start with insularity and consider a thought-
experiment suggested by the previous argument. Imagine two groups,
I and D, of equal size (say each accounts for 12 percent of the
population). The members of one group, the I's, are distributed in
an insular way, concentrated in a single massive island within the sea
of American life; the D's, on the other hand, are diffused evenly
throughout the sea. Is it really so clear that, by virtue of their
diffusion throughout American life, the D's will gain systematic ad-
vantages over the I's in the normal course of pluralist politics?
Hardly. To begin with the basics, a political interest gains a great
advantage if its proponents can form a well-organized lobby to press
their cause in the corridors of power. Yet the construction of a
pressure group is no easy task. The main obstacle is the familiar free-
rider problem. 20 Simply because a person would find his interests
advanced by the formation of a pressure group, it does not follow
that he will spend his own scarce time and energy on political orga-
nization. On the contrary, from each individual's selfish viewpoint,
abstaining from interest-group activity is a "heads-I-win-tails-you-lose"
proposition. If only a few people adopt the do-nothing strategy, the
do-nothings will free-ride on the successful lobbying effort of others.
If free-riding becomes pervasive, things will not improve much if a
single member of the group adds his money and time to the flounder-
ing political effort. Either way, it pays for a selfish person to remain
a free rider even if he has a lot to gain from concerted lobbying. For
this reason, many interests remain ineffectively organized even in
pressure-group America.2 1 How, then, does a minority's insularity
affect the probability that it will break through the free-rider barrier
and achieve organizational effectiveness?
Far from being a patent disadvantage, insularity can help I-groups
in at least four different ways, all of which depend upon a single
sociological assumption that we should identify at the outset: however
oppressed the I's may be in other respects, they have not been pre-
vented from building up a dense communal life for themselves on
20 The classic statement of this problem is to be found in M. OLSON, THE LOGIC OF
COLLECTIVE ACTION 5-65 (1965).
21 See id. at 165-67.
 CAROLENE PRODUCTS

their tight little island. Thus, wherever an I looks, he will find himself
in businesses and churches, schools and labor unions, composed
largely of people speaking in distinctively I-accents about the daily
problems of. social life. This fundamental fact will generate a whole
series of advantages for I-members who seek to organize for political
purposes.
First, insularity will help breed sentiments of group solidarity.
Given an I's daily immersion in social realities that reaffirm his group
identity, the typical I will conceive his I-ness as something much more
than an incidental fact about himself. Instead, I-ness will serve as a
fundamental feature of self-identity - one that will encourage each I
to view the political activities of the group from a perspective that
transcends the purely instrumental. Thus, when a black or a Jew
gives $25 to the NAACP or the Anti-Defamation League, he is not
merely, or even principally, gambling that his small bit of money will
perceptibly increase his chance of enjoying the fruits of future lobbying
victories. Rather, the contribution is a means by which the donor can
symbolize the seriousness of his own commitment to his I-ness. By
contributing to the group cause, I demonstrate to myself, as well as
others, that I am serious about the values I profess to hold. Here, at
last, is one commodity - group identification - that is immune from
the free-rider problem: for if I do not give even a few dollars to the
group cause, can I plausibly say, even to myself, that I take my I-
22
ness seriously?
But insularity does more than engender the sentiment of group
solidarity that encourages symbolic contributions - contributions
that, when multiplied thousands of times, add up to very substantial
resources for the interest group receiving them. It also aids the I-
group in a second way by providing it with a new range of social
sanctions to impose upon would-be free riders. An I who refuses to
contribute to his interest group cannot expect this fact to be kept
secret from his fellow I's - news travels fast along the grapevine in
an insular community. Thus, the shirker cannot reasonably hope to
evade the moral disapproval of other I's when they learn that he has
failed to do his fair share to support the group cause. This stigma
may not only lead to embarrassment, but may also result in very
concrete material disadvantages in an I's day-to-day dealings with his
fellow group members. 23 Only by uprooting himself from his insular
community can an I hope to escape the opprobrium his free-riding
may engender. In contrast, a member of a diffuse D-group need not
suffer such severe dislocation in order to avoid the disapproval of his
fellow group members. Instead, he may insulate himself from their

22 For a related observation, see A. HIRSCHMAN, SHIFTING INVOLVEMENTS 85-91 (1982).

23 For an interesting effort to model phenomena: of this kind, see Boorman & Levitt, The
Cascade Effect: An Essay in Disequilibrium Theory, in RESEARCH METHODS IN SocIAL NET-
WORKS ANALYSIS (L. Freeman, A. Romney & D. White eds. forthcoming 1985).
 HARVARD LAW REVIEW [VOL. 98:713

displeasure by assimilating into the majoritarian mainstream - un-

doubtedly a costly process, but typically less costly than the social
stigma heaped on the free-riding 1. In short, as compared to an I, a
D is both less likely to view his group membership as fundamental to
his self-identity and less likely to suffer severe sanctions at the hands
of his fellow group members if he fails to do his fair share to support
their common objectives.

B. OrganizationalCosts
It follows, then, that the average I is more likely to contribute his
time and money to the group cause than is an otherwise comparable
D. Yet this conclusion tells only half the'story: not only will an I-
group receive more resources from its constituency, but I's will also
find it cheaper to organize themselves for effective political action.
First, the dense communications network generated by insularity dra-
matically reduces one of the heaviest costs involved in effective polit-
ical lobbying: the cost of communicating with a mass membership.
To get its messages out to its constituency, an insular political group
can often avail itself of the communications channels already estab-
lished by the group's churches, businesses, or labor unions. In con-
trast, a D-group must somehow locate and reach people who interact
with one another much less frequently and who have fewer channels
already established for the cheap transmission of D-group concerns.
Second, the organic character of insular life greatly reduces the
costs of selecting credible political leaders. The I-group can draw
upon a pool of people who have already earned the respect of their
fellow I's in other communal contexts: ministers and rabbis, successful
lawyers, businessmen, union leaders. In contrast, even if D-group
members manage to overcome the communications barrier, they must
often take the risk of selecting political leaders who have not been
tested and observed in other leadership settings.

C. Insularity and CongressionalInfluence

So insularity is an asset, both in increasing a group's political
resources and in reducing its organizational costs. Yet before we
render an overall assessment, we must consider the way insularity is
likely to affect the attitude of the people upon whom all the bargaining
pressure must ultimately be brought to bear: politicans seeking to
gain, and retain, elective office. Here we must refine our definitions
if we are to make analytic progress. Thus far, I have used "insularity"
to refer to the tendency of group members to interact with great
frequency in a variety of social contexts.2 4 If the term is conceived

