Shapiro v. McManus

Filing 202

MEMORANDUM Before NIEMEYER, Circuit Judge, and BREDAR and RUSSELL, District Judges.(jnls, Deputy Clerk)

IN THE UNITED STATES DISTRICT COURT
FOR THE DISTRICT OF MARYLAND
THREE-JUDGE COURT

O. JOHN BENISEK, et al.,

*

Plaintiffs

*

v.

*

LINDA H. LAMONE, et al.,

*

Defendants
*

*

CIVIL NO. JKB-13-3233

*
*

*

*

*

*

*

*

*

*

*

MEMORANDUM
Before NIEMEYER, Circuit Judge, and BREDAR and RUSSELL, District Judges.
BREDAR, District Judge. On May 31, 2017, Plaintiffs O. John Benisek, et al. (“Plaintiffs”)
filed a Rule 65(a) Motion for a Preliminary Injunction and to Advance and Consolidate the Trial
on the Merits or, in the Alternative, for Summary Judgment. (ECF No. 177.) The State responded
on June 30, 2017, with a Cross-Motion for Summary Judgment. (ECF No. 186.) Both motions
have been briefed. On June 28, 2017, this three-judge Court set in a hearing on Plaintiffs’
preliminary injunction motion. On its own motion, the Court directed the parties to also address
whether further proceedings in this case should be stayed pending the Supreme Court’s decision
in Gill v. Whitford, No. 16-1161, a political gerrymandering case set to be argued in the
forthcoming Term. A hearing on both matters was held on July 14, 2017.1

1

In a pre-hearing scheduling order, the Court made clear that the only matters it would take up at the July 14 hearing
were Plaintiffs’ motion for preliminary injunctive relief and the Court’s sua sponte request for argument on the
propriety of a stay. (ECF No. 190.) The Court did not then, and does not now, rule on the pending cross-motions for
summary judgment. Nor has the Court advanced the trial on the merits under Rule 65(a)(2).

For the reasons explained below, the Court now DENIES Plaintiffs’ preliminary injunction
motion and STAYS this case pending the outcome of Whitford. As set forth in Part II.B, Judge
Bredar concludes that such action is necessary because the justiciability of political
gerrymandering claims remains in doubt, but the Supreme Court will likely resolve or clarify this
threshold jurisdictional matter in its Whitford decision. As set forth in Part II.C, Judges Bredar
and Russell conclude that Plaintiffs have not made an adequate preliminary showing that they will
likely prevail on the causation element of their First Amendment retaliation claim. While the Court
by no means excludes the possibility that Plaintiffs may ultimately prevail, Plaintiffs have not
demonstrated that they are entitled to the extraordinary (and, in this case, extraordinarily
consequential) remedy of preliminary injunctive relief. A stay pending further guidance in
Whitford is appropriate at this juncture.
As set forth in his dissenting opinion, Judge Niemeyer would grant Plaintiffs’ motion for
preliminary injunctive relief.
I.

Procedural History
A review of the recent history of this redistricting case may prove helpful. Following a

remand from the Supreme Court on a procedural issue, see Shapiro v. McManus (Shapiro I), 136
S. Ct. 450 (2015), the case was assigned to a three-judge panel composed of Circuit Judge
Niemeyer and District Judges Bredar and Russell. (ECF No. 42.) On March 3, 2016, Plaintiffs
filed a Second Amended Complaint challenging Maryland’s 2011 congressional districting map
as an unconstitutional political gerrymander. (ECF No. 44.) The State moved to dismiss pursuant
to Rule 12(b)(6) of the Federal Rules of Civil Procedure. (ECF No. 51.)
On August 24, 2016, the Court denied the State’s motion to dismiss in a 2–1 decision, with
Judge Bredar dissenting. See Shapiro v. McManus (Shapiro II), 203 F. Supp. 3d 579 (D. Md.

2

2016). In its ruling, the panel majority held that Plaintiffs’ Second Amended Complaint stated a
justiciable claim for relief. The majority went on to endorse a standard for assessing political
gerrymandering claims under the First Amendment:
When applying First Amendment jurisprudence to redistricting, we
conclude that, to state a claim, the plaintiff must allege that those responsible for
the map redrew the lines of his district with the specific intent to impose a burden
on him and similarly situated citizens because of how they voted or the political
party with which they were affiliated. In the context of redistricting, this burden is
the injury that usually takes the form of vote dilution. . . . [T]o establish the injury
element of a retaliation claim, the plaintiff must show that the challenged map
diluted the votes of the targeted citizens to such a degree that it resulted in a tangible
and concrete adverse effect. . . . Finally, the plaintiff must allege causation—that,
absent the mapmakers’ intent to burden a particular group of voters by reason of
their views, the concrete adverse impact would not have occurred.
When a plaintiff adequately alleges the three elements of intent, injury, and
causation . . . he states a plausible claim that a redistricting map violates the First
Amendment and Article I, § 2. Of course . . . the State can still avoid liability by
showing that its redistricting legislation was narrowly tailored to achieve a
compelling government interest.
Id. at 596–97.2
Following the Court’s decision at the pleading stage, the parties entered a contentious
period of discovery, which resulted in voluminous procedural rulings that need not be reviewed
here. At the conclusion of this discovery period, the parties filed their pending motions. (ECF
Nos. 177, 186.)
As explained more fully in Part II, the Court concludes that preliminary injunctive relief is
inappropriate at this stage because Plaintiffs have not shown that they can likely prevail on each
of the three elements of their First Amendment claim. Moreover, any further proceedings—

Judge Bredar disagreed that Plaintiffs had identified a workable standard because (1) “the Supreme Court has
expressed some degree of tolerance for partisanship in the districting context, but that tolerance creates intractable
line-drawing problems”; and (2) courts are ill-equipped to “ascertain those unusual circumstances in which
redistricting inflicts an actual, measurable burden on voters’ representational rights,” yet that is “precisely what the
Supreme Court has required.” Shapiro II, 203 F. Supp. 3d at 601 (Bredar, J., dissenting). Ultimately, Judge Bredar
concluded, there is no reliable, administrable standard for “distinguishing electoral outcomes achieved through
political gerrymandering from electoral outcomes determined by the natural ebb and flow of politics.” Id. at 606.
2

3

whether in relation to the pending cross-motions for summary judgment or at a bench trial—would
be premature because the Supreme Court is poised to consider issues that go to the heart of
Plaintiffs’ gerrymandering case. Until the Supreme Court speaks, prudence compels this Court to
stay further proceedings.
II.

Analysis
A. Standard of Decision
1. Preliminary Injunction
Plaintiffs seek preliminary injunctive relief in the form of an order barring the State from

enforcing the 2011 redistricting plan and requiring the State to implement a new map in advance
of the 2018 midterm elections. To prevail on their motion for such relief, Plaintiffs must show (1)
that they are likely to succeed on the merits of their political gerrymandering claim, (2) that they
will likely suffer irreparable harm in the absence of preliminary relief, (3) that the balance of
equities tips in their favor, and (4) that an injunction would serve the public interest. WV Ass’n of
Club Owners & Fraternal Servs., Inc. v. Musgrave, 553 F.3d 292, 298 (4th Cir. 2009) (citing
Winter v. NRDC, 555 U.S. 7, 20 (2008)). “A preliminary injunction is an ‘extraordinary remed[y]
involving the exercise of very far-reaching power’ and is ‘to be granted only sparingly and in
limited circumstances.’” Int’l Refugee Assistance Project v. Trump, 857 F.3d 554, 588 (4th Cir.
2017) (alteration in original) (quoting MicroStrategy Inc. v. Motorola, Inc., 245 F.3d 335, 339 (4th
Cir. 2001)), cert. granted, 137 S. Ct. 2080 (2017).
Rule 52(a)(2) of the Federal Rules of Civil Procedure provides that in “granting or refusing
an interlocutory injunction, the court must . . . state the findings and conclusions that support its
action.” See Greenhill v. Clarke, 672 F. App’x 259, 260 (4th Cir. 2016) (per curiam) (“Rule
52(a)(2) . . . requires that the district court make particularized findings of fact supporting its

4

decision to grant or deny a preliminary injunction; such findings are necessary in order for an
appellate court to conduct meaningful appellate review.”); accord Booker v. Timmons, 644 F.
App’x 219 (4th Cir. 2016) (mem.). Because Judge Bredar’s discussion in Part II.B, concerning
justiciability, involves a pure question of law, no findings are enumerated in that Part. However,
the opinion of the Court in Part II.C, concerning the causation element of Plaintiffs’ First
Amendment theory, includes findings germane to that issue as well as separately stated
conclusions of law. Such findings and conclusions are, given the procedural posture of this case,
preliminary, and they will not bind the Court in any future proceedings. See Blake v. Balt. Cty.,
662 F. Supp. 2d 417, 421 (D. Md. 2009) (citing Univ. of Tex. v. Camenisch, 451 U.S. 390, 395
(1981)).
2. Stay of Proceedings
The Supreme Court has long recognized that the “power to stay proceedings is incidental
to the power inherent in every court to control the disposition of the causes on its docket with
economy of time and effort for itself, for counsel, and for litigants.” Landis v. N. Am. Co., 299
U.S. 248, 254 (1936); see also Williford v. Armstrong World Indus., Inc., 715 F.2d 124, 127 (4th
Cir. 1983) (recognizing that courts enjoy the inherent authority to grant a stay “under their general
equity powers and in the efficient management of their dockets”). The decision to stay an action
“calls for the exercise of judgment, which must weigh competing interests and maintain an even
balance.” Landis, 299 U.S. at 254–55; see also Rogler v. Fotos, Civ. No. WDQ-14-228, 2015 WL
7253688, at *13 (D. Md. Nov. 17, 2015), aff’d, 668 F. App’x 462 (4th Cir. 2016) (mem.); Cutonilli

5

v. Maryland, Civ. No. JKB-15-629, 2015 WL 5719572, at *4 (D. Md. Sept. 28, 2015), appeal
dismissed, 633 F. App’x 839 (4th Cir. 2016) (mem.).
In deciding whether to stay proceedings, a court should consider the likely impact of a stay
on each party as well as the “judicial resources that would be saved by avoiding duplicative
litigation if the case is in fact stayed.” Mitchell v. Lonza Walkersville, Inc., Civ. No. RDB-123787, 2013 WL 3776951, at *2 (D. Md. July 17, 2013) (citing Yearwood v. Johnson & Johnson,
Inc., Civ. No. RDB-12-1374, 2012 WL 2520865, at *3 (D. Md. June 27, 2012)).
B. Justiciability
At the pleading stage in Shapiro II, the panel majority recognized “the justiciability of a
claim challenging redistricting under the First Amendment and Article I, § 2, when it alleges intent,
injury, and causation.” 203 F. Supp. 3d at 598. Judge Bredar disagreed, writing that because (1)
Plaintiffs had “not shown that their framework would reliably identify those circumstances in
which voters’ representational rights have been impermissibly burdened” and (2) no “acceptable
alternative framework” had been identified, Plaintiffs’ claim must be treated as nonjusticiable. Id.
at 601–02 (Bredar, J., dissenting). Despite the disagreement among the members of the panel on
this threshold issue, the majority opinion remains the law of the case absent reconsideration by at
least two judges or intervention by the Supreme Court. This Memorandum does nothing to unsettle
that prior decision.
However, this case has long since passed the pleading stage.

Plaintiffs now seek

preliminary injunctive relief in the form of an order that, if entered, would cause an unprecedented
disruption in Maryland’s legislative and districting process. In granting such relief, the Court
would enjoin enforcement of a map that was duly enacted by the General Assembly of Maryland,
see Md. Code Ann., Elec. Law §§ 8-701 et seq., and that survived a voter referendum by a wide

6

margin. The remedy would require emergency action by the legislature. The time and resources
necessary to implement a new map would surely have the effect of scuttling other legislative
priorities in advance of the 2018 session. The remedy would be highly consequential.
In the arena of legislative and congressional districting, unelected federal judges should
exercise great caution before declaring unconstitutional the work product of the people’s elected
representatives. Cf. Davis v. Bandemer, 478 U.S. 109, 145 (1986) (O’Connor, J., concurring in
the judgment) (“The opportunity to control the drawing of electoral boundaries through the
legislative process of apportionment is a critical and traditional part of politics in the United States,
and one that plays no small role in fostering active participation in the political parties at every
level. Thus, the legislative business of apportionment is fundamentally a political affair, and
challenges to the manner in which an apportionment has been carried out . . . present a political
question in the truest sense of the term.”).
The preliminary injunction mechanism under Rule 65(a) of the Federal Rules of Civil
Procedure does not authorize a federal court to grant such an extraordinary remedy haphazardly.
Rather, the court must be confident, among other things, that the plaintiff has shown it is likely to
prevail on the merits of its claim. Winter, 555 U.S. at 20. That assessment is quite different from
the plaintiff-friendly evaluation of the pleadings under Rule 12(b)(6) and the Supreme Court’s
decisions in Ashcroft v. Iqbal, 556 U.S. 662 (2009), and Bell Atlantic Corp. v. Twombly, 550 U.S.
544 (2007). A court that has made a preliminary legal determination in the plaintiff’s favor must
decide at the Rule 65(a) stage whether the plaintiff has carried its burden to show it will likely

7

succeed on the merits. Intervening developments in the law and, in particular, signals from
appellate courts, must inform this analysis.
In this case, an intervening development casts a cloud over the panel majority’s prior ruling
as to the justiciability of Plaintiffs’ political gerrymandering claim. On June 19, 2017, the Supreme
Court agreed to hear argument in Gill v. Whitford, No. 16-1161, a direct appeal from a decision by
a three-judge panel that enjoined a Wisconsin legislative map as an unconstitutional political
gerrymander. Argument is calendared for October 3, 2017. The decision below in Whitford v.
Gill, 218 F. Supp. 3d 837 (W.D. Wis. 2016), is fairly remarkable in that it is the first district court
opinion since the Supreme Court’s splintered ruling in Vieth v. Jubelirer, 541 U.S. 267 (2004), to
(1) endorse a standard for adjudicating political gerrymandering claims, (2) apply that standard to
rule in the plaintiff’s favor, and then (3) order the state to draw a new map.3
In a 5–4 order, the Supreme Court stayed the district court’s judgment pending disposition
of the appeal. The Court declined to note probable jurisdiction, ordering instead that “[f]urther
consideration of the question of jurisdiction is postponed to the hearing of the case on the merits.”
Plaintiffs in this case brush aside the justiciability question in Whitford as the “last of the five
questions presented” in that appeal (ECF No. 193 at 2), and the dissent makes no mention of
Whitford. Yet the Supreme Court’s decision to hold over the jurisdictional question for argument
is a strong signal that the question remains unsettled in the minds of the Justices.
That should come as no surprise. The justiciability of political gerrymandering claims has
plagued the Court for decades. As the panel majority observed in Shapiro II, six Justices
acknowledged in Bandemer that such claims are theoretically justiciable, 478 U.S. at 125, but the
Court fractured on the standard for adjudicating these claims. Conversely, Chief Justice Burger

3

The Whitford panel addressed the remedy separately in an unpublished opinion, see Whitford v. Gill, No. 15-cv-421bbc, 2017 WL 383360 (W.D. Wis. Jan. 27, 2017).

