University of Chicago Law School

Chicago Unbound
Journal Articles Faculty Scholarship

1998

Against Constitutional Theory

Richard A. Posner

Follow this and additional works at: http://chicagounbound.uchicago.edu/journal_articles

Part of the Law Commons

Recommended Citation
Richard A. Posner, "Against Constitutional Theory," 73 New York University Law Review 1 (1998).

This Article is brought to you for free and open access by the Faculty Scholarship at Chicago Unbound. It has been accepted for inclusion in Journal
Articles by an authorized administrator of Chicago Unbound. For more information, please contact unbound@law.uchicago.edu.
 NEW YORK UNIVERSITY
LAW REVIEW
VOLUME_ 73 APRIL 1998 NurwER 1

AGAINST CONSTITUTIONAL THEORY

RIcHARD A. POSNER*

In this Madison Lecture, Chief Judge Posner advocates a pragmaticapproach to

constitutionaldecisionmaking, criticizingconstitutionaltheoristswho conceal their
normative goals in vague and unworkable principles of interpretation. After dis-
cussingspecific constitutionaltheories as well as the legal academy's increasingreli-
ance on theory in genera Posner demonstrates the ineffectuality of constitutional
theory, using the Supreme Court's decisions in United States v. Virginia and
Romer v. Evans as examples. He argues not that these cases were necessarily
wrongly decided, but that the opinions lack the empiricalsupport diat is crucialto
sound constitutionaladjudication. Posnerurges law professorsto focustheirschol-
arship on forms of inquiry that will actually prove usefid to judges and concludes
by askingthat judges themselves recognize and acknowledge the limitationsof their
empiricalknowledge

INTRODUCrION

Constitutional theory, as I shall use the term, is the effort to de-

velop a generally accepted theory to guide the interpretation of the
Constitution of the United States. It is distinct on the one hand from
inquiries of a social scientific character into the nature, provenance,
and consequences of constitutionalism-the sort of thing one associ-
ates mainly with historians and political scientists, such as Charles
Beard, Jon Elster, and Stephen Holmes-and on the other hand from
commentary on specific cases and doctrines, the sort of thing one as-
sociates with legal doctrinalists, such as Kathleen Sullivan, Laurence
Tribe, and William Van Alstyne. A number of scholars straddle this
* Chief Judge, United States Court of Appeals for the Seventh Circuit; Senior Lec-
turer, University of Chicago Law SchooL This is the revised text of the twenty-ninth James
Madison Lecture on Constitutional Law delivered at New York University School of Law
on October 21, 1997. I thank Sorin Feiner and Andrew Trask for helpful research assist-
ance, and faculty and students at NYU for many helpful comments.

HeinOnline -- 73 N.Y.U. L. Rev. 1 1998

Imaged with the Permission of N.Y.U. Law Review
 NEW YORK UNIVERSITY LAW REVIEW [Vol. 73:1

divide, such as Ronald Dworkin and Lawrence Lessig, and although I

mean to keep to one side of it in this lecture, the straddle is no acci-
dent. Constitutional theorists are normativists; their theories are
meant to influence the way judges decide difficult constitutional cases;
when the theorists are law-trained, as most of them are, they cannot
resist telling their readers which cases they think were decided consist-
ently with or contrary to their theory. Most constitutional theorists,
indeed, believe in social reform through judicial action. Constitu-
tional theory that is strongly influenced by moral theory has addi-
tional problems, as I have discussed recently and will not repeat here.1
I must stress at the outset the limited domain of constitutional
theory. Nothing pretentious enough to warrant the name of theory is
required to decide cases in which the text or history of the Constitu-
tion provides sure guidance. No theory is required to determine how
many Senators each state may have. Somewhat more difficult inter-
pretive issues, such as whether the self-incrimination clause should be
interpreted as forbidding the prosecutor to comment on the defen-
dant's failure to take the stand,2 can be resolved pretty straightfor-
wardly by considering the consequences of rival interpretations. Were
the prosecutor allowed to argue to the jury that the defendant's re-
fusal to testify should be taken as an admission of guilt, it would be
extremely difficult for defense counsel to counter with some plausible
explanation consistent with his client's being innocent. So allowing
comment would pretty much destroy the privilege-at least as it is
currently understood. That is an important qualification. It has been
strongly argued that the current understanding is incorrect, that the
purpose of the privilege is merely to prevent improper methods of
interrogation; and if this is right then there is no basis for the rule of
no comment.3 Maybe, as this example suggests, when fully ventilated
no issue of constitutional law not founded on one of the numerical
provisions of the Constitution is beyond contestation. But as a practi-
cal matter there are large areas of constitutional law that the debates
over constitutional theory do not touch and that consequently I shall
ignore.
Constitutional theory in the sense in which I am using the term is
at least as old as the Federalist papers. And yet after more than two
centuries no signs of closure or even, it seems to me, of progress, are

1 See Richard A. Posner, The Problematics of Moral and Legal Theory, 111 Harv. L.
Rev. (forthcoming May 1998).
2 As held in Griffin v. California, 380 U.S. 609 (1965).
3 See Albert W. Alschuler, A Peculiar Privilege in Historical Perspective, in R.H.
Helmholz et al., The Privilege against Self-Incrimination: Its Origins and Development
181, 203 (1997).
HeinOnline -- 73 N.Y.U. L. Rev. 2 1998
Imaged with the Permission of N.Y.U. Law Review
April 1998] AGAINST CONSTITUTIONAL THEORY

visible. The reason is that constitutional theory has no power to com-

mand agreement from people not already predisposed to accept the
theorist's policy prescriptions. It has no power partly because it is
normative, partly because interpretation, the subject of constitutional
theory, is not susceptible of theoretical resolution, and partly because
normativists in general and lawyers (and as I said most constitutional
theorists are lawyers, albeit professors of law rather than practicing
lawyers) do not like to be backed into a comer by committing them-
selves to a theory that might be falsified by data, just as no practicing
lawyer wants to take a position that might force him to concede that
his client has no case. Neither type of lawyer wants the validity of his
theory to be a hostage to what a factual inquiry might bring to light.
But as a result, constitutional theory, while often rhetorically power-
ful, lacks the agreement-coercing power of the best natural and social
science.
An even more serious problem is that constitutional theory is not
responsive to, and indeed tends to occlude, the greatest need of con-
stitutional adjudicators, which is the need for empirical knowledge, as
I shall argue using as illustrations the Supreme Courts recent deci-
sions forbidding the Virginia Military Institute to exclude women 4 and
forbidding Colorado to ban local ordinances that protect homosexuals
from discrimination on the basis of their sexual orientation. s I know
that just getting the facts right can't decide a case. There has to be an
analytic framework to fit the facts into; without it they can have no
normative significance. Only I don't think that constitutional theory
can supply that framework. Nor that the design of the framework, as
distinct from fitting the facts into it, is the big problem in constitu-
tional law today. The big problem is not lack of theory, but lack of
knowledge-lack of the very knowledge that academic research,
rather than the litigation process, is best designed to produce. But it is
a different kind of research from what constitutional theorists
conduct.
The leading theorists are intelligent people, and it is possible that
their lively debates have a diffuse but cumulatively significant impact
on the tone and texture and occasionally even on the outcomes of
constitutional cases. (Whether it is a good impact is a different ques-
tion, and one that cannot be answered on the basis of existing knowl-
edge.) If the theorists do not have a large audience among judges, and
I do not think they do, they have a large audience among their own
students and hence among the judges' law clerks, whose influence on