24 Although this definition serves well enough for purposes of the present argument, it
contains two distinct ideas. The first - call it breadth - involves the number of different
19851 CAROLENE PRODUCTS

in this sociological way, people who live far apart from one another
may still be members of a single I-group, especially under modern
conditions. Conversely, it is easy for people living cheek by jowl to
fail to qualify as an insular minority-from a sociological point of view.
We have reached a point, however, where it is necessary to intro-
duce an explicitly geographic concept of insularity into the discussion
- for the simple reason that geography is of the first importance in
assessing a group's influence within the American political system.
For present purposes, it will suffice to restrict our speculations to two
simple geographic alternatives. On the one hand, our sociologically
insular minority might also be geographically insular: concentrated in
a relatively small number of places in the United States. On the other
hand, geographic insularity might not accompany sociological insular-
ity. Indeed, at the limit, the I-group might be evenly spread over the
fifty states and 435 congressional districts. For heuristic purposes, let
us begin, with the alternative that is empirically less common, but
analytically more tractable. Suppose that an I-group is distributed in
a geographically diffuse way: if it contains 12 percent of the national
population, it accounts for 12 percent of each congressional district.
Now compare this geographically diffuse I-group with a D-group that
is both sociologically and geographically diffuse. Other things being
equal, which group is more likely to succeed in influencing Congress-
men?
The previous analysis suggests that the I-group will probably have
greater influence. Such a group is more likely to form a political lobby
peopled by credible leaders who remain in close touch with the insular
constituency they represent. When such lobbyists threaten a Con-
gressman with electoral retribution, they can expect a respectful hear-
ing. Even if the interest-group leaders can influence only 1o to 20
percent of their 12 percent of the population, no sensible politician
would lightly forfeit i to 2 percent of the vote. Of course, if it happens
that the I-group's interests are diametrically opposed to those of other
groups within a Congressman's electoral coalition, a reelection-maxi-
mizing politician might decide to ignore the I-group's demands. Yet
his reluctance to forsake the group will be greater than it would be if
he were dealing with a comparable D-group. The D-group is less
likely to have a well-organized lobby to press its cause. It is also less
likely to have the communications network necessary for the lobby's
leaders credibly to threaten Congressmen with the prospect of electoral

sociological settings in which a group interacts. A group is more insular in this respect as it
interacts in more settings: schools, churches, unions, and so forth. The second idea - call it
intensity - involves the importance members attach to any particular social setting. Thus,
some insular groups may enjoy ties of great breadth but little intensity, whereas others may
experience only a few narrow, but intense, sociological bonds. Although it is not necessary to
emphasize the distinction between breadth and intensity in this essay, the distinction may become
quite important in future work. See infra note 28.
 HARVARD LAW REVIEW [Vol. 98:713

retribution. In short, even if the I-group is distributed evenly through-

out the nation, it has a greater ability to exert political influence
through the ultimate currency of democratic politics: votes on election
day.
This conclusion is reinforced when we turn to the more realistic
case in which the middling I-group is distributed very unevenly
throughout the country. In this scenario, a middling minority could
reasonably expect to be a local majority - or at least a decisive voting
bloc - in 20 to 30 congressional districts. For the representatives of
these districts, the support of the I-group amounts to nothing less
than the stuff of political survival. In fact, for all our Carolene talk
about the powerlessness of insular groups, we are perfectly aware of
the enormous power such voting blocs have in American politics. The
story of the protective tariff is, I suppose, the classic illustration of
insularity's power in American history. 25 Over the past half-century,
we have been treated to an enormous number of welfare-state varia-
tions on the theme of insularity by the farm bloc, the steel lobby, the
auto lobby, and others too numerous to mention.2 6 In this standard
scenario of pluralistic politics, it is precisely the diffuse character of
the majority forced to pay the bill for tariffs, agricultural subsidies,
and the like, that allows strategically located Congressmen to deliver
the goods to their well-organized local constituents. Given these fa-
miliar stories, it is really quite remarkable to hear lawyers profess
concern that insular interests have too little influence in Congress.
Instead, the American system typically deprives diffuse groups of their
rightful say over the course of legislative policy. If there is anything
to Carolene Products, then, it cannot be a minority's insularity, taken
by itself - something more must be involved.

IV. DISCRETE AND INSULAR MINORITIES?

Could that something be the "discreteness" of a Carolene minority?
I begin with a question because it is not obvious whether most
constitutional lawyers endow the word "discrete" with independent
25 The classic study on this subject is E. SCHATTSCHNEIDER, POLITICS, PRESSURES AND
THE TARIFF (1935).
26 For a thoughtful review of the political science literature, see G. WILSON, INTEREST
GROUPS IN THE UNITED STATES (I98i). For some recent work by political scientists extending
the scope of pluralist analysis, see R. ARNOLD, CONGRESS AND THE BUREAUCRACY (979); M.
FIORINA, CONGRESS - KEYSTONE OF THE WASHINGTON ESTABLISHMENT (1977); and J. WIL-
SON, THE POLITICS OF REGULATION (1980). The role of interest groups in politics has also
been a growing concern of economists. See, e.g., R. NOLL & B. OWEN, THE POLITICAL
ECONOMY OF DEREGULATION: INTEREST GROUPS IN THE REGULATORY PROCESS (1983); Kalt
& Zupan, Capture and Ideology in the Economic Theory of Politics, 74 Am. ECON. REV. 279
(1984); B. PASHIGIAN, ENVIRONMENTAL REGULATION: WHOSE SELF-INTERESTS ARE BEING
PROTECTED? (Center for the Study of the Economy and the State, Working Paper No. 022,
Feb. 1984).
1985] CAROLENE PRODUCTS

significance in their understanding of the Carolene doctrine. 27 None-

theless, we can conceive the term in a way that adds something
important to the overall formula. I propose to define a minority as
"discrete" when its members are marked out in ways that make it
relatively easy for others to identify them. For instance, there is
nothing a black woman may plausibly do to hide the fact that she is
black or female. Like it or not, she will have to deal with the social
expectations and stereotypes generated by her evident group charac-
teristics. In contrast, other minorities are socially defined in ways
that give individual members the chance to avoid easy identification.
A homosexual, for example, can keep her sexual preference a very
private affair and thereby avoid much of the public opprobrium at-
tached to her minority status. It is for this reason that I shall call
homosexuals, and groups like them, "anonymous" minorities and con-
trast them with "discrete" minorities of the kind paradigmatically
exemplified by blacks.
This way of defining terms allows us to complement our analysis
of insularity in a natural way. While the insularity-diffuseness con-
tinuum measures the intensity and breadth of intragroup interaction,
the discreteness-anonymity continuum measures the ease with which
people outside a group can identify group members. It should be
plain that these two continua are not invariably associated with one
another. Blacks, for example, are both discrete and insular, whereas
women are discrete yet diffuse; homosexuals are anonymous but may
be somewhat insular, 2 8 whereas the poor are both relatively anony-
mous and diffuse. Because there is no necessary correlation between

27 A notable exception, however, is the person who witnessed footnote four's birth, Professor
Louis Lusky. In one discussion, Lusky adopts a conception of discreteness very similar to the
one advanced here. Defining it to mean "separate or distinct," Lusky rejects Graham v. Rich-
ardson's finding that aliens constitute a "discrete and insular minority," 403 U.S. 365, 372 (1971),
on the ground, anong others, that "many of them, who are Anglophones, pass unnoticed."
Lusky, Footnote Redux: A Carolene Products Reminiscence, 82 COLUM. L. REV. 1093, 1105
n.72 (1982). Although Lusky is right to say that this deprives Anglophones of "discreteness," I
cannot agree with his further suggestion that English-speaking aliens should be deprived of
Carolene protection. As I stressed at the outset, see supra p. 7M7, this essay considers the
relative disadvantage of only those groups that are participating actively in the normal political
process, not those categorically excluded from participation. Aliens, of course, are the prototype
of the latter group. In Carolene terms, the justification for their judicial protection appears in
footnote four's second paragraph, see supra note i, which I accept as valid, not the third
paragraph, whose future is the subject of this essay.
28 At this point the distinction between intensity and breadth, see supra note 24, may be
helpful in refining the text's qualifying "somewhat." Although homosexuals do not character-
istically share a broad range of social settings in which they interact as homosexuals, a few
social contexts do serve as loci for an intense reaffirmation of homosexual identities - most
notably, the network of homosexual bars and restaurants found in major American cities.
Predictably, this network provided an important organizational focus for the recent political
movement on behalf of homosexual rights. See J. D'EMILIo, SEXUAL POLITICS, SEXUAL COM-
MUNITIES 129-250 (1983).
 HARVARD LAW REVIEW [Vol. 98:713