8

and Justices O’Connor and Rehnquist would have held that political gerrymandering claims “raise
a nonjusticiable political question that the judiciary should leave to the legislative branch as the
Framers of the Constitution unquestionably intended.” Id. at 144 (O’Connor, J., concurring in the
judgment).
Eighteen years later, the Court revisited the question in Vieth, where four Justices (Chief
Justice Rehnquist and Justices Scalia, O’Connor, and Thomas) would have held “that political
gerrymandering claims are nonjusticiable and that Bandemer was wrongly decided.” 541 U.S. at
281. Justice Kennedy, the swing vote, declined to sign on to the plurality opinion that would have
overruled Bandemer, but he sounded sharp notes of caution, writing that there are “weighty
arguments for holding cases like these to be nonjusticiable; and those arguments may prevail in
the long run.” Id. at 309 (Kennedy, J., concurring in the judgment); see also id. at 317 (“The
failings of the many proposed standards for measuring the burden a gerrymander imposes on
representational rights make our intervention improper.”).
While the dissent in the instant case states that “five Justices in Vieth concluded that the
[political gerrymandering] issue remained justiciable,” post, at 44–45, Justice Kennedy’s opinion
was more guarded than that: it was so guarded, in fact, that the plurality characterized it as a
“reluctant fifth vote against justiciability at district and statewide levels—a vote that may change
in some future case but that holds, for the time being, that this matter is nonjusticiable.” Id. at 305
(plurality opinion) (emphasis added); see also Michael S. Kang, When Courts Won’t Make Law:
Partisan Gerrymandering and a Structural Approach to the Law of Democracy, 68 Ohio St. L.J.
1097, 1111 (2007) (“Justice Kennedy’s ambivalence leaves it bizarrely unclear where the law of
partisan gerrymandering stands. The plurality in Vieth, as a result, argued that Justice Kennedy’s
vote ought to be understood effectively, if not expressly, as ‘a reluctant fifth vote against

9

justiciability.’” (footnotes omitted)). Hardly a resounding triumph for those who would ask federal
courts to adjudicate political gerrymandering disputes, Vieth was the last case in which the Court
squarely confronted the question.4
The Supreme Court’s willingness to consider and reconsider the justiciability question is
understandable, given how fundamental that question is to the exercise (and even the legitimacy)
of federal judicial power. Justiciability is a threshold matter that courts are required to evaluate,
sua sponte if necessary, before reaching the merits of a case. “Justiciability concerns ‘the power
of the federal courts to entertain disputes, and . . . the wisdom of their doing so.’” Republican
Party of N.C. v. Martin, 980 F.2d 943, 950 (4th Cir. 1992) (alteration in original) (quoting Renne
v. Geary, 501 U.S. 312, 316 (1991)); see also Hamilton v. Pallozzi, 848 F.3d 614, 619 (4th Cir.
2017) (“Justiciability is an issue of subject-matter jurisdiction, and we have an independent
obligation to evaluate our ability to hear a case before reaching the merits of an appeal.”); Proctor
v. Prince George’s Hosp. Ctr., 32 F. Supp. 2d 820, 824 (D. Md. 1998) (“It is appropriate for a
district court to raise issues of justiciability sua sponte.”).
Merely because the Supreme Court has agreed to hear argument in Whitford and has
deferred the jurisdictional question, it does not necessarily follow that the Court will clear up the
ambiguity next Term. The composition of the Court has changed dramatically since Vieth, as that
case was decided before Chief Justice Roberts and Justices Alito, Sotomayor, Kagan, and Gorsuch
took their seats. Nonetheless, it is conceivable that the Justices could again divide as the Court did
in Vieth, with a majority declining to agree on a standard but with at least five votes for the

4

In League of United Latin American Citizens (LULAC) v. Perry, 548 U.S. 399, 414 (2006), a majority of Justices
declined to address the question of justiciability. Chief Justice Roberts and Justice Alito stressed in a separate opinion
that they took “no position on that question, which has divided the Court.” Id. at 492–93 (Roberts, C.J., concurring
in part, concurring in the judgment in part, and dissenting in part). Justices Scalia and Thomas reiterated their view
that political gerrymandering claims are nonjusticiable. Id. at 512 (Scalia, J., concurring in the judgment in part and
dissenting in part).

10

proposition that some standard might yet exist. Or perhaps the Justices will endorse the standard
recognized by the three-judge court in Whitford, or some other standard; or perhaps they will rule
finally that federal courts may not adjudicate these types of political questions. It would be idle to
speculate as to the outcome of a case that has yet to be heard.
But with due respect to the other members of this panel, it would be irresponsible to grant
a drastic remedy on the basis of a claim that the Supreme Court may invalidate in a matter of
months. We know now that the Court is poised to consider the justiciability question. Guidance
of some sort (maybe dispositive guidance) is forthcoming. Accordingly, to suggest that Plaintiffs
are likely to prevail on the merits of their claim and to award injunctive relief on that basis would
place the cart far ahead of the horse.
This is particularly so in light of a case to which neither party has devoted much attention
and which, once again, the dissent does not mention. That case is Cooper v. Harris, 137 S. Ct.
1455 (2017), a racial gerrymandering case decided late last Term. In a separate opinion, Justice
Alito—joined by Chief Justice Roberts and, strikingly, Justice Kennedy—took a dim view on the
justiciability of political gerrymandering:
We have repeatedly acknowledged the problem of distinguishing between racial
and political motivations in the redistricting context. . . . As we have acknowledged,
“[p]olitics and political considerations are inseparable from districting and
apportionment,” and it is well known that state legislative majorities very often
attempt to gain an electoral advantage through that process.
Partisan
gerrymandering dates back to the founding, and while some might find it
distasteful, “[o]ur prior decisions have made clear that a jurisdiction may engage in
constitutional political gerrymandering . . . .”
Id. at 1488 (Alito, J., concurring in the judgment in part and dissenting in part) (citations omitted).
Justice Alito stressed that the Court’s cases require “extraordinary caution” any time the state has
“articulated a legitimate political explanation for its districting decision.” Id. at 1504 (internal
quotation marks and citation omitted). He added that “if a court mistakes a political gerrymander
11

for a racial gerrymander, it illegitimately invades a traditional domain of state authority, usurping
the role of a State’s elected representatives.” Id. at 1490 (emphasis added).
Justice Alito’s remarks are non-majority dicta in a case involving a different (though
analogous) claim. These remarks should not be treated as proof that any member of the Supreme
Court has prejudged the issues on appeal in Whitford. But see Crowe v. Bolduc, 365 F.3d 86, 92
(1st Cir. 2004) (“[C]arefully considered statements of the Supreme Court, even if technically
dictum, must be accorded great weight and should be treated as authoritative.” (citation omitted));
Jordon v. Gilligan, 500 F.2d 701, 707 (6th Cir. 1974) (“Even the Court’s dicta is of persuasive
precedential value.”); Fouts v. Md. Cas. Co., 30 F.2d 357, 359 (4th Cir. 1929) (“[C]ertainly dicta
of the United States Supreme Court should be very persuasive.”). However, these remarks are
further evidence that the justiciability question is far from settled and will likely be a focal point
at the October 2017 argument.
Nothing about this discussion should be taken to suggest that Judge Bredar has decided, as
a matter of law, that political gerrymandering claims are nonjusticiable. Indeed, two members of
this panel have already decided that such claims are justiciable pursuant to the First Amendment
framework that Justice Kennedy contemplated in Vieth, and the Supreme Court has not—to date—
overruled Bandemer or held that partisan gerrymandering presents a nonjusticiable political
question. Nor has the Court rejected Justice Kennedy’s First Amendment theory, though that
theory remains nothing more (or less) than a “theory put forward by a Justice of th[e] Court and
uncontradicted by the majority in any . . . cases,” Shapiro I, 136 S. Ct. at 456.5
The dissent simply is incorrect when it states that Judge Bredar advocates “judicial
abdication from partisan gerrymandering cases,” post, at 48. Far from it. A final decision by a

The dissent seems to suggest that political gerrymandering claims must be justiciable lest “unacceptable results”
obtain, such as a “pointillistic” map that assigns voters to various districts “regardless of their geographical location.”
5

12

majority of Justices instructing lower courts to apply a particular standard to resolve partisan
gerrymandering claims would be a welcome development in the law. See Shapiro II, 203 F. Supp.
3d at 600 (Bredar, J., dissenting) (“This opinion is not a defense of the State’s authority to segregate
voters by political affiliation so as to achieve pure partisan ends: such conduct is noxious and has
no place in a representative democracy.”). The point of this discussion is not to suggest that
political gerrymandering claims are not or should not be justiciable; rather, it is to call attention to
the uncertainty in the law, an uncertainty that was amplified two months ago when the Court
granted argument in Whitford. Pausing these proceedings to await further guidance from the
Supreme Court is not abdication: it is an expression of prudence, judicial restraint, and respect for
the role of a district court that must scrupulously adhere to the instructions of appellate authorities.
Because Plaintiffs are unable at this time to demonstrate that they will likely prevail on the
threshold question of justiciability, and because the Supreme Court is poised to act and in so doing
may change the legal landscape, Plaintiffs’ preliminary injunction motion should be denied and
their case stayed pending the Supreme Court’s decision in Whitford.
C. Causation
1. Preliminary Injunction
Apart from any doubts as to justiciability, and assuming without deciding that Plaintiffs
have adduced sufficient evidence to show that the State crafted the 2011 redistricting plan (and the

Post, at 28 (emphasis omitted). This case, of course, does not involve any such extreme practices. Whatever else
might be said, Maryland’s congressional districts generally adhere to traditional districting principles such as
contiguity and the preservation of communities of interest. Should a state legislature ever attempt to implement a
pointillistic map, a reviewing court could simply establish a bright line rule requiring some degree of contiguity on
the theory that pointillism subverts the framers’ intentions as expressed in Article I, § 2. A rule barring pointillism
would be easy to administer, would not require courts to predict voter behavior, and would not present the thorny linedrawing problems at issue in the typical political gerrymandering case. Pointillism would be the proverbial “easy
case” in this context, and the Court would be fortunate indeed to be confronted with such a simple challenge. It is not,
though, and we should not oversimplify the challenge of adjudicating the claim that is actually before us on the basis
of a hypothetical that has little to do with that claim.

13

Sixth District in particular) with the “specific intent to impose a burden” on Plaintiffs and similarly
situated citizens through vote dilution, Shapiro II, 203 F. Supp. 3d at 596, it is unclear whether
any such nefarious plan was and remains effective. This Court is not now persuaded that Plaintiffs
will likely prove that “absent the mapmakers’ intent to burden a particular group of voters by
reason of their views, the concrete adverse impact would not have occurred.” Id. at 597. Put more
simply, the Court is not yet persuaded that it was the gerrymander (versus a host of forces present
in every election) that flipped the Sixth District and, more importantly, that will continue to control
the electoral outcomes in that district. Voter decisions are mutable and subject to change, despite
voting history and party affiliation. As discussed below, the razor’s-edge Sixth District race in
2014 is evidence that suggests significant party-crossover voting and calls into doubt whether the
State engineered an effective gerrymander.
Trial testimony and other evidence, including thorough cross-examination, may yet
establish that Plaintiffs have met their burden of proof with respect to causation, but the Court is
not persuaded that they have done so now, at least not to the high standard set for the granting of
preliminary injunctions. Since but-for causation is an element of Plaintiffs’ First Amendment
claim, it follows that if Plaintiffs are unable to prove this element, their claim will collapse on its
merits. At this stage, the Court cannot say that it is likely that Plaintiffs will prevail on this

14

element—only that they might. For that reason, the Court must deny Plaintiffs’ request for the
extraordinary remedy of preliminary injunctive relief.
a. Findings of Fact
Strictly for purposes of deciding whether to enter a preliminary injunction, the Court makes
the following findings of fact, see Fed. R. Civ. P. 52(a)(2), corresponding to the causation element
of Plaintiffs’ First Amendment claim:
1. Maryland’s 2011 redistricting process involved two parallel procedures: a
public-facing procedure led by the Governor’s Redistricting Advisory
Committee and an internal procedure involving Maryland’s congressional
delegation and a consulting firm called NCEC Services, Inc. (ECF No. 177–4
at 36:4–13; ECF No. 177–5 ¶ 18.)
2. NCEC in turn designated analyst Eric Hawkins to review the State’s
redistricting plan and prepare sample maps using voter demographic data
(including party affiliation and voting history) and a computer program called
“Maptitude for Redistricting.” (ECF No. 177–4 at 36:18–37:17.)
3. In performing his analysis, Hawkins relied on a proprietary metric called the
Democratic Performance Index (DPI), a weighted average of candidate
performance that takes account of voting history. (Id. at 24:5–19.) A higher
DPI signals a greater statistical likelihood of Democratic candidate success
based on past performance.
4. Hawkins created between ten and twenty draft maps. He analyzed six maps
alongside proposals submitted by third parties. Each of the six maps would
have produced a federal DPI of 52% or greater for the Sixth District, while the
third-party submissions would have produced much lower DPIs. (Id. at 38:2–
9; ECF No. 177–34; ECF No. 177–35 at 31–32.)
5. There is no evidence that Hawkins personally created the final map that was
enacted into law. (ECF No. 177–1 at 13 n.9; ECF No. 186–1 at 11.) Former
governor Martin O’Malley testified that legislative director Joe Bryce and staff
from the Maryland Department of Planning likely created the final document.
(ECF No. 177–3 at 53:12–54:7.)
6. The map as enacted had the effect of transferring 360,368 Marylanders out of
the Sixth District and 350,179 Marylanders into the Sixth District. (ECF No.
177–19 at 12.) In the process, 66,417 registered Republicans were removed

