Defacing Democracy: The Prominence of As-Applied Challenges in the Supreme
Court’s Recent Election Law Decisions
Nathaniel Persily*
Jennifer Rosenberg**
In its first three years, the Roberts Court has issued a series of important election
law decisions on topics ranging from gerrymandering and voter identification to
regulation of political parties and campaign finance.1 The substance of those decisions
has been dramatic enough, but the decisions also illustrate the evolution of important
constitutional and election law procedural doctrines concerning facial and as-applied
challenges. The Court has clarified its strong preference for as-applied challenges in the
election law context as in others, such as abortion cases. However, this largely procedural
debate has spilled into other doctrinal territory, with the consequence that a variety of
different types of deficiencies in voting rights plaintiffs’ claims are now caught in the
ambit of the as-applied/facial challenge jurisprudence.
This Article examines this doctrinal evolution and evaluates the costs and benefits
of the Roberts Court’s preference for as-applied challenges to election laws. We argue
that the renewed emphasis on as-applied challenges masks other strategies and arguments
concerning how inconvenient precedent can be overturned and how the Court should stay
its hand when the factual record supporting a challenge requires further development.
*
Professor of Law and Political Science, Columbia Law School
Voting Rights Fellow, Brennan Center for Justice at NYU School of Law. The views expressed herein
are those only of the authors and not reflective of the organizations with which they are affiliated.
1
Cite LULAC v. Perry, Randall v. Sorrell, Wisconsin Right to Life, Riley v. Kennedy, Lopez-Torres,
Washington Grange, FEC v. Davis.
**
1
Part I sets forth the basic doctrine concerning as-applied and facial challenges. Part II
sketches out two exceptions to the general rule concerning facial challenges: First
Amendment and abortion rights. Part III discusses the Roberts Court’s recent election
law cases, which have dealt with this issue to a surprising extent. Part IV considers
whether election law cases should be one more arena in which the requirements for a
facial challenge should be relaxed. Part V presents our conclusions.
I.
The Basic Doctrine Concerning As-Applied and Facial Challenges
We begin, however, with the basics. A plaintiff can challenge the
constitutionality of a statute in two principal ways. The more ambitious approach – a
facial challenge – requires that a plaintiff prove that the statute is unconstitutional in all
(or nearly all) of its applications. The remedy for such a violation is to void the statutory
provision so as to make it unenforceable against anyone. The less ambitious, and
therefore often more successful approach – an as-applied challenge – alleges that the
statute is unconstitutional as applied to a particular plaintiff given the facts of her case.
The remedy for an as-applied challenge will vary somewhat depending on the nature of
the allegation, but, if doing so is consistent with the meaning and intent of the statute, a
Court will excise the plaintiff and those similarly situated from the statute’s constitutional
reach by effectively severing the unconstitutional applications of the statute from the
unproblematic ones. These definitions gloss over the controversies explored later in this
Article, but this basic, even if assumed, distinction serves a good jumping-off point.
2
The preference for as-applied challenges, which is hardly unique to the Roberts
Court,2 arises from concerns about judicial restraint and respect for the work of politically
accountable branches. The strongest weapon a federal court can wield is its power to
declare an act of a legislative body unconstitutional. A series of doctrines counsels
against the use of that power unless absolutely necessary. Federal courts avoid
constitutional questions unless answering them is essential to the challenge to the law and
even then, sometimes abstain if important enough interests counsel against resolution of
the issue with a particular case. The “case or controversy” requirement leads federal
courts to avoid deciding cases that are moot or unripe. The preference for as-applied
challenges derives from a similar impulse: to strike down as little of a law as possible so
as to salvage the constitutional parts of the law for which the people or their
representatives voted. In some cases, deleting (in effect) a few words from the statute is
the most scalpel-like approach to curing a constitutional defect; whereas in others carving
out an exception for a particular plaintiff serves those values. When the statute is not
amenable to such a construction or a given plaintiff in a particular factual context is not
The Court’s reticence toward facial challenges can be traced back to Yazoo & Mississippi Valley R.R. v.
Jackson Vinegar Co., in which the Court rejected a facial challenge predicated entirely on a hypothetical
circumstance. 226 U.S. 217 (1912). In Yazoo, a railroad launched a facial attack against a Mississippi
statute that imposed penalties on corporations for failing to promptly settle claims for goods that were lost
or damaged during freight. Id. The railroad contended that the statute violated equal protection and due
process because it failed to contain an exception for claims that were excessive or frivolous. The railroad
conceded, however, that the particular claim at presently in issue was neither excessive nor frivolous.
Emphasizing that its role was to “deal with the case in hand, and not with imaginary ones,” the Court
rejected the facial challenge outright, holding that that as-applied to the case before it, the statute must
survive as written. Id. For commentary on the repercussions of Yazoo, see Michael C. Dorf, Facial
Challenges to State and Federal Statutes, 46 Stan. L. Rev. 235, 243 (1994) (arguing that Yazoo’s result
seems “contradictory to the right to be judged by a valid rule of law”); Richard H. Fallon, Making Sense of
Overbreadth, 100 Yale L.J. 853, 861 (1991) (describing Yazoo’s ruling as “harsh” and asserting the Court
had “bypass[ed] . . . future opportunit[ies] to consider the permissibility of statutory policy and, if it found
an injustice, to end it”). CITE CHECK.
2
3
characteristically different from anyone else who might challenge the law, voiding the
law on its face may be the only appropriate remedy to vindicate the plaintiff’s rights.
Of course, the key question remains when, if ever,3 a facial challenge is
appropriate – that is, what does a plaintiff need to prove to have a law voided on its face?
The so-called Salerno standard – which itself did not purport to create new law –
explained that “the challenger must establish that no set of circumstances exists under
which the Act would be valid.”4 In addition to what appear to be acknowledged
exceptions to the Salerno rule discussed in the next section, the Court has also looked the
other way in a variety of constitutional contexts when striking down laws on their face
without considering if they would, in fact, be unconstitutional in all circumstances. 5 It
has done so in challenges brought under the Commerce Clause,6 Equal Protection,7 Fifth
Amendment due process.8
3
4
Cite to Fallon
Salerno, 481 U.S. 739, X (1987).
5
6
See, e.g., Gonzales v. Raich, 125 S. Ct. 2195 (2005); U.S. v. Lopez, 514 U.S. 549 (1995); U.S. v.
Morrison, 529 U.S. 598 (2000). For a comprehensive treatment facial challenges in the Commerce Clause
context, see David L. Franklin, Facial Challenges, Legislative Purpose, and the Commerce Clause, 92
Iowa L. Rev. 41 (2006) (arguing that the Rehnquist Court demonstrated “strong preference” for facial over
as-applied review of statutes implicating the Commerce Clause), and Nathaniel Stewart, Note, Turning the
Commerce Clause Challenge “On Its Face”: Why Federal Commerce Clause Statutes Demand Facial
Challenges, 55 Case W. Res. L. Rev. 161 (2004). ADD SUPPORT – STRUVE.
7
See, e.g., Carey v. Population Servs. Int’l, 431 U.S. 678 (1977) (facially invalidating, on equal protection
and SDP grounds, a New York ordinance that prohibited the distribution of contraceptives by persons other
than licensed pharmacists). It is generally agreed that wherever equal protection challenges succeed against
legislative classifications meriting rational basis review, the statutes are struck down on their face. See,
e.g., Bowers v. Hardwick (X) (striking down Texas’s sodomy statute on its face); Romer v. Evans, 517 U.S.
620 (1996) (X). But see City of Cleburne v. Cleburne Living Ctr., 473 U.S. 432, 476 (using as-applied
review to invalidate a city ordinance that classified mentally retarded persons). In Cleburne, Justice
Marshall, concurring in the judgment and dissenting in part, wrote that “to [his] knowledge, the Court has
never before treated an equal protection challenge to a statute on an as-applied basis.”). See also Matthew
Adler, Rights Against Rules: The Moral Structure of American Constitutional Law, 97 Mich. L. Rev. 1, X
Adler (regarding City of Cleburne as the exception proving the rule); Marc E. Isserles, Overcoming
Overbreadth: Facial Challenges and the Valid Rule Requirement, 48 Am. U. L. Rev. 359, X (1998) (same)
CITE CHECK.
The Court embraced facial challenges in seminal voting rights cases. See, e.g., Harper v. Virginia
Board of Elections, 383 U.S. 663 (1966) (wholesale invalidation of state poll tax); Louisiana v. U.S., XX
(1965) (wholesale invalidation of state literacy requirement).