discreteness and insularity, I shall treat discreteness as a distinct sub-

ject for analysis and consider how a group's place on the discreteness-
anonymity continuum can be expected to add to, or detract from, its
probable political influence.
Carolene takes a straightforward position on this question. In its
view, discreteness is a political liability. Once again, however, the
only thing that is obvious is that this is not obvious. The main reason
why has been elegantly developed in Albert Hirschman's modern
classic, Exit, Voice and Loyalty. 29 The book's title refers to three
nonviolent ways of responding to an unsatisfactory situation: if you
dislike something, you may try to avoid it (exit), you may complain
about it (voice), or you may grin-and-hope-for-improvement (loyalty).
Although these three responses may be related to one another in a
number of ways, the relationship between two of them - exit and
voice - is of special relevance here. People do not respond to a bad
situation by engaging in a random pattern of avoidance and protest.
Instead, according to Hirschman, an inverse relationship obtains: the
more exit, the less voice, and vice versa. The reason for this is
straightforward: the easier it is to avoid a bad situation, the less it
will seem worthwhile to comDlain, and vice versa.
This inverse relationship holds significant implications for the rel-
ative political strength of minorities at different points on the discrete-
ness-anonymity scale. 30 If you are a black in America today, you
know there is no way you can avoid the impact of the larger public's
views about the significance of blackness. Because exit is not possible,
there is only one way to do something about disadvantageous racial
stereotypes: complain about them. Among efficacious forms of com-
plaint, the possibility of organized political action will surely rank
high.
This is not to say, of course, that individual blacks, or members
of other discrete minorities, will necessarily lend their support to
interest-group activity. They may, instead, succumb to the tempta-
tions of free-riding and thus deprive the group of vital political re-
sources. But even if discreteness is no cure-all for selfishness, it does
free a minority from the organizational problem confronting an anon-
ymous group of comparable size. To see my point, compare the
problem faced by black political organizers with the one confronting
organizers of the homosexual community. As a member of an anon-
ymous group, each homosexual can seek to minimize the personal
harm due to prejudice by keeping his or her sexual preference a tightly
held secret. Although this is hardly a fully satisfactory response,

29 A. HIRSCHmAN, EXIT, VOICE AND LOYALTY (1970).

3o Although Hirschman explores the way his insight illuminates a variety of social contexts,
see id. at 44-54, he has not himself considered the application developed in the paragraphs that
follow.
19851 CAROLENE PRODUCTS

secrecy does enable homosexuals to "exit" from prejudice in a way

that blacks cannot. This means that a homosexual group must con-
front an organizational problem that does not arise for its black coun-
terpart: somehow the group must induce each anonymous homosexual
to reveal his or her sexual preference to the larger public and to bear
the private costs this public declaration may involve.
Although some, perhaps many, homosexuals may be willing to pay
this price, the fact that each must individually choose to pay it means
that this anonymous group is less likely to be politically efficacious
than is an otherwise comparable but discrete minority. For, by defi-
nition, discrete groups do not have to convince their constituents to
''come out of the closet" before they can engage in effective political
activity. So it would seem that Carolene Products is wrong again: a
court concerned with pluralist bargaining power should be more, not
less, attentive to the claims of anonymous minorities than to those of
discrete ones.

V. PREJUDICE
But surely it is time to stop playing Hamlet without the Prince.
The whole point of Carolene's concern with "discrete and insular
minorities" cannot be understood, I am sure you are thinking, without
grasping the final term of the formula: prejudice. Indeed, it has been
one of my aims to provoke precisely this reaction. By detailing all
the ways discrete and insular minorities gain political advantage over
diffuse and anonymous groups, I have meant to emphasize how heavy
a burden the idea of prejudice must carry in the overall argument for
Carolene Products. The burden is of two kinds: one empirical, the
other conceptual. To take them one at a time, I shall defer all
problems involved in conceptualizing prejudice 3 ' so that we may first
focus upon the empirical side of the matter.

A. Questions of Fact
Carolene's empirical inadequacy stems from its underinclusive con-
ception of the impact of prejudice upon American society. It is easy
to identify groups in the population that are not discrete and insular
but that are nonetheless the victims of prejudice, as that term is
commonly understood. Thus, the fact that homosexuals are a rela-
tively anonymous minority has not saved the group from severe prej-
udice. Nor is sexism a nonproblem merely because women are a
diffuse, if discrete, majority. Prejudice is generated by a bewildering
variety of social conditions. Although some Carolene minorities are
seriously victimized, they are not the only ones stigmatized; nor is it

31 See infra pp. 737-40.

 HARVARD LAW REVIEW [Vol. 98:713

obvious that all Carolene minorities are stigmatized more grievously

than any other non-Carolene group. Why should the concern with
"prejudice" justify Carolene's narrow fixation upon "discrete and in-
sular" minorities?
The answer seemed easy in a world in which members of the
paradigmatic Carolene minority group - blacks - were effectively
barred from voting and political participation. Something is better
than nothing: whatever the organizational problems engendered by
anonymity and diffuseness, surely they are not nearly so devastating
as total disenfranchisement. 32 As we turn toward the future, however,
it is far less clear that such selective perception makes constitutional
sense. Nonetheless, I shall give Carolene the benefit of the doubt by
sketching a "pariah" model of the political process in which Carolene's
emphasis on the fate of discrete and insular minorities will still seem
empirically plausible. 3 3 As we move beyond the pariah model, how-
ever, anonymous or diffuse minorities will increasingly emerge as the
groups that can raise the most serious complaints of pluralist disem-
powerment.
i. The Pariah Model. - Assume a polity in which middling
minorities - in the 1o to 20 percent range - attain majority status
in a significant number of congressional districts because of the way
their insularity interacts with the geographic biases of the American
political system. Nonetheless, the minority representatives these
groups elect are entirely ineffective in Congress - because all re-
maining Congressmen refuse to bargain with them in any way. Thus,
imagine that you are a nonminority Congressman trying to get a
legislative coalition together to support a bill of central importance to
your district. After extensive wheeling and dealing, you come within
20 votes of your goal, but the only possible source of support remain-
ing is a bloc of 25 minority Congressmen. Despite your fervent desire
to pass the bill for your district, you do not even try to interest them
in joining your legislative coalition. Defeat of your priority bill is
preferable to victory with the aid of congressional pariahs. And so
you grin and bear it as your bill goes down to defeat, thanks to your
refusal to deal.
Sound implausible? Nonetheless, there are conditions in which a
nonminority Congressman, concerned with maximizing his chances of
reelection, might treat his minority brethren like complete pariahs.
Imagine, for example, that a majority Congressman's constituency