15

from the district and 24,460 registered Democrats were added to the district.
(Id. at 6.)
7. After the 2011 plan was implemented, a plurality (44.8%) of voters in the Sixth
District were registered Democrats, while 34.4% of voters were registered
Republicans. 20.8% of voters were registered with neither major political party.
(ECF No. 186–19 at 5–6.)
8. The “Cook Partisan Voting Index” promulgated by the Cook Political Report
formerly rated the Sixth District as a safe Republican seat. As a consequence
of the 2011 redistricting, the Sixth District is now rated as a “likely” Democratic
seat. (ECF No. 177–52 at 8.)
9. In the 2012 congressional election (the first held in the new Sixth District),
Democrat John Delaney defeated incumbent Republican congressman Roscoe
Bartlett by a 20.9% margin. (ECF No. 177–5 ¶ 54.) However, in the U.S.
Senate election conducted that same cycle, Democrat Ben Cardin carried the
Sixth District by just 50% of the vote, despite winning 56% of the vote
statewide. (ECF No. 186–19 at 10; ECF No. 186–42 PDF at 2.)
10. Congressman Delaney won reelection in 2014 and 2016 by margins of 1.5%
and 14.4%, respectively. (ECF No. 177–5 ¶¶ 55–56.)
11. While Plaintiffs have produced expert reports predicting, based on party
affiliation and other demographic data, that Democratic candidates will likely
fare better under the 2011 plan than under the former plan, Plaintiffs have
conducted no statistical sampling and have adduced no individual voter data
showing how displaced and current residents of the Sixth District actually voted
in 2012, 2014, and 2016.
12. Plaintiffs have not surveyed voters to determine (1) whether former supporters
of Congressman Bartlett who remained in the Sixth District after the 2011
redistricting voted for Congressman Delaney instead, (2) whether such voters
switched party affiliation or simply selected a different candidate on an ad hoc
basis, and (3) the reasons underlying these voters’ decisions. Nor have
Plaintiffs amassed data concerning the voting behavior and preferences of
former Sixth District residents who now reside in other congressional districts.
13. Congressman Bartlett underperformed the other seven members of Maryland’s
congressional delegation in fundraising leading up to his defeat in the 2012
election. (ECF No. 104–13 at 2/2.)
14. In 2014, Republican challenger Dan Bongino nearly unseated Congressman
Delaney even though Bongino resided outside the Sixth District (ECF No. 186–
20 at 18:15–20) and operated at a financial disadvantage vis-à-vis Delaney (id.
at 36:21–37:10). Also in 2014, Republican gubernatorial candidate Larry
16

Hogan won 56% of the vote in the Sixth District, besting his Democratic rival
by 14 percentage points. (ECF No. 186–19 at 10.)
b. Conclusions of Law
In denying Plaintiffs’ preliminary injunction motion, the Court reaches the following
conclusions of law:
1. Under Winter v. NRDC, a plaintiff seeking preliminary injunctive relief must
demonstrate that plaintiff is likely to prevail on the merits of its claim. 555 U.S.
at 20.
2. In Shapiro II, this Court held that, to state a claim for First Amendment
retaliation via gerrymandering, Plaintiffs must allege not only that the
gerrymander diluted votes of targeted citizens “to such a degree that it resulted
in a tangible and concrete adverse effect” but also that “absent the mapmakers’
intent to burden a particular group of voters by reason of their views, the
concrete adverse impact would not have occurred.” 203 F. Supp. 3d at 597.
3. In other words, the First Amendment framework that the Shapiro II majority
endorsed requires proof that but for the gerrymander, the challenged effect
(here, the switch in political power in the Sixth District) would not have
happened.
4. The dissent complains that “the majority’s new First Amendment standard
depends on an election’s results, not on the adverse impact of dilution on the
targeted voters.” Post, at 59. In the dissent’s view, “the adverse effect is the
dilution of votes—and the corresponding burdening of expression by voters—
regardless of how the election turned out.” Post, at 59. However, the Shapiro
II majority recognized that “vote dilution is a matter of degree, and a de minimis
amount of vote dilution, even if intentionally imposed, may not result in a
sufficiently adverse effect on the exercise of First Amendment rights to
constitute a cognizable injury.” 203 F. Supp. 3d at 596–97. The dissent offers
no yardstick to measure vote dilution that exceeds a “de minimis amount” yet
falls short of altering electoral outcomes. Nor have Plaintiffs shown that they
suffered any tangible First Amendment burden other than, perhaps, their
inability to elect their preferred candidate. A political gerrymander that
imposes nothing more than an abstract “burden” without actually affecting
tangible voter rights or interests surely is not justiciable, even pursuant to the
framework two judges endorsed in Shapiro II.
5. The dissent frets that “under the majority’s new standard, no redistricting map
could be challenged before an election.” Post, at 60. To whatever extent this
critique is accurate, it is a consequence of adjudicating political gerrymandering
claims according to the standard adopted in Shapiro II. There may be some
other, as-yet unidentified standard that would enable courts to enjoin
17

implementation of a map prior to the first election conducted thereunder, but
neither Plaintiffs nor the dissent have proffered any such workable standard
here. Strictly prospective relief is relatively uncommon in the law, and courts
are far more likely to be tasked with curing or vindicating a prior harm than
with anticipating and forestalling a potential one.
6. Citing a handful of First Amendment cases that do not deal with election law,
the dissent proposes to import into the political gerrymandering context the
burden-shifting framework of Mt. Healthy City School District Board of
Education v. Doyle, 429 U.S. 274 (1977). Post, at 56–57. The Court declines
to do so, at least at this preliminary stage. As the dissent explains, Mt. Healthy
stands for the proposition that “where the government takes an injurious action,
an injured party need not show that the government would never have taken the
same action anyway.” Post, at 57. Mt. Healthy assumes an injury has occurred
and focuses on questions of motive and intent. The problem is that in the
redistricting context, the government’s “action” is only “injurious” if it actually
alters the outcome of an election (or otherwise works some tangible, measurable
harm on the electorate). In other words, the question of but-for causation is
closely linked to the very existence of an injury: if an election result is not
engineered through a gerrymander but is instead the result of neutral forces and
voter choice, then no injury has occurred.
7. For this reason, the dissent’s poisoning hypothetical, post, at 60–61, is beside
the point. If a victim sips poison, or trains collide, or an employee is fired, or a
homeowner’s request for a zoning variance is denied, there is no question that
an injury of one sort or another has occurred. The question for courts to resolve
in such cases is whether that injury was caused by some illicit action (or
inaction) of the defendant and whether the defendant has an adequate defense
to the charge. But if Roscoe Bartlett loses to John Delaney, voters are thereby
injured if but only if that loss is attributable to gerrymandering or some other
constitutionally suspect activity. If the loss is instead a consequence of voter
choice, that is not an injury. It is democracy.
8. But-for causation—not some metaphysical, could-be burden—is the standard
that controls in this case, and Plaintiffs bear the burden to prove this element is
satisfied. Assuming that Maryland’s former congressional map provides an
acceptable benchmark for assessing the 2011 map, this but-for causation
requirement would be satisfied only if Roscoe Bartlett would have won
reelection in 2012 had the prior map remained intact (with minor adjustments
to account for demographic changes reflected by the 2010 Census). Plaintiffs
admit as much: “[O]ur burden is to show that the purposeful dilution of
Republican votes in the Sixth District was a but-for cause of the routing of

18

Roscoe Bartlett in 2012 and of the Republican losses in 2014 and 2016.” (ECF
No. 191 at 13.)6
9. The fact that John Delaney defeated Roscoe Bartlett by an impressive 20.9%
margin in 2012 may shed some light on the effectiveness of the alleged
gerrymander. However, even a much smaller victory by Delaney would have
shifted the Sixth District seat from Republican to Democratic control. The
dispositive question is whether the shift would have occurred absent the alleged
gerrymander—that is, whether Delaney would have prevailed (even if by a
much smaller margin) absent the State’s reliance on NCEC’s DPI and
demographic data.
10. Upon the record, the briefs, and the hearing, the Court cannot now conclude
that the likely outcome of this litigation is a finding that, but for the alleged
gerrymander, the Republican Party would have retained control of the Sixth
District congressional seat. Plaintiffs have not produced voter sampling or
statistical data, affidavits, or other evidence of a sufficient quantity to
demonstrate how and why voters who would have been included in a neutrally
drafted Sixth District voted in the 2012, 2014, and 2016 elections. Without
such data, the Court cannot reverse-engineer those elections and is unprepared
to assume, at this preliminary stage, that enough such voters would have voted
for the Republican candidate so as to preserve Republican control.
11. While Plaintiffs have adduced some persuasive predictive evidence through the
Cook Partisan Voting Index and expert reports and testimony, the Court is
unconvinced, certainly by the standard governing the issuance of a preliminary
injunction, that such evidence is determinative of but-for causation. In
particular, the Court is not convinced that such predictive evidence accurately
accounts for subjective factors such as evolving political temperament and the
personal strengths or weaknesses of individual candidates. The surprising
results of various elections in 2016 illustrate the limitations of even the most
sophisticated predictive measures. Experience teaches that voter preferences
are mutable and that American democracy is characterized by a degree of
volatility and unpredictability. See Bandemer, 478 U.S. at 160 (O’Connor, J.,
concurring in the judgment) (“To allow district courts to strike down
apportionment plans on the basis of their prognostications as to the outcome of

But see Shapiro II, 203 F. Supp. 3d at 606 (Bredar, J., dissenting) (“Because of the inherent mutability of political
affiliation, the Court cannot simply compare the results of an election conducted pursuant to Map X with those of a
subsequent election conducted pursuant to Map Y and blame any shift in power on redistricting: each election cycle
is unique, and voter behavior is as unpredictable as the broader societal circumstances that may make one candidate,
or one party, more appealing than the other to particular voters and communities. For that matter, treating a prior map
as a baseline for measuring the constitutionality of a subsequent map assumes that the prior map was itself free of
impermissible manipulation—yet we know, as a practical matter, that gerrymandering is widespread in our political
system and as old as the Republic.”); LULAC, 548 U.S. at 446 (Kennedy, J.) (“There is no reason . . . why the old
district has any special claim to fairness.”).
6

19

future elections or future apportionments invites ‘findings’ on matters as to
which neither judges nor anyone else can have any confidence.”).
12. The Court is especially reluctant at this preliminary stage, absent more concrete
voter data, to find an effective gerrymander given that Congressman Delaney
nearly lost control of his seat in 2014 in a race against a candidate burdened
with undisputed geographic and financial limitations.
13. Indeed, this recent near defeat raises serious doubts about whether Plaintiffs’
alleged injury is likely to recur. The most relevant question in a case involving
a claim for solely injunctive relief is not whether a harm may have occurred in
the past but whether the harm is presently occurring or very likely to recur. If
the injury, if any, has long since concluded, there is nothing to enjoin. See
Bloodgood v. Garraghty, 783 F.2d 470, 475 (4th Cir. 1986) (“An injunction is
a drastic remedy and will not issue unless there is an imminent threat of illegal
action. ‘[An i]njunction issues to prevent existing or presently threatened
injuries. One will not be granted against something merely feared as liable to
occur at some indefinite time in the future.’” (quoting Connecticut v.
Massachusetts, 282 U.S. 660, 674 (1931))); cf. Beck v. McDonald, 848 F.3d
262, 277 (4th Cir. 2017) (“‘[A]bsent a sufficient likelihood that [Plaintiffs] will
again be wronged in a similar way’ . . . past events, disconcerting as they may
be, are not sufficient to confer standing to seek injunctive relief.” (alteration in
original) (citations omitted)); Bryant v. Cheney, 924 F.2d 525, 529 (4th Cir.
1991) (“The courts should be especially mindful of th[e] limited role
[prescribed by Article III] when they are asked to award prospective equitable
relief . . . for a concrete past harm, and a plaintiff’s past injury does not
necessarily confer standing upon him to enjoin the possibility of future
injuries.”).
14. Despite the Court’s present doubt as to Plaintiffs’ proof on the causation prong
of their First Amendment claim, the Court does not hold that Plaintiffs cannot
prevail on their claim. Any such holding would be every bit as premature as
the extraordinary relief that Plaintiffs have requested and that the dissent urges.
The Court simply concludes that Plaintiffs have not carried their burden to show
they are likely to prevail on the merits, and so preliminary injunctive relief is
not proper.
15. The Court remains open to the possibility that the evidence Plaintiffs have
adduced, when subject to robust cross-examination and the development that
only a trial can bring, may satisfy Plaintiffs’ burden of proof. The Court also is
willing to entertain requests by either party to reopen discovery (subject to the
stay discussed immediately below) to address the evidentiary gaps and deficits
or potential deficits flagged in this Memorandum. Regardless whether either