4 See United States v. Virginia, 116 S. Ct. 2264 (1996).

5 See Romer v. Evans, 116 S. Ct. 1620 (1996).

HeinOnline -- 73 N.Y.U. L. Rev. 3 1998

Imaged with the Permission of N.Y.U. Law Review
 NEW YORK UNIVERSITY LAW REVIEW [Vol. 73:1

constitutional law, though small, is not completely negligible. Yet the

real significance of constitutional theory is, I believe, as a sign of the
increased academification of law school professors, who are much
more inclined than they used to be to write for other professors rather
than for judges and practitioners. The causes of this academification
are beyond the scope of this Article, but a particularly mundane cause
is simply that there are so many more law professors than there used
to be that it has become possible for them to have a nonnegligible
audience for their work even if their work is read only by other law
professors, as I believe is largely the case with regard to constitutional
theory. In addition, as constitutional theory becomes more "theoreti-
cal," less tethered to the practice of law, it becomes increasingly trans-
parent to professors in other fields, such as political theory and moral
philosophy; and by this means the ranks of the constitutional theorists
grow to the point of self-sufficiency. Constitutional theory today cir-
culates in a medium that is largely opaque to the judge and the prac-
ticing lawyer.

I
THE HISTORY AND VARIETIES OF CONSTITUTIONAL THEORY

The problem in political theory to which constitutional theory is

offered as a solution is that our judicially enforceable Constitution
gives the judges an unusual amount of power. This was seen as prob-
lematic long before the democratic principle became as central to our
concept of government as it is now. Hamilton's solution to the prob-
lem, drawing on what was already an age-old formalist tradition
stretching back to Cicero and shortly to be echoed by John Marshall,
was to assert that it was the law that was supreme, not the judges,
since judges are (in Blackstone's phrase, but it is also Hamilton's
sense) just the oracles, the mouthpieces, of the law. 6
After a century of judicial willfulness, this position was difficult to
maintain with a straight face. The Constitution had obviously made
the judges a competing power center. James Bradley Thayer argued
in the 1890s that this was bad because it sapped the other branches of
government of initiative and responsibility. He urged courts to en-
force a constitutional right only when the existence of the right, as a
matter of constitutional interpretation, was clear beyond a reasonable
doubt.7 He thought, in other words, that the erroneous grant of a con-
6 See The Federalist No. 78 (Alexander Hamilton); Osborn v. Bank of United States,
22 U.S. (9 Wheat.) 738, 866 (1824) (Marshall, C.J.) ("Courts are the mere instruments of
the law, and can will nothing.").
7 See James B. Thayer, The Origin and Scope of the American Doctrine of Constitu-
tional Law, 7 Harv. L. Rev. 129 (1893).
HeinOnline -- 73 N.Y.U. L. Rev. 4 1998
Imaged with the Permission of N.Y.U. Law Review
Apri 1998] AGAINST CONSTITUTIONAL THEORY

stitutional right was a more serious error than the erroneous denial of
such a right, in just the same way that the criminal justice system as-
sumes that the erroneous conviction of an innocent person is a more
serious error than the erroneous acquittal of a guilty person. But
Thayer didn't explain where he had gotten his weighting of constitu-
tional errors or why it was the correct weighting.
Thayer is the father of the "outrage" school of constitutional in-
terpretation, whose most notable practitioner was Holmes. Holmes's
position was not identical to Thayer's; nor were Cardozo's and
Frankfurter's positions identical to Holmes's, though there are broad
affinities among all four. This school teaches that to be justified in
trying to stymie the elected branches of government it shouldn't be
enough that the litigant claiming a constitutional right has the better
of the argument; it has to be a lot better; the alleged violation of the
Constitution has to be certain (Thayer's position),s or stomach-turning
(Holmes's "puke" test),9 or shocking to the conscience (Frankfurter's
test),o or, a synthesis of the positions (one supported by Holmes's
dissent in Lochner11 ), the sort of thing no reasonable person could
defend. The school of outrage is almost interchangeable with the doc-
trine of judicial self-restraint when that doctrine is understood as
seeking to minimize the occasions on which the courts annul the ac-
tions of other branches of government. The judge who is self-re-
strained in this sense wishes to take a back seat to the other branches
of government, but is stirred to action if his sense of justice is suffi-
ciently outraged.
I own to considerable sympathy with this way of approaching
constitutional issues. And when the outrage approach is tied, as I
have just suggested it can be, to the doctrine of judicial self-restraint-
a doctrine that is founded on reasons12 -the approach is no longer so
purely visceral as my initial description may have suggested. But I
cannot pretend that outrage or even self-restraint furnishes much in
the way of guidance to courts grappling with difficult issues. And I
could defend the approach convincingly only by showing, what may be
impossible as a practical matter to do, that decisions invalidating stat-
utes or other official actions as unconstitutional, when the decision

8 See id. at 144.

9 On Holmes's position, see Richard A. Posner, Pragmatic Adjudication, 18 Cardozo
L. Rev. 1, 2 (1996).
10 See, e.g., Rochin v. California, 342 U.S. 165, 172 (1952) (Frankfurter, J.).
11 See Lochner v. New York, 198 U.S. 45, 74-76 (1905) (Holmes, J., dissenting).
12 Which I summarize and elaborate in Richard A. Posner, The Federal Courts. Chal-
lenge and Reform 304-34 (1996).

HeinOnline -- 73 N.Y.U. L. Rev. 5 1998

Imaged with the Permission of N.Y.U. Law Review
 NEW YORK UNIVERSITY LAW REVIEW [Vol. 73:1

could not have been justified under Thayer's or Holmes's or Car-

dozo's or Frankfurter's approach, have done more harm than good.
Hamilton-style formalism now has a defender in Justice Scalia. 13
But he lacks the courage of his convictions. For he takes extreme lib-
ertarian positions with respect to such matters as affirmative action
and freedom of speech on the ground that these positions are dictated
not by the Constitution but by the cases interpreting the Constitu-
tion.' 4 Take away the adventitious operation of stare decisis and
Scalia is left with a body of constitutional law of remarkable meager-
ness-which is not an objection but which requires a greater effort at
justification than he has been able to offer. Indeed he has offered
little by way of justification other than bromides about democracy.
Complaining that the Supreme Court is undemocratic begs the ques-
tion.' 5 The Court is part of the Constitution, which in its inception
was rich in undemocratic features, such as the indirect election of the
President and of the Senate, and a highly restricted franchise. The
Constitution still has major undemocratic features. They include the
method of apportionment of the Senate, which results in weighting the
votes of people in sparsely populated states much more heavily than
the votes of people in densely populated states; the election of the
President on the basis of electoral rather than popular votes, which
could result in the election of a candidate who had lost the popular
vote; the expansion of constitutional rights brought about by the Bill
of Rights and the Fourteenth Amendment, which curtails the powers
of the elected branches of government; and, of course, lifetime ap-
pointment of federal judges who exercise considerable political power
by virtue of the expansion of rights to which I just referred. The
Supreme Court is certainly undemocratic in a sense, but not in a sense
that makes it anomalous in the political system created by the Consti-
tution, given the other "undemocratic" features that I have men-
tioned. A further drawback to Scalia's approach is that it requires
judges to be political theorists, so that they know what "democracy"