32 See A. BICKEL & B. SCHMIDT, THE JUDICIARY AND RESPONSIBLE GOVERNMENT 908-
go (Oliver Wendell Holmes Devise History of the Supreme Court of the United States No. 9,
1984) (recounting the Supreme Court's response to the disenfranchisement of blacks).
33 Even within this "pariah" model, however, it is possible that some anonymous or diffuse
groups might be as politically disadvantaged by prejudice as any discrete and insular minority.
The point of sketching the pariah model is simply to describe a political system in which
Carotene's narrow focus, although not necessarily correct, becomes "empirically plausible."
1985] CAROLENE PRODUCTS

were so prejudiced against the discrete and insular minority that no

legislative benefit the Congressmen might deliver would, in their eyes,
compensate them for the ideological affront they would suffer from
seeing their Representative soliciting minority support. The magni-
tude of this prejudice must be very great indeed to preempt concern
for practical matters. Moreover, this condition, let us call it "preemp-
tive prejudice," must pervade a very substantial number of congres-
sional districts before the pariah model is applicable. Yet it is only
by indulging in something like these strong empirical assumptions that
Carolene can claim that the effects of prejudice plainly outweigh the
political advantages enjoyed by minorities that are discrete and insu-
lar. 34
2. Beyond the Pariah Model. - Once we deny the general em-
pirical validity of the pariah model, our assessment of the political
impact of a discrete and insular group will invariably be more com-
plex. Such a minority may be expected to bring to the bargaining
process an asset that many diffuse and anonymous groups lack -
namely, 20 to 25 Representatives devoted to the energetic pursuit of
the minority's interests. 35 Whether this substantial bargaining advan-
34 By introducing the idea of monitoring costs, it is possible to weaken the preemptive-
prejudice condition and still come up with the pariah result. In the monitoring-cost model, we
may assume that if the Representative had a chance to provide his constituents with all the
facts about his bargaining situation, he would be able to convince most of them that bargaining
with the pariahs was worthwhile. The trouble is that most of the Congressman's constituents
do not have the time or inclination to follow the kind of blow-by-blow account of the goings-
on in Washington that might make their Congressman's deal with the pariahs seem rational.
Given the existence of monitoring costs, it is conceivable that constituents might follow a per
se rule against deals with pariahs in the belief that such deals are so rarely justified that it is
not worthwhile evaluating them individually.
Nonetheless, although this response is one way constituents might react to heavy monitoring
costs, I do not think it a likely response. Most of us are perfectly aware that politics often
makes for strange bedfellows, and that our Congressman cannot be expected to vote on all
matters precisely the way we want him to. Except under exceptional circumstances, we will
not turn against a Congressman on the basis of a few votes.
If this observation is true, introducing monitoring costs into the model might make the pariah
result seem less likely, rather than more so. Monitoring costs may enable majority Congressmen
to bargain with their minority colleagues in ways they hope their prejudiced constituents will
readily forgive - or will not even discover. Thus, even after including monitoring costs, the
pariah model will begin to seem a plausible description of political reality only if society is
plagued by prejudice that is both pervasive and exceptionally intense.
3S Twenty-one blacks served in the House of Representatives in x984. The number of black
Congressmen has increased steadily over the past three decades: from two Representatives in
the early I95os, to four in 196o, to nine Representatives and one Senator (Edward Brooke of
Massachusetts) in 1970, to 17 Representatives in ig8o. See BUREAU OF THE CENSUS, U.S.
DEP'T OF COMMERCE, THE SOCIAL AND ECONOMIC STATUS OF THE BLACK POPULATION IN
THE UNITED STATES 154, x56 (1979) (P-23, No. 80); T. CAVANAGH & D. STOCKTON, BLACK
ELECTED OFFICIALS AND THEIR CONSTITUENCIES 2 (1983). Hispanic representation in Con-
gress has undergone a similar increase. One Hispanic Congressman was elected in i96o, six in
197o, and eleven in 1984; this last figure includes two Senators, Dennis Chavez and Joseph
Montoya, both of New Mexico. Telephone interview with Lisa Navarrete, spokesperson for the
 HARVARD LAW REVIEW [Vol. 98:713

tage is completely negated by prejudice will depend upon many par-

ticular institutional factors that will change from decade to decade, if
not day to day - facts, -moreover, that the Supreme Court of the
United States would have special difficulties elaborating upon in a
judicial opinion.
It is within this context that I turn to a very different approach
to Carolene presented by John Hart Ely in his important work, De-
mocracy and Distrust.36 Among its many virtues, the book explicitly
recognizes that the conventional Carolene wisdom about the power-
lessness of discrete and insular minorities is "in need of some reex-
amination." ' 3 7 Yet Dean Ely does not attempt the interest-group anal-
ysis that has thus far engaged our energies. Instead, he relies
exclusively on a social-psychological approach to the legislative pro-
cess. On his view, the critical thing about prejudice is the way it
allows legislators to stereotype "discrete and insular" minorities. It is
this legislative propensity to divide the world into "we-they" categories
that, according to Ely, lies at the core of Carolene's concern: "we"f
legislators will both overestimate the dangers posed by the "they"
group and underestimate its similarities to the "we" group. As a
consequence, legislation that disadvantages "they" groups will be
based on an intolerably distorted perception of social reality. 38
Even if we were to accept a "we-they" view of legislative psy-
chology, 39 Ely's failure to recognize the limits of the pariah model

Congressional Hispanic Caucus, Inc. (Sept. 6, 1984). According to the Census Bureau, Hispanics
accounted for 6.4% of the American population in i98o. See BUREAU OF THE CENSUS, supra
note 14, at 1-21.
36 J. ELY, supra note 6.
37Id.at 152.
38 See id. at 157-6o.
39 I, for one, do not find Ely's social psychology persuasive. No doubt many people are
trapped by "we-they" stereotypes of the kind Ely describes. Yet Ely is not concerned with
humankind in general, but with the particular people who serve as legislators. When we narrow
our focus to this small group, I do not believe that Ely's simplistic psychology is the norm. No
good politician is eager to antagonize any portion of his constituency unnecessarily - if only
because he can never tell whether, in five or ten years, he will desperately require the support
of a group that is now electorally unnecessary. Rather than alienate groups of voters, most
would-be politicians - right, left, and center - make the rounds of all significant social and
ethnic groups in our society. In this endless round of labor-union breakfasts, bankers' lunches,
and Knights of Columbus dinners, would-be Congressmen may be expected to be extremely
attentive to currents of opinion - especially among groups who are not their natural allies -
in an effort to discover ways to reach out to otherwise indifferent or hostile constituencies. If
a Congressman learns anything from these countless hours on the hustings, it is that people
very different from himself are people too, and that he can understand distinctive points of
view even if he has not lived his whole life with every group in America.
Of course, a politician may have to make some hard decisions about the particular interests
he will champion in forthcoming elections. And, in the end, he may well decide that an all-
out appeal to prejudice is in his best electoral interest. Yet it is naive to assume that most
successful politicians make such decisions on the basis of simplistic "we-they" stereotypes rather
than hard-headed electoral calculations. Certainly this is not the consensus of modern political
 CAROLENE PRODUCTS

serves as an independent ground for questioning his conclusions.