20

party seeks additional discovery, the parties may find it helpful to take account
of the Court’s discussion here in any future briefs or oral presentations.
2. Stay of Proceedings
The Court’s concerns about Plaintiffs’ proof with respect to the causation element of their
First Amendment claim compel the Court not only to deny preliminary injunctive relief but also
to stay proceedings pending the Supreme Court’s further guidance in Whitford.
While Plaintiffs argue vociferously that “[t]his case and the Wisconsin case are
fundamentally different” (ECF No. 193 at 4), this Court disagrees. Fundamentally, these cases are
two sides of the same coin: both propose a standard by which federal courts might adjudicate
claims of unlawful political gerrymandering. Both cases invoke the First Amendment as a source
of constitutional authority. And the standard that the Western District of Wisconsin has endorsed
is remarkably similar to the standard endorsed by the majority in Shapiro II: “We conclude,” the
Wisconsin court wrote, “that the First Amendment and the Equal Protection clause prohibit a
redistricting scheme which (1) is intended to place a severe impediment on the effectiveness of the
votes of individual citizens on the basis of their political affiliation, (2) has that effect, and (3)
cannot be justified on other, legitimate legislative grounds.” Whitford, 218 F. Supp. 3d at 884.
True, the cases differ in their particulars. The Wisconsin case is a statewide challenge to
state legislative districts, based in part on partisan asymmetry (the so-called “efficiency gap”); the
Maryland case is a single-district challenge to a congressional district, grounded in a retaliation
theory. For plaintiffs in either case to prevail, however, they would have to show that the
gerrymander about which they complain actually inflicted a constitutional injury on them, one that
is sufficiently personal so as to satisfy the threshold requirements of Article III and sufficiently
definite and clear so as to justify the drastic remedy of an injunction against enforcement of an
otherwise lawfully enacted map. In determining whether a constitutional injury has occurred, the
21

court invariably must reach the question of causation, for if election outcomes (whether in a single
district or across the state) arise not from political machinations at the statehouse but instead from
neutral forces or the “natural ebb and flow of politics,” Shapiro II, 203 F. Supp. 3d at 606 (Bredar,
J., dissenting), no injury has occurred and no remedy may issue. While the Supreme Court’s
decision in Whitford may not prove dispositive of Benisek, the Court’s analysis undoubtedly will
shed light on critical questions in this case, and the parties and the panel will be best served by
awaiting that guidance.
D. Additional Practical Considerations Supporting the Decision to Stay Proceedings
Two practical considerations bolster the Court’s conclusion that a stay is appropriate at this
time.
First: this Court is in no position to award Plaintiffs the remedy they have requested on
the timetable they have demanded. For the reasons explained in Part II.C, two members of this
panel are unconvinced that Plaintiffs will likely prevail on the causation element of their First
Amendment claim. Plaintiffs therefore are not entitled to preliminary injunctive relief. This case
will likely require a full trial on the merits, where witnesses for both parties will be subject to
cross-examination and where the Court will be equipped to make detailed findings and credibility
determinations. But a trial—particularly one requiring the coordination of three judges and their
respective chambers staff—is a substantial undertaking.
Plaintiffs have indicated that a revised districting plan must be enacted no later than
December 19, 2017, to allow orderly implementation in advance of the 2018 midterms. (ECF No.
177–1 at 31.) Plaintiffs also have suggested that an injunction should issue no later than August
18, 2017, to accommodate legislative mapmaking or, if necessary, a judicially imposed map. (Id.
at 32.) Despite the Court’s diligence in ruling on the pending preliminary injunction motion (which

22

has been a priority for each member of this panel), that August date has already come and gone.
Since the Court cannot deliver the remedy Plaintiffs have requested, Plaintiffs’ opposition to a stay
pending Whitford loses considerable force. It is unclear what hardship Plaintiffs will suffer by
waiting a few months if, as a practical matter, the Court would have been unable to cure any
constitutional ill in advance of the 2018 midterms even had it scheduled a trial at the earliest
opportunity.7
Second: while the Supreme Court no doubt benefits from the efforts of lower courts in
resolving difficult legal issues, it is not clear how additional proceedings in this case would aid the
Court’s resolution of Whitford. The threshold justiciability question that the Court must again
confront in Whitford is hardly a novel one, and this panel has rigorously analyzed that threshold
question in the separate opinions in Shapiro II. The Whitford litigants and the Justices will have
access to those opinions during the forthcoming proceedings. Further, as the divergent opinions
in Vieth illustrate, the Justices are not bound to decide Whitford along the lines that the Western
District of Wisconsin found persuasive. If the First Amendment theory that Plaintiffs here have
proposed and that two members of this panel have recognized as justiciable strikes one or more of
the Justices as workable, the Justices certainly may adopt, co-opt, modify, or otherwise incorporate
elements of that theory into a framework for decision or a possible framework for future cases.
Here is the bottom line: a stay in these proceedings will not preclude the Supreme Court
from taking advantage of the important legal work that has been done in this case, and the marginal

7

Plaintiffs alternatively propose that the Court should enter a permanent injunction and then stay enforcement of that
injunction so that the parties may expeditiously take their appeal. (ECF No. 193 at 3.) The Court declines to do so.
The Court will not abandon its duty to conscientiously resolve this years-long dispute so that the parties may squeeze
their case onto the Supreme Court’s fall calendar. Nor will the Court make the findings that would support a permanent
injunction—including that Plaintiffs have suffered an irreparable injury and that, “considering the balance of hardships
between the plaintiff and defendant, a remedy in equity is warranted,” eBay Inc. v. MercExchange, L.L.C., 547 U.S.
388, 391 (2006)—only to then stay that equitable remedy. Rather, the Court will enjoin the State to implement a new
map if but only if it becomes persuaded that Plaintiffs have proved each element of their First Amendment claim to
the requisite degree of certainty.

23

gains—if any—that further fact-finding might offer the Justices would be greatly outweighed by
the efficiency costs of charging ahead only to later learn that Plaintiffs must return to square one
(or, perhaps, that their action is no longer viable).
III. Conclusion
Though the members of this panel differ in their views concerning the implications of
Supreme Court precedent, the evidence Plaintiffs have thus far adduced, and the efficient
management of this complicated and important case, all agree that political gerrymandering is a
noxious and destructive practice. The segregation of voters by political affiliation so as to achieve
purely partisan ends is repugnant to representative democracy. See Ariz. State Legislature v. Ariz.
Indep. Redistricting Comm’n, 135 S. Ct. 2652, 2658 (2015). This Court will not shrink from its
responsibility to adjudicate any viable claim that such segregation has occurred in Maryland. But
in order to correctly adjudicate such a claim, the Court must first insure that it is proceeding on
the correct legal foundation—that in measuring the legality and constitutionality of any
redistricting plan in Maryland it is measuring that plan according to the proper legal standard.
Following the Supreme Court’s decision in Whitford, this panel will be better equipped to make
that legal determination and to chart a wise course for further proceedings.
For the foregoing reasons, an Order shall enter DENYING Plaintiffs’ preliminary
injunction motion, and a separate Order shall enter HOLDING IN ABEYANCE the pending crossmotions for summary judgment and STAYING further proceedings pending the Supreme Court’s
decision in Whitford.
Judge Russell joins all but Part II.B of this Memorandum and joins the accompanying
Orders. Judge Niemeyer joins neither the Memorandum nor the Orders.

24

NIEMEYER, Circuit Judge, dissenting:
In denying the plaintiffs’ motion for a preliminary injunction, the majority overlooks the
obvious and relies on abstract notions of the causal relationship between intent and effect that bear
no relationship to the real world evidence regarding the conduct at issue or to the First Amendment
standard adopted in this case. Its entire reason for denying the injunction rests on a bizarre notion
of causation that requires the exclusion of all possible alternative explanations, however remote
and speculative. When that effort inevitably fails, it concludes that causation has not been
established, despite extraordinarily strong evidence of the connection between intent and effect. I
believe that the record could not be clearer that the mapmakers specifically intended to dilute the
effectiveness of Republican voters in the Sixth Congressional District and that the actual dilution
that they accomplished was caused by their intent. Accordingly, the motion should be granted.
The record demonstrates, without any serious contrary evidence, that the Maryland
Democrats who were responsible for redrawing congressional districts in 2011 specifically
intended to dilute the votes of Republicans in the Sixth District and in fact did so. They identified
likely Republican voters and moved them in large numbers into the Eighth District, which had a
safe margin of Democratic voters. They simultaneously replaced these Republican voters with
Democratic voters from the Eighth District. More specifically, they moved 360,000 persons
(roughly one-half of the District’s population) out of the former Sixth District — when only 10,000
had to be moved in response to the 2010 census — and simultaneously moved 350,000 into the
“new” Sixth District. And critically, in making those moves, they focused on voting histories and
party registration to move 66,400 registered Republicans out of the Sixth District and replace them
with 24,400 registered Democrats, creating a Democratic voter majority in the new Sixth District
of 192,820 Democrats to 145,620 Republicans. Prior to the massive shuffle, the Sixth District had

25

208,024 Republicans and 159,715 Democrats. This 2011 shuffle accomplished the single largest
redistricting swing of one party to another of any congressional district in the Nation.
Consistent with this evidence, the State’s Democratic leadership stated that their
reshuffling of voters by voting history was specifically intended to flip the Sixth District from
Republican to Democratic so as to create a 7 to 1 Democratic congressional delegation. For
example, Maryland Governor Martin O’Malley, who led the effort to develop a new congressional
map after the 2010 census, stated that he wanted to redraw the lines of the Sixth District to “put
more Democrats and Independents into the Sixth District” and ensure “the election of another
Democrat.” He added, “Yes, this was clearly my intent.” And other Democrats involved in the
process similarly revealed their intent with statements indicating, for example, that the Sixth
District was redrawn to “minimize the voice of the Republicans” and to “hit[]” Republican
Congressman Roscoe Bartlett from the Sixth District “pretty hard.” Moreover, the firm hired to
draw the map was given only two instructions — to come up with a map (1) that protected the six
incumbent Democrats and (2) that would produce a 7 to 1 congressional delegation.
Republican voters affected by the redrawing of the Sixth District commenced this action,
contending that they were targeted, based on the way they voted in the past, with the intent to
dilute their vote and diminish their representational rights, in violation of the First Amendment.
On the State’s motion to dismiss the complaint under Federal Rule of Civil Procedure 12(b)(6),
we held that the plaintiffs stated a cause of action and would succeed in their challenge of the Sixth
District’s gerrymander if they were to demonstrate (1) that the mapdrawers redrew the district lines
with the specific intent to impose a burden on voters because of how they voted in the past or
because of the political party with which they were affiliated; (2) that the targeted voters suffered
a tangible, concrete burden on their representational rights; and (3) that the mapdrawers’ intent to

26

burden a particular group of voters by reason of their views was the but-for cause of the concrete
effect. Simply, the standard requires a showing of (1) specific intent, (2) concrete effect, and
(3) causation between the first two requirements. See Shapiro v. McManus, 203 F. Supp. 3d 579,
596–97 (D. Md. 2016).
Following the completion of extensive discovery, the plaintiffs filed a motion for a
preliminary injunction with a request to advance the trial on the merits under Rule 65(a)(2) so as
to obtain a final injunction ordering a redrawing of the lines defining the Sixth District without the
use of data that reveal how voters registered or voted in the past.
*

*

*

The widespread nature of gerrymandering in modern politics is matched by the almost
universal absence of those who will defend its negative effect on our democracy. Indeed, both
Democrats and Republicans have decried it when wielded by their opponents but nonetheless
continue to gerrymander in their own self interest when given the opportunity. The problem is
cancerous, undermining the fundamental tenets of our form of democracy. Indeed, as Judge Bredar
has observed in this case, gerrymandering is a “noxious” practice with “no place in a representative
democracy.” Shapiro, 203 F. Supp. 3d at 600 (Bredar, J., dissenting).
The Supreme Court has joined the chorus of voices recognizing the potential ills inflicted
on our democracy by gerrymandering.

Accepting the general proposition that partisan

gerrymandering, when sufficiently extreme, violates the Constitution, the Justices have
nonetheless yet to agree on a standard for determining when the practice crosses the line. See Vieth
v. Jubelirer, 541 U.S. 267, 292 (2004) (plurality opinion); id. at 308 (Kennedy, J., concurring in
the judgment). For this reason, a minority of the Justices have indicated that the issue of whether
partisan gerrymandering violates the Equal Protection Clause is not justiciable. See id. at 305.

27

But a categorical rule that would abandon efforts at judicial review surely cannot be
accepted lest it lead to unacceptable results. For instance, in Maryland, which has a voting
population that historically votes roughly 60% for Democrats and 40% for Republicans, the
Democrats, as the controlling party, could theoretically create eight safe Democratic congressional
districts by assigning to each district six Democrats for every four Republicans, regardless of their
geographical location. Citizens residing in Baltimore City, others residing in Garrett County in
the western portion of the state, and yet others residing in the suburbs of Washington, D.C., could
all be assigned to a single district so that the Democrats would outnumber Republicans by a margin
of 60% to 40%. Under such a map, no district would have a single boundary, nor indeed any
relationship to geography or to the communities that constitute the State, and neighbors would
have different Representatives. Such a pointillistic map would, of course, be an absurd warping
of the concept of representation, resulting in the very “tyranny of the majority” feared by the
Founders. Yet, such an extreme possibility would be open to the most politically ambitious were
courts categorically to abandon all judicial review of political gerrymandering.
I believe that the First Amendment standard previously adopted by us in this case does not
allow for such a possibility. Building on the Supreme Court’s previous holdings that ensure “one
person, one vote” and that prevent racially motivated gerrymanders, we held earlier in this case
that when district mapdrawers target voters based on their prior, constitutionally protected
expression in voting and dilute their votes, the conduct violates the First Amendment, effectively
punishing voters for the content of their voting practices. See Shapiro, 203 F. Supp. 3d at 595–96.
This First Amendment test focuses on the motive for manipulating district lines, and the effect the
manipulation has on voters, not on the result of the vote. It is therefore sufficient in proving a
violation under this standard to show that a voter was targeted because of the way he voted in the

28

past and that the action put the voter at a concrete disadvantage. The harm is not found in any
particular election statistic, nor even in the outcome of an election, but instead on the intentional
and targeted burdening of the effective exercise of a First Amendment representational right.
Recent comments of Supreme Court Justices made both in this case and in Vieth have suggested
that this standard is available for assessing the constitutionality of a gerrymander. And under this
standard, I respectfully conclude, the plaintiffs have succeeded in carrying their burden.
The majority instead expresses doubts as to whether the earthquake upheaval in the
political landscape of the Sixth District was attributable to the fulfillment of the Democrats’
gerrymandering plan, positing that the flip of the Sixth District might have been attributable to
changes in voting preferences or other demographics. But this view reflects nothing more than an
effort to skirt around the obvious — that the Democrats set out to flip the Sixth District; that they
made massive shifts in voter population based on registration and voting records to accomplish
their goal; and that they succeeded.
The plaintiffs have not only made the requisite showing that they are likely to succeed on
the merits, they have actually succeeded well in demonstrating that the State’s gerrymandering
violated their First Amendment rights. I would accordingly issue the injunction requested and
require the redrawing of the Sixth District’s boundaries without the use of information about how
citizens voted in the past.

I
A. Facts of Record
The historical facts of record are not disputed. Following the 2010 census, the State of
Maryland was required to redraw the lines of its eight congressional districts to ensure that each

29

district had an equal share of the State’s population. This action focuses on the boundaries that
the State chose to draw for the Sixth District.
Historically, the Sixth District included western Maryland and much of north-central
Maryland, and after the Supreme Court’s announcement of the “one person, one vote” rule in
Wesberry v. Sanders, 376 U.S. 1 (1964), the Sixth District had always included all of the State’s
five most northwestern counties — Garrett, Allegany, Washington, Frederick, and Carroll
Counties. After the 2002 redistricting, the District also included a small northern portion of
Montgomery County and larger portions of Baltimore and Harford Counties, as shown.

