Vesilind v. Va. State Bd. of Elections

PRESENT: All the Justices
RIMA FORD VESILIND, ET AL.
OPINION BY
v. Record No. 170697 JUSTICE S. BERNARD GOODWYN
May 31, 2018
VIRGINIA STATE BOARD OF ELECTIONS,
ET AL.
FROM THE CIRCUIT COURT OF THE CITY OF RICHMOND
W. Reilly Marchant, Judge
In this appeal, we consider whether the circuit court erred in confirming the constitutional
validity of Virginia General Assembly legislative districts that were allegedly drawn in violation
of the compactness requirement expressed in Article II, § 6 of the Constitution of Virginia.
PROCEDURAL BACKGROUND
On September 14, 2015, Rima Ford Vesilind and 13 other “citizens of the
Commonwealth of Virginia, who live in legislative districts that are alleged to have been drawn
in violation of the compactness requirement” contained in Article II, § 6 of the Constitution of
Virginia (Challengers), filed a complaint in the Circuit Court of the City of Richmond against the
Virginia State Board of Elections (VSBE) and several officers of the VSBE, in their official
capacities (collectively, the Board). The complaint sought a declaratory judgment that the “State
House of Delegates and Senate districting plans, and specifically House of Delegates districts 13,
22, 48, 72, and 88, and Senate districts 19, 21, 28, 29, 30, and 37 [(the Challenged Districts)]
violate the Constitution of the Commonwealth of Virginia,” as those districts were drawn in the
2011 redistricting plan (Enacted Plan or 2011 Plan) adopted by the General Assembly. The
districts in the 2011 Plan were established in H.B. 5005, Va. Gen. Assem. (Spec. Sess. 2011),
which was codified in Code § 24.2-303.3 (Senate districts) and Code § 24.2-304.03 (House
districts).
In their complaint, the Challengers state that legislative districts must be drawn so as to
comport with certain mandatory requirements. The requirements include those in Article II, § 6
of the Constitution of Virginia, which provides that legislative districts “shall be composed of
contiguous and compact territory and shall be so constituted as to give, as nearly as is
practicable, representation in proportion to the population of the district,” as well as federal
requirements of “one person, one vote” and the Voting Rights Act, 52 U.S.C. § 10101 et seq.
(VRA). 1 The complaint asserts that “[w]hile the legislature may consider other rational public
policy considerations, the mandates of the United States and Virginia Constitutions can never be
subordinated to those considerations.”
The Challengers allege that the General Assembly “subordinated” the mandatory
compactness requirement to other public policy considerations, and ignored compactness in
favor of “nonconstitutional considerations,” such as “favor[ing] partisan interests” and
“protect[ing] particular incumbent[s],” “with the result that the Challenged Districts are not
within any acceptable objective measures of compactness.” They claim that “when the General
Assembly drew the 2011 House and Senate district plans . . . it did not make a good-faith effort
to draw compact districts and . . . numerous districts in the adopted plans are not in fact compact
as required by the Virginia Constitution.”
In their complaint, the Challengers contend that “[a]s measured by quantitative and
qualitative tests, objective and reasonable persons could not reach any conclusion but that the
Challenged Districts do not meet the Virginia Constitution’s compactness requirement.”
1
After the 2011 redistricting, the Supreme Court of the United States ruled that
preclearance by the Department of Justice under Section 5 of the VRA is no longer determined
by the coverage formula in Section 4(b) of the VRA. Shelby County v. Holder, 570 U.S. 529,
556-57 (2013).
2
Accordingly, they request a judgment declaring the Challenged Districts, and thus the 2011 Plan,
in violation of the Constitution of Virginia.
On October 21, 2015, the circuit court entered a consent order permitting the Virginia
House of Delegates and the Honorable William J. Howell, then the Speaker of the House of
Delegates (the House), to intervene as defendants. 2 The Board and the House filed answers and
defenses, denying that the Challenged Districts were constitutionally infirm.
The case subsequently proceeded to a three-day bench trial beginning on March 13, 2017.
The Challengers, the Board, and the House presented evidence.
On March 31, 2017, the circuit court issued its final order and opinion. Weighing the
evidence on both sides, the circuit court concluded that the evidence would “lead reasonable and
objective people to differ,” as to whether the Challenged Districts were compact, and so it was
“fairly debatable” whether the Challenged Districts violated the compactness requirement
expressed in Article II, § 6. Therefore, the circuit court denied the Challengers’ request that the
Challenged Districts be declared to be in violation of the Constitution of Virginia.