13 See Antonin Scalia, Common-Law Courts in a Civil-Law System: The Role of

United States Federal Courts in Interpreting the Constitution and Laws, in A Matter of
Interpretation: Federal Courts and the Law 3, 23-25 (Amy Gutmann ed., 1997).
14 See Antonin Scalia, Response, in id. at 129, 138-39. He says, "Where originalism will
make a difference is not in the rolling back of accepted old principles of constitutional law
but in the rejection of usurpatious new ones." Id. at 139. But on his understanding of
proper constitutional interpretation, most of the "accepted old principles" were themselves
"usurpatious" when first announced, and some of them were first announced within the
last few decades on the basis of just the kind of nonoriginalist interpretation that he consid-
ers usurpative.
15 Cf. Ronald Dworkin, Freedom's Law: The Moral Reading of the American Consti-
tution 75 (1996).
HeinOnline -- 73 N.Y.U. L. Rev. 6 1998
Imaged with the Permission of N.Y.U. Law Review
April 1998] AGAINST CONSTITUTIONAL THEORY

is, and also to be historians, because it takes a historian to reconstruct

16
the original meaning of centuries-old documents.
Most constitutional theorizing in this century has taken a
nonformalistic direction, unlike that of a Hamilton or a Scalia. We
may begin with Learned Hand's argument that the Bill of Rights pro-
vides so little guidance to judges that it ought to be deemed (largely)
nonjusticiable, 17 and move on to Herbert Wechsler's prompt riposte
that constitutional law can be stabilized by judicial evenhandedness,
what he called "neutral principles," soon recognized as merely princi-
ples and since principles can be bad as well as good, Wechsler's riposte
failed.' 8 Focus then shifted to an effort to identify good principles to
guide constitutional decisionmaking. Leading candidates include John
Hart Ely's principle of "representation reinforcement" 1 9 and Ronald
Dworkin's principle of egalitarian natural justice 20 These are sub-
stantive political principles, and they founder on the authors' lack of
steady interest in and firm grasp of the details of public policy. I have
complained elsewhere about the egregious underspecialization of con-
stitutional lawyers and theorists,2 1 and I don't want to repeat myself.
People who devote most of their lives to the study of political theory
and constitutional doctrine do not thereby equip themselves to formu-
late substantive principles designed to guide decisionmaking across
the vast range of difficult issues that spans affirmative action and ex-
clusionary zoning, legislative apportionment and prison administra-
tion, telecommunications and euthanasia, the education of alien
children and the administration of capital punishment, to name just a
few current and recent issues in constitutional law.
The constitutional theories propounded by the formalists, by
Thayer and his followers, and by Wechsler, were procedural in the
sense of offering a method of analysis rather than a master substantive
16 The shortcomings of lawyers and judges, even of law professors, as legal historians
have been noted often. For recent discussions, citing the earlier literature, see Martin S.
Flaherty, History 'Lite' in Modem American Constitutionalism, 95 Colum. L Rev. 523
(1995); Barry Friedman & Scott B. Smith, The Sedimentary Constitution (Oct. 30, 1997)
(unpublished manuscript, on file with author).
17 See Learned Hand, The Bill of Rights 30 (1958).
18 See Herbert Wechsler, Toward Neutral Principles of Constitutional Law, 73 Harv. L
Rev. 1 (1959). Criticisms of the article are summarized in Richard A. Posner, Overcoming
Law 71-75 (1995).
19 See John Hart Ely, Democracy and Distrust: A Theory of Judicial Review 87-104
(1980).
20 Expounded in many places, most recently in his article, Ronald Dworkin, In Praise
of Theory, 29 Ariz. St. L. 353 (1997).
21 See, e.g., Posner, supra note 18, at 207-14. See id. at 198-207 on the shortcomings of
Ely's theory, and Michael W. McConnell, The Importance of Humility in Judicial Review.
A Comment on Ronald Dworkin's "Moral Reading" of the Constitution, 65 Fordham L
Rev. 1269 (1997), on the shortcomings of Dworkin's theory.
HeinOnline -- 73 N.Y.U. L. Rev. 7 1998
Imaged with the Permission of N.Y.U. Law Review
 NEW YORK UNIVERSITY LAW REVIEW [Vol, 73:1

principle in the style of Ely or Dworkin. The formulation of proce-

dural theories has continued. Examples are Bruce Ackerman's "con-
stitutional moments" approach, 22 Lawrence Lessig's "translation"
approach, 23 John Rawls's "public reason" approach, 24 and Cass
Sunstein's counter to Rawls-his "incompletely theorized" or "judi-'
cial minimalism" approach.25 Ackerman argues that courts should
identify political watersheds, such as the New Deal, and accord them
the same authority for changing constitutional law as they would ac-
cord a formal amendment. This approach requires judges to have the
skills of historians, political theorists, and political scientists, so it is
open to some of the same objections as Scalia's otherwise quite dis-
similar approach. It is also rather too "legal realist," one might even
say realpolitikisch,in inviting judges to bend law to powerful currents
of public opinion, such as those that welled up during the New Deal
and are now understood to have been to a considerable degree
deeply, even tragically, misinformed.
Lessig argues that just as a good translation is not necessarily a
literal one, so keeping faith with the intended meaning of the Consti-
tution's framers may require rulings that depart from the framers' lit-
eral meanings. But whether a literal translation is good depends on
the purpose of the translation; for some purposes, literal translations
are best. Then, too, fidelity to original meanings need not be the sov-
ereign virtue of constitutional interpretation. The real significance of
Lessig's approach is that it turns the tables on Scalia by showing that
originalism is compatible with what Scalia would think an impermissi-
ble flexibility of interpretation.
Rawls is not and does not pretend to be well informed about con-
stitutional law or judicial practice. But his prestige in academic circles
is such that his rather offhand suggestion that judges in interpreting
the Constitution should confine themselves to what he calls "public
reason," defined as the set of considerations that every reasonable
person would consider admissible to resolve issues of public policy, 26