4
The Court’s cognitive dissonance on the propriety of facial review has inspired a
great deal of scholarship. Most commentators now agree that the facial/as-applied
distinction, while tempting for its simplicity, is ill-conceived and ultimately obfuscates
more than it elucidates.9
Most efforts to parse out a coherent facial challenge doctrine recognize that
notwithstanding Salerno, the categories of facial and as-applied challenges are not
mutually exclusive.10 The dichotomy itself is false, for all statutory challenges actually
fall somewhere along a continuum bookended by facial challenges that ask courts to
invalidate entire statutes, and “pure” as-applied challenges that ask courts invalidate a
fact-specific instance of enforcement.11 Professor Fallon has famously argued that every
litigant challenging a statute is asserting it cannot be enforced against them and therefore
A host of recent decisions invalidating statutory provisions under Section Five of the Fourteenth
Amendment declare them unconstitutional on their face. See, e.g., Board of Trs. v. Garrett, 531 U.S. 356,
373-74 (2001); Kimel v. Florida Board of Regents, 528 U.S. 62 (2000) (invalidating parts of the ADEA
without considering whether the state employers had in fact violated the EP Clause in any of the cases
before the Court); Florida Prepaid Postsecondary Edu. Expense Board v. College Savings Bank, 527 US
627, 639-41 (1999) (Stevens, J., dissenting on the ground that the statute should be upheld as-applied); City
of Boerne v. Flores, 521 U.S. 507, 511, 532-33 (1997). But see Tennessee v. Lane, 541 U.S. 509, 530-31
(2004); Gillian E. Metzger, Facial Challenges and Federalism, 105 Colum. L. Rev. 873, 894-97 (2005)
(arguing that what other scholars have identified as Section Five facial invalidations are actually less facial
than they initially appear).
8
The Court has facially invalidated statutes on vagueness grounds. See, e.g., City of Chicago v. Morales,
527 U.S. 41 (1999) (striking down municipal ordinance that criminalized “loitering”); Kolender v. Lawson,
U.S. 352 (1983) (X); Lanzetta v. New Jersey, 306 U.S. 451, 453 (1939) (“If on its face the challenged
provision is repugnant to the due process clause, specification of details of the offense intended to be
charged would not serve to validate it.”).
9
See, e.g., Fallon, As-Applied, 113 Harv. L. Rev. at X (wondering whether “the Justices are not only
divided, but also conflicted or even confused about when statutes should be subject to facial invalidation”);
Dorf, Facial Challenges, 46 Stan. L. Rev. at 236.
10
See, e.g., Fallon, As-Applied, 113 Harv. L. Rev. at X (“[T]here is no single distinctive category of facial,
as opposed to as-applied, litigation.”). ADD SUPPORT.
11
See, e.g., Dorf at 294 (criticizing Salerno’s “unconstitutional in all applications” rule as vastly inefficient
in light of the “wide gulf [that] separates the statute that might operate unconstitutionally under some
conceivable set of circumstances from one that operates unconstitutionally under all circumstances”); ADD
SUPPORT.
5
all constitutional challenges “are in an important sense, as-applied.”12 Matthew Adler, on
the other hand, has gone so far as to contend that practically all challenges to statutes are
facial and that there is “no such thing as a true as-applied constitutional challenge.”13
A more salient distinction is between two types of facial attacks.14 First, there
are overbreadth facial attacks, which we discussed below in the context of the First
Amendment and abortion jurisprudence. These overbreadth challenges posit that a
statute must be invalidated because it has too many unconstitutional applications.15
Second, there are “valid rule” facial challenges, which argue that a statute has a defect
other than overbreadth which renders it unconstitutional in all of its applications.16 This
defect is structural and most commonly arises out of a statute’s failure to pass muster
under an applicable substantive doctrinal test.17
Marc Isserles argues that Salerno correctly states the standard governing “valid
rule facial challenges,” but leaves it “an open question” whether overbreadth facial
12
Fallon, As-Applied at 1324.
Matthew D. Adler, Rights Against Rules, 97 Mich. L. Rev. 1, 157 (1998). See also Matthew D. Adler,
Rights, Rules, and the Structure of Constitutional Adjudication: A Response to Professor Fallon, 113 Harv.
L. Rev. 1371 (X). For a helpful synopsis of the Adler-Fallon debate, see David M. Driesen, Standing for
Nothing: The Paradox of Demanding Concrete Context for Formalist Adjudication, 89 Cornell L. Rev.
808, 859-862 (2004).
14
As Stuart Buck notes, “[i]f all facial challenges fell into the overbreadth model, then the third-party
standing rule would never apply to facial challenges—an even more anomalous result.” Stuart Buck,
Salerno v. Chevron: What to do About Statutory Challenges, 55 Admin. L. Rev. 427, 430.
15
Marc E. Isserles, Overcoming Overbreadth: Facial Challenges and the Valid Rule Requirement, 48 Am.
U.L. Rev. 359, 363 (1998).
16
See, e.g., Dorf, Facial Challenges, 46 Stan. L. Rev. at 269 (asserting that overbreadth doctrine generally
extends to all “nonlitigation fundamental rights”); Isserles, 48 Am. U.L. Rev. at 363.
17
See Isserles, 48 Am. U.L. Rev. at 423; Fallon, As-Applied, 113 Harv. L. Rev. at 1324, 1368 (“[T]he
availability of facial challenges varies on a doctrine-by-doctrine basis….The crucial, mediating variables
are the general principles or doctrinal tests. . . Some [of which] require courts to engage in processes of
reasoning with the potential to mark a statute as invalid on its face, not merely as applied.”); Dorf, Facial
Challenges, 46 Stan. L. Rev. at 294 (arguing that facial/as-applied distinction should be cast aside, for “the
proper approach to a constitutional case typically turns on the applicable substantive constitutional doctrine
and the institutional setting”). A favorite example of a doctrinal test inherently measuring a statute on its
face begins with a legislative classification motivated by discriminatory intent. Regardless of how such a
classification is challenged, its invidious purpose condemns it to facial invalidation on equal protection
grounds.
13
6
challenges should be permitted.18 David Gans and others have made compelling
arguments against the efficacy of as-applied adjudication in various contexts, noting that
the gradualism and myopic character of case-by-case adjudication makes it much harder
to protect many individual rights.19 In addition, the Salerno principle ignores the
strategic justifications for preferring facial over as-applied invalidation in certain cases.20
Three of these justifications are particularly strong: (1) facial invalidations can guard
against a law’s chilling effect” in a manner that case-by-case litigation inherently
cannot;21 (2) facial challenges are more effective prophylactics against discrimination
because they allowing courts to root out statutes that confer “excessive discretion” on law
enforcement or regulatory officials who could abuse that discretion to violate others’
rights in situations that may never be uncovered by as-applied adjudication;22 (3) facial
18
See 48 Am. U.L. Rev. at 422-23. Isserles contends that both those who favor of a liberal use of facial
invalidation, as well as those who favor as-applied invalidation, base their preference on the “practical
effects” of each type of adjudication. Id. This focus on practical effects distracts from the more salient
point that whether a facial challenge is appropriate in any given case often depends on structural factors,
such as the underlying doctrinal test. Id. This misguided tendency to rely on an “effects test” is particularly
prevalent in abortion cases, where the problem of multiple doctrinal tests complicates and often
overshadows the underlying valid rule inquiry. Id.
19
See, e.g., Craig v. Boren, 429 U.S. 190 (1976) (sustaining a facial challenge on the ground that “a
decision by us to forgo consideration of the constitutional merits in order to await the initiation of a new
challenge to the statute by injured third parties would be impermissibly to foster repetitive and timeconsuming litigation under the guise of caution and prudence.”). See David H. Gans, Strategic Facial
Challenges, 85 Boston U.L. Rev. 1333, 1336 (2005) (X). ADD SUPPORT.
20
See Dorf, Facial Challenges, 46 Stan. L. Rev. at 276 (asserting that the same principles that justify First
Amendment overbreadth doctrine justify its extension to other fundamental rights, and that the
substantiality requirement used to limit the scope of First Amendment overbreadth has also been used “for
quite some time” to limit overbreadth in other contexts, including abortion law).
21
See, e.g., Casey, X U.S. at X (at least #X Justices finding it persuasive that the law would deter a
“significant number” of women from procuring abortions); Stenberg (finding that the Nebraska law
impermissibly chilled second trimester abortions).
22
Gans, Strategic Facial Challenges, 85 Boston U.L. Rev. at X. Gans illustrates this point with Louisiana
v. United States, in which the Court struck down Louisiana’s literacy test on the ground that it subjected
voters to the “passing whim or impulse of an individual registrar.” 380 U.S. 145, 153 (1965). See also
Casey, X U.S. at X (finding that the challenged statute created a risk of arbitrary or discriminatory
enforcement by state actors, a factor which weighed in favor of facial invalidation).
7
invalidation is a superior way to vitiate stigmas associated with laws that convey
inequality or prejudice.23
Adherence to the facial / as-applied distinction can also be a stopgap for
consideration of doctrines that are not only more logical, but also more effective tools of
judicial restraint.24 Most significantly, there is a deeply rooted assumption in American
constitutional law that most statutes are severable and that a statute’s unconstitutional
applications can be quarantined from valid applications.25 Since canons of judicial
restraint and constitutional avoidance ordinarily favor partial over wholesale
See, e.g., Bowers v. Hardwick, X U.S. X (facially invalidating Texas’s sodomy law). ADD SUPPORT.