The Challengers appeal, asserting that the circuit court erred by finding that the evidence
produced at trial sufficed to make the constitutionality of the General Assembly’s redistricting
decision, regarding the compactness of the Challenged Districts, fairly debatable.
2
By order dated March 23, 2018, the Hon. M. Kirkland Cox, current Speaker of the
Virginia House of Delegates, was substituted for the Hon. William J. Howell as an appellee in
this appeal.
3
TRIAL
A. Evidence Provided by the Challengers
The Challengers presented two witnesses, Nicholas Mueller (Mueller) and Dr. Michael
McDonald (Dr. McDonald). First, Mueller testified that he used software called Maptitude
“under [Dr. McDonald’s] direction to draw [the] alternate [Virginia legislative district] maps” to
compare against the 2011 Plan and the Challenged Districts.
Mueller stated that the goal of the “Alternative Plan 1” House and Senate maps he
created was to “make the [alternative legislative] districts as compact as possible,” while
“preserv[ing] the majority/minority [VRA] districts exactly as they are in the current plan,” and
also meeting the continuity and “same equal population” criteria mandated by federal and state
law.
Mueller also created “Alternative Plan 2” House and Senate maps at Dr. McDonald’s
direction. Alternative Plan 2 kept the VRA districts unchanged and has contiguous districts with
the same equal population criteria, but it also incorporates two discretionary criteria: (1)
refraining from splitting political subdivisions and voting districts, and (2) refraining from
pairing incumbents to the extent possible. Mueller authenticated the materials he produced.
Thereafter, Dr. McDonald was presented as an expert in redistricting. He testified that he
has been consulting for political redistricting matters in 14 states since the 1980s. Dr. McDonald
testified that he was asked to determine if priority was given to the constitutional requirement of
compactness in the Challenged Districts or whether other discretionary criteria, not mandated by
federal or state law, predominated over compactness. He explained that “Required Criteria”
means those criteria required by the federal Constitution, Constitution of Virginia, or the VRA,
4
and “Discretionary Criteria” refers to all other criteria that the legislature could conceivably have
considered in drawing the districts.
Dr. McDonald testified about his methodology for measuring constitutional compactness
and his conclusions. He explained that he chose the Reock, Polsby-Popper, and Schwartzberg
scores to measure compactness of the legislative districts because Virginia measures
compactness based on spatial parameters, rather than population-based parameters, and because
the Office of the Attorney General of Virginia used all three measures in its report to the
Department of Justice under Section 5 of the VRA. 3 He stated that he directed Mueller to
produce compactness scores for the alternative map districts so that those scores could be
compared to the compactness scores of the districts in the 2011 Plan maps.
Dr. McDonald stated that in contemplating how to approach this case, he “struck upon
this idea that we could use predominance” to demonstrate how compactness had been
subordinated to discretionary criteria. He theorized that predominance could be proven by
determining if the compactness of an ideally compact district (such as in Alternative Plan 1) was
3
The Reock, Polsby-Popper and Schwartzberg tests are three different statistical
calculations used by the social science community to measure the compactness of legislative
districts. According to Dr. McDonald, the Polsby-Popper score “is a perimeter measure that
compares the area of a district to the area of a circle with the same perimeter.” The Reock score
“is an area or dispersion measure that compares the area of a district to the area of the small
circle that can be drawn around the district.” The Schwartzberg measure “is a perimeter-based
measure that uses a simplified version of the district’s perimeter that is designed to smooth [out]
the jaggedness in the perimeter of a district that may be caused by natural geography such as
rivers or coast lines or sometimes the map drawer’s choices.” The Schwartzberg measure “then
compares the simplified perimeter of the district to the perimeter of the circle with the same area
as the original district.” The Reock and Polsby-Popper measures range from 0 to 1, with 1 being
the most compact. The Schwartzberg measure initially results in scores that increase from 1, but
Dr. McDonald converted those scores to result in scores that are in the same range of 0 to 1, with
1 representing the most compact form. Dr. McDonald acknowledged that there are
approximately 50 such measures developed and used by social scientists.
5
reduced by more than 50 percent when discretionary criteria were considered in redrawing that
ideally compact district. He concluded that if the compactness score was reduced more than 50
percent when discretionary criteria were considered, then it “must be that [the] discretionary
criteria had predominated over the state constitutional requirement of compactness.”