22 See 1 Bruce Ackerman, We the People: Foundations (1991).

23 See, e.g., Lawrence Lessig, Fidelity in Translation, 71 Tex. L. Rev. 1165, 1166-73
(1993); Lawrence Lessig, Fidelity and Constraint, 65 Fordham L. Rev. 1365, 1371-76
(1997).
24 See John Rawls, Political Liberalism 212-54 (1993).
25 See, e.g., Cass R. Sunstein, Legal Reasoning and Political Conflict 35-48 (1996); Cass
R. Sunstein, The Supreme Court, 1995 Term-Foreword: Leaving Things Undecided, 110
Harv. L. Rev. 4, 6-10 (1996); Cass R. Sunstein, Judicial Minimalism (Aug. 1997) (unpub-
lished manuscript, on file with author).
26 "[Citizens are to conduct their fundamental discussions within the framework of
what each regards as a political conception of justice based on values that the others can
reasonably be expected to endorse and each is, in good faith, prepared to defend that
HeinOnline -- 73 N.Y.U. L. Rev. 8 1998
Imaged with the Permission of N.Y.U. Law Review
April 1998] AGAINST CONSTITUTIONAL THEORY

has received respectful attention from constitutional theorists. The

suggestion, if adopted, would confine judges to a level of generality so
void of operational content as to deny them the tools they need to
decide cases.
Sunstein takes almost the opposite tack from Rawls, pointing out
that people often converge on the resolution of a particular issue
though incapable of agreeing on the principles that determine that
resolution. This is importantly true of judges, a majority of whom
have to agree on a resolution even if they can't agree on a broad
ground that would resolve a host of other issues as well. Sunstein fur-
ther points out that a "minimalist" approach that eschews broad
grounds will reduce the magnitude of the judges' inevitable errors. I
like Sunstein's approach, but I see it more as sounding cautionary
notes about constitutional theory, in much the fashion of theoretically
oriented constitutional commentators who are not themselves pro-
pounders of constitutional theories, such as Jack Balkin and Sanford
Levinson, than as a theory itself.27 I have to acknowledge, however,
that Sunstein's is close to my own preferred stance, which I call "prag-
matic." Pragmatism may seem just another theory, in which event I
am contradicting myself in withholding the name of theory from Sun-
stein's approach. But while in one sense pragmatism is indeed a the-
ory and a constitutional theory when applied to constitutional law, in
an equally valid and more illuminating sense it is an avowal of skepti-
cism about various kinds of theorizing, including the kind that I am
calling constitutional theorizing.
Although Sunstein's and my approaches are similar, we fre-
quently disagree at the level of application to particular cases. He
commends recent decisions by the Supreme Court, including the
Romer and VI decisions, as commendably minimalist because they
avoid (Romer more clearly) announcing principles that might over-
turn a lot of other laws. I consider them wedge decisions, in which the
Court takes a first tentative step toward a new abyss, as when the
Court moved, and quickly too, without much thought, from the bare
holding in Baker v. Carr- that legislative malapportionment is justici-
able to a rigid rule ("one man, one vote") founded on a naive concep-
tion of democracy. The decisions that Sunstein commends are
minimalist when compared to hypothetical decisions holding that all
governmental discrimination against homosexuals is unconstitutional

conception so understood." Rawls, supra note 24, at 226; see also John Rawls, The Idea of
Public Reason Revisited, 64 U. Chi. L. Rev. 765 (1997).
27 See, e.g., J.M. Balkin, Agreements with Hell and Other Objects of Our Faith, 65
Fordham L. Rev. 1703 (1997).
28 369 U.S. 186 (1962).

HeinOnline -- 73 N.Y.U. L. Rev. 9 1998

Imaged with the Permission of N.Y.U. Law Review
 NEW YORK UNIVERSITY LAW REVIEW [Vol. 73:1

and likewise all segregation of the sexes (in public restrooms, in mili-
tary units, in college dorms). But from another standpoint they are
uninformed adventures in judicial activism; and that is the view I shall
be defending.
Sunstein's politics, and I believe his conception of where he
would like to see constitutional law heading, are similar to those of
Ackerman, Ely, Dworkin, and Lessig. What he understands better
than they is that judges, with only a few exceptions, are put off by
constitutional theory. Their background is usually not in any kind of
theoretical endeavor even if they are former law professors, as a grow-
ing fraction of appellate judges are. For even today most law profes-
sors are analysts of cases and legal doctrines rather than propounders
of general theories of political or judicial legitimacy, the class of theo-
ries to which constitutional theory belongs. And even if the judge's
background is theory, a theoretical perspective is very difficult to
maintain when one is immersed in deciding cases as part of a commit-
tee. (This may have been a factor in Robert Bork's resignation from
the D.C. Circuit.) The rise of constitutional theory has less to do with
any utility that such theory might have for judges than, as I suggested
at the outset, with the growing academification of legal scholarship.
When Wechsler was crossing swords with Learned Hand, law profes-
sors still thought of themselves as lawyers first and professors second
and saw their role in relation to the judiciary as a helping one. Nowa-
days many law professors, especially the most prestigious ones at the
most prestigious schools, think of themselves primarily as members of
an academic community engaged in dialogue with the other members
of the community and the judges be damned.
I am exaggerating. Constitutional theorists want to influence
constitutional practice. One cannot read Ely and Dworkin and the
others without sensing a strong desire to influence judicial decisions or
even (in Dworkin's case) the composition of the Supreme Court-for
one remembers his polemic against the appointment of Bork. 29 And
Scalia is on the Supreme Court. But to get the richest rewards avail-
able within the modern legal academic community a professor has to
do "theory," and this tends to alienate the professors from the judges.
Sunstein's anti-theory is more likely to move judges, but he suffers
guilt by association; increasingly judges believe that legal academics
are not on the same wavelength with them, that the academics are not
interacting with judges but instead are chasing their own and each

29 See Ronald Dworkin, The Bork Nomination, New York Review of Books, Aug. 13,
1987, at 3, 10, reprinted as Bork: The Senate's Responsibility, in Dworkin, supra note 15,
at 265, 267, 273.
HeinOnline -- 73 N.Y.U. L. Rev. 10 1998
Imaged with the Permission of N.Y.U. Law Review
April 1998] AGAINST CONSTITUTIONAL THEORY

other's tails. I do not think that Justice Scalia's active participation in

the debates over constitutional theory is inconsistent with my claims.
He is plainly unmoved by the academics' criticisms of his position; and
most of them plainly regard him as an unsophisticated, because aca-
demically superannuated, antagonist, one who among other things
tacks between theory and practice, using the constraints of his judicial
role (for example, the constraint of stare decisis) to bevel his sharp-
edged theoretical stance.

II
ToVARD A NEW APPROACH

I would like to see an entirely different kind of constitutional the-

orizing. It would set itself the difficult-although, from the perspec-
tive of today's theorists, the intellectually modest-task of exploring
the operation and consequences of constitutionalism. It would ask
such questions as, what difference has it made for press freedom and
police practices in the United States compared to England that we
have a judicially enforceable Bill of Rights and England does not?
How influenced are judges in constitutional cases by public opinion?
How influenced is public opinion by constitutional decisions? Are
constitutional issues becoming more complex, and if so, what are the
courts doing to keep abreast of the complexities? Does intrusive judi-
cial review breed constitutionally dubious statutes by enabling legisla-
tors to shift political hot potatoes to the courts? What is the effect of
judicial activism on judicial workloads and is there a feedback loop
here, activism producing heavy workloads that in turn cause the
judges to become restrained in order to reduce the number of cases
and thus alleviate the workload pressures? Does the Court try to pre-
vent the formation of interest groups that might obtain constitutional
amendments that would curtail the Court's power or abrogate some of
its doctrines, or to encourage the formation of interest groups that will
defend the Court's prerogatives? And what role do interest groups
play in constitution-making and -amending? In the appointment of
Supreme Court Justices? In the reception of Supreme Court decisions
by the media and through the media the public? Above all, what are
the actual and likely effects of particular decisions and doctrines? Did
Brown v. Board of Education30 improve the education of blacks? Did
Roe v. Wade3 ' retard abortion law reform at the state level? What
effect have the apportionment cases had on public policy? Did the
Warren Court's decisions expanding the constitutional rights of crimi-

30 347 U.S. 483 (1954).