Quite simply, our efforts in bargaining theory have led us to expect
that "middling minorities" of the "discrete and insular" kind will elect
a significant number of Representatives - say 20 to 25 - who are
extremely responsive to their interests. As long as these politicians
are not treated like pariahs, they can become a potent legislative force -
- trading votes with other legislators to further the objectives of their
own constituents. Thus, if minority politicians also harbor "we-they"
prejudices, the legislative dynamic will inevitably be far more complex
than Ely allows. Rather than remorselessly reflecting a monochro-
matic view of social reality, legislation will be the joint product of
different "we-they" prejudices held by the different politicians sitting
around the legislative bargaining table. This is more than diffuse or
anonymous minorities can expect: their representatives may not even
be at the bargaining table. 40 As soon as he moves beyond the narrow
confines of the pariah model, Dean Ely cannot rehabilitate Carolene's
exclusive focus on discrete and insular minorities through "we-they"
psychology alone.
Dean Ely seems aware of all this. Although Democracy and Dis-
trust does not contain a fully developed analysis of minority legislative
power, it does hint at an approach different from the one I advance
here. Dean Ely suggests that minority politicians may suffer from a
distinctive psychological affliction: while other Congressmen act on
"we-they" prejudices in favor of their own constituents, minority pol-
41
iticians may accept the very stereotypes they should be challenging.

scientists. To the contrary, an important strand of recent work explicitly tries to model American
Congressmen as rational vote-maximizers. See, e.g., M. FiORINA, supra note 26, at 39-49; D.
MAYHEW, CONGRESS: THE ELECTORAL CONNECTION 3-5, 13-38 (1974). Dean Ely does not
discuss this literature in Democracy and Distrust.
40 This is not to say that all anonymous or diffuse minorities are worse off than all discrete
and insular ones. On the one hand, prejudice against a discrete and insular minority may, as
in the pariah model, entirely offset the group's other organizational advantages. See supra pp.
732-33. On the other hand, the organizational disadvantages suffered by an anonymous and
diffuse group may sometimes be offset by a "third-party beneficiary" effect. In such cases, a
group does not gain power through its own direct organizational efforts, but is instead the happy
beneficiary of political deals struck by other, better organized interests. This third-party effect
may operate so powerfully that it more than compensates the group for its weaknesses in direct
bargaining. When this is so, I do not mean to suggest that a diffuse or anonymous group may
rightly claim systematic disadvantage in the ongoing pluralist process.
My point is simply that some anonymous or diffuse groups - homosexuals, women, the
poor - cannot in fact expect more from the third-party beneficiary effect than can many classic
discrete and insular minorities. In such cases, the organizational weaknesses of anonymous or
diffuse groups make them more disadvantaged in pluralist bargaining, all else remaining equal,
than the groups upon whom Carolene courts have focused their concern. Advocates for such
diffuse or anonymous interests, then, should no longer be embarrassed by their groups' evident
differences from the classic discrete and insular minority. For it is precisely these differences
that make their constitutional case stronger, not weaker.
41 See J. ELY, supra note 6, at 165-66.
 HARVARD LAW REVIEW [Vol. 98:713

If this point were conceded, Ely's argument would take on a self-

sealing quality: no matter how actively minority representatives par-
ticipated in the bargaining process, they would only reinforce, and
never challenge, prevailing prejudices.
Dean Ely shows great restraint in dealing with this suggestion of
minority "false consciousness." While he says that "[t]he general idea
is one that in some contexts has merit,"' 42 his book does not, in fact,
spend very much space defending and elaborating it. 43 I believe,
moreover, that an appeal to "false consciousness" cannot be elaborated
in a way that makes constitutional sense.
The first question to ask about "false consciousness" is an empirical
one: will the rising generation of minority politicians in fact passively
accept debasing stereotypes? I see no reason to project such a grim
image upon our future. To the contrary, the classic prejudices are
under vigorous challenge by powerful voices emerging from a broad
range of discrete and insular communities. This is not to say, of
course, that minority-group representatives will unanimously agree
about matters of public policy. But I do not believe it useful to
analyze these inevitable - and often reasonable - disagreements by
labeling one or another group of disputants as the victims of "false
consciousness."
The second question is: even if some social psychologist could
"prove" the existence of false consciousness, should the Supreme Court
transform this social phenomenon into an assumption of constitutional
law? We are dealing here not with an academic scientific inquiry, but

42 Id. at 165.
43 Indeed, Dean Ely devotes most of his brief discussion to explaining why he does not think
that women as a class have internalized an unquestioning acceptance of male stereotypes. See
id. at i65-66. In contrast, he spends less time explaining the conditions under which the
problem of false consciousness might rise to constitutional significance. All he does is quote
with approval a brief passage from a concurring opinion by Justice Marshall: "'Social scientists
agree that members of minority groups frequently respond to discrimination and prejudice by
attempting to disassociate themselves from the group, even to the point of adopting the majority's
negative attitudes towards the minority.'" Id. (quoting Castaneda v. Partida, 430 U.S. 482, 503
(1977) (Marshall, J., concurring)). The great studies cited by Justice Marshall, however, date
from the 194os and 195os, see, e.g., G. ALLPORT, THE NATURE OF PREJUDICE 150-53 (1954);
A. ROSE, THE NEGRO'S MORALE 85-95 (1949); Bettelheim, Individual and Mass Behavior in
Extreme Situations, 38 J. ABNORMAL & SOC. PSYCHOLOGY 417 (1943), and their conclusions
should not be projected mechanically onto the psyches of the next generation's minority politi-
cians. Moreover, the case in which Justice Marshall made his remarks involves a problem
different from the one that concerns us here. The question in Castaneda was simply whether
the fact that Spanish speakers were the "governing majority" in a county should lead to a
change of the rules establishing a prima facie case of intentional discrimination against the
selection of Mexican-American grand jurors. In holding that the elements of a prima facie case
were constitutionally unaffected, the Court was simply recognizing that bias in the administration
of the law can outlive the political conditions that gave rise to it. This conclusion in no way
implies, however, that minority politicians will systematically suffer from a pernicious form of
false consciousness.
 CAROLENE PRODUCTS

with a question of institutional relationships. In branding minority

politicians as victims of "false consciousness" on the pages of the
United States Reports,4 4 the Supreme Court would be consigning them
to a peculiarly demeaning constitutional status. Henceforth, they -
and they alone - would be deemed constitutionally incapable of
discharging the representative functions of democratically elected leg-
islators. Such a declaration would make a mockery of Carolene's
promise. Rather than attempting to approximate the results of a
perfect pluralist democracy, the Court would be protecting minority
rights by emphatically impugning the capacity of these very same
minorities to engage in democratic politics at all.
Once we reject the appeal to false consciousness, I see no way to
avoid reformulating Carolene's implications if it is to serve the needs
of a polity moving beyond the pariah model. This doctrinal reorien-
tation, it bears repeating, does not suppose that we have reached a
point in our history in which prejudice against discrete and insular
minorities has become a thing of the past. Instead, it simply recog-
nizes that many of these groups can deal with the problem politically
in ways that other victims of prejudice may be powerless to match.
It is the members of anonymous or diffuse groups who, in the future,
will have the greatest cause to complain that pluralist bargaining
exposes them to systematic - and undemocratic - disadvantage.