Dr. McDonald’s Alternative Plan 1 did not simply maximize compactness in the
Challenged Districts, but tried “to maximize the compactness of all the districts,” because
changing the borders of one district has a “ripple effect” for other districts. Alternative Plan 2
was developed to “show how the legislature could pursue other discretionary goals . . . without
failing that predominance test,” and Dr. McDonald selected the two discretionary criteria of not
splitting counties, cities, and voting districts and pairing incumbents because they “were easily
measurable.”
Dr. McDonald testified regarding each of the Challenged Districts that the degradation in
compactness between the Enacted Plan and Alternative Plan 1 compactness scores was more
than 50 percent, and so he opined that “discretionary criteria predominated over the
constitutional requirement of compactness” in each of the Challenged Districts.
For example, he compared maps of House District (HD) 72 as enacted and as designated
in Alternative Plan 1, and compared the composite average of the Reock, Polsby-Popper, and
Schwartzberg measures for the enacted version of HD 72 and for the HD 72 in Alternative
Plan 1. Based on his comparisons, he explained:
the degradation is the composite score [of the alternative plan], .71, minus the
composite score of .21 [of the enacted district], that difference, and then it’s
divided through by the composite score, .71. It tells us what percentage decrease
that we had in the composite score from Alternative Plan 1 to the current plan.
And here we see that the degradation was 70.42 percent, and that’s above 50
6
percent. And so my opinion, then, is that for this district, that discretionary
criteria predominated over the state constitutional requirement of compactness. 4
Dr. McDonald found that there was “only a degradation of 19.72 percent . . . in terms of
compactness” for HD 72 between Alternative Plan 1 and Alternative Plan 2, which showed “an
attempt to draw districts that are going to conform to local political boundaries that will respect
voting precincts and will attempt to minimize the pairing of incumbents.”
Dr. McDonald did not apply his predominance test to VRA districts, because “[t]hose are
districts where there are other considerations that are at play, and the test that we have devised is
one that can only be applied to nonvoting rights districts.” He stated that those considerations
“are not present in any of the [C]hallenged [D]istricts except for House District 72,” which
shares a small border with a VRA district in the Enacted Plan, but not in Alternative Plan 1.
Dr. McDonald acknowledged that there was “no universal measure of what constitutes
the best measure for compact districts,” and that “[i]n the social science community, there’s no
threshold for when a district is compact or noncompact” using these measures. He stated that a
difference of 0.01 or 0.02 in compactness scores, and even a difference of 0.03, was not “a large
difference.” On cross-examination, he conceded that he had not previously used or published the
predominance test, and was not aware of another expert, court opinion, or state using the
“predominance test” that he used to develop degradation percentages regarding compactness of
the legislative districts.
4
Dr. McDonald initially used a composite score of all three measures, but later he
focused on composite scores based only on the Reock and Polsby-Popper measures, because
those were the measures used by the legislature.
7
B. Stipulations
The parties stipulated to the number of counties, cities, and precincts split in the 2001 and
2011 redistricting plans, as well as in Alternative Plans 1 and 2 developed by Dr. McDonald.
They stipulated to the compactness scores for the 2001 and 2011 redistricting plans, as well as
the scores for the districts in Dr. McDonald’s Alternative Plans 1 and 2. They also stipulated to
the Reock, Polsby-Popper, and Schwartzberg compactness scores of the districts challenged in
Jamerson v. Womack, 244 Va. 506, 423 S.E.2d 180 (1992), and Wilkins v. West, 264 Va. 447,
571 S.E.2d 100 (2002), because the scores were not reported in those opinions. Based on these
scores as stipulated, the lowest Reock score in the relevant legislative districts was 0.12 in
Jamerson and 0.16 in Wilkins, and the lowest Polsby-Popper score in both cases was 0.10. For
the Challenged Districts, the lowest Reock score was 0.15 (SD 28), and the lowest Polsby-
Popper score was 0.08 (SD 28 and HD 72). The Reock scores of all of the Challenged Districts
exceeded the minimum score in Jamerson, and the Polsby-Popper scores of nine of the eleven
Challenged Districts exceeded the minimum score in Jamerson and Wilkins.
The parties also stipulated to documents submitted by the Attorney General of Virginia to
the Department of Justice for preclearance under the VRA, which used the Reock, Polsby-
Popper, and Schwartzberg measures, and to the Criteria Resolutions adopted by the House and
the Senate of Virginia (Senate) (collectively, General Assembly or legislature) Privileges and
Elections Committees regarding the 2011 redistricting. The Criteria Resolutions listed six
criteria used to draw the districts in the 2011 Plan.5
5
Criteria I was that the “population of each district shall be as nearly equal to the
population of every other district as practicable,” within “plus-or-minus two percent” for the
Senate and “plus-or-minus one percent” for the House, and Criteria II was to draw the districts in
accordance with the VRA.