31 410 U.S. 113 (1973).

HeinOnline -- 73 N.Y.U. L. Rev. 11 1998

Imaged with the Permission of N.Y.U. Law Review
 NEW YORK UNIVERSITY LAW REVIEW [Vol. 73:1

nal defendants contribute to the increase in the crime rate in the 1960s
and 1970s and provoke a legislative backlash, increasing the severity
of sentences? These questions have not been entirely ignored,32 but
the literature on them is meager, and law professors have contributed
very little to it. Exploring these questions would be a more fruitful
use of academic time and brains than continuing the 200-hundred-
year-old game of political rhetoricizing that we call constitutional the-
ory. Some of these questions might actually be answerable, and the
answers would alter constitutional practice more than theorizing has
done or can do. Thus I am in radical disagreement with Dworkin,
who insists that cases in which facts or consequences matter to sound
'33
constitutional decisionmaking are "rare.
Which brings me to the VMI and Romer decisions. I will not
claim to have picked these as data for testing my critical and construc-
tive theses by a random process, but it would be easy to pick equally
good illustrations from any term of the Supreme Court. What these
cases illustrate is that the Court does not base its constitutional deci-
sions on fact. If this is right, it makes it unlikely that what the Court
needs is theory, unless telling the Court to pay more attention to so-
cial realities can count as a theoretical assertion.
I am not advocating the transformation of litigation into a setting
for generating or marshaling social scientific data and for testing social
scientific hypotheses. The capability of the courts to conduct scientific
or social scientific research is extremely limited, and perhaps nil. But
their assimilative powers are greater. I would like to see the legal
professoriat redirect its research and teaching efforts toward fuller
participation in the enterprise of social science, and by doing this
make social science a better aid to judges' understanding of the social
problems that get thrust at them in the form of constitutional issues.
What the judges should do until the professoriat accepts this challenge
and makes real progress in the study of race relations, sexual activity,
euthanasia, education theory, and the other areas of social life that are
generating constitutional issues these days is an issue that I shall defer
until I have explained what seem to me to be the unfortunate conse-
quences of judicial ignorance of the social realities behind the issues
with which they grapple.

32 See, e.g., Gerald N. Rosenberg, The Hollow Hope: Can Courts Bring About Social
Change? (1991); Donald J. Boudreaux & A.C. Pritchard, Rewriting the Constitution: An
Economic Analysis of the Constitutional Amendment Process, 62 Fordham L. Rev. 111
(1993); Isaac Ehrlich & George D. Brower, On the Issue of Causality in the Economic
Model of Crime and Law Enforcement: Some Theoretical Considerations and Experimen-
tal Evidence, 77 Am. Econ. Rev. Papers & Proc. 99 (May 1987).
33 See Ronald Dworkin, Reply, 29 Ariz. St. L.J. 432, 433 (1997).

HeinOnline -- 73 N.Y.U. L. Rev. 12 1998

Imaged with the Permission of N.Y.U. Law Review
April 1998] AGAINST CONSTITUTIONAL THEORY

The Virginia Military Institute is a public college the mission of

which is to produce "citizen-soldiers" by bullying methods (the "ad-
versative method," as it is euphemistically called) modeled on the
well-known brutalities of the English public schools ("The Battle of
Waterloo was won on the playing fields of Eton") and of the tradi-
tional Army or Marine boot camp, all being institutions designed to
forge male solidarity viewed as the condition of effective military ac-
tion. VMI refused to admit women (who of course had also been ex-
cluded from the institutions on which it was modeled), precipitating
the suit. The Court begins its opinion in United States v. VirginiaM by
commending "the school's impressive record in producing leaders,"
but accompanies this bit of polite fluff with the unsubstantiated asser-
tion that "neither the goal of producing citizen-soldiers nor VMI's im-
plementing methodology is inherently unsuitable to women."35 How
does the Court know? And even if the methodology were suitable for
women, it wouldn't follow that the school's goal would not be imper-
led; one would have to consider the effect of mixing the sexes. Men
and women both use toilets, but it doesn't follow that unisex public
restrooms are just as appropriate as sex-segregated ones.3
The Court's essential reasoning, in invalidating VMI's exclusion
of women, is that in the past, men, and many women for that matter,
entertained false beliefs about the capacities of women relative to
those of men. In ridiculing the mistakes of past generations, however,
the Court ignored the possibility that our ancestors' false beliefs about
women, whatever the motivation, were the best interpretation of the
then-existing scientific knowledge-a point that has been made about
Aristotle's belief that a child is (in modem terminology) the clone of
its father, the mother being merely an incubator.3 7 Moreover, some of
the discredited beliefs about women's educational and occupational
capacities may well have been true in the then-existing circumstances.
When a woman must be pregnant throughout her fertile years in order
to have a reasonable assurance of producing a few children who will
survive to adulthood, and when most jobs in the economy require
brawn, equal employment opportunities for women are not going to

34 116 S. CL 2264 (1996).

35 Id. at 2269. The Fourth Circuit, in an earlier stage of the case, had based this conclu-
sion on a non-sequitur: the success of women's colleges, which are not military and do not
employ the adversative method. See United States v. Virginia, 976 F.2d 890, 897 (4th Cir.
1992).
36 The lower court had found "that VMI's mission can be accomplished only in a single-
gender environment and that changes necessary to accommodate coeducation would tear
at the fabric of VMI's unique methodology." United States v. Virginia, 976 F.2d at 897.
37 See Johannes Morsink, Was Aristotle's Biology Sexist?, 12 J. Hist. Biology 83,110-12
(1979).