See Stuart Buck and Mark L. Rienzi, Federal Courts, Overbreadth and Vagueness: Guiding Principles
for Constitutional Challenges to Uninterpreted State Statutes, 2002 Utah L. Rev. 381, X (2002) (arguing
that the Court is often pulled in opposite directions by its abstention, avoidance, and severability doctrines,
on one side, and its overbreadth and vagueness doctrines on the other).
Standing. Facial overbreadth challenges are often viewed as assertions the rights of non-parties,
and in this respect they raise third-party or jus tertii standing considerations. See, e.g., Sabri v. United
States, 541 U.S. 600, 601 (2004) (“Facial challenges . . . are to be discouraged because they invite
judgments on fact-poor records and entail a departure from the norms of federal-court adjudication by
calling for relaxation of familiar standing requirements to allow a determination that the law would be
unconstitutionally applied to different parties and different circumstances from those at hand.”) (citing
Chicago v. City of Morales, 527 U.S. 41, 55-56, n.22). But see Henry Paul Monaghan, Third Party
Standing, 84 Colum. L. Rev. 277, XX (1984) (asserting that everyone has a personal right, independent of
standing, to challenge the enforcement of a statute, since the Constitution forbids the imposition of
sanctions except pursuant to a valid rule of law).
According to Fallon, a perception that overbreadth challanges undermine the usual rule that one
may not assert the rights of another explains, at least in part, the Court’s reticence towards them. See
Fallon, As-Applied and Facial, 113 Harv. L. Rev. at X. This theory is supported by caselaw. See, e.g., City
of Chicago v. Morales, 119 S. Ct. 1849, 1859 n.22 (1999) (“[T]he threshold for facial challenges is a
species of third party (jus tertii) standing, which we have recognized as a prudential [discretionary]
doctrine.”). Moreover, although many facial challenges appear at first blush to raise traditional third-party
standing concerns and have been analyzed under that framework, they are properly distinguished from
“pure” third-party standing cases, in which no litigant contests the validity of a statutory rule of law.
25
The notion of severability presupposes that statutes are comprised of independent sub-rules, and a
presumption of statutory severability applies. For thorough discussions of the tension between facial
challenges and severability doctrine, see, e.g., Fallon, As-Applied, 113 Harv. L. Rev. at X (arguing that
valid subrules can typically be severed from invalid ones, and thus “it is often unnecessary for a court to
adjudge the validity of a statute ‘on its face’; it is enough to determine whether a valid subrule applies to a
particular case”); and Dorf, Facial Challenges to State and Federal Statutes, 46 Stan. L. Rev. 235, 250
(1994); Monaghan, Overbreadth, 1981 Sup. Ct. Rev. 1, 6-7. .
23
24
8
invalidation,26 “the debate regarding the availability of facial challenges [becomes], at
bottom, fundamentally a debate about severability.27
Of course, severability is ultimately a question of statutory interpretation. Unless
a statute contains a severability clause,28 avoidance principles instruct that “every
reasonable construction must be resorted to, in order to save a statute from
unconstitutionality.”29 Only where a statute cannot be saved by a narrowing construction
does the inquiry become whether its fatal provisions can be severed. In assessing a
statute’s severability, the critical inquiry is “would the legislature have preferred what is
left of its statute to no statute at all?”30
26
The Court has directed that wherever possible, severability is the required course. See, e.g., Ayotte, X
U.S. at X (“It is axiomatic that where a statute has a constitutional flaw, the Court should, if possible, sever
and “enjoin only [its] unconstitutional applications . . . while leaving other applications in force,” thereby
“sever[ing] its problematic portions while leaving the remainder intact.”); Brocket v. Spokane Arcades,
Inc., 472 U.S. 491 (stating that the usual rule is partial rather than facial invalidation).
27
Gillian E. Metzger, Facial Challenges and Federalism, 105 Colum. L. Rev. 873, 896-97 (2005); Alfred
Hill, Some Realism About Facial Invalidation of Statutes, 30 Hofstra L. Rev. 647, 664 (2002) (arguing that
there has never been an instance of the Court striking down an entire statute that had both valid and invalid
components).
28
John Nagle has argued that the presence or absence of a severability clause will not necessarily lead a
court to reach a particular result in determining whether a legislature intended for a statutory provision to
stand independently. See John C. Nagle, Severability, 72 N.C. L. Rev. 203, 234-40 (1993).
29
Some Justices have accused the Court of purposefully ignoring established rules of construction in cases
where heeding them would have saved a socially undesirable statute. See, e.g., Stenberg v. Carhart, X U.S.
at X (“The Court at times employed an antagonistic canon of construction under which in cases involving
abortion, a permissible reading of a statute [was] to be avoided at all costs.’” (quoting Stenberg, at 977,
(Kennedy, J., dissenting) (internal quotations omitted)). For a discussion of various machinations of
constitutional avoidance, see Justice Brandeis’s concurring opinion in Ashwander v. Tennessee Valley
Authority, 297 U.S. 288, 345-48 (1936). See also Michael C. Dorf, “Chief Justice Roberts Advocates the
Passive Virtues, Even as the Supreme Court’s Docket Reveals Their Subtle Vices,” 779 PLI/Lit 47 (August
2, 2007) (discussing the role of metholodgical conservatism and quoting a favorite aphorism of Justice
Roberts: “If it is not necessary to decide an issue to resolve a case, then it is necessary not to decide that
issue.”).
30
Ayotte, 546 U.S. at 330. See John Copeland Nagle, Severability, 72 N.C.L. Rev. 203, 246-249 (1993)
(X). Where the statute is a state statute, the task of divining the intention of the legislature falls primarily
on the state court. See Dorchy v. Kansas, 264 U.S. 286 (1924). See also David H. Gans, Severability as
Judicial Lawmaking, 76 Geo. Wash. L. Rev. 639, 641 (2008) (cautioning that the legislative intent test
gives courts too much latitude in fashioning remedies, thereby encroaching on the province of legislatures).
Gans argues that wholesale facial challenges can be preferable because they can curb judicial legislation
while inciting elected legislators into action. Id. Granted, this call to legislative action is only likely to
result where invalidation would have “momentous consequences,” such as where an entire sentencing
regime or redistricting plan would be nullified. Id. at 692 (offering examples of instances where the Court
9
II.
Exceptions to the General Rule Concerning Facial and As-Applied Challenges
a. First Amendment
The Court has long recognized First Amendment overbreadth doctrine as an
exception to its purported proscription on facial challenges.31 Under First Amendment
overbreadth doctrine, litigants may seek facial invalidation of a statute that restricts
expression, regardless of whether the litigant herself was engaging in protected speech at
the time the statute was enforced against her.32 In other words, the doctrine allows
litigants to assert the rights of a hypothetical third party, without any need to establish
that the statute at issue was unconstitutional as-applied.33
There are three common rationales for First Amendment overbreadth doctrine.
Under the first, the doctrine is necessary because overly broad laws restricting speech
tend to “chill” or deter other from engaging in protected expression.34 According to the
stayed facial invalidation of a statute in order to give lawmakers an opportunity to fix its constitutional
defect).
31
Indeed, the Court first conceived of this special standing component of First Amendment overbreadth as
an exception to the normal rules governing facial challenges. See Salerno, 481 U.S. at X. ADDITIONAL
SUPPORT.
32
Compare Henry Paul Monaghan, Overbreadth, 1981 Sup. Ct. Rev. 1, 21-30 (arguing that First
Amendment overbreadth flows automatically from substantive first amendment law, rather than from any
special standing rules or nonseverability presumption) with Richard Fallon, Making Sense of Overbreadth,
100 Yale L. J. 853, 867-75 (asserting the opposite view). CITE CHECK.
33
See, e.g., Coates v. City of Cincinnati, 402 U.S. 611 (1971) (holding an ordinance prohibiting
“annoying” speech unconstitutional on its face); see Salerno 481 U.S. 739, 745 (1987) (citing X).
34
See, e.g., Massachusetts v. Oakes, 491 U.S. 576 (1989) (refusing to apply the overbreadth doctrine to a
statute that had since been amended because any subsequent ruling would not serve the purpose of the
doctrine, which is to avert chill); R.A.V. v. City of St. Paul, 112 S. Ct. 2538, 2553 (1992). See Note, the
First Amendment Overbreadth Doctrine, 83 Harv. L. Rev. 844 (1970) (advancing the position that overly
broad restrictions on speech have chilling effects); Fallon, Making Sense of Overbreadth, 100 Yale L. J. at
867-75 (maintaining that the degree of chill varies with the particular speech being regulated – i.e., statutes
that prohibit forms of spontaneous expression have a lower chill factor than statutes regulating more
deliberate, prolonged forms of expression, such as picketing); Michael C. Dorf, Facial Challenges, 46 Stan.