At the time of the 2010 congressional election — the last held prior to the 2011 redistricting
— 47% of the District’s 446,000 eligible voters were registered Republicans, 36% were registered
Democrats, and 16% were registered Unaffiliated, making the District the most Republican in the
State. Joint Stipulations ¶ 10 & Ex. 2 at 2 (ECF No. 104). Representative Roscoe Bartlett, a
30

Republican, had continuously represented the District since 1993, and he won reelection in 2010
by a margin of 28%. Id. ¶ 8.
The 2010 census showed that the Sixth District had grown somewhat, having 10,186
residents more than the ideal adjusted population of 721,529 for a Maryland congressional district,
a variation of only 1.4%. Joint Stipulations ¶¶ 9, 52. Nonetheless, the Democratic mapdrawers
responsible for the 2011 redistricting plan redrew the District’s boundaries far more dramatically
than was necessary to move 10,186 voters from the District. Indeed, the new Sixth District retained
only 51% of its original population, retaining the residents of Garrett, Allegany, Washington
Counties, and a portion of the residents of Frederick County and moved the other half — roughly
360,000 residents — to other districts. Approximately 60% of these residents — those from
Frederick County and more than half the population of Carroll County — were shifted into the
Eighth District, which had previously been confined almost entirely to the heavily Democratic
Montgomery County. In the place of the removed residents, the plan added to the new Sixth
District approximately 350,000 residents from Montgomery County, most of whom had previously
been assigned to the Eighth District. The final 2011 map for the Sixth District was as follows:

31

The area removed from the former Sixth District was predominately Republican, while the
area added was predominately Democratic. Specifically, in the precincts removed from the Sixth
District, there were on average approximately 1.5 times as many registered Republicans as
Democrats.

By contrast, in the precincts added to Sixth District, registered Democrats

outnumbered Republicans by more than 2 to 1. In total, the reshuffling of the Sixth District’s
boundaries resulted in a net reduction of more than 66,000 registered Republicans and a net
increase of some 24,000 registered Democrats, for a swing of about 90,000 voters. See Opening
Expert Report of Dr. Peter A. Morrison ¶ 134 & tbl. 1 (ECF No. 177-35); Opening Expert Report
of Prof. Michael P. McDonald at 12 (ECF No. 177-19).
Not surprisingly, this major reshuffling of the Sixth District’s population directly affected
the District’s political complexion. At the time of the 2012 congressional election (the first held
under the new map), the major parties’ respective shares of the District’s registered voters roughly
reversed compared to just two years before. Of the new District’s roughly 437,000 eligible voters,
32

33% were registered Republicans, 44% were registered Democrats, and 22% were registered as
Unaffiliated. Joint Stipulations ¶ 53 & Ex. 19. In the 2012 election, Democratic candidate John
Delaney, a newcomer to politics, defeated Republican incumbent Bartlett by a 21% margin, and
he was elected again in 2014 and 2016. Id. ¶ 54.
The parties have stipulated that “[o]ne widely understood consequence of the Plan was that
it would make it more likely that a Democrat rather than a Republican would be elected as
representative from the [Sixth] District.” Joint Stipulations ¶ 31. But the record demonstrates
even more. Far from being an incidental, though anticipated, byproduct of achieving some other
set of redistricting goals, the Maryland Democrats who controlled the 2011 redistricting process
sought to assure themselves of a 7 to 1 Democratic delegation by flipping the Sixth District to
Democratic control.
Governor O’Malley, who was both “the leader of [Maryland’s] Democratic Party,”
O’Malley Dep. 46:20–21 (ECF No. 177-3), and “directly in charge of running the congressional
redistricting process,” id. at 30:19–20, agreed that he “set out to draw the borders in a way that
was favorable to the Democratic Party,” id. at 9:22–10:2. As he later testified:
[T]hose of us in leadership positions in our party, the Speaker, the Senate President,
the Democratic Dean of the Delegation, myself, Lieutenant Governor, we all
understood that, while our — while we must fulfill our responsibility on
redistricting, must be mindful of constitutional guidelines, restrictions, case law,
statutes, it was also — part of our intent was to create a map that was more favorable
for Democrats over the next ten years and not less favorable to them. Yes, that was
clearly one of our many [goals].
Id. at 81:1–11. Specifically, O’Malley wanted to use the redistricting process to change the overall
composition of the U.S. House Delegation to seven Democrats and one Republican by flipping
either the First District, on the eastern shore of Maryland, or the Sixth District, in Western
Maryland. Id. at 22–27. Because altering the political makeup of the First District, the only other
Maryland district represented by a Republican, would have required awkwardly “jump[ing] the
33

Chesapeake Bay and draw[ing] a line in such a way that [would] put[] . . . more Democratic voters
[in] the Eastern Shore [district],” id. at 24:16–19, he stated that “a decision was made to go for the
Sixth,” id. at 27:3–4.
Following the customary process in Maryland, Governor O’Malley pursued two courses
for developing a revised congressional map. For one, he created the public-facing “Governor’s
Redistricting Advisory Committee,” and for the other, he “asked Congressman [Steny] Hoyer, . .
. the dean of the [U.S.] House delegation,” to “lead the effort . . . to inform the [Committee] about
congressional redistricting” and “come up with a map that a majority of the congressional
delegation supports.” O’Malley Dep. 47:20–48:5; see also Willis Dep. 185–88 (ECF No. 177-14)
(agreeing that, historically, “[t]he process starts with the [Democratic] members of Congress,” who
“[e]ndeavor to come to a consensus,” “and then it flows to the governor and legislators,” who “do
their best to respect the wishes . . . of the congressional delegation”). Consistent with this
customary procedure, the record shows that the work performed on behalf of the Democratic
members of Maryland’s congressional delegation largely shaped the contours of the new Sixth
District that the Advisory Committee ultimately recommended to Governor O’Malley. See Miller
Dep. 97:19 (ECF No. 177-15) (testifying that the map “primarily was drawn by the congressional
people”).
The Advisory Committee held public hearings across the State from July through
September 2011 and received comments from members of the public. Joint Stipulations ¶ 22. At
hearings conducted in western Maryland, residents provided suggestions regarding potential
changes to the shape of the Sixth District. Several of these residents testified about various
connections between Frederick County and Montgomery County — including Interstate 270 (“I270”), a 35-mile highway running between the City of Frederick and southern Montgomery

34

County — and advocated for replacing part of the Sixth District with territory from Montgomery
County. None of the speakers contemplated a map that would remove much of Frederick County
itself, which had been included in its entirety in the Sixth District since 1872. See, e.g., Public
Hearing Testimony (ECF No. 186-3) at MCM 000029–31 (“[T]he start of the Sixth District is
pretty easy, with Garrett, Allegany, Washington, and Frederick, you’ve got a nucleus there . . . .
Once you start with those four counties, . . . your orientation should be to go east into either
Howard, or go southeast into Montgomery Counties, to the greatest extent possible, and . . . leave
Harford, Baltimore, and even portions of Carroll for a Baltimore-oriented district”).
While the Advisory Committee was holding public hearings across the State, the
Democratic members of Maryland’s U.S. House Delegation — led by Representative Hoyer, a
self-described “serial gerrymanderer,” ECF No. 191-3 — had already begun to redraw the State’s
congressional map. Indeed, around the time that the results of the 2010 census became available
in late February/early March 2011 — months before the Advisory Committee was even created —
Hoyer and the other Maryland Democrats in the House retained NCEC Services, Inc., a political
consulting firm that provides “electoral analysis, campaign strategy, political targeting, and GIS
[geographic information system] services” to Democratic organizations. ECF No. 177-17; see
also Hawkins Dep. 28–31 (ECF No. 177-4); ECF No. 177-18. NCEC was specifically charged
with drawing a map that maximized “incumbent protection” for Democrats and changed the
congressional delegation from 6 Democrats and 2 Republicans to 7 Democrats and 1 Republican,
and it was given no other instruction as how to draw the map. Hawkins Dep. 40–42, 47–49.
The primary NCEC analyst assigned to the task, Eric Hawkins, analyzed various
congressional redistricting plans to inform the Democratic members of the Maryland delegation
how “different options would change their districts,” and he personally prepared between 10 and

35

20 different draft congressional maps using a GIS computer software program called Maptitude
for Redistricting. Hawkins Dep. 36–38. Maptitude allows users to “[c]reate districts using any
level of geography,” “[a]dd political data and election results,” and “[u]pdate historic results to
new political boundaries.” Joint Stipulations ¶ 28. With Maptitude, “data reflecting . . . citizens’
political party affiliation and voting histories[] can be used to determine how the outcome of
historical elections would have changed . . . if the proposed plan had been in place in prior years,”
id. ¶ 30, thus enabling users to accurately predict the likely outcome of future elections.
Hawkins specifically used a proprietary metric created by NCEC called the Democratic
Performance Index (the “DPI”), which indicates how a generic Democratic candidate would likely
perform in a particular district. As Hawkins explained, the DPI “is an average of how statewide
candidates perform over time in competitive elections” that is “weighted differently for different
election years,” and which “take[s] into account past voting history in a state or a district.”
Hawkins Dep. 24:12–18. NCEC also calculated separate versions of the DPI specific to federal
and state races — with the federal DPI “only us[ing] federal races” and the state DPI “only us[ing]
state races” — to better account for “ticket splitting.” Id. at 25.
Hawkins used the DPI to meet the dual “goals” given to NCEC — namely, to draw a map
that would maximize “incumbent protection” for the Democrats currently representing Maryland
districts in Congress and that would “chang[e] the make-up of Maryland’s U.S. House delegation
from six Democrats and two Republicans to seven Democrats and one Republican.” Hawkins
Dep. 40–42; see also id. at 47–49. With respect to this 7 to 1 goal, Hawkins’ efforts focused on
redrawing the Sixth District’s lines to increase its federal DPI, which Hawkins calculated under
the preexisting map as standing at 37.4%, indicating low Democratic performance and
correspondingly strong Republican performance. Over the course of working with Maryland’s

36

Democratic House Delegation and their staff, Hawkins prepared several different draft maps under
which the Sixth District would have a 51% federal DPI. In preparing these maps, Hawkins
considered neither “any measure of compactness,” id. at 126:12–13, nor whether “there was a
community of interest related to the I-270 corridor,” id. at 128:19–20. Rather, “[t]he intent was to
see if there was a way to get another Democratic district in the state.” Id. at 230:19–20.
Maps were also proposed by third-party entities, but those maps resulted in a far smaller
federal DPI for the Sixth District. For example, a map proposed by the Maryland Legislative Black
Caucus would have resulted in a federal DPI of 39% for the Sixth District, ECF No. 177-34, a
proposal a senior congressional staffer worried would be “a recipe for 5–3, not 7–1,” ECF No.
177-36. Needless to say, these proposals did not influence the maps submitted by Hawkins to the
Democratic House Delegation.
Ultimately, Maryland’s Democratic members of the U.S. House Delegation proposed and
forwarded to the state Democratic leadership at least two maps prepared by Hawkins. The shape
of the Sixth District in one of these maps, which had a DPI of 51.36%, was very similar to the plan
that was ultimately adopted. See Decl. of Dr. Michael McDonald at 4 & fig. 5 (ECF No. 191-5).
After Maryland’s U.S. House Democrats submitted their proposals, further work was done
by a group of senior staffers of O’Malley, Maryland Senate President Thomas Miller, and
Maryland House Speaker Michael Busch. These senior staffers were equipped with a laptop
loaded with the Maptitude software; “party registration data and voter turnout data,” including at
the census block level, the smallest geographic unit used by the U.S. Census Bureau; and a “data
file[] that contained Democratic Performance Index information at the precinct level,” the smallest
geographic unit in Maryland (averaging around 3,000 people) at which election results are
reported. Weissman Decl. ¶¶ 3–5 (ECF No. 186-11). These state Democratic officials thus

37

continued to use the DPI — as well as other information about how local groups of citizens had
previously voted and the political party with which they were affiliated — to finalize a map for
the Advisory Committee.
The Advisory Committee publicly released a proposed congressional redistricting map on
October 4, 2011, with the Committee’s lone Republican casting the sole dissenting vote against
the plan. Joint Stipulations ¶ 32. The Committee’s map had a federal DPI of 53% in the Sixth
District, which was greeted as “good news” by the man who was widely expected to be the
Democratic nominee to represent the newly redrawn Sixth District in the upcoming 2012 election.
ECF No. 177-25.
Members and staff of the Advisory Committee briefed a joint session of the state House
and Senate Democratic Caucuses about their recommended congressional plan on October 3, 2011.
Joint Stipulations ¶ 35. Talking points prepared for Senate President Miller’s introductory remarks
encouraged him to emphasize that “[e]ven though the map isn’t pretty, it accomplishes a few
important goals,” including “creat[ing] an opportunity for Montgomery County to control two
congressional districts”; “preserv[ing] all six incumbent Democrats in ‘safe’ districts,” none of
which would have “less than 58% Democratic performance”; and “giv[ing] Democrats a real
opportunity to pick up a seventh seat in the delegation by targeting Roscoe Bartlett.” ECF No.
177-23. The talking points continued, “In the face of Republican gains in redistricting in other
states around the nation, we have a serious obligation to create this opportunity.” Id.
Following Senate President Miller’s remarks, Chairwoman Jeanne Hitchcock delivered a
PowerPoint presentation that stated that the Sixth and Eighth Districts had been “[c]onfigured to
reflect the North-South connections between Montgomery County, the I-270 Corridor, and western
portions of the State.” Joint Stipulations, Ex. 6. The record suggests that those in attendance were