8
Additionally, the parties stipulated to letters from Governor Robert F. McDonnell
regarding his veto of the original 2011 redistricting plan, H.B. 5001, and his signing of H.B.
5005, Va. Gen. Assem. (Spec. Sess. 2011), which established the 2011 Plan. 6
C. Motion to Strike
At the conclusion of the Challengers’ evidence, the Board and the House moved to strike,
on the grounds that the Challengers failed to present evidence to support their claims that
Criteria III stated, in relevant part: “Districts shall be contiguous and compact in
accordance with the Constitution of Virginia as interpreted by the Virginia Supreme Court in the
cases of [Jamerson] and [Wilkins].” Criteria IV required that “[a]ll districts shall be single-
member districts,” and Criteria V stated: “Districts shall be based on legislative consideration of
the varied factors that can create or contribute to communities of interest,” which “may include
. . . economic factors, social factors, cultural factors, geographic features, governmental
jurisdictions and service delivery areas, political beliefs, voting trends, and incumbency
considerations.” Criteria VI stated the rules of priority:
All of the foregoing criteria shall be considered in the [redistricting] process, but
population equality among districts and compliance with federal and state
constitutional requirements and the Voting Rights Act of 1965 shall be given
priority in the event of conflict among the criteria. Where the application of any
of the foregoing criteria may cause a violation of applicable federal or state law,
there may be such deviation from the criteria as is necessary, but no more than is
necessary, to avoid such violation.
(Emphasis added.)
6
Governor McDonnell vetoed H.B. 5001, noting, among other concerns, that “it is
apparent that districts proposed in the Senate plan are not compact, as required in the
Constitution of Virginia,” and that “[u]sing the most commonly recognized tools of compactness
scoring, the Reock and Polsby-Popper methods, the plan adopted by the Senate has less compact
districts than the existing House or Senate districts or other plans that have been proposed.” In a
statement marking his signature of H.B. 5005, the Governor noted that the revised “plan retains
more geographic and municipal boundaries, contains districts that are somewhat more compact,
and passed the Senate on a strong bipartisan vote.” The Governor acknowledged that “[w]hile
additional improvements in measures of compactness and preservation of communities of
interest would have been ideal, and no plan is perfect, the Constitution of Virginia tasks the
General Assembly with drawing lines . . . .”
9
compactness was ignored and that the Challenged Districts were not compact. Instead, the
Challengers only presented “their new compactness test and the alternative maps that it’s based
on.” They argued that the Challengers’ evidence showed that compactness was considered, and
that the “legislature’s value judgment of the relative degree of compactness required, when
reconciling multiple concerns of apportionment, is fairly debatable.” The Challengers responded
that their evidence showed that the districts were not constitutionally compact because
discretionary factors predominated over the constitutional requirement of compactness. The
court denied the motions to strike.
D. Evidence Presented by the Board and the House
The Board presented evidence concerning the Senate districts from the 2011 Plan, and the
House presented evidence concerning the House districts from the 2011 Plan. The Board moved
into evidence floor speeches from several Senators during the 2011 redistricting process. The
Senators stated that H.B. 5001 “meets all federal and state legal requirements,” and that “all
districts are contiguous and compact in accordance with the constitution of Virginia as
interpreted by the Virginia Supreme Court in the cases of Jamerson v. Womack and Wilkins v.
West.” Senators also made similar comments regarding H.B. 5005 and the Enacted Plan.
Senator Jeremy S. McPike (Senator McPike) of the 29th Senate District testified that it was not
hard for him to travel around his district, and that, during the campaign, the “shape or geographic
extent of [his] district” did not cause any problem in communicating with constituents because
“[m]ost of it’s social media now.”
Dr. M.V. “Trey” Hood, III (Dr. Hood) testified as an expert for the Board. He stated that
there was no consensus on the best measure of compactness or a threshold for when a district
was not compact, and that the different scores “measure compactness in different ways,” but he
10
does not average different types of compactness scores. Dr. Hood stated that there was a
decrease in the average compactness scores overall from the 2001 plan to the 2011 Plan, but the
decline in compactness did not “on its face, mean that these districts are necessarily uncompact.”