10
second view, the standing allowance is a logical outgrowth of substantive First
Amendment doctrine, namely the requirement that laws regulaing speech be the least
restrictive means of accomplishing their stated goals.35 The third view is simply that
constitutionally protected speech possesses “transcendent value to all society” and
therefore merits special protections.36
In Broadrick v. Oklahoma the Court announced a “substantially requirement”:
that a statute’s overbreadth “must be not only real, but substantial as well, judged in
relation to the statute’s plainly legitimate sweep.”37 Where the statute was enacted by a
state legislature, federal courts must assess the statute as interpreted by state courts, rather
than as written. For example, in Osborne v. Ohio, the Court held that the state court
could offer a narrowing construction in the course of enforcing a statute that, if read
literally, would be unconstitutionally overbroad.38 With respect to federal statutes,
federal courts must construe them in a manner consistent with principles of constitutional
avoidance: a reasonable construction that would raise constitutional problems must yield
to one that does not, unless of course the latter is “plainly contrary to the intent of
L. Rev. at 264-271 (arguing that the chilling effect justifies applying the overbreadth doctrine to all
fundamental rights, including free speech). But see Martin Redish, The Warren Court, The Burger Court,
and the First Amendment Overbreadth Doctrine, 78 Nw. U.L. Rev. 1031 (1983) (critiquing the chilling
effect theory).
35
See Henry Paul Monaghan, Overbreadth, 1981 Sup. Ct. Rev. 1 at X (1981). According to Monaghan,
First Amendment overbreadth doctrine embodies one of the most deeply rooted governing principles of
American constitutional law, which is that all litigants have a right to be judged according to a
constitutionally valid rule of law. See id. But see CRITIQUES OF 1ST PARTY VIEW.
36
See Gooding v. Wilson, 405 U.S. 518 521 (1972). ADD SUPPORT.
37
Broadrick v. Oklahoma, 413 U.S. 601, 615 (1973). See also Osborne v. Ohio, 495 U.S. at 122 (referring
to First Amendment overbreadth doctrine as “strong medicine” that ought to be applied rarely). The Court
has invalidated several statutes under Broadrick’s more stringent standard. See, e.g., Board of Airport
Commissioners v. Jews for Jesus, X U.S. X (1987); New York v. Ferber, 458 U.S. 747 (1982).
38
See 495 U.S. 103, 119-20 (1990) (affirming state court’s authority to provide a narrowing construction).
See generally Hart & Weschler, The Federal Courts and the Federal System, 5th ed. At 188-191 (discussing
prevailing narrowing constructions doctrine and its interplay with judicial efforts to enjoin the enforcement
of overly broad state statutes).
11
Congress.”39 In cases where the statute cannot be saved by a narrowing construction, the
question becomes whether its unconstitutional provisions can be severed.40 Because
severability determinations necessarily involve looking at the particular conduct or
speech being regulated, in practice the severability requirement means that courts
considering facial challenges must a prior consider the statutes as-applied.41
b. Abortion rights
The Salerno Court maintained that it had never recognized an overbreadth
doctrine outside the limited scope of the First Amendment.42 While other commentators
have established the patent inaccuracy of this claim across various contexts,43 for
purposes of laying foundation for our discussion of election law overbreadth in Part III, it
is useful to flesh out the Court’s parallel creation of an overbreadth exception in the
39
Edward J. DeBartolo Corp. v. Florida Gulf Coast Bldg. & Constr. Trades Council, 485 U.S. 568, 575
(1988). But see Reno v. ACLU, 521 U.S. 844, 884 (1997) (refusing to narrowly construe a federal statute
where Congress gave “no guidance whatsoever for limiting” it).
40
See Dorf, Facial Challenges, 46 Stan. L. Rev. at 288-93. Although we address severability in greater
detail below, it suffices to say here that where a statute is severable, First Amendment overbreadth doctrine
cannot be used to invalidate it wholesale. See, e.g., Brockett v. Spokane Arcades, Inc., 472 U.S. 491 (1985)
(holding that because a presumption of severability existed under Washington State law, the lower court
should have voided the statute only insofar as it reached protected speech).
41
See Board of Trustees, State Univ. of N.Y. v. Fox, 492 U.S. 469, 484-85 (1989) (stating that the Court’s
“usual” practice is to consider the as-applied challenge first, and only where necessary delve into “the
overbreadth question,” which is “ordinarily more difficult to resolve . . . since it . . . requires consideration
of many more applications than those immediately before the Court”). But see Hart & Weschler at 191
(listing a handful of cases where the Court entertained a facial overbreadth challenge without first ruling on
the statute as-applied).
42
See Salerno, 481 U.S. at X.
43
Some Justices expressly endorsed an overbreadth doctrine outside the First Amendment context. Justice
O’Connor, for instance, interpreted the Court’s precedent as clearly facial attacks wherever statutes reach
“a substantial amount of constitutionally protected conduct.” Kolender v. Lawson, 461 U.S. at 359 n. 8.
See also, e.g., See Sabri v. United States, 541 U.S. 600, 609-10 (2004) (acknowledging that the Court has
“recognized the validity of facial attacks alleging overbreadth (though not necessarily using that term) in
relatively few settings,” including the Fifth Amendment right to travel and abortion) (citing, inter alia,
Aptheker v. Secretary of State, 378 U.S. 500 (1964); Stenberg v. Carhart, 530 U.S. 914 (2000). See
generally Dorf, Facial Challenges, at 271-272 (observing that the Court has often employed overbreadth
analysis in fundamental rights cases and that the Roe v. Wade “exemplifies overbreadth analysis”). But see
Hart & Weschler at 195 (questioning whether Aptheker is more appropriately viewed as a First Amendment
overbreadth case).
12
special context of abortion law. As Professor Fallon has observed, prior to Chief Justice
Roberts’s tenure, “[v]irtually all of the abortion cases decided after Roe v. Wade . . .
involved facial attacks on state statutes, and the Court, whether accepting or rejecting the
challenges on the merits . . . typically accepted this framing of the question presented.”44
In cases where these challenges succeeded on the merits, the usual result was an
injunction that gutted the entire statute, regardless of whether it had any constitutional
applications.45
For example, in Planned Parenthood Planned of Southeastern Pa. v. Casey, the
Court reached the merits of a facial challenge without any mention of Salerno and
facially invalidated portions of a Pennsylvania statute where it would have operated as an
undue burden in “a large fraction” of cases.46 More recently, in Stenberg v. Carhart, the
Court struck down Nebraska’s entire so-called “partial birth abortion” statute on the
ground that its lack of a health exception for the mother would lead to numerous
44
Fallon, Making Sense of Overbreadth, at 859 n. 29 (citing cases). See also Stuart Buck, Salerno v.
Chevron: What to do About Statutory Challenges, 55 Admin L. Rev. 427, 430 (2003) (recognizing that
“[a] doctrine parallel to that of First Amendment overbreadth has emerged in the abortion context” (citing
Planned Parenthood v. Casey, 505 U.S. 833, X (1992))).
45
See, e.g., Planned Parenthood v. Casey, 505 U.S. at X; ADD SUPPORT. But see H.L. v. Matheson, 450
U.S. 398 (1981) (finding that the plaintiff, an unmarried 15-year old girl who lived with her parents, could
not facially attack a state statute requiring parental notification).
46
505 U.S. 833 at X. Casey also underscores the need for litigants alleging non-First Amendment
overbreadth to provide evidentiary support for their challenge. In Casey, the Court assessed the factual
record associated with each challenged provision of the Pennsylvania statute. Since litigants failed to
proffer sufficient evidence demonstrating that the statute’s informed consent provision created an “undue
burden,” the Court sustained that provision as written. By contrast, the litigants submitted a detailed factual
record proving that the spousal consent provision would unduly burden a critical mass of women, and
accordingly the Court sustained the facial attack on that provision. Id. at 461-62. See also Ohio v. Akron
Ctr. For Reprod. Health, 497 U.S. 502 (1990) (X); Hodgson v. Minnesota, 497 U.S. 417 (1990) (X);
Webster v. Reproductive Health Services, 492 U.S. 490 (1989) (X); Planned Parenthood of Mo. v.
Danforth, 428 U.S. 52, 81-84 (1976) (upholding facial challenge to state statute regulating abortion).