38

skeptical that the I-270 corridor justified dramatically redrawing both the Sixth and the Eighth
Districts. For example, immediately after Hitchcock’s presentation, Democratic Delegate Curt
Anderson told a reporter, “It reminded me of a weather woman standing in front of the map saying,
‘Here comes a cold front,’ and in this case the cold front is going to be hitting Roscoe Bartlett
pretty hard.” Joint Stipulations ¶ 46 & Ex. 13. And, while listening to Hitchcock give a similar
presentation earlier in the day, one senior congressional aide who had been intimately involved in
the redistricting process wrote to another, “This is painful to watch. . . . I’m not sure I buy the
themes they are selling. Hopefully they have some better ones for the public face of it.” ECF No.
177-58.
On October 15, 2011, Governor O’Malley announced that he was submitting a map to the
General Assembly “that was . . . substantially the same as” the Advisory Committee’s proposal,
Joint Stipulations ¶ 33, and two days later, on October 17, Senate President Miller introduced the
Governor’s proposed redistricting map as Senate Bill 1 at a special legislative session. With only
minor technical amendments, Senate Bill 1 was signed into law on October 20, 2011, three days
after it had been introduced. Id. ¶ 34; see Md. Code Ann., Elec. Law §§ 8-701 to -709.
“No Republican Senator or Delegate voted for Senate Bill 1 in committee or on the floor
in recorded roll call votes.” Joint Stipulations ¶ 36. Moreover, while the legislation was
progressing rapidly through the General Assembly, numerous legislators made comments
reflecting their clear understanding that the massive redrawing of the Sixth District was designed
primarily to give the eventual Democratic nominee a distinct electoral advantage over the
Republican nominee. For example, one Delegate bluntly stated in a floor speech that he supported
the map because it meant “more Democrats in the House of Representatives.” Id. ¶ 44. Another
Delegate stated in an October 17 interview that, “What we’re doing is we are trying to get more,

39

in terms of — currently we have two Republican districts and six Democratic Congressional
districts and we’re going to try to move that down to seven and one, with the additional
Congressional district coming more out of Montgomery county and going into western Maryland
that would give the Democrats more.” Id. ¶ 47. One Democratic Senator who voted for the bill
nonetheless lamented in a floor speech that partisan gerrymandering was a problem across
America, adding that “it’s a process where we dress up partisan and political ambition on both
sides of the aisle in high principal, but we can all tell what’s really going on.” Joint Stipulations
¶ 43(a) (emphasis added). And the only Democratic Senator to vote against the bill stated in an
October 14 interview, “[W]hen you look at the way these districts are drawn, they’re absolutely
drawn with one thing in mind. . . . [I]t’s certainly drawn so that you can minimize the voice of the
Republicans.” ECF No. 177-41 at 16 (emphasis added).
The effect of the Sixth District’s wholesale recomposition was precisely as intended.
Tellingly, in October 2012, the Cook Political Report released an analysis of “all 435 newly
redrawn Congressional districts in the country” using its Partisan Voter Index (“Cook PVI”), ECF
No. 177-52 at 1, a well-respected “measurement of how strongly a United States congressional
district or state leans toward the Democratic or Republican Party, compared to the nation as a
whole,” ECF No. 177-51; see also Lichtman Dep. 131 (ECF No. 177-49) (testimony of State’s
expert witness that the Cook PVI is a “well respected” and “well regarded” metric). The Cook
Report specifically examined “which districts underwent the most dramatic alterations in
redistricting” and found that Maryland’s Sixth District experienced the single largest redistricting
swing of any district anywhere in the Nation. ECF No. 177-52 at 6–8. Specifically, before the
2011 redistricting, the Sixth District had a Cook PVI of “R+13” and a “Solid Republican” label;
after redistricting, the District received a Cook PVI of “D+2” and a “Likely Democratic” label.

40

Id. at 8. An academic analysis that looked at the accuracy of the Cook Report’s forecasting helps
unpack the significance of this swing.

When the Cook Report has rated a district “Solid

Republican” on the eve of a congressional election, the Republican candidate has won the race
99.7% of the time; when a district has been rated as “Likely Democratic,” the Democratic
candidate has won 94% of the time. See James E. Campbell, The Seats in Trouble Forecast of the
2010 Elections to the U.S. House, 43 Pol. Sci. & Politics 627, 628 (2010) (ECF No. 191-8).
Moreover, the Cook Report’s analysis of the effect of redistricting on the Sixth District was
corroborated by NCEC’s own data. According to NCEC, in the 2016 congressional election cycle,
“Democrats [nationwide] won only four districts where DPI was below 50 percent”; in none of
those districts was the DPI below 40%, as it was in the Sixth District prior to redistricting. ECF
No. 191-7. Conversely, among the 160 districts across the country with a DPI above 50%, all but
12 were won by the Democratic candidate. Id. Both Cook’s and NCEC’s data confirmed that the
Democrats held a clear electoral advantage and that Republican voices had indeed been minimized.

B. Proceedings
Three Maryland citizens, acting pro se, commenced this action in November 2013, naming
as defendants the Chair and the Administrator of the State Board of Elections and alleging that the
2011 redistricting plan violated their rights under the First Amendment and Article I, § 2, of the
U.S. Constitution. A single district court judge granted the State’s motion to dismiss, Benisek v.
Mack, 11 F. Supp. 3d. 516 (D. Md. 2014), and the Fourth Circuit summarily affirmed, Benisek,
584 F. App’x 140 (4th Cir. 2014). The Supreme Court reversed, however, concluding that the
plaintiffs’ constitutional challenge was not “wholly insubstantial” and that therefore it had to be
decided by a district court composed of three judges, as required by 28 U.S.C. § 2284. Shapiro v.
McManus, 136 S. Ct. 450, 456 (2015) (quoting Bell v. Hood, 327 U.S. 678, 682 (1946)). In doing
41

so, the Court observed that the theory underlying the plaintiffs’ First Amendment claim had
originally been suggested by Justice Kennedy in Vieth and was “uncontradicted by the majority in
any of [the Court’s] cases.” Id.
After remand, the plaintiffs, now represented by counsel, filed a second amended
complaint, adding six additional plaintiffs and refining the theory underlying their constitutional
challenge. Two of the original plaintiffs later agreed to their dismissal from the action, leaving
seven plaintiffs, all of whom are registered Republicans who lived in the Sixth District prior to the
2011 redistricting. Three of these plaintiffs still reside in the Sixth District, while four of them
now live in the Eighth District as a result of the redistricting.
The plaintiffs’ second amended complaint alleged that those responsible for the 2011
congressional map “purposefully and successfully flipped [the Sixth District] from Republican to
Democratic control by strategically moving the [D]istrict’s lines by reason of citizens’ voting
records and known party affiliations.” Second Am. Compl. ¶ 1. They alleged that “[t]he drafters
of the Plan focused predominantly on the voting histories and political-party affiliations of the
citizens of the State in deciding how to” redraw the Sixth District’s lines and that they “did so with
the clear purpose . . . of diluting the votes of Republican voters.” Id. ¶ 6. They alleged further that
the plan achieved its intended effect, imposing a significant burden on the former Sixth District’s
Republican voters and preventing them in 2012 and 2014 “from continuing to elect a Republican
representative . . . , as they had in the prior ten congressional elections.” Id. ¶ 7(b). And they
maintained that “the State cannot justify the cracking of the [Sixth] District by reference to
geography or compliance with legitimate redistricting criteria.” Id. ¶ 7(c). Based on these
allegations, the plaintiffs claimed in essence that the plan’s redrawing of the Sixth District’s
boundaries constituted unlawful retaliation in violation of their rights under the First Amendment.

42

In an opinion issued August 24, 2016, this three-judge court denied the State’s motion to
dismiss, concluding that the plaintiffs had adequately alleged a justiciable claim for relief.
Shapiro, 203 F. Supp. 3d at 586, 600. We held that to succeed on their claim, the plaintiffs would
have to prove three elements: first, “that those responsible for the map redrew the lines of [their]
district with the specific intent to impose a burden on [them] and similarly situated citizens because
of how they voted or the political party with which they were affiliated”; second, “that the
challenged map diluted the votes of the targeted citizens to such a degree that it resulted in a
tangible and concrete adverse effect”; and third, “that, absent the mapmakers’ intent to burden a
particular group of voters by reason of their views, the concrete adverse impact would not have
occurred.” Id. at 596–97.
Following the completion of extensive discovery, the plaintiffs filed a motion for a
preliminary injunction and requested, pursuant to Rule 65(a)(2), that the trial on the merits be
advanced and consolidated with a hearing on their motion. Briefing on the plaintiffs’ motion for
a preliminary injunction was completed, with the parties presenting a robust evidentiary record of
more than 80 exhibits, and on July 14, 2017, we conducted a half-day hearing on the motion.

II
This court is clearly of one mind that, as a general matter, partisan gerrymandering is
noxious to our form of democracy. And if we read correctly the public sentiment, that view is
widely shared. Indeed, the Supreme Court, with no disagreement from any Justice, has concluded
that severe partisan gerrymandering is incompatible with democratic principles. Yet, Judge
Bredar, writing only for himself, expresses doubts as to whether claims of partisan gerrymandering
are justiciable.

43

To be sure, drawing the lines of congressional districts is a political process. But the
Supreme Court has repeatedly recognized that the political nature of redistricting does not
immunize the process from claims that are based on violations of particular provisions of the
Constitution. Thus, in Baker v. Carr, 369 U.S. 186 (1962), the Court recognized that when
redistricting denies citizens equal protection, the issue is justiciable because “the equal protection
clause is not diminished by the fact that the discrimination relates to political rights,” id. at 210
(quoting Snowden v. Hughes, 321 U.S. 1, 11 (1944)). Similarly, in Wesberry v. Sanders, 376 U.S.
1 (1964), the Court concluded that the political nature of redistricting does not “immunize state
congressional apportionment laws which debase a citizen’s right to vote from the power of courts
to protect the constitutional rights of individuals from legislative destruction, a power recognized
at least since our decision in Marbury v. Madison . . . . The right to vote is too important in our
free society to be stripped of judicial protection,” id. at 6–7. In a similar vein, the Court has found
justiciable an equal protection redistricting claim where “race was the predominant factor
motivating the legislature’s decision to place a significant number of voters within or without a
particular district.” Ala. Legislative Black Caucus v. Alabama, 135 S. Ct. 1257, 1267 (2015)
(emphasis added) (quoting Miller v. Johnson, 515 U.S. 900, 916 (1995)). And in circumstances
more analogous to those presented in this case, the Court in Davis v. Bandemer, 478 U.S. 109,
113, 118–27 (1986), held that a claim alleging an unconstitutional dilution of votes of one political
party’s members was justiciable.
While claims alleging violations of individual constitutional rights are justiciable and have
been so since Marbury v. Madison, the Court has been unable to find a standard by which to
conclude that suspect districts, although equal in population, violate the Equal Protection Clause
based on extreme partisanship. See Vieth, 541 U.S. at 281–301. Even so, five Justices in Vieth

44

concluded that the issue remained justiciable. Moreover, Justice Kennedy, canvassing the Court’s
decisions, appropriately recognized that claims asserting other constitutional rights, such as a
violation of the First Amendment, could be reviewable. As he stated:
First Amendment concerns arise where a State enacts a law that has the purpose
and effect of subjecting a group of voters or their party to disfavored treatment by
reason of their views. In the context of partisan gerrymandering, that means that
First Amendment concerns arise where an apportionment has the purpose and effect
of burdening a group of voters’ representational rights.
Id. at 314 (Kennedy, J., concurring in the judgment). He went on to point out that “[i]f a court
were to find that a State did impose burdens and restrictions on groups or persons by reason of
their views, there would likely be a First Amendment violation.” Id. at 315. Indeed, in this very
case, a unanimous Supreme Court expressly invited our consideration of the First Amendment
theory articulated by Justice Kennedy, noting that the theory remains “uncontradicted by the
majority in any of our cases.” Shapiro, 136 S. Ct. at 456. And we concluded, from a fuller review
of the Supreme Court’s First Amendment jurisprudence, that a First Amendment theory is viable
and justiciable. See Shapiro, 203 F. Supp. 3d at 594–97.
To begin, it is “axiomatic” that the government violates the First Amendment when it
regulates speech “based on its substantive content or the message it conveys.” Rosenberger v.
Rector & Visitors of Univ. of Va., 515 U.S. 819, 828 (1995). As the Court noted, while restrictions
based on content presumptively offend the First Amendment, “[w]hen the government targets not
subject matter, but particular views taken by speakers on a subject, the violation of the First
Amendment is all the more blatant.” Id. at 829 (emphasis added). As a result, “[t]he government
must abstain from regulating speech when the specific motivating ideology or the opinion or
perspective of the speaker is the rationale for the restriction.”

Id.

Moreover, viewpoint

discrimination is no more constitutional when the offending restriction does not explicitly mention

45

any individual viewpoint. Rather, facially neutral restrictions are nonetheless subject to strict
scrutiny when they “were adopted by the government ‘because of disagreement with the message
[the speech] conveys.’” Reed v. Town of Gilbert, 135 S. Ct. 2218, 2227 (2015) (alteration in
original) (quoting Ward v. Rock Against Racism, 491 U.S. 781, 791 (1989)).
Moreover, the government may not suppress one viewpoint even in spheres of activity
where it can lawfully restrict the categories of speech permitted and the time, place, and manner
in which it is conveyed. Rosenberg, 515 U.S. at 829 (“The necessities of confining a forum to the
limited and legitimate purposes for which it was created may justify the State in reserving it for
certain groups or for the discussion of certain topics”); see also Perry Educ. Ass’n v. Perry Local
Educators’ Ass’n, 460 U.S. 37, 45–49 (1983). So, for example, while the government may
lawfully exercise significant control over its employees, it may not fire someone solely because he
belongs to a disfavored political party, as this would amount to blatant “government discrimination
based on the viewpoint of one’s speech or one’s political affiliations.” Bd. of Cty. Comm’rs v.
Umbehr, 518 U.S. 668, 683 (1996); see also Pickering v. Bd. of Educ., 391 U.S. 563, 568–70
(1968) (barring discharge of public-school teacher for writing letter critical of school board).
Indeed, even where the government is allowed, or even required, to consider the viewpoint
of expression that it regulates, this does not give it permission to intentionally advance one
viewpoint over the other. Thus, in Board of Education v. Pico, 457 U.S. 853 (1982), the Court
evaluated a First Amendment challenge to a school board’s removal of certain library books from
school libraries.