B. Questions of Value
But Carolene's failure to recognize the political predicament of
anonymous or diffuse groups that are victims of prejudice is only half
the problem; the other half is more conceptual, but no less troubling.
The idea of "prejudice" is simply unequal to the task assigned it
within the overall Carolene analysis. Recall that Carolene's promise
is a form of argument that allows a court to say that it is purifying
the democratic process rather than imposing its own substantive val-
ues upon the political branches. And yet it is just this process ori-
entation that is at risk when a Carolene court undertakes to identify
the prejudices that entitle a group to special protection from the
vagaries of pluralist politics. One person's "prejudice" is, notoriously,
another's "principle." How, then, do we identify a group for Carolene

44 It would be no more acceptable for a court to excise the offensive language from its
opinion and covertly use the offensive principle as an operating premise of its constitutional
approach to minority rights. See Monaghan, Taking Supreme Court Opinions Seriously, 39 MD.
L. REV. 1, 2-3, 19-22 (1979). Indeed, I believe such an approach would be even more
disrespectful of minority groups than would a candid declaration of constitutional principle.
Whatever may be said in support of mystification in other contexts, see G. CALABREsI, A
COMMON LAw FOR THE AGE OF STATUTES I72-81 (1982); G. CALABREsI & P. BOBBITT, TRAGIC
CHOICES 17-28, 195-99 (1978), I cannot believe that thoughtful lawyers would embrace it in
this one.
 HARVARD LAW REVIEW [Vol. 98:713

protection without performing the substantive analysis of constitu-

tional values that Carolene hopes to avoid?
The kind of answer required is clear enough. To redeem Caro-
lene's promise, the judicial identification of a prejudice cannot depend
upon the substance of the suspect view, but must turn on the way in
which legislators come to hold their belief. The process-oriented ar-
gument goes something like this: although each of us cannot always
expect to convince our legislators, we can at least insist that they treat
our claims with respect. At the very least, they should thoughtfully
consider our moral and empirical arguments, rejecting them only after
conscientiously deciding that they are inconsistent with the public
interest. If a group fails to receive this treatment, it suffers a special
wrong, one quite distinct from its substantive treatment on the mer-
its. 45 And it is this purely processual kind of prejudice that constitutes
the grievance Carolene courts may endeavor to remedy without en-
gaging in the suspect task of prescribing substantive values.
Of course, no one imagines that it will be easy for the courts to
act effectively on behalf of the victims of purely processual prejudice.
To the contrary, a rich and provocative literature describes the diffi-
culties involved in legislative mind reading. 4 6 For present purposes,
I shall assume that the partisans of the process approach can solve
these problems in one way or another. My own objections to the
enterprise arise only after these threshold difficulties have been over-
come. Thus, I shall assume that judges can accurately gauge the
quality of legislative deliberation behind a statute, and I shall ask you
to speculate about what they would find if they deployed their high-
powered techniques on a representative range of legislation.
To begin with the obvious: judges would find that a lot of purely
processual prejudice does exist in the case of classic discrete and
insular minorities. There are plenty of racial and religious bigots who
have never stopped for a moment to consider the arguments and
interests on the other side. But this obvious fact is hardly sufficient
to justify Carolene's selective focus on discrete and insular minorities.

45 For insightful elaborations of this theme, see J. ELY, supra note 6, at 81-86, and Sunstein,
Naked Preferences and the Constitution, 84 COLUM. L. REv. 1689, 1689-98 (1984).
46 The Court's uncertainties on this subject, compare Palmer v. Thompson, 403 U.S. 217,
224-26 (197r) (statute's effect, not legislature's motive, is the central issue for equal protection
analysis), with Washington v. Davis, 426 U.S. 229, 238-48 (976) (discriminatory intent is
necessary to establishing equal protection violation), has generated an unending stream of
commentary. Among the works I have found most enlightening are J. ELY, supra note 6, at
136-45; Brest, Palmer v. Thompson: An Approach to the Problem of UnconstitutionalLegislative
Motive, 1971 SuP. CT. REv. 95; Ely, Legislative and Administrative Motivation in Constitu-
tional Law, 79 YALE L.J. 1205 (1970); Schnapper, Perpetuation of Past Discrimination, 96
HARv. L. REV. 828 (1983); Simon, Racially Prejudiced Governmental Actions: A Motivation
Theory of the ConstitutionalBan Against Racial Discrimination, 15 SAN DIEGO L. REV. 1041
(1978); and Note, Reading the Mind of the School Board: Segregative Intent and the De Factol
De Jure Distinction, 86 YALE L.J. 317 (1976).
1985] CAROLENE PRODUCTS

The critical question, instead, is whether purely processual prejudice

is more characteristic in this context than in the political treatment of
other interests and opinions. What of the prejudice middle-class leg-
islators may have toward the poor? Heterosexuals toward homosex-
uals? More fundamentally, are we right to assume that only those
opposed to "progressive" causes can be processually prejudiced?
Let me propose a test case. Imagine that, after reading Herbert
Wechsler's famous essay, 47 a group of conservative legalists becomes
sincerely convinced that Brown v. Board of Education could not in
fact be based on neutral principles and so does not deserve its place
as a cornerstone of our constitutional law. Acting on this conviction,
the group begins a campaign advocating a constitutional amendment
to repeal Brown and generates some modest interest among conser-
vatives across the country. Arriving in Washington, D.C., with their
legal process arguments elaborately developed, the group proceeds to
the lobbies of Congress. How do you think the group would be
received? Would most Representatives be willing and able to confront
the Wechslerian arguments with a thoughtful defense of our consti-
tutional commitment to equality? Or would they respond in a pro-
cessually prejudiced fashion - peremptorily brushing aside the
Wechslerians' arguments with a catch-phrase or two that fails to join
issue?
This is, in principle, an empirical question - though, like many
others, it will never get a good empirical answer. Nonetheless, if my
study of politics has taught me anything, I would not expect the
agitating Wechslerians to receive a processually unprejudiced response
on Capitol Hill. As far as I can tell, any large representative assembly
will contain a bewildering variety of human types - from the elab-
orately thoughtful to the superficially unquestioning. 48 It is simply
self-congratulatory to suppose that the members of our own persuasion
have reached their convictions in a deeply reflective way, whereas
those espousing opinions we hate are superficial. Instead, a thoughtful
judge can expect to find an abundance of stereotype-mongers and
knee-jerks on all sides of every important issue - as well as many
who have struggled their way to more considered judgments. Given
the complexity of the human comedy, a judge is bound on a fool's
errand if he imagines that the good guys and bad guys of American
politics can be neatly classified according to the seriousness with which
they have considered opposing points of view. Processual prejudice
is a pervasive problem in the American political system.
But if this is right, Carolene cannot justify its concern with discrete
and insular minorities without calling on judges to engage in a very

47 Wechsler, Toward Neutral Principles of ConstitutionalLaw, 73 HARv. L. REv. i (I959).

48 For a revealing sketch of the diverse foundations of congressional judgment, see J.
KINGDON, CONGRESSMEN'S VOTING DECISIONS 227-40 (2d ed. i98i).
 HARVARD LAW REVIEW [Vol. 98:713

different kind of judgment, one dealing with the substance of racial

and religious prejudice. In doing so, the judge need not try to play
the elaborate psychological and political guessing game required to
assess the extent to which a statute is the product of a prejudiced
refusal to give a respectful hearing to disfavored interests and opin-
ions. Instead, she proceeds to a more familiar judicial inquiry into
the nature of the substantive reasons that might plausibly justify the
legislature's assertion of authority. If the only plausible reasons for
the statute's enactment offend substantive constitutional principles,
the groups aggrieved by the statute are declared victims of "prejudice";
if not, not. Although this judicial inquiry into the rational foundations
of a statute may sometimes require a focused inquiry into the data
available to, or even the subjective opinions of, particular public
officials, 49 the critical legal question is of a very different kind: why
are the political principles endorsed by some groups judicially recog-
nized as vindicating the constitutionality of a statute, while others are
viewed as inadmissible "prejudices" delegitimating a statute's claim to
constitutionality?
If Carolene somehow hoped to find a shortcut around this sub-
stantive inquiry into constitutional values, its journey was fated to
fail from the outset. The difference between the things we call "prej-
udice" and the things we call "principle" is in the end a substantive
moral difference. And if the courts are authorized to protect the
victims of certain "prejudices," it can only be because the Constitution
has placed certain normative judgments beyond the pale of legiti-
macy.5 0