13
unconstitutional applications of the statute. Looking through the other end of the
telescope, these rulings are outright rejections of the Salerno principle.47
The Roberts Court’s general hostility toward facial challenges is perhaps nowhere
as pronounced as it is in recent rulings that disavow any continuation of the abortion
overbreadth exception. Beginning with Ayotte v. Planned Parenthood, the Court
breathed new life into Salerno by rejecting, as a procedural matter, a facial overbreadth
challenge to a New Hampshire statute regulating abortions.48 The Court avoided the
merits of the question—whether New Hampshire could require parental notification
without accommodating immediate threats to the mother’s health—and instead turned its
attention to the remedy being sought. Writing for the majority, Justice Stevens took
wholesale invalidation off the table, reaffirming that the ‘normal rule’ is that “partial,
rather than facial, invalidation is the required course . . . .”49 Justice Stevens also
attempted to distinguish Stenberg, writing that the only reason why the Court had
invalidated the entire Nebraska statute was simply because no litigant had asked for, and
the Court “did not contemplate . . . relief more finely drawn.”50
47
Not all Justices approved of these applications of overbreadth, and Justice Scalia in particular railed the
Court’s reliance on it in abortion cases, disparaging it as a regrettable “political correctness” exception to
the Court’s usual (and correct) restrictions on facial challenges. City of Chicago v. Morales, 527 U.S. at 81.
48
See also Richard H. Fallon, Jr., Daniel J. Meltzer, & David L. Shapiro, Hart and Wechsler’s The Federal
Courts and the Federal System 12 (Supp. 2005) (questioning whether there is “any principled explanation
of when the Court will entertain overbreadth facial challenges and when it will not”).
49
546 U.S. 320, 329 (2006).
50
Id. at X. Stuart Buck observes that although the parties in Stenberg may not have asked the Court to save
the Nebraska statute by merely severing or narrowly construing its problematic provisions, at least one
amicus brief did request such relief. See Stuart Buck, “The Ayotte Case,”
http://stuartbuck.blogspot.com/2006/01/ayotte-case.html, posted 1/20/2006 (stating that an amicus brief
filed by Feminists for Life had in fact asked the Court for a narrow interpretation of the statute that would
have included a health exception)
14
A year after Ayotte, the Court reaffirmed its commitment to rolling back the
abortion overbreadth exception.51 In Gonzales v. Carhart, a 5-4 majority reversed the
Eighth and Ninth Circuit’s wholesale invalidations of the Partial Birth Abortion Act of
2003.52 Both circuits had struck down the Act as unconstitutionally broad because it
failed to include a provision that would have permitted the controversial procedure in
circumstances necessary to protect the health of the mother.53 In overturning the circuit
courts, the Court declared the Act impervious to facial challenges, given that the
“latitude” afforded First Amendment facial challenges was “inapplicable.”54 As such, the
facial challenges at issue “should not have been entertained in the first instance.”55 The
Court also rejected the notion that all restrictions on abortion procedures require health
exceptions, directing that wherever a factual dispute exists as to whether or not a statute
poses significant health risks to women, the proper mode of judicial review is case-bycase.56 With respect to the evidence proffered to establish that the law created an “undue
burden,” the Court observed that Respondents failed to demonstrate the Act “would be
unconstitutional in a large fraction of relevant cases.”57 This statement arguably suggests
51
See also Hartnett, Modest Hope for a Modest Roberts Court, 59 SMU L. Rev. at X (arguing that the
comparative competence of courts strongly cautions against facial adjudication and predicting that with the
succession of Justice O’Connor by Justice Alito, the Court would move further in that direction).
52
See 550 U.S. X, 127 S.Ct. 1610, X (2007). Justice Ginsburg, writing for the dissenters, voiced alarm at
the majority’s blatant “refusal to take . . . seriously” precedent that not only countenanced the facial
invalidation of statutes lacking health exceptions, but also endorsed the general availability of facial
challenges in the abortion setting. Id. (citations omitted).
53
The Ninth Circuit also found that the Act impermissibly vague and unduly burdensome on a woman’s
right to choose a second trimester abortion. See Planned Parenthood v. Gonzales, 435 F.3d 1163 (9th Cir.
2006).
54
Gonzales v. Carhart, 127 S.Ct. at X.
55
Id.
56
See id at X. Although the Court let the door open for as-applied challenges to the Act, it appears as
though none followed. See Edward Whelan, “The Mystery of the Missing Lawsuits: One year after the
Supreme Court’s partial-birth-abortion ruling,” National Review Online, 4/18/08 (arguing that no asapplied challenges have been brought because medical experts now agree that the procedure contested in
Gonzales is never necessary to protect the mother’s health).
57
127 S.Ct. at X. (citing Casey, 112 S. Ct. at 2791). In light of the Roberts Court’s determination to roll
back the abortion overbreadth exception, one commentator has suggested that reproductive rights advocates
15
that where a statute’s challengers can proffer a detailed factual record establishing a
critical mass of unconstitutional applications, the Court would be willing to invoke
severability.58
III.
The Special Context of Election Law
The emphasis on as-applied challenges as the preferred method of challenging
unconstitutional state action has been particularly salient in recent election law cases. In
the context of campaign finance, regulation of political parties, and voter identification
requirements, the controlling opinions for the Court have discussed the as-applied/facial
issue to a degree never before seen in an election law case. These cases illustrate the
tension in the evolution of the relevant constitutional jurisprudence, while at the same
time forcing us to ask whether the election law context, like the First Amendment and
abortion rights contexts, should be treated as special for some reason.
would fare better by bringing lawsuits predicated on the “purpose” prong of Casey’s undue burden test.
See Note: After Ayotte: The Need to Defend Abortion Rights With Renewed ‘Purpose,’ 119 Harv. L. Rev.
2552 (2006) (focusing on language in Casey stating that an undue burden is created wherever a law “has
the purpose or effect of placing a substantial obstacle” in a woman’s path).
58
This is similar to what happened in Seaboard Air Line R. Co. v. Blackwell, in which the Court sustained
an as-applied challenge and struck down an entire statute, whereas previously that same statute had
survived a facial attack premised on the same constitutional argument. See 37 S.Ct. 640, X (1917)
(explaining that in contrast to the earlier facial challenge, the as-applied challenge was supported with a
specific factual record evidencing the statute’s unreasonableness).
16
a. Wisconsin Right to Life
Perhaps the largest transformation in election law jurisprudence during the
Roberts Court has occurred in the area of campaign finance. The replacement of Chief
Justice William Rehnquist and Justice Sandra Day O’Connor with Chief Justice John
Roberts and Justice Samuel Alito flipped a fragile 5-4 majority that was deferential to
campaign finance reforms to a similar one that appears to be more aggressive in striking
such measures down as abridging First Amendment rights. In fact, the Roberts Court has
struck down all three campaign finance laws it has considered.59 “Struck down” might
overstate what it has done, because in its most recent and significant campaign finance
decision, the Court vindicated an as-applied challenge that effectively gutted the relevant
provisions of the law.
In FEC v. Wisconsin Right to Life (WRTL), the Court struck down Title II of the
Bipartisan Campaign Reform Act (BCRA) as applied to an advertisement that asked
voters in Wisconsin to call their Senators to tell them to confirm Bush’s judicial
nominees.60 The law prohibits the airing of such advertisements, funded by corporate
treasury funds, within 60 days of an election if they refer to a clearly identified candidate
for federal office. WRTL paid for the ad with corporate funds and the ad mentioned by
name Senator Russ Feingold, who was up for reelection,61 so it was captured by the law.
However, the Court considered the ad protected issue advocacy, instead of less-protected
electioneering communication, and also held that the law was unconstitutional as-applied
to any similar ad unless the ad was “susceptible of no reasonable interpretation other than
59
Discuss Randall v. Sorrel and FEC v. Davis.
60
61
It also mentioned Senator Herb Kohl, who was not up for reelection.
17
as an appeal to vote for or against a specific candidate.”62 The Court split into three
camps: Chief Justice Roberts’ controlling opinion joined by Justice Alito struck down the
law as-applied63; Justice Scalia’s concurrence (joined by Justices Thomas and Kennedy)
would have struck down the law on its face64; and Justice Souter’s dissent (joined by
Justices Stevens, Ginsburg and Breyer) would have upheld the law both as-applied and
on its face.65
Unlike the other opinions we soon discuss, WRTL is notable for its vindication of
an as-applied challenge, as opposed to its reservation of as-applied challenges en route to
denying a facial challenge. Generally, we think of as-applied challenges as carving out
exceptions to a largely constitutional statute. However, the WRTL decision strikes down
the law as applied to most of the advertisements to which it was directed. Most
corporate-funded advertisements run within 60 days of an election that refer to candidates
running for office are capable of some other interpretation than an appeal to vote for or
against a candidate.66 Indeed, the reason Congress passed such a sweeping corporate
speech code was that the “magic words” standard of Buckley v. Valeo67 proved so easy to
get around.68 Ads would end with an exhortation to “call your member of Congress to
tell them how you feel”, instead of an appeal to vote a particular way, but the message
was still clear.69
62
cite
Justice Alito also wrote separately to emphasize that if the standard in the controlling opinion proved
unworkable, he would also strike down the law on its face. See cite
63
64
65
66
Cite to dissent and related materials.