HeinOnline -- 73 N.Y.U. L. Rev. 13 1998

Imaged with the Permission of N.Y.U. Law Review
 NEW YORK UNIVERSITY LAW REVIEW [Vol. 73:1

be in the cards even if a few exceptional women might be able to take

advantage of them. Indignation about historical injustice often re-
flects ignorance of history-of the circumstances that explain and, yes,
sometimes justify practices that in the modem state of society (com-
fortable, rich, scientifically advanced, pushbuttony) would be arbitrary
and unjust.
It is flattering to think of ourselves as being the moral superiors
of our predecessors, but it is false. And it is sheer illogic (it is the
fallacy of naive induction) to argue that if in the past the biological
differences between the sexes, so far as those differences bear on apti-
tudes for various jobs, were exaggerated, those differences must be
zero. Not that the Court went that far; but it does convey the impres-
sion that it thinks the only differences between men and women are
physical. Although the biological differences between men and
women in relation to a variety of professional activities were indeed
exaggerated at a time when biological science was far less developed
than it is today and social conditions far different, the conclusion that
there are no relevant differences not only does not follow from the
history, but is no better than an article of faith. Until recently we did
not realize that dolphins communicate with each other by something
quite like speech; it doesn't follow that with greater educational op-
portunities and perhaps a pinch of affirmative action they could learn
to speak French. The fact that biology used to be riven with mistake,
superstition, and ideology doesn't mean that it's still riven with mis-
take, superstition, and ideology.
Once the advance of science is conceded, it becomes appropriate
to observe that like many articles of faith the "no difference" claim is
contradicted by modern science. Modem science suggests that there
are inherent differences between the average man and the average
woman with respect to aggressiveness, competitiveness, the propensity
to take risks, and the propensity to resort to violence 38-characteris-
tics that, along with the acknowledged differences in physical strength,
are relevant to military fitness and performance. When judges are
faced with creationists' challenges to the theory of evolution, they re-
veal themselves to be resolutely scientific in their outlook. But when
they are faced with evolutionary biologists' challenge to the pieties of
political correctness and radical egalitarianism, they turn pietistic.
I said "average" man and "average" woman. Within each sex,
there is a distribution of characteristics, and the two distributions, the
male and the female, overlap. Because some women are more aggres-

38 See Kingsley R. Browne, Sex and Temperament in Modem Society: A Darwinian

View of the Glass Ceiling and the Gender Gap, 37 Ariz. L. Rev. 971, 1016-64 (1995).
HeinOnline -- 73 N.Y.U. L. Rev. 14 1998
Imaged with the Permission of N.Y.U. Law Review
April 1998] AGAINST CONSTITUTIONAL THEORY

sive, competitive, etc., than some men, the adversative methods used
by VMI may be more suitable for some women than they are for some
of the men admitted to VI. It could be argued that these excep-
tional women should be given a chance. But there are two objections.
First, the prevalence-the near universality-of qualifying examina-
tions and other set requirements for entry into private schools sug-
gests that a policy of giving everyone a chance to prove himself or
herself, in lieu of a preliminary screening for likelihood of success,
would be highly inefficient. If only a minute percentage of women,
relative to men, are qualified to undergo adversative training, individ-
ual consideration of women's applications would yield few benefits.
Second, a concern with the consequences of mixing the sexes in the
unusual setting of a military academy is unrelated to whether women
are able to function as well in that setting as men are.
The Court in the VMI case was much taken with the analogy be-
tween sex-segregated and race-segregated public educational institu-
tions. Judges can rarely resist analogies, a form of "evidence" (if it
can be called that) that is generated by ingenuity rather than by
knowledge. Analogies are typically, as here, inexact and often, as
here, misleading. Racial segregation was demonstrably a component
of an exploitative social system descended from slavery and seeking to
preserve its essential characteristics. Sex segregation has a more com-
plex history, one that is not free from elements of oppression but that
is also bound up with a desire to limit sexual contact between young
people and to tailor education to the difference in life roles between
men and women--differences reflecting, as I suggested earlier, funda-
mental conditions of society that were not less real for having largely
dissipated today. Yet even today we do not consider single-sex
restrooms to present the same issue that single-race restrooms
would? 9
Even if the history of society's treatment of women is as oppres-
sive and unjust as a majority of today's Supreme Court Justices appear
without adequate reflection or inquiry to believe, and is not just a
function of limited knowledge or different material conditions of so-
cial life, it would not follow that a specific "discrimination," for exam-
ple in military training, was oppressive and unjust. I would be very
surprised to learn that any Justice of the Supreme Court believes that
the maintenance of sex-segregated public restrooms violates the Con-
stitution. This means that public segregation of the sexes has to be

39 I am not arguing that because single-sex restrooms are lawful, VMI should be enti-
tled to exclude women. That would be as illegitimate a use of analogy as the ones that I am
criticizing.

HeinOnline -- 73 N.Y.U. L. Rev. 15 1998

Imaged with the Permission of N.Y.U. Law Review
 NEW YORK UNIVERSITY LAW REVIEW [Vol. 73:1

evaluated case by case and therefore that the Court can get little mile-
age from ridiculing, as it did at such length, the former exclusion of
women from the practice of law and medicine.
Thousands of words into its opinion the Court finally gets to the
issue, but lingers there only briefly, for one short, and evasive, para-
graph. The issue, as it would appear to a disinterested student of pub-
lic policy unburdened by commitment to any of the constitutional
theories, is whether excluding women from VMI is likely to do more
harm to women-whether material, psychological, or even just sym-
bolic (and so perhaps indirectly or eventually material or psychologi-
cal)-than including them would do to the mission of training citizen-
soldiers. The Court says nothing about the first point, as if it were
obvious that the exclusion of women from one obscure though distin-
guished military academy would be the kind of insult to women that
forbidding black people to attend military academies would be to
blacks or that the exclusion of male homosexuals from the armed
forces is to homosexuals by branding them as unmanly. That the
equal status of women depends to even a trivial degree on their gain-
ing admission to the Virginia Military Institute would be a laughable
suggestion, which may be why the Court passed over the question in
silence. And for the handful of women who might want to attend
VMI the state had set up a parallel institution-a "separate but equal"
school that was not in fact equal, as the Court pointed out, ignoring
however the fact that it could not be equal, because so few Virginia
women want to attend a quasi-military college that it would not make
any sense to establish a women's parallel institution as richly sup-
ported and maintained as the men's.
If many other public institutions of learning wanted to exclude
women, and a decision in favor of VMI would be a precedent enabling
them to do so, the harm to women would be greater. But as far as I
know or the Court says, no other public institution wants to exclude
women. Still, it could be argued that a decision in favor of VMI would
be a precedent for the exclusion of women from other military acade-
mies and from the combat branches of the armed forces, the branches
most likely to favor the "adversative" style of college education. Yet
if the national government decided to reduce the percentage of
women in the armed forces, it is unthinkable that the Court would
stand in its way. The Court always and properly has been timid about
intruding into military and diplomatic affairs. These are areas in
which the Court is either aware of the limitations of its knowledge and
the costs of error or convinced that it lacks the political authority to
make intervention stick. It is, as it were, the military irrelevance of

HeinOnline -- 73 N.Y.U. L. Rev. 16 1998

Imaged with the Permission of N.Y.U. Law Review
Apri 1998] AGAINST CONSTITUTIONAL THEORY