VI. FROM CRITIQUE TO RECONSTRUCTION

Paradoxically, it is by reflecting upon this last mistake that we
may begin to reorganize Carolene in a way that will renew its promise
for the next generation. Our discussion suggests that, in responding
to "prejudice against discrete and insular minorities," Carolene courts
have in fact been trying to force a single formula to express two very
different insights. Carolene's first insight is that some groups suffer
from systematic disadvantages in pursuing their interests in the plu-
ralist bargaining process normally central to American politics. On
this view, the Court appears as a perfecter of pluralist democracy. It

49 For the most recent judicial exploration of this quagmire, see the concurring and dissenting
opinions of Justices Brennan and Rehnquist in Kassel v. Consolidated Freightways Corp., 450
U.S. 662, 679, 687 (198x).
SOI do not wish to suggest that this is a new discovery. For thoughtful elaborations of
parallel themes, see Brest, The Substance of Process, 42 OHIo ST. L.J. 131, 134-37 (ig8x), and
Tribe, supra note 6, at 1072-77.
1985] CAROLENE PRODUCTS

corrects political results generated by unfair bargaining advantages

but does not question the substantive values pursued by the partici-
pants.
Carolene's emphasis on "prejudice," however, announces a second,
quite different, conception of judicial review. Here the courts do not
enter as the perfecters of pluralist democracy, but as pluralism's ulti-
mate critics. In exercising this critical function, courts insist that there
are certain substantive principles - Carolene calls them "prejudices"
- that pluralist politicians are simply not allowed to bargain over in
normal American politics. It is only when statutes emerging from the
pluralist process do not offend these constraining constitutional values
that they have the force of law in our political system.
By collapsing the perfecting and critical functions of judicial re-
view into a single formula, Carolene poses a formidable intellectual
problem for lawyers in the years ahead. Indeed, I believe the future
vitality of both functions will significantly depend upon the success
with which constitutional lawyers manage to disentangle the two
themes from one another - and so permit each to receive the sus-
tained doctrinal elaboration it deserves.
Consider first the Supreme Court's function as a perfecter of plu-
ralist democracy. Here Carolene's fundamental concern seems more
salient today than when first announced a half-century ago. In 1938,
Americans were only just beginning to move beyond a political system
that limited interest-group struggle over federal policy to a few classic
pork-barrel issues: the tariff, internal improvements, land distribution,
and the like. Yet Carolene was remarkably prescient in recognizing
that the downfall of the Old Court's laissez-faire jurisprudence had
transformed the structure of pluralist bargaining into an issue of prime
constitutional importance. In the system of activist government in-
augurated by the New Deal, the course of pluralist bargaining would
have a profound and pervasive impact upon the shape of every Amer-
ican's life. Within this setting, the existence of systematic bargaining
disadvantage erodes the perceived legitimacy of our constitutional
regime in the eyes of broad segments of the American population. If
Carolene had not already impressed a concern with the integrity of
the pluralist process onto our constitutional consciousness, another
case would inevitably have made the same point.
The problem comes only in the way in which Carolene elaborates
its fundamental insight in terms of judicial doctrine. Even in this
regard it would be wrong to judge Carolene's focus upon "discrete
and insular minorities" too harshly. The case was handed down in
the shadow of the infamous Nuremberg laws, which had stripped
German Jews of all their previously established civil rights. This
served to recall, in the starkest way, the grim process by which black
Americans had been stripped of their civil rights in the aftermath of
 HARVARD LAW REVIEW [Vol. 98:713

Reconstruction. 5 ' In both cases, the dreadful consequences of political

exclusion were plain to all who cared to look. Moreover, it was -
and remains - obvious that the political decision to disenfranchise
these groups was made vastly easier by virtue of their discreteness
and insularity. It was precisely these characteristics that permitted
German Nazis and American white supremacists to portray Jews and
blacks as aliens within the body politic, to be used and abused in any
way the master race desired.
Against this historical background,5 2 the Carolene Court was ab-
solutely right to emphasize the special vulnerability of discrete and
insular minorities, as well as the fundamental importance of ensuring
their effective participation in the democratic process. After a gen-
eration of renewed struggle for civil rights, however, it no longer
follows that the discreteness or insularity of a group will continue to
serve as a decisive disadvantage in the ongoing process of pluralist
bargaining. Rather than find this fact embarrassing, constitutional
lawyers ought to be proud of it. It suggests that, despite the racial
and religious prejudices that still haunt our society, Americans have
made some progress toward a more just polity.
It will be a tragedy, however, if the progress we have made serves
to justify a refusal to develop and extend Carolene's concern with the
integrity of pluralist process to contemporary conditions. Long after
discrete and insular minorities have gained strong representation at
the pluralist bargaining table, there will remain many other groups
who fail to achieve influence remotely proportionate to their numbers:
groups that are discrete and diffuse (like women), or anonymous and
somewhat insular (like homosexuals), or both diffuse and anonymous
(like the victims of poverty). If we are to treat Carolene as something
more than a tired formula, constitutional lawyers must develop para-
digms that detail the systematic disadvantages that undermine our
system's legitimacy in dealing with the grievances of these diffuse or
anonymous groups.
At the same time that we enrich the capacity of constitutional law
to perfect pluralist democracy, we must also reaffirm a second fun-
damental mission for judicial review: to expound the ultimate limits
imposed on pluralist bargaining by the American constitutional
system. In the exercise of this critical function, the courts insist that,
for all our plural differences, We the People of the United States do
have a set of fundamental commitments that bind us together in ways
that our interest-group representatives are not normally elected to
modify. It is this idea of higher law that must be taken with renewed
seriousness if we are to sustain judicial protection for racial and
religious minorities in the coming generation. Although, as we have

S See A. BICKEL & B. SCHMIDT, supra note 32, at 721-990.

52 For further elaboration, see Cover, supra note 3, at 1294-1304.
1985] CAROLENE PRODUCTS

seen, the Carolene effort to protect minorities ultimately required the

elaboration of substantive constitutional principles, the Carolene tra-
dition's reliance on bad political science has made it seem possible to
avoid the sustained inquiry into democratic theory that substantive
judicial review entails. More particularly, Carolene's focus on pluralist
bargaining has subtly encouraged the belief that pluralism is the alpha
and the omega of the American constitutional system, and that any
effort by the courts to challenge the substantive values generated by
legislative compromise is necessarily antidemocratic.
We must repudiate this reduction of the American Constitution to
a simple system of pluralist bargaining if we are to reassert the legit-
imacy of the courts' critical function. Although the bargaining model
captures an important aspect of American politics, it does not do
justice to the most fundamental episodes of our constitutional history.
We make a mistake, for example, to view the enactment of the Bill
of Rights and the Civil War Amendments as if they were outcomes
of ordinary pluralist bargaining. Instead, these constitutional achieve-
ments represent the highest legal expression of a different kind of
politics - one characterized by mass mobilization and struggle that,
after experiences like the Revolution and the Civil War, yielded fun-
damental principles transcending the normal processes of interest-
group accommodation. It is only by reasserting the relevance of this
tradition of constitutional politics, as I have called it,5 3 that we shall
gain the necessary perspective to put pluralist bargaining in its place
as one - but only one - form of American democracy, and the lesser
form at that.
Not that the Carolene tradition - or Carolene itself - is entirely
oblivious to the limits of pluralist bargaining. Indeed, it was just this
issue that initially provoked Chief Justice Hughes to press for a re-
vision of Carolene in the opinion-writing process. While Justice
Stone's early draft had focused exclusively on the pluralist perfection
rationale, the Chief Justice believed that something essential was miss-
ing in the case for judicial review. In response to this expression of
concern, Justice Stone added a first paragraph to footnote four that
takes our higher-law tradition more explicitly into account. Thus,
before addressing the pluralist themes we have considered here, Car-
olene noted that "the presumption of constitutionality" may also be
overcome "when legislation appears on its face to be within a specific
prohibition of the Constitution, such as those of the first ten amend-
ments. '"54
In calling the Bill of Rights "specific," Justice Stone doubtless
wished to emphasize that the Court had learned its lesson in I937 and