cite
68
cite
69
cites
67
18
In McConnell v. FEC,70 the Court upheld Title II (the primary definition) on its
face but left open the possibility of as-applied challenges in a footnote.71 The ad at issue
in WRTL presented such a challenge, but the Court’s decision goes much farther than
merely holding that particular ad protected. It established the “susceptible of no
reasonable interpretation” standard, described above, for as-applied challenges to the
BCRA going forward that blew a huge hole in the law’s primary definition of
electioneering communications. It did so by reaffirming precedent concerning the anticorruption interest necessary to justify corporate treasury-funded ad bans72 and by
(putatively) not revisiting McConnell’s holding concerning the facial validity of the
statute.73
The decision and reasoning of WRTL highlight the use of the as-applied/facial
challenge distinction as a strategy to accomplish other doctrinal goals. It is very difficult
to reconcile the holding of WRTL with the Court’s rejection of the facial challenge to
Title II in McConnell. Perhaps it is so obvious that it need not be said, but this
vindication of the as-applied challenge can best be explained as Chief Justice Roberts and
Justice Alito’s rough compromise between fidelity to precedent (stare decisis) and their
fundamental disagreement with McConnell’s holding. The as-applied holding allows
them to carve an exception to the law that is as large as the legislative record justifying it:
most of the “objectionable” ads that formed the core justification of Title II’s sweeping
70
cite
See cite
72
The Court reconciles Austin v. Michigan Chamber of Commerce, Bellotti and McConnell so as to come
to the conclusion that
71
in a way that ends up with
by coming to the conclusion that the unque anti corruption interest
73
19
regulation turn out to be constitutionally protected. By striking down this law as applied
to the facts of this ad and all others satisfying the “no reasonable interpretation standard”,
the Court can say that BCRA Title II is constitutional in theory, but rarely in practice.
Moreover, as Justice Alito’s separate opinion states, the new as-applied standard allows
for the possibility that further challenges will demonstrate its unworkability and prove the
necessity of overturning the law on its face.
Completely unmentioned in the controlling opinion is the fact that Title II of the
BCRA had its own built-in way to accommodate Roberts and Alito’s misgivings: the
backup definition of electioneering communications, which would be triggered if the
primary definition were held unconstitutional. In the event the primary definition of
electioneering communications were declared unconstitutional, Title II provided a
secondary definition:
any broadcast, cable, or satellite communication which
promotes or supports a candidate for that office, or attacks
or opposes a candidate for that office (regardless of
whether the communication expressly advocates a vote for
or against a candidate) and which also is suggestive of no
plausible meaning other than an exhortation to vote for or
against a specific candidate74
However, because it limits itself to the law as applied to those plaintiffs, the WRTL
holding does not technically trigger the backup definition.
74
cite
20
From the standpoint of as-applied doctrine or jurisprudence, the Court’s decision
to redefine the primary definition of electioneering, rather than strike it down and trigger
the backup, is peculiar along several dimensions. The remedy to an as-applied challenge
depends on severability of the relevant statute – that is, the court carves out an exception
for the plaintiff while remaining true, if possible, to the purpose of the statute. However,
WRTL guts the primary definition of electioneering and substitutes its own definition,
which is very close, but not precisely the same, as the backup definition.75 In doing so,
the Court rewrites a law that does not need to be rewritten and does so in a way that
Congress specifically avoided. The normal preference for as-applied challenges arising
from considerations of judicial restraint is completely inapposite here when Congress has
clearly expressed its preference in legislation and the Court’s sweeping exception
swallows the legislative rule while providing for a third option for which no one voted.
Given the similarities in impact between the Court’s new standard and the backup
definition, it becomes even more difficult to characterize the decision as truly an asapplied holding. One cannot think of an ad that could be regulated under the Court’s
definition of electioneering (i.e., one capable of no reasonable interpretation other than an
exhortation to vote for or against) but would not be regulated under the statute’s backup
definition (which covers all ads that both are incapable of no other “plausible meaning”
other than an exhortation to vote for or against and also promote, support, attack or
oppose a candidate). If that is right, then the Court actually struck down the primary
definition on its face by carving out all possible applications of the primary definition that
would not have been covered by the secondary definition.
75
Note slight differences here.
21
Finally, there is no mention in the controlling opinion at all that the First
Amendment standard for facial challenges differs than that for the run-of-the mill
constitutional case. As noted above, “substantial overbreadth” is all that would be
necessary to strike down a speech restriction on its face. It is almost as if campaign
finance (or perhaps election law more generally) were viewed as an exception to the
First Amendment exception to facial challenges. Surely, the logic of the opinion and a
cursory look at the legislative record suggest that, at a minimum, a “substantial number”
of the applications of the primary definition of electioneering are unconstitutional,
“judged in relation to the statute’s plainly legitimate sweep.”76
b. Washington Grange
This last critique of WRTL -- that it ignores the uniqueness of the facial/as-applied
distinction in the First Amendment context – can also be lodged at the Court’s 7 to 2
decision in Washington State Grange v. Washington.77 There the Supreme Court rejected
a facial challenge to Washington’s modified blanket primary system, which allowed any
candidate to express a “party preference” on a unified primary ballot. The parties argued
that this represented forced association in violation of the First Amendment because
voters would assume the party had endorsed, nominated or merely associated itself with
such a candidate. The Supreme Court rejected the facial challenge but left open the door
to future as-applied challenges once the state had crafted an actual ballot in which a
plaintiff could show voter confusion.
76
77
New York v. Ferber, 458 U.S. 747, 769-771 (quoting Broadrick v, Oklahoma, 413 U.S. 601 (1973)).
128 S. Ct. 1184 (2008).
22
Washington Grange illustrates the interrelationship (or misunderstanding)
between the as-applied doctrine and the doctrines of ripeness and constitutional
avoidance. The deficiency in the facial challenge brought by the parties there was not
that it failed, per se, to prove the law was unconstitutional in all its applications. Rather,
the challenge failed because the law had not yet been applied at all so it was unclear if it
would be unconstitutional in all or any of its applications. This is a subtle but important
difference. At this preenforcement stage, the Court could not strike down the law,
because it did not know what the law actually meant nor could it discern whether a
constitutional violation would occur once it was enforced. Unlike the paradigmatic
rejection of a facial challenge on the grounds that some applications of the law would be
constitutional, here the Court suggested that the law might be unconstitutional in all of its
applications (that is, if it caused voter confusion), but such an injury was mere
speculation at the time the lawsuit was filed.78
Perhaps more than in any other case – certainly any other election law case – the
Washington Grange opinion explains the virtues of avoiding facial invalidations if at all
possible. The Court warns that claims of facial invalidity because they “often rest on
speculation . . . . raise the risk of ‘premature interpretation of statutes on the basis of
factually barebones records.”79 It also extols the virtue of a bias against facial
invalidation as arising from general principles of judicial restraint (i.e., not deciding a
constitutional issue unless absolutely necessary and even then crafting as narrow a rule as
78
Cite. One other possible interpretation would be that the Court, once voter confusion would be shown,
would only strike down the law as applied to the party which had been injured by the confusion. That
seems implausible given that such relief would then never cure the constitutional injury. At a minimum,
the Court might prospectively immunize the party from having candidates use its name as a party
preference. But it is difficult to see why one party would be susceptible to voter confusion as to who it
supports while another would not be.
79
Grange, 128 S. Ct. at 1191 (quoting Sabri v. United States, 541 U.S. 600, 609 (2004)).
23
possible) and respect for the democratic process.80 Therefore, while a law that
authentically confused voters as to whether a candidate was in fact the party’s nominee or
even a member might be unconstitutional, one that merely allowed candidates to state a
party preference was not. The Court did not yet know which law the plaintiffs were
challenging, and needed to await both the state’s interpretation and enforcement of the
law and the concomitant confusion, if any, that it would produce.
Rather than describing this decision as one upholding the law on its face while
reserving the right to strike it down as applied to some later plaintiff (a la McConnell), it
seems more appropriate to say the Court (despite express language to the contrary)
merely refused to pass judgment on the law because the factual record was inadequate to
establish a constitutional violation. One could describe this as a argument concerning
ripeness, standing, abstention, or simply the failure of the plaintiffs to provide an
adequate factual record to demonstrate an “injury in fact.” But whatever one calls it, the
“upholding” of the law against a facial attack was really more about dismissing the claim
because of a lack of information about the law and its effects.
Perhaps the best way to reveal what the Court was really doing is to envision
what the next case would look like. Suppose a David Duke-like character runs in the
primary and expresses a preference for the Republican Party. And suppose that
preference is expressed by putting an “R” next to his name on the ballot, but the ballot in
small print at the bottom of the first page (or even worse, in some separate ballot
pamphlet) indicates that party preference designations do not imply any association,
membership or endorsement of the political party. And suppose a later poll or other
study reveals that most voters thought Duke was endorsed by the Republican Party. The
80
Id.