The Court recognized that, although the local school board “possess[ed]

significant discretion to determine the content of their school libraries,” its discretion could “not
be exercised in a narrowly partisan or political manner.” Id. at 870. As the Court observed, “If a
Democratic school board, motivated by party affiliation, ordered the removal of all books written

46

by or in favor of Republicans, few would doubt that the order violated the constitutional rights of
the students denied access to those books.” Id. at 870–71.
In cases where some regulation of expression is inevitable, such as in Pico, assessing a
constitutional claim “depends upon the motivation behind [the government’s] actions.” Pico, 457
U.S. at 871 (emphasis added). In assessing the school board’s removal of books in that case, the
Court explained, “If petitioners intended by their removal decision to deny respondents access to
ideas with which petitioners disagreed, and if this intent was the decisive factor in petitioners’
decision, then petitioners have exercised their discretion in violation of the Constitution.” Id. And
the Court defined “decisive factor” to mean “a ‘substantial factor’ in the absence of which the
opposite decision would have been reached.” Id. n. 22 (citing Mt. Healthy City Sch. Dist. Bd. of
Educ. v. Doyle, 429 U.S. 274, 287 (1977)).
Given these stringent limitations on the government’s ability to advance ideological
motives by regulating speech, it would be strange indeed if a State’s administration of elections
were not similarly limited. In fact, the Court has noted specifically that “in exercising their powers
of supervision over elections and in setting qualifications for voters, the States may not infringe
upon basic constitutional protections.” Kusper v. Pontikes, 414 U.S. 51, 57 (1973). Thus, because
an election campaign is “an effective platform for the expression of views on the issues of the
day,” Anderson v. Celebrezze, 460 U.S. 780, 788 (1983), the Court has deemed justiciable
challenges to laws that threaten “the right of qualified voters, regardless of their political
persuasion, to cast their votes effectively,” id. at 787 (quoting Williams v. Rhodes, 393 U.S. 23, 30
(1968)). Similarly, while a State may constitutionally choose locations for polling places, even
though some voters may be more inconvenienced by a location than others, see Lee v. Va. State
Bd. of Elections, 843 F.3d 592, 601 (4th Cir. 2016), the Constitution would obviously not permit

47

the State to locate a polling place specifically to make it more difficult for voters of a particular
party to vote.
Against the backdrop of this First Amendment jurisprudence, the plaintiffs’ First
Amendment claim is readily justiciable. We previously concluded that plaintiffs stated a claim in
alleging that the defendants drew district lines in order to dilute and thus diminish the effectiveness
of their expression. The allegation that district lines were drawn with the intent to suppress the
effectiveness of one political party’s voters is essentially no different from the familiar claims of
adverse employment action due to protected political speech, see, e.g., Mt. Healthy, 429 U.S. at
283–84, or claims that a government has taken an otherwise permissible action with the
impermissible motive of silencing one side of a political debate, see, e.g., Pico, 457 U.S. at 871.
Moreover, judicial abdication from partisan gerrymandering cases, as advocated by Judge
Bredar, would have the most troubling consequences.* If there were no limits on the government’s
ability to draw district lines for political purposes, a state might well abandon geographical districts
altogether so as to minimize the disfavored party’s effectiveness. In Maryland, where roughly
60% of the voters are Democrats and 40% Republicans, the Democrats could create eight safe
congressional districts by assigning to each district six Democrats for every four Republicans,
regardless of the voters’ geographical location. In a similar vein, a Republican government faced

Judge Bredar protests that it “is incorrect” to state that he concludes that partisan gerrymandering claims are not
justiciable. Ante at 12. Yet he expressly relies on what he considers to be the plaintiffs’ failure to establish
justiciability as a basis for denying their motion for a preliminary injunction. Ante at 13–14; see also ante at 2.
The standing law is that partisan gerrymandering claims are justiciable. See Bandemer, 478 U.S. at 113. To
be sure, various Supreme Court Justices continue to debate the question, but they have not held otherwise. Judge
Bredar relies on comments by Justices who were not speaking for the Court to conclude that justiciability has been
cast into doubt. And he further speculates that any rule of justiciability previously recognized may be changed in Gill
v. Whitford, No. 16-1161, now pending before the Court. Lower courts are admonished, however, to follow the
Supreme Court’s existing law until the Court changes it and not to speculate on changes. See Bosse v. Oklahoma, 137
S. Ct. 1, 2 (2016) (per curiam) (“Our decisions remain binding precedent until we see fit to reconsider them, regardless
of whether subsequent cases have raised doubts about their continuing vitality”); Rodriguez de Quijas v. Shearson/Am.
Express, Inc., 490 U.S. 477, 484 (1989) (“If a precedent of this Court has direct application in a case, yet appears to
rest on reasons rejected in some other line of decisions, the Court of Appeals should follow the case which directly
controls, leaving to this Court the prerogative of overruling its own decisions”).
*

48

with these same voters could create a map in which two districts consisted entirely of Democrats,
leaving six that would be 53% Republican. Such a paradigm would be strange by any standard.
A congressman elected in such a system could have constituents in Baltimore City, others in
Garrett County, and yet others in the suburbs of Washington, D.C., preventing him from
representing any of his constituents effectively. Similarly, members of a single household could
be assigned to different congressional districts, and neighbors would be denied the ability to
mobilize politically. Such partisan gerrymandering, at its extreme, would disrupt the “very essence
of districting,” which “is to produce a different . . . result than would be reached with elections at
large, in which the winning party would take 100% of the legislative seats.” Gaffney v. Cummings,
412 U.S. 735, 753 (1973).
Drawing on traditional First Amendment jurisprudence, which includes “well-established
standards for evaluating ordinary First Amendment retaliation claims,” Shapiro, 203 F. Supp. 3d
at 596, we thus previously held that a plaintiff states a claim for unconstitutional gerrymandering
when he demonstrates that (1) “those responsible for the map redrew the lines of his district with
the specific intent to impose a burden on him and similarly situated citizens because of how they
voted or the political party with which they were affiliated,” (2) “the challenged map diluted the
votes of the targeted citizens to such a degree that it resulted in a tangible and concrete adverse
effect,” and (3) “the mapmakers’ intent to burden a particular group of voters by reason of their
views” was a but-for cause of the “adverse impact.” Id. at 596–97. And that is the standard that
we must now apply.

49

III
To grant a preliminary injunction, we must conclude that the plaintiffs are likely to succeed
on the merits; that without the injunction, they are likely to suffer irreparable harm; that the balance
of equities favors them; and that the injunction would be in the public interest. See Winter v. Nat.
Res. Def. Council, Inc., 555 U.S. 7, 20 (2008). In this case, the only issue seriously disputed is
whether the plaintiffs are likely to succeed on the merits. I address the remaining requirements in
Part IV.
Under the standard established in this case for a First Amendment claim, the plaintiffs must
show (1) intent, (2) concrete adverse effect, and (3) causation. Shapiro, 203 F. Supp. 3d at 596–
97. If the plaintiff makes a sufficient showing of these requirements, “the State can still avoid
liability by showing that its redistricting legislation was narrowly tailored to achieve a compelling
government interest.” Id. at 597.
First, with respect to the mapmakers’ intent, the process described in the record admits of
no doubt. Maryland Democratic officials worked to establish the congressional district boundaries
in 2011 with a narrow focus on diluting the vote of Republicans in the Sixth District, so as to
ensure the election of an additional Democratic representative. Governor O’Malley, who was
responsible for the redistricting process, asked Congressman Hoyer to begin the redistricting
effort, and Hoyer retained NCEC to draw up district maps that protected Democratic incumbents
and flipped the Sixth District from Republican to Democrat. Hawkins, an NCEC analyst, prepared
district maps using NCEC’s proprietary DPI metric to assess the likelihood that a district would
elect a Democratic candidate. He homed in on maps using data that predicted a Democratic victory
in the Sixth District, unlike maps submitted by third parties, which had sub-50% DPI values for
the Sixth District. Hawkins submitted several maps, each with a higher DPI in the Sixth District,

50

to the Democratic members of Maryland’s congressional delegation. The delegation, in turn,
culled NCEC’s proposed maps down to a handful where the DPI for the District was approximately
51% and submitted those to the Advisory Committee. The Advisory Committee’s staffers then
used those maps, the DPI information, and their data on party registration and voter turnout to
finalize a map with a 53% DPI for the Sixth District, which the General Assembly thereafter
adopted.
The Advisory Committee’s reliance on the DPI was essential to satisfying the Committee’s
intent to flip the Sixth District from safely Republican to securely Democratic. Notes prepared for
Senate President Miller’s remarks to the state House and Senate Democratic Caucuses about the
redistricting plan emphasized that the map “create[d] an opportunity for Montgomery County to
control two congressional districts”; “preserve[d] all six incumbent Democrats in ‘safe’ districts,”
none of which would have “less than 58% [DPI]”; and “g[ave] Democrats a real opportunity to
pick up a seventh seat in the delegation by targeting Roscoe Bartlett.” ECF No. 177-23. Governor
O’Malley admitted that his Advisory Committee sought to “create a district” that “would be more
likely to elect a Democrat than a Republican.” O’Malley Dep. 82:16–18; see also id. at 27:12–15
(describing aim of “put[ting] more Democrats and Independents into the Sixth District” to ensure
“the election of another Democrat”). Senate Majority Leader Garagiola admitted that “one of the
purposes[] [was] to make the Sixth Congressional District have 53 percent Democratic
performance.” Garagiola Dep. 27:4–9 (ECF No. 177-24). These sorts of statements, particularly
by delegates and state senators during the General Assembly’s abbreviated consideration of the
proposed map, are legion. See, e.g., Joint Stipulations ¶¶ 40–51.
The State’s argument that its officials intended only “to allow Democrats to have an equally
effective voice in the election of a representative” in the Sixth District — an intent that it argues

51

“cannot be equated with an intent to burden [Republicans’] representational rights” — is hollow.
Defs’ Memo. at 31. Even if the intent to make one party “more competitive” were constitutionally
permissible, the record shows something materially different.

Members of the Advisory

Committee, with the help of NCEC, worked to craft a map that would specifically transform the
Sixth District into one that would predictably —that is, by a 94% chance — elect a Democrat by
removing Republicans from the District and adding Democrats in their place.
More fundamentally, the State’s argument misunderstands the law. If the government uses
partisan registration and voting data purposefully to draw a district that disfavors one party, it
cannot escape liability by recharacterizing its actions as intended to favor the other party. The
First Amendment does not distinguish between these intents. A school board, for example, cannot
manipulate its stock of library books for “narrowly partisan” reasons, whether its conduct is
described as removing all books written by Republicans or as constructing a library full of books
written by Democrats. Pico, 457 U.S. at 870–71. Where the government singles out a person or
class of persons based on their political affiliation and voting and acts so as to hamper their ability
to effectively engage in future expression, it has run afoul of the First Amendment no matter how
it characterizes its intent.
The State also argues that its officials did not act with impermissible intent because they
did not target specific voters based on their individual party affiliation or voting history. This
argument, too, is based on a misunderstanding of First Amendment jurisprudence. Here, the
plaintiffs have shown that they were targeted for disfavored treatment because of a shared marker
of political belief — their Republican party affiliation. The fact that the State moved Republican
voters out of the Sixth District en masse, based on precinct-level data, and did not examine each
voter’s history with care before taking that punitive action does not make its action less culpable

52

under the First Amendment. Cf. Miller, 515 U.S. at 920 (condemning State’s targeting of areas
with “dense majority-black population” for inclusion in district); Sweezy v. Wyman, 354 U.S. 234,
250 (1957) (“Any interference with the freedom of a party is simultaneously an interference with
the freedom of its adherents”). If anything, the First Amendment is more skeptical where the
government uses peoples’ nominal party alignment as a proxy for their actual expression. See,
e.g., Elfbrandt v. Russell, 384 U.S. 11, 19 (1966).
Thus, because State officials have admitted that they intended “to create a district where
the people would be more likely to elect a Democrat than a Republican” and that they removed
likely Republican voters from the Sixth District specifically to achieve that aim, the plaintiffs have
established that the State acted with constitutionally impermissible intent.
Second, with respect to the adverse effect element, the plaintiffs have shown that the
redrawn Sixth District did, in fact, burden their representational rights. At the threshold, it is
important to reiterate that, under the standard set forth in our denial of the motion to dismiss, a
plaintiff who has shown that the State acted with impermissible retaliatory intent need not show
that the linedrawing altered the outcome of an election — though such a showing would certainly
be relevant evidence of the extent of the injury. See Shapiro, 203 F. Supp. 3d at 598. And, contrary
to the State’s argument, the plaintiffs need not show that the new Sixth District was certain to
produce a Democratic congressman. See Kusper, 414 U.S. at 58 (explaining that, while restriction
on primary voting did not “deprive [voters] of all opportunities to associate with the political party
of their choice,” it was nevertheless “a ‘substantial restraint’ and a ‘significant interference’ with
the exercise of the constitutionally protected right of free association”); Rutan v. Republican Party
of Ill., 497 U.S. 62, 74 (1990) (holding that adverse employment actions not amounting to
discharge may nevertheless violate the First Amendment). Rather, the plaintiffs must show only

53

that their electoral effectiveness was meaningfully burdened — and, of course, that it was
intentionally burdened for partisan reasons.