53 I consider this tradition at length in Ackerman, supra note 12.

54 For the complete text of the footnote's first paragraph, see note i. For an authoritative
account of the footnote's genesis, see Lusky, supra note 27, at io96-iioo.
 HARVARD LAW REVIEW [Vol. 98:713

would not use the Constitution's grand abstractions to revive the

laissez-faire capitalism of the Lochner era. Nonetheless, by framing
its pledge of judicial restraint in this way, Carolene added a distortion
of its own. For it intimated that the judicial process of articulating
the nature of our higher law values can be reduced to a mechanical
effort to apply "specific" constitutional rules to predetermined facts.
Such a position requires judges to repudiate the main line of modern
American legal thought, which - from Pound to Dworkin 55 - is one
long elaboration of the inadequacies of mechanical jurisprudence.
Even more fundamentally, it trivializes the nature of the American
people's higher-law achievement. Our Constitution does not even
attempt to provide a detailed set of rules that might suggest the
possibility of pseudomechanical application. Instead, our higher law
tradition gains its distinctive character precisely by speaking in ab-
stract and general terms about the nature of our basic rights. Hence,
in endowing the Bill of Rights with a false "specificity," Carolene not
only proffered a misleading and unattainable picture of responsible
judicial decisionmaking. It also diverted us from the main question:
having cleared away the laissez-faire debris of the Lochner era, can
we still reconstruct, out of authoritative sources, a legally cogent set
of higher-law principles that can continue to govern the pluralist
process in the name of We the People?
The point of this essay is not to answer this question, but to
convince you that it needs asking if we are to preserve the constitu-
tional rights of discrete and insular minorities during the coming
decades. I do not believe that the weaknesses in Carolene's defense
of minority rights will long remain a professional secret locked in the
pages of the Harvard Law Review. Instead, Carolene's errors will
become increasingly apparent on the surface of American political life.
Thanks largely to the achievements of the generation that looked to
Carolene for inspiration, black Americans today are generally free to
participate in democratic politics - and do so by the millions in every
national election. 56 Moreover, the predicted consequences of the dis-
creteness and insularity of black voters are beginning to be obvious
at every level of American government. From City Hall to Capitol
Hill, black politicians now aggressively represent their constituencies

ss See Dworkin, The Model of Rules, 35 U. CHI. L. REv. 14, 22-32 (1967) (distinguishing
legal principles from legal rules); Pound, Mechanical Jurisprudence, 8 COLUM. L. Rv. 605
(i9o8).
56 In the congressional elections of 1982, voting participation among whites was 5o%; among
blacks, 43%. SeeBUREAU OF THE CENSUS, U.S. DEP'T OF COMMERCE, VOTING AND REGIS-
TRATION IN THE ELECTION OF NOVEMBER 1982, at vi (1983) (P-2o, No. 383). Only 25% of
the Spanish-speaking population voted. The Census Bureau concludes that this disparity was
"primarily due to the large proportion of this population who reported that they were not
citizens." Id. at v. If we restrict our attention to those who were registered, 72% of Hispanics
voted, compared to 73% of blacks and 76% of whites. Id. at vi.
1985] CAROLENE PRODUCTS

in the citadels of power.5 7 Similarly, religious organizations are in-

58
creasingly involved in pressure-group politics.
I am not suggesting that America is on the way to becoming a
religious and racial utopia. Despite their political gains, blacks still
suffer under the weight of grossly disproportionate economic, educa- 59
tional, and social disadvantage, as well as sheer racial prejudice.
In light of these facts, it is far too early to say that we have redeemed
the promise of the thirteenth and fourteenth amendments. In contrast
to black political mobilization, the heightened involvement of organ-
ized religion can readily undermine our substantive constitutional leg-
acy - threatening the very values of religious toleration and free
exercise to which our higher law is committed.
Yet as long as we use Carolene rhetoric to express our constitu-
tional concerns with racial equality and religious freedom, we will
find ourselves saying things that are increasingly belied by political
reality. While constitutional lawyers decry the political powerlessness
of discrete and insular groups, representatives of these interests will
be wheeling and dealing in the ongoing pluralist exchange - winning
some battles, losing others, but plainly numbering among the organ-
ized interests whose electoral power must be treated with respect by
their bargaining partners and competitors. Gradually, this clash be-
tween constitutional rhetoric and political reality can have only one
result. As time goes by, the constitutional center will not hold: the
longer Carolene remains at the core of the constitutional case for
judicial review, the harder lawyers will find it to convince themselves,
let alone others, that judicial protection for the rights of "discrete and
insular minorities" makes constitutional sense.
For those who are constitutional conservatives in the deepest sense,
and who look upon our tradition of civil liberties as one of the greatest
achievements of American law, the challenges are clear. On the one
hand, if we are to remain faithful to Carolene's concern with the
fairness of pluralist politics, we must repudiate the bad political sci-
ence that allows us to ignore those citizens who have the most serious
complaints: the anonymous and diffuse victims of poverty and sexual
discrimination who find it most difficult to protect their fundamental
interests through effective political organization. On the other hand,

57 "Between 1965 and 1982, the number of black elected officials increased tenfold, from
about 5oo to more than 5,1oo . . . . Blacks have now been elected to every major category of
public office except the presidency, vice presidency, and governorship . . . ." T. CAVANAGH &
D. STOCKTON, supra note 35, at i. Blacks have made the most substantial gains at the local
level: in 1982, there were 465 black county officials, 2,451 elected blacks at the muncipal level,
and 563 in the judiciary or in law enforcement. Id. at 2. For the data on black Representatives
in Congress, see note 35.
58 See THE NErw CHRISTIAN RIGHT (R. Liebman & R. Wuthnow eds. 1983).
S9 See Wilson, The Urban Underclass, Social Dislocation, and Public Policy, in AMERICAN
MINORITIES AND CIVIL RIGHTS (L. Dunbar ed. forthcoming 1985).
746 HARVARD LAW REVIEW [Vol. 98:713

we must explain to our fellow Americans that there are constitutional

values in our scheme of government even more fundamental than
perfected pluralism - most notably, those that bar prejudice against
racial and religious minorities. If we persist in holding these rights
hostage to pluralist theory, we shall only end up mocking the proud
role that Carolene has played in the pursuit of constitutional values
over the past half-century. By failing to adapt Carolene's constitu-
tional theory to a changing political reality, we shall have passively
allowed the Constitution's profound concern for racial equality and
religious freedom to be trivialized into a transparent apologia for the
status quo.