24
Republican Party brings suit, now having already suffered the injury (suppose, even
worse, that Duke wins office so the injury is long-lasting or irreparable). If it launches a
successful challenge to the law, can the Court strike down the law on its face based on
these new developments or could it, at most, strike it down as-applied to the Republican
Party and any other party victimized by similar confusion?
In the traditional way of considering facial invalidity, it would appear that the law
should be struck down on its face. The law is not unconstitutional because of the specific
and unique harms befalling the Republican Party; we simply know more about the
meaning of the law and the extent of the potential injury than we did when the statute
stood unenforced. Given that no remedy the Court could order could retroactively undo
the harm of the confusion that led to Duke’s election, it makes no sense to suggest (as is
often the case when considering unconstitutional applications) that the law was
unconstitutional as applied to the Republican Party in the context of a Duke candidacy.
Any relief must be of a kind that holds this law, as enforced and interpreted, is
unconstitutional as applied to every party for every candidate based on the evidence
gleaned from the Duke experience.81
The Court’s decision rejecting the facial challenge conflates two problems: the
absence of a ballot (or state court interpretation) implementing the law and the absence of
actual confusion arising from whatever ballot format the state employs. The difference
between those two deficiencies in this case is important for cases going forward. If the
81
One might say that the Court should strike down, on its face, the interpretation of the law by the state
courts or administrative bodies while upholding the statute on its face because it could have been
interpreted or enforced differently. There are two responses to this objection. First, the meaning of a state
law for purposes of federal court review includes the state’s definition and interpretation of the law.
Second, even if one believes that the Court was, in a technical sense, upholding the statute but striking
down its interpretation, it is doing both on their face. The Court’s decision will apply to all potential
plaintiffs because the injury will extend to them based on this new knowledge of how the law works.
25
principal problem was the Court’s not knowing what the law would mean and what the
ballot would look like, then the decision has a minor effect in the run-of-the-mill case in
which the meaning of a law is clearer. Moreover, a challenge to the law need only wait
until the ballot is printed, not until the election of a candidate objectionable to the party
arising from voter confusion. If the crux of the Court’s decision rests on the absence of
voter confusion on the record before it, then the implications may be more dramatic and
harmful. The result would be that a party (or rather, the polity) must first suffer the
injury of voter confusion in an actual election before it (or any other party) can prove the
laws unconstitutionality.
This is more than a mere academic or semantic point. The effect of the Court’s
decision is to force plaintiffs to suffer irreparable harm (in this case, confuse voters in an
election) in order to generate evidence as to the law’s unconstitutionality. Chief Justice
Roberts’ concurrence took a different approach, which illustrates the problem with the
majority’s couching the case in the terms of as-applied or facial invalidity. For him, the
only dispositive question was what the ballot would look like. “If the ballot is designed in
such a manner that no reasonable voter would believe that the candidates listed there are
nominees or members of, or otherwise associated with, the parties the candidates claimed
to “prefer,” the I-872 primary system would likely pass constitutional muster. I cannot
say on the present record that it would be impossible for the State to design such a ballot.
. . . On the other hand, if the ballot merely lists the candidates' preferred parties next to
the candidates' names, or otherwise fails clearly to convey that the parties and the
candidates are not necessarily associated, the I-872 system would not survive a First
26
Amendment challenge.”82 If the judge considers confusion to be a likely consequence of
the ballot design, Roberts would have him strike it down on its face without having to
wait for actual confusion at the polls. “Nothing in my analysis,” Roberts wrote, “requires
the parties to produce studies regarding voter perceptions on this score, but I would wait
to see what the ballot says before deciding whether it is unconstitutional.”83
c. Crawford v. Marion County
The Washington Grange case seemed like an exotic brand of election law disputes
until a plurality of the Court used its holding concerning as-applied challenges as
precedent for the Court’s most significant election case of the term: Crawford v. Marion
County.84 The Court there upheld Indiana’s voter identification requirement against a
facial challenge, while expressly reserving the possibility that it might be unconstitutional
as applied to particular plaintiffs in some later case. Like Washington Grange, however,
the opinion is somewhat unclear as to whether the problem with the plaintiffs’ case was a
bad set of facts, an inadequate record, poor lawyering, or premature adjudication.
The Court split three ways in Crawford. Justice Stevens wrote the controlling
opinion for himself, Justice Kennedy and Chief Justice Roberts, rejecting the facial
challenge, but leaving open the possibility that a person who found it very difficult to get
a photo ID could launch an as-applied challenge.85 Justice Scalia concurred, joined by
Justices Thomas and Alito, arguing that the law should be upheld on its face and future
82
128 S. Ct. 1197 (Roberts, C.J., concurring).
Id. at 1197 (Roberts, C.J., concurring)..
84
See 128 S. Ct.1610, 1622.
85
cite
83
27
as-applied challenges should be foreclosed.86 Justice Souter, joined by Justice Ginsburg,
submitted a dissent, as did Justice Breyer.87
The controlling opinion is somewhat inexact in its assessment of the severity of
the burden on voting rights posed by the photo ID law and of the number of people for
whom it will present a severe burden.88 In part, this is a result of what Justice Stevens
finds to be insufficient facts in the record on these critical points.89 The plaintiffs did not
produce a single person who would say he or she would not vote because of the ID
requirement.90 The number of Indianans without ID was uncertain and contested,91 and
as to the difficulties faced by the elderly and indigent voters in their attempts to obtain
ID, “the record says virtually nothing.”92 “Even assuming an unjustified burden on some
voters,” the controlling opinion maintains (citing Ayotte and Washington Grange), those
challenging the law had not demonstrated that facial invalidation was the proper
remedy.93
Left hanging in the opinion is the question what type of evidence might have been
sufficient to warrant striking down the law on its face or what future plaintiffs would
need to demonstrate to prove the law unconstitutional as applied to them. If the opinion
really is about the deficiencies in the record, then it seems to leave open the possibility
that the law is both unconstitutional on its face and as-applied to some subset of voters
who have difficulty getting photo ID. At the time the district court had heard the
86
cite
cite
88
128 S. Ct. at 1623 n20.
89
“[O]n the basis of the evidence in the record it is not possible to quantify either the magnitude of the
burden on this narrow class of voters or the portion of the burden imposed on them that is fully justified.”
128 S.Ct. at 1622.
90
Cite to posner appellate ct opinion.
91
128 S. Ct. at 1622.
92
128 S. Ct. at 1622.
93
128 S. Ct. at 1623.
87
28
challenge to the ID law, no statewide election had taken place, and as mentioned before,
no voter came forward to say they would not be able to vote because of the law. If in a
subsequent election it turns out that a large number94 of voters cannot vote because of the
ID requirement, maybe we will learn that the law is unconstitutional on its face because
the state’s anti-fraud interests do not justify such an impediment for such a large number
of people. Or, even if the law is constitutional on its face, perhaps a small group of
voters (or maybe even just an individual voter) has sufficient difficulty getting an ID that
they can prove the law unconstitutional as applied to them. Such might have been the
case for the now-famous group of nuns that had difficulty voting in the Indiana
presidential primary.95
Of course, these options are more theoretical than real, at least in the context of
the Indiana law. The number of voters as to whom one could definitively prove that
acquiring a photo ID is severely burdensome such that they will not vote almost certainly
does not rise to a level that would justify striking the law down on its face.96 This is
especially true given the credence and significance the Court attributed to the state’s antifraud interests. And for those voters who face a special burden navigating the
bureaucratic hurdles to acquire an ID, launching a federal as-applied lawsuit would
hardly seem to be an easier path to the voting booth.
The lack of clarity in the meaning of the Crawford decision illustrates unique
problems concerning the as-applied/facial distinction in voting rights cases. In the
ordinary constitutional case governed by the Salerno standard, the demonstration of a
Of course, “how large a number of voters” and “how severe the burden” will remain the critical
questions.
95
cite
96
Michael Pitts estimates that approximately 400 voters in the Indiana presidential primary election did not
have their provisional ballots counted because of a failure to present ID.
94
29
severe burden on the rights of a subgroup disparately affected by the law (all else equal)
would lead a court to find the law unconstitutional as applied to that group but not those
for whom the law does not present a severe burden. However, the constitutional test that
comes from Anderson v. Celebrezze97 and which the Court applied in Crawford
acknowledges the possibility that a severe burden on a minority of people affected by the
law could still justify facial invalidation of the law.98 In Crawford, the controlling
opinion held that the petitioners failed to prove that the law posed “‘excessively
burdensome requirements’ on any class of voters.”99 Therefore, if the plaintiffs in a
future case could demonstrate a severe burden on at least some voters, the Court would
then need to answer whether the class of severely burdened voters was sufficiently large
to outweigh the state’s antifraud interests and whether the proper remedy would be
invalidating the statute or simply protecting the burdened voters by striking it down as
applied to them.