That is, a voter must have experienced a

“demonstrable and concrete adverse effect” on his “right to have ‘an equally effective voice in the
election’ of a representative.’” Shapiro, 203 F. Supp. 3d at 598 (quoting Reynolds v. Sims, 377
U.S. 533, 565 (1964)); see also Balt. Sun Co. v. Ehrlich, 437 F.3d 410, 416 (4th Cir. 2006) (“[W]e
have recognized a distinction between an adverse impact that is actionable, on the one hand, and
a de minimis inconvenience, on the other”).
The plaintiffs here have made such a showing. By several measures, the new Sixth District
map severely disfavors Republican voters. In creating the map, the State removed over 66,000
registered Republicans from the Sixth District and added some 24,000 registered Democrats, such
that Republican voters went from outnumbering Democrats 1.3 to 1 (47% of the district’s
registered eligible voters being Republicans and 36% Democrats) to nearly the exact inverse (44%
Democrats, 33% Republicans). Joint Stipulations ¶¶ 10, 53. According to the DPI metric used by
the mapmakers and the Cook PVI metric endorsed by the State’s expert, Republican voters in the
new Sixth District were, in relative terms, much less likely to elect their preferred candidate than
before the 2011 redistricting, and, in absolute terms, they had no real chance of doing so. Indeed,
the Cook report deemed the district’s swing — from “Solid Republican” (R+13) to “Likely
Democratic” (D+2) — the largest of any district in the country. ECF No. 177-52 at 6–8. And,
historically, “Likely Democratic” districts elect a Democrat 94% of the time. See Campbell, supra,
at 628.
Moreover, while the State’s linedrawing need not change the outcome of an election to be
culpable, the fact that a Democratic candidate was elected in the three elections following the 2011
redistricting supports the fact that the Republican voters have suffered constitutional injury. In

54

other words, the Democratic officials who drew the map achieved what they aimed to do — to
make Republican voters in the Sixth District less effective.
The State argues that the plaintiffs have not adequately shown that the new Sixth District
map actually “chilled” their protected expression. Defs’ Memo. at 38. This argument has two
flaws. First, a First Amendment injury need not take the form of “chilling” or “deterring” speech.
Rather, a plaintiff may claim retaliation if his expression is “adversely affected,” Suarez Corp.
Indus. v. McGraw, 202 F.3d 676, 686 (4th Cir. 2000), and surely the government’s reduction of
the effectiveness of expression qualifies as an adverse effect. Second, there is no requirement an
individual plaintiff show that the government’s action has specifically deterred him from engaging
in protected conduct. The Supreme Court’s patronage cases, which are rooted in retaliation
principles, have expressly repudiated any requirement “that dismissed employees prove that they,
or other employees, have been coerced into changing, either actually or ostensibly, their political
allegiance.” Branti v. Finkel, 445 U.S. 507, 517 (1980). Rather, “[t]he determination of whether
government conduct or speech has a chilling effect or an adverse impact is an objective one.” Balt.
Sun, 437 F.3d at 416. Thus, to evaluate whether the State’s conduct caused a First Amendment
injury, we assess only whether its purposeful dilution of Republicans’ electoral power would
adversely affect the protected expression of a reasonable person situated similarly to the plaintiffs.
See id.
The State’s action here would impair a reasonable Republican voter’s exercise of his First
Amendment rights. Republicans in the Sixth District faced a severe political disadvantage after
the 2011 redistricting. This itself is a constitutional injury. Moreover, it is not hard to see how
the dilution of Republican voters’ effectiveness could deter reasonable voters from full
participation in the political process. A committed Republican voter who finds himself in the

55

minority may well lose interest in voting or in supporting candidates for a legislative office that,
realistically, they are unlikely to fill. A different Republican voter in the new Sixth District might
choose to abandon his party, finding his energy better spent supporting moderate candidates in
Democratic primaries. See Rutan, 497 U.S. at 73 (“[E]mployees who have been laid off may well
feel compelled to engage in whatever political activity is necessary to regain regular paychecks
and positions corresponding to their skill and experience”); Elrod v. Burns, 427 U.S. 347, 355
(1976) (“[A] pledge of allegiance to another party, however ostensible, only serves to compromise
the individual’s true beliefs”).

Here, there was direct evidence of chilled expression, as

participation in the Sixth District’s Republican primaries dropped substantially between 2010 and
2014, which supports the notion that partisan manipulation deterred the robust exercise of
representational rights. There is also anecdotal evidence of Republicans not voting after the
redistricting because of confusion or loss of interest. Of course, voters have no constitutional right
to be successful in electing the candidate they favor, and voters regularly lose interest in politics
or switch parties for reasons unrelated to gerrymandering. But this does not answer the relevant
First Amendment question. In short, the purposeful reduction of one party’s effectiveness may
well chill the protected expression of that party’s voters, even if no individual plaintiff establishes,
as a factual matter, that he was so chilled.
Finally, as to causation, the plaintiffs have established that, absent the State’s retaliatory
intent, the Sixth District lines would not have been drawn to dilute the electoral power of
Republican voters to the same extent. The framework governing our inquiry into causation is set
forth in Mt. Healthy, 429 U.S. 274. Specifically, once the plaintiffs have established that the
government’s constitutionally impermissible intent “was a ‘motivating factor’ in [its] decision,”
the burden shifts to the State to show that, even absent the forbidden intent, “it would have reached

56

the same decision.” Id. at 287 (quoting Vill. of Arlington Heights v. Metro. Hous. Dev. Corp., 429
U.S. 252, 270 (1977)). In other words, assuming that the State intended to burden the plaintiffs’
representational rights, we must then determine “if this intent was the decisive factor in [their]
decision” to do so. Pico, 457 U.S. at 871; see also Vill. of Arlington Heights, 429 U.S. at 270 n.21
(explaining that, where “the same decision would have resulted even had the impermissible
purpose not been considered,” then “there would be no justification for judicial interference with
the challenged decision”). Under the Mt. Healthy framework, therefore, where unlawful intent in
fact drove the State to its decision, the State cannot escape liability by “hypothesiz[ing] that it
might have employed lawful means of achieving the same result.” Regents of Univ. of Cal. v.
Bakke, 438 U.S. 265, 320 n.54 (1978) (opinion of Powell, J.) (emphasis added) (declining to allow
a remand because it “would result in fictitious recasting of past conduct”).
The State rejects Mt. Healthy’s burden-shifting framework for causation, contending that
it applies only in the context of public employment. But there is simply no support for the State’s
cramped reading of that case. On the contrary, Mt. Healthy stands for a general, common-sense
principle applicable in all retaliation-based First Amendment claims — that, where the government
takes an injurious action, an injured party need not show that the government would never have
taken the same action anyway. The Supreme Court has accordingly relied on the Mt. Healthy
framework in several types of claims unrelated to public employment, and indeed in allegations of
constitutionally forbidden intent beyond those related to protected expression. See Pico, 457 U.S.
at 870, 871 n.22 (school board’s removal of books from school library); Wilkie v. Robbins, 551
U.S. 537, 558 n.10 (2007) (Bureau of Land Management’s intimidation of landowner to induce
his grant of an easement); Texas v. Lesage, 528 U.S. 18, 20–21 (1999) (per curiam) (university’s

57

rejection of application under race-conscious admissions program); Vill. of Arlington Heights, 429
U.S. at 270 n.21 (village’s denial of rezoning request). The framework is no less applicable here.
As already noted, the record demonstrates that the State intended to burden the plaintiffs’
representational rights, which leaves the question of whether the State has shown that, absent this
intent, it would have drawn lines that similarly burdened Republican voters in the Sixth District.
While it probably would be impossible for the State to show that it would have drawn the exact
same district lines absent the impermissible intent, to satisfy its end of the burden-shifting inquiry,
it would at least have to show that it would have drawn lines that similarly burdened the plaintiffs’
representational rights.
Even this, however, the State cannot do. It points to two primary objectives that it claims
justify the Sixth District’s reconfiguration in 2011 — preventing the new First District from
crossing the Chesapeake Bay and grouping residents of the I-270 corridor together in one district.
But the evidence of intent in this case is overwhelming and undisputed that the State drew the lines
of the Sixth District to flip the District from Republican to Democratic control, and it is implausible
that consideration of these other objectives would have led to a map that similarly burdened
Republican voters. Again, in tasking Hawkins with drawing a map, Democratic officials provided
him with only two goals — protecting Democratic incumbents and obtaining a seventh Democratic
seat. Hawkins was not instructed to consider whether “there was a community of interest related
to the I-270 corridor.” Hawkins Dep. 128:19–20. The record shows no invocation of I-270 as a
justification for the shapes of the Sixth and Eighth District’s until Jeanne Hitchcock’s presentation
of the nearly final map to the joint session of House and Senate Democratic Caucuses and,
unsurprisingly, even Democratic delegates found it a flimsy justification for the dramatic
reshuffling of the two districts. See, e.g., Joint Stipulations ¶ 46.

58

The majority, in finding that the plaintiffs failed to show a likelihood of success on the
causation element, commits two significant errors. First, it mischaracterizes our previous holding
on the causation element to adopt a new standard that is inconsistent with First Amendment
jurisprudence. Second, it applies the new standard to the facts in a confusing and inherently
inconsistent manner.
The majority begins correctly by stating the causation standard from our previous holding
— that the gerrymander must create a tangible, adverse impact that would not have occurred but
for the unconstitutional intent of the mapmakers. See ante at 17. But then it leaps from this correct
statement of the causation standard to its own newly created standard by requiring “proof that but
for the gerrymander, the challenged effect (here, the switch in political power in the Sixth District)
would not have happened.” Id. (second emphasis added). Explaining its new standard further, the
majority states that the causation element “would be satisfied only if [the evidence showed that]
Roscoe Bartlett would have won reelection in 2012 had the prior map remained intact.” Ante at
18. Indeed, it expressly contemplates that voters’ injury takes the form of Bartlett’s loss to
Delaney, “but only if that loss is attributable to gerrymandering or some other constitutionally
suspect activity. If the loss is instead a consequence of voter choice, that is not an injury.” Id.
These arguments, however, represent a failure to understand First Amendment jurisprudence,
which focuses not on who wins but on the burden imposed on First Amendment rights — here, on
the right to cast an undiluted vote. In short, the majority’s new First Amendment standard depends
on an election’s results, not on the adverse impact of dilution on the targeted voters. Under the
applicable First Amendment framework, however, the adverse effect is the dilution of votes —
and the corresponding burdening of expression by voters — regardless of how the election turned
out.

59

Under the majority’s standard requiring an altered election outcome, critical First
Amendment violations could never be remedied. For instance, claims that the party in control of
State government deliberately attempted to suppress political speech before an election or
deliberately located polling places to inconvenience the other party could never be pursued under
the majority’s standard, because the plaintiffs would be unable to show that the election results
were tipped as a result of the unconstitutional conduct. More to the point, under the majority’s
new standard, no redistricting map could be challenged before an election. Any standard of
causation that would so arbitrarily limit our ability to redress constitutional injuries must be
rejected.
In applying its new standard to the facts in the record, the majority’s analysis is yet more
confusing. The majority accepts that the defendants here did in fact intend to retaliate against
voters who had previously voted for Republican candidates in the Sixth District by drawing a map
that moved over 66,000 Republicans from the old Sixth District and introduced some 24,000 new
Democrats to diminish the Republicans’ ability to express their political viewpoint. The majority
also accepts, as it has to, that this map was in fact adopted and that, under this new map, the
Republicans’ voice was diminished and the Democrats achieved unprecedented electoral success
in the Sixth District. I submit that only one conclusion can be drawn from these accepted facts —
that a degree of vote dilution significant enough to place Republican voters at a concrete electoral
disadvantage was caused by the conduct that the State specifically intended. Yet, somehow, the
majority holds that these actions did not cause the retaliatory harm that the State intended. The
majority somehow concludes that the State’s plan was ineffective, despite its intended effect
coming to pass. Such a view of causation necessarily embraces the bizarre notion that other,
unnamed factors might have coincidently caused those effects. Under such reasoning, a defendant

60

who intentionally poisons a victim’s drink could not be found to cause the death because the victim
might have died from a heart attack anyway. Yet this is the argument that the majority embraces.
Moreover, applying a causation standard that seeks to eliminate all possible but unproved
factors, however remote and speculative, is directly contrary to the causation standard that the
Supreme Court has established for retaliation claims. In Mt. Healthy, the Court required only a
showing that the constitutionally impermissible intent was a motivating factor, such that the State
cannot escape liability by hypothesizing some remote or speculative cause. See Mt. Healthy, 429
U.S. at 287.
In sum, the record amply proves that the State violated the First Amendment under the
standard we previously adopted in this case. Indeed, on this record, there is no way to conclude
otherwise, even as a possibility. A fortiori, it follows that the plaintiffs have demonstrated the
likelihood of success on the merits, as required for entering a preliminary injunction.

IV
The other three factors governing our issuance of a preliminary injunction do not require
extensive discussion. Absent an injunction, the plaintiffs are likely to suffer irreparable harm.
Because the State’s construction of the Sixth District in 2011 likely violated the plaintiffs’ First
Amendment rights, the plaintiffs are experiencing ongoing constitutional injury without a new
map. See League of Women Voters of N.C. v. North Carolina, 769 F.3d 224, 247 (4th Cir. 2014)
(“Courts routinely deem restrictions on fundamental voting rights irreparable injury”). Moreover,
the plaintiffs seek a preliminary injunction now so as to have a new map in place for the 2018
congressional election cycle. We must be mindful of the fact that, “once the election occurs, there
can be no do-over and no redress. The injury to these voters is real and completely irreparable if
nothing is done to enjoin this law.” Id.
61

The balance of the equities here also favors the plaintiffs. To be sure, requiring Maryland
to redraw the Sixth District’s boundaries is no trivial matter. But where, as here, plaintiffs establish
a strong likelihood of success on the merits and irreparable injury, they have generally shown that
the equities work in their favor. Direx Israel, Ltd. v. Breakthrough Med. Corp., 952 F.2d 802,
812–13 (4th Cir. 1991). And though there is no doubt that the State would have to expend
resources in redrawing district lines that comply with our injunction, the fact that the State
regularly creates new legislative maps so as to comply with other constitutional requirements, like
“one person, one vote,” suggests that the burden will not be unduly onerous. Indeed, our discussion
of the merits reveals the ease and precision with which lines can be drawn using mapmaking
software, and we are confident that the State, with the aid of such software, will have little trouble
devising an alternative map that complies with the law. The plaintiffs in fact have offered an
alternative map in this case where only the line between the Sixth and Eighth Districts had to be
redrawn.
Finally, it is obvious that an injunction here will serve the public interest. An injunction
will not only redress a serious, ongoing constitutional injury, but will also enable the plaintiffs and
those similarly situated to them — a large portion of Maryland voters — to more fully participate
in congressional elections.
In sum, this fulsome record overwhelmingly shows the plaintiffs’ satisfaction of our First
Amendment standard, and the ongoing harm can only be rectified by the entry of an injunction. I
would therefore grant the plaintiffs’ motion in full.

V
If the plaintiffs were to appeal the denial of their motion for an injunction, see 28 U.S.C.
§ 1253, I would have no objection to the entry of a stay. Failing that, however, the mere pendency
62

of Gill v. Whitford, No. 16-1161, in the Supreme Court does not justify delaying a final decision
in this case alleging a serious breach of an important constitutional right. The nature of the claim
in Gill, as well as the facts supporting the claim, is materially different from the nature of the claim
before us. Gill centers on an Equal Protection claim relating to statewide redistricting, while this
case involves a First Amendment claim arising from the line-drawing of a single district.
Accordingly, at this juncture, I do not join the majority’s sua sponte entry of a stay.

63