VIVIl that enabled the Court to invalidate a form of military sex

discrimination.
As for the possibility that VMI's program would be impaired if
women participated, all the Court said is that "[w]omen's successful
entry into the federal military academies, and their participation in the
Nation's military forces, indicate that Virginia's fears for the future of
VMI may not be solidly grounded." 40 In the word "may" lies a note-
worthy concession to reality. No one knows what effect incorporating
large numbers of women into the nation's armed forces will have on
military effectiveness. It is an experiment the results of which may not
be known until the nation is challenged in a major war. It is not as if
the armed forces had wanted or welcomed the influx of women. The
influx was forced upon them by the civilian leadership of the military,
responding to political pressure. This does not make it a bad thing.
Military professionals, like other professionals (notably including law-
yers and judges), tend to be narrow, parochial, and reflexively resis-
tant to change. The racial integration of the armed forces was
accomplished in 1948 by civil initiative over military objections, and
has been a success. The performance of women in the Gulf War of
1991 was by all accounts excellent. But since then the percentage of
women in the armed forces has grown, more and more combat slots
have been opened to them, new tensions have arisen, and there is in-
creased grumbling in military and national-security circles. Maybe
this explains that telltale "may." But if simple prudence requires cau-
tion about dismantling every vestige of sex segregation in the military,
I find it difficult to understand by what rational process the Court
could conclude that Virginia was violating the Constitution by exclud-
ing women from VMI. The harm to women from the exclusion seems,
as I have said, trivial-the entire harm being the difference in the
value of a VMI education and the education in the substitute program
that the state had created for women multiplied by the very small
number of women who would like to attend VMI-and the Court had
no basis either theoretical or empirical for thinking that VMI's educa-
tional program would not be seriously impaired, disproportionately to
the harm to women from exclusion, by the admission of women.
We live in a period of profound peace-or, rather, that is how it
appears to people for whom not only the world wars, but the cold war,
are a rapidly fading memory. It is difficult in such a period to take the
needs of national defense completely seriously against claims emanat-
ing from more contemporary social issues. In such a period the Vir-
ginia Military Institute can only seem a quaint vestige and hence an

4D United States v. Virginia, 116 S. CL 2264, 2281 (1996) (footnotes omitted).

HeinOnline -- 73 N.Y.U. L. Rev. 17 1998

Imaged with the Permission of N.Y.U. Law Review
 NEW YORK UNIVERSITY LAW REVIEW [Vol. 73:1

appropriate subject for social experimentation. It seems to me that

this is about the sum and substance of the Court's thinking in the VMI
case.
But it may be objected that in suggesting that the Court should
have weighed the harm to women from exclusion against the harm to
VMI's educational program from their compelled inclusion, I am pro-
pounding my own constitutional theory, one utilitarian or even eco-
nomic in character, and thus inviting the same criticisms that I have
made of other theorists. Am I not covertly acknowledging that an
atheoretical approach to constitutional decisionmaking is impossible?
But I never meant to suggest that it is possible to approach constitu-
tional issues free from any predispositions, free, that is, from an ap-
proach, or if you will, a theory. I happen to belong to what I earlier
described as the school of "outrage," and it is natural for the members
of that school to ask about the balance of harms; it is when a govern-
mental policy inflicts severe and seemingly gratuitous injury on a
group (women, say, or blacks) that the juices of outrage are likely to
flow. I would be inconsistent only if I tried to show that the school of
outrage had a truer view of the Constitution than its rivals. I have not
tried to show that, and I do not believe that the intellectual tools exist
for establishing which of the competing theories of constitutional deci-
sionmaking is the soundest, although it is possible to point out the
weaknesses in each theory.
What I do concede is that the Court could not actually have
weighed the harm to women from exclusion against the harm to
VMI's educational program from their inclusion. The data are miss-
ing. The fault, in part anyway, lies with constitutional theory, which
claims to offer the courts a data-free method of deciding cases, rather
than helping in the discovery and analysis of the relevant data. The
first thing the courts have to learn is how little they know. What to do
in the face of radical uncertainty is a separate issue, one that I shall
come back to.
Justice Scalia's dissent in the VMI case has a different focus from
my criticisms of the majority opinion; his focus is on the implications
of the Court's decision for single-sex education in general, apart from
the military or quasi-military setting. A Court taken with the crude
analogy of sexual to racial segregation is unlikely to look with favor
on any kind of single-sex education, unless perhaps if the sex is fe-
male-and it may be willing to sacrifice the benefits of single-sex edu-
cation for women on the altar of perceived neutrality. It seems to me
that the courts are as poorly equipped to evaluate sex-segregated edu-
cation in nonmilitary as in military settings. Judges who do not have a
military background doubtless think they know more about education
HeinOnline -- 73 N.Y.U. L. Rev. 18 1998
Imaged with the Permission of N.Y.U. Law Review
April 1998] AGAINST CONSTITUTIONAL THEORY

than they do about war and are therefore less willing to cut the polit-
ical branches of government slack when dealing with educational is-
sues. But do they know enough more about education to make
intelligent decisions? Little is known about what makes for effective
education. The role of resources, of class size, of curriculum, of racial
or other demographic sorting or mixing, of extracurricular activities,
of technology, of standardized testing, of family structure, of home-
work-the significance and interaction of these elements of the educa-
tional process remain largely unknown. Judges can certainly be
forgiven for not knowing what people who devote their lives to a spe-
cialized field do not know; it is less easy to forgive them for not know-
ing that they don't know. Part of a sense of reality, of an empirical
sense, of just the kind of sense that constitutional theory does not cul-
tivate, is knowing which areas of social life are charted and which are
not, and being willing to follow the chart where there is a chart and to
acknowledge when one is embarking on uncharted seas. If even the
experts know very little about education, and this after two and a half
millennia of serious reflection (beginning with Plato), this implies that
we should welcome continued experimentation and diversity.
Brown v. Board of Education is increasingly considered a flop
when regarded as a case about education, which is how the Court pre-
tended (presumably for political reasons) to regard it. For there is no
solid evidence that it led to an improvement in the education of blacks
or even to substantial public-school integration. 41 It is better viewed
as a case about racial subordination, whereas the exclusion of women
by the Virginia Military Institute cannot be regarded with a straight
face as the warp or woof of a tapestry of sex subordination, given the
political and economic power of American women.
I shall end with a few remarks about the Romer case. This is the
second scrape that the Supreme Court has had with homosexuality,
the first being of course Bowers v. Hardwick,42 and the most remarka-
ble thing about both judicial performances is the Court's unwillingness
or inability to talk realistically about the phenomenon. The majority
opinion in Bowers and Chief Justice Burger's concurrence treat it as

41 See, e.g., Sonia R Jarvis, Brown and the Afrocentric Curriculum, 101 Yale L.J. 1285,
1289-91 (1992) (explaining that there has been minimal progress in equalization of educa-
tional opportunity for black children, and that efforts to integrate schools-predominantly
through busing-have been met with widespread resistance, white flight, and ultimate
resegregation); Steven Spiegel, Race, Education, and the Equal Protection Clause in the
1990s: The Meaning of Brown v. Board of Education Re-examined in Light of Milwau-
kee's Schools of African-American Immersion, 74 Marq. L Rev. 501, 503-07 (1991) (find-
ing principles articulated in Brown problematic when applied to Milwaukee's attempt to
improve education of black children).
42 478 U.S. 186 (1986).