The potential effect of the court’s facial holding was not lost on Justice Scalia,
who thought the only constitutional question was the impact of the photo ID requirement
on voters generally, not any subgroup in particular.100 He derided the controlling opinion
as relying on a “record-based resolution of these cases.”101 For him, the fact that the law
applied to everyone and had no invidious discriminatory purpose was the end of the
matter.102 The burden on all voters was the same, even though some voters, because of
97
cite
128 S. Ct. 1610, 1623
99
128 S. Ct. at 1623 (“In sum, on the basis of the record that has been made in this litigation, we cannot
conclude that the statute imposes ‘excessively burdensome requirements’ on any class of voters.”)
(citations omitted, emphasis added).
100
128 S. Ct. at 1625 (Scalia, J., concurring).
101
128 S. Ct. at 1627.
102
To hold otherwise “would effectively turn back decades of equal-protection jurisprudence,” Scalia
argued, because “[t]he Fourteenth Amendment does not regard neutral laws as invidious ones, even when
98
30
their personal situations, felt different impacts from the law. Although he did not weigh
in on whether the law might be unconstitutional as applied to voters suffering from
unique impacts, his discomfort with the specter of “constant litigation” where the
“potential allegations of severe burden are endless” suggests even those voters would be
out of luck.103
IV.
An Election Law Exception to the Doctrine Governing Facial Challenges?
Justice Scalia’s concurrence did not end with his objection to the “record-based
resolution” of the voter ID case. What he next suggested provides a global argument as
to why in the election law arena the court should prefer facial challenges to as-applied
ones. “This is an area where the dos and don’t need to be known in advance of the
election, and voter-by-voter examination of the burdens of voting regulations would
prove especially disruptive.”104 Ordinarily one thinks of Justice Scalia as pushing in the
direction of as-applied challenges,105 but in all three of the election law cases discussed
above he would have either upheld (Crawford) or struck down (Washington Grange and
WRTL) the laws on their face. Even if one disagrees with where he would come out, his
concern as to the costs of as-applied challenges in election law cases should also worry
those who would be more receptive to regulations of campaign finance and party reform
their burdens fall disproportionately on a protected class.” 128 S.Ct. at 1626 (emphasis in original). This
is a peculiar argument to make in the voting rights context given that it derives from the fundamental
interest prong of equal protection, not the suspect classification prong. Indeed, this approach would itself
turn back the clock – effectively overruling a series of cases from Reynolds v. Sims (one person, one vote)
to Bush v. Gore (vague intent of the voter standard deemed unconstitutional under Equal Protection clause).
Justice Scalia appears to make an exception for poll taxes and candidate filing fees because they were
“[]related to money,” 128 S.Ct. 1610, 1626 n*, which is a way to distinguish those cases, but not explain
them.
103
128 S.C.t 1626.
104
128 S.Ct at 1626.
105
cite
31
and more concerned about voter ID laws. Indeed, as a general matter, the asapplied/facial challenge doctrine in the election law arena (1) ought to facilitate the
development of clear pre-election rules, (2) should recognize the irreparability of the
injury that often occurs if the results of an election must serve as the record for
establishing an individual’s voting rights claim; and (3) should acknowledge that the
costs to the individual in bringing a voting rights lawsuit almost always outweigh the
benefits of vindicating her individual right to vote.
We should also reiterate that the as-applied/facial doctrine has grown to mean a
lot more in the election law context than deciding the type or category of a constitutional
challenge. As WRTL, Washington Grange and Crawford demonstrate in different ways,
doctrines of ripeness, standing, severability, abstention and justiciability have begun to
overlap with doctrine concerning as-applied and facial challenges. As much as our first
preference would be a crisper, cleaner and clearer demarcation between the different
constitutional questions, the confusion highlights the need for our recommended
principles just mentioned. However one categorizes these various inquiries the Court
undertakes en route to striking down a statute in whole or in part, we still would urge an
adaptation of the inquiry toward one that recognizes the importance of clear preelection
rules that do not depend on evidence from an electoral meltdown or on piecemeal
litigation from individual plaintiffs, few of whom have a real incentive to bring such
claims.
The controlling opinions in all three of the election law cases discussed here
violate these principles. Leaving aside the point made above as to the lack of restraint
embodied in the WRTL Court’s move to ignore the backup definition of electioneering
32
Congress placed in the statute, the decision’s reliance on as-applied challenges under a
vague standard force potential corporate speakers into a Hobson’s choice. On the one
hand, they could allow the Court’s vague standard to chill their pre-election speech. On
the other, they could run their advertisement and risk being fined by the FEC or could
bring their own preenforcement action, neither of which is likely to be worth it for the
average corporate speaker and especially for non-profits. Similarly, no individual who
has trouble getting an ID (such as one of the elderly Indiana nuns) is likely to launch a
federal lawsuit to have the Indiana law struck down as unconstitutional as applied. This
mismatch between the litigation costs and the benefits does not apply to the political
party plaintiffs in any future challenge to the Washington primary. However, there is
great risk that the Court’s decision in Washington Grange will be read to require them to
suffer irreparable harm – the election of an objectionable candidate bearing their name –
before the necessary evidence of voter confusion (and therefore constitutional injury)
materializes.
Nothing in these arguments should be read as endorsing a particular result in any
of the cases the Court decided. (Indeed, the coauthors of this article do not agree between
themselves which way each case should go.) We would agree, however, that the Court
should have been more aggressive in deciding whether BCRA Title II, the Washington
primary law, and the Indiana voter ID law were unconstitutional on their face. If the case
is not ripe or other reasons exist for a federal court to stay its hand (e.g., if, as in
Washington Grange, the actual meaning of the law is unknown), then the court should
dismiss the case, not uphold the law. When it decides the merits of a case, it should
33
reverse the normal presumptions of constitutional adjudication and decide the facial
constitutionality of a statute unless compelling reasons suggest otherwise.
What we are arguing for here is akin to expansion of “substantial overbreadth”
analysis to the election law context. Indeed, as we wondered earlier in this Article, it
remains a puzzle why in the First Amendment election contexts of campaign finance and
party associational rights, courts do not seem to conduct such an analysis. In a case
challenging barriers to voting, something like overbreadth analysis seems unavoidable.
The inquiry in Crawford, as in any similar case, boils down to whether the state can
justify the likelihood that a certain number of voters will find it difficult or impossible to
vote because of the law. Although “how many voters?” and “how difficult?” remain the
important questions under any inquiry, the Court must decide whether the state interests
justify the burden placed on the few, but that do not affect the majority. That question
cannot be answered by saying that the burdened voters ought to bring as-applied
challenges: poll taxes are not unconstitutional as applied only to poor people, and literacy
tests would not be unconstitutional (if at all) as applied only to illiterate people. The
prerequisite to voting is either justifiable – in the abstract and given the recognized
potential to prevent some from voting – as a proper means of serving some state interest
or not.
V.
Conclusions
Like the doctor who sees the disease but is unsure what therapy to prescribe, we
are more confident in our descriptive account of what is happening in the Roberts Court
election law decisions than our normative recommendation that the Court should change
34
the normal presumptions governing facial challenges for this constitutional arena. Still,
this story of evolution in the meaning of as-applied and facial challenges is an important
one to tell. It recounts both strategic moves and doctrinal expansion that may have
significant consequences both for the type of election-related challenges we will see in
the near future and their chances of success.
The strategic side to this shift in doctrinal meaning has two components to it. The
first is the critique often lodged with respect to as-applied challenges: that a bias against
facial challenges makes civil rights litigation more costly and less likely. Impact
litigation is rarely worthwhile if the only remedy plaintiff lawyers can achieve is a narrow
exception for their clients and others similarly situated. The second is the use of the asapplied/facial challenge doctrine as an exit strategy from disputed precedent. As the
WRTL case demonstrates, a series of as-applied challenges can allow the Court to
overturn even recent precedent but methodically punching holes in it until the exception
becomes the rule.
The story as to doctrinal expansion is both a semantic point and a cautionary note.
The as-applied/facial distinction is now being used to describe problems formerly labeled
justiciability, standing, ripeness, abstention, and constitutional avoidance. Were this only
a doctrinal relabeling it would barely be worth noting. However, under those other
doctrines, a Court usually stays its hand and chooses not to decide a constitutional issue.
In all of the contexts described here, the Court notes similar deficiencies as it would when
it labeled the identical problem something else, but then decides the case on the merits.
The Washington Grange case is the best example of that. There the Court admits not
knowing what the law means (usually cause for abstention) or whether the plaintiff has
35
yet to be injured (usually a concern for standing or ripeness), but nevertheless upholds the
Washington primary system on its face.
It remains to be seen whether the developments in the doctrine governing asapplied and facial challenges are a way station toward greater clarity in election law or a
permanent fixture in how such cases will need to be litigated. The having-your-cake-andeating-it-too quality of the new doctrine is a difficult benefit for the Court to give up. For
those of us who seek clarity and finality in the rules governing elections, the costs of this
doctrinal confusion exceed its benefits.
36