HeinOnline -- 73 N.Y.U. L. Rev. 19 1998

Imaged with the Permission of N.Y.U. Law Review
 NEW YORK UNIVERSITY LAW REVIEW [Vol. 73:1

an uncontroversially reprobated horror, like pedophilia, while the dis-

sents in Bowers and the majority opinion in Romer treat it as a so-
cially irrelevant innate condition like being left-handed, and Justice
Scalia's dissent in Romer treats homosexual rights as a sentimental
charitable project of the intelligentsia, like the protection of harp
seals. The majority opinion in Romer finds, sensibly enough, that the
constitutional amendment under challenge, which barred local gov-
ernments from forbidding discrimination against homosexuals, was
motivated by hostility toward homosexuality. The Court then holds
that hostility is not an adequate justification for treating one class of
people differently from another. And that is just about all there is in
the opinion. Ignored are the questions that an ordinary person, his
mind not fogged by legal casuistry, would think central: why there is
hostility to homosexuality and whether the challenged amendment
was a rational expression of that hostility.
Many religious people, Christian and Jewish, believe that homo-
sexual activity is morally wrong. There is no way to assess the validity
of this belief, and what weight if any such a belief should be given in a
constitutional case seems to me an equally indeterminate question.
The belief in equality that informs the VMI opinion is as much an
article of faith as the Judeo-Christian antipathy to homosexuality, 43
and to suppose that securing equality for homosexuals is part of the
meaning of the Equal Protection Clause is equally a leap of faith. In
any event, most Americans, whether religious or not, dislike homosex-
uality and in particular do not want their children to become homo-
sexuals. 44 They are not sure whether homosexuality is acquired or
innate, but, unconvinced that it is purely the latter, they worry about
their children becoming homosexual through imitation or seduction.
They also worry about AIDS spreading from the homosexual to the
heterosexual population (although this fear has abated with the peak-
ing of the epidemic). For these and other reasons, most people dislike
the flaunting of homosexual relationships and activities. They particu-
larly do not want government to endorse homosexuality as a way of
life entitled to the same respect that we accord to heterosexual rela-
tionships particularly within marriage. An ordinance forbidding dis-

43 Sanford Levinson remarks (following Michael Perry) upon the double standard that
prevails in discussions of the legitimate scope of judicial reasoning: the nonreligious are
permitted to make almost any argument they want in support of the positions they take,
but the religious are not permitted to make religious arguments in support of their posi-
tions. See Sanford Levinson, Abstinence and Exclusion: What Does Liberalism Demand
of the Religiously Oriented (Would Be) Judge?, in Religion and Contemporary Liberalism
76, 79 (Paul J. Weithman ed., 1997).
44 See the summary of polling data in Stephen Zmansky, Colorado's Amendment 2 and
Homosexuals' Right to Equal Protection of the Law, 35 B.C. L. Rev. 221, 245-46 (1993).

HeinOnline -- 73 N.Y.U. L. Rev. 20 1998

Imaged with the Permission of N.Y.U. Law Review
April 1998] AGAINST CONSTITUTIONAL THEORY

crimination in housing, employment, or public accommodations on

the basis of sexual orientation is naturally viewed as a form of public
endorsement of homosexuality.
My own view is that there is compelling scientific evidence that
homosexual preference is genetic or at least congenital, and not ac-
quired 4 5 so that the fear of homosexual "contagion" from flaunting or
public endorsement of the homosexual way of life is groundless. And
it is as likely that increasing the rights of homosexuals would reduce
46
AIDS-producing sex among homosexuals as decreasing them would.
No allusion to the scientific and social scientific evidence bearing on
the phenomenon of homosexuality was made in the Romer opinion,
however, so that as it stands the Court seems prepared to forbid dis-
crimination against homosexuals even if the Colorado ban on protec-
tive legislation for homosexuals is entirely rational discrimination-
the equivalent of "discriminating" against airline pilots who have the
misfortune to be old or infirm and as a result are grounded against
their will.
There are analogies, which may have been in the minds of some
of the Justices, between hostility to homosexuals and other, now dis-
credited hostilities, such as anti-Semitism, just as there is an analogy
between racial and sexual segregation of public facilities. But analo-
gies, to repeat an earlier point less contentiously, invite inquiry into
difference and similarity; they should not be permitted to elide in-
quiry. Hostility to homosexuals is plainly a different phenomenon
from anti-Semitism and has to be analyzed on its own terms, which the
Court has refused to do. Some manifestations of that hostility may be
so egregious, hurtful, mean-spirited, even barbarous that the courts
should invalidate them without waiting to find out a lot about the phe-
nomenon. But merely barring local governments from making efforts
to prevent peaceable private discrimination and by doing so to be seen
as endorsing the homosexual way of life falls far short of savagery.
My point is not so much that Romer and the VI case were de-
cided incorrectly as that the decisions are so barren of any engage-
ment with reality that the issue of their correctness scarcely arises. It
is the lack of an empirical footing that is and always has been the
Achilles heel of constitutional law, not the lack of a good constitu-
tional theory. But this raises the question of what the courts are to do
in difficult constitutional cases when their ignorance is irremediable,

45 See Posner, supra note 18, at 572; Richard A. Posner, The Economic Approach to
Homosexuality, in Sex, Preference, and Family- Essays on Law and Nature 173, 186, 191
n.26 (David M. Estlund & Martha C. Nussbaum eds., 1997).
46 See Tomas J. Philipson & Richard A. Posner, Private Choices and Public Health:
The AIDS Epidemic in an Economic Perspective 179-80 (1993).

HeinOnline -- 73 N.Y.U. L. Rev. 21 1998

Imaged with the Permission of N.Y.U. Law Review
 NEW YORK UNIVERSITY LAW REVIEW [Vol. 73:1

though one hopes only temporarily so. Judges don't yet know enough
about the role of women in the military, or about the causes of homo-
sexual orientation, to base decisions in cases such as Romer and VMI
on the answers to these empirical questions. Inevitably, the judge's
vote in such a case will turn on his values and temperament. Those
judges who believe (a belief likely to reflect a judge's values and tem-
perament rather than a theory of judicial review) in judicial self-re-
straint, in the sense of wanting to minimize the occasions on which the
courts annul the actions of other branches of government, will con-
sider ignorance of the consequences of a challenged governmental
policy that is not completely outrageous a compelling reason for stay-
ing the judicial hand in the absence of sure guidance from constitu-
tional text, history, or precedent. (An important qualification: many
constitutional issues can be resolved on the basis of these conven-
tional legal materials.) Activists will plow ahead. These poles will not
meet until much more is known about the consequences of judicial
activism and judicial self-restraint. So one thing that we may hope for
through the application of the methods of scientific theory and empiri-
cal inquiry to constitutional law is the eventual accumulation of
enough knowledge to enable judges at least to deal sensibly with their
uncertainty about the consequences of their decisions. Ultimately
many of the uncertainties may be dispelled. Until that happy day ar-
rives, the most we can realistically ask of the judges is that they be
mindful of the limitations of their knowledge. And I do not mean
knowledge of constitutional theory.

HeinOnline -- 73 N.Y.U. L. Rev. 22 1998

Imaged with the Permission of N.Y.U. Law Review