Dr. Hood listed among his “criticisms” of Dr. McDonald’s approach that he should have
compared the whole redistricting plan to other plans, rather than compare district-to-district,
because the districts “are not drawn in isolation from one another” and because the districts in
the alternative plans and the 2011 Plan did not exactly overlap geographically, so “sometimes
it’s not easy to compare one district from one plan to another district from another plan.” His
“criticisms” also included that Dr. McDonald should not have averaged compactness scores, and
that there “could be” another Alternative Plan 1, although neither he nor anyone else had
proposed such a plan. 7 Dr. Hood also had “questions about how this [the predominance test]
would actually be implemented if it was ever put in place,” because there would be “a fight” in
the legislature over drawing alternative plans, because there is “no best way” to “completely
maximiz[e] compactness.” Also, he claimed Dr. McDonald did not account for how the
legislature would attempt to balance all of the factors involved in redistricting.
The House then presented its evidence. Delegate Chris Jones (Delegate Jones) testified
that he was the chief patron of the 2011 redistricting legislation in the House, and he was also an
architect of the 2001 redistricting plan. He stated that there was a “very similar” increase in
population in both the 2000 census that informed the 2001 plan (a “13 or 14 percent population
increase”) and the 2010 census that informed the 2011 Plan (“about a 13 percent increase in
population”). He explained that he thought the districts in the 2011 Plan were more compact
7
Dr. McDonald acknowledged that he could not say that Alternative Plan 1 was “the
most compact map,” and agreed that “Alternative Plan 1 is an ideal, not the ideal compactness
plan for all purposes and all possibilities.”
11
than those in the 2001 plan. For example, even though the compactness scores showed that HD
13 was less compact in 2011 than it was in 2001, he explained that, “[f]rom [my] perspective,” it
was more compact in 2011, because both the population and geographic area of the district were
reduced, and “because of the ease of going back and forth between the districts.”
Delegate Jones stated that the House hired John Morgan (Morgan) and Dale Oldham
(Oldham) to assist in the 2011 redistricting, and explained the criteria he used in the map
drawing process, including equal population, the VRA, and communities of interest. He stated
that he did not believe that the VRA districts could be treated differently in terms of being
compact and contiguous as required by the Constitution of Virginia.
Delegate Jones testified that he believed that the 2011 Plan complied with the Criteria
Resolutions regarding priority of constitutional considerations over discretionary factors. He
stated that priority was given to state constitutional criteria, and that, “if there was an actual
conflict between compactness and [discretionary criteria],” compactness prevailed.
On cross examination, Delegate Jones testified that, during the 2011 redistricting process,
he would draw a map, and Oldham and Morgan “would look at the compactness and contiguity
scores” with Reock and Polsby-Popper measures. He said that he “was confident that the map
that I had drawn, the districts that were there that were contained within, met the criteria for
compactness and contiguity per” Wilkins and Jamerson.
Delegate Jones asserted that the criteria of compactness was “never subordinated,”
because the “top three criteria [the federal Constitution, Constitution of Virginia, and the VRA]
always were met within the range established by Wilkins.” He stated that “my assumption is that
when we ran the plans, that if a score was better than [the score] that was affirmed by the
Supreme Court [in Jamerson and Wilkins], then we would . . . be in a good state,” but he “did not
12
bother myself with that information. I was more drawing the districts.” The court clarified that
Delegate Jones was “assuming there was some numerical score from those compactness tests,
and you just relied on counsel to tell you that you were meeting them.”
Morgan testified that he worked with Delegate Jones in the 2011 redistricting using the
House Criteria Resolution. He stated that compactness was “implemented” in the 2011 Plan
because they ran “compactness reports” using the Reock and Polsby-Popper measures, and the
compactness scores “were within the acceptable range as defined by the Jamerson and Wilkins
cases.” He stated that compactness was not ignored or given pro forma consideration, but that it
was “looked at during the process” in 2011.
Dr. Thomas B. Hofeller (Dr. Hofeller) testified as an expert in redistricting who had been
in the fields of redistricting and census information for nearly 50 years, and had drawn or
advised the drawing of hundreds of redistricting maps in “almost all the states.” He compiled the
minimum and maximum compactness scores for all of the districts in Virginia’s 1991, 2001, and
2011 plans, and stated that “Jamerson and Wilkins spoke to . . . the minimum scores of the
districts that [] complied with the compactness criteria in the state constitution. So it forms a
floor. You can characterize it as a floor.”
Dr. Hofeller stated that he did not think that Dr. McDonald’s test is “a proper and useful
way to determine compliance” with the compactness requirement. He stated that Dr.
McDonald’s test was done six years after the redistricting, so his report “could have never related
to the processes of one [redistricting] unfolding.” He also opined that Dr. McDonald “kind of
snatched complexity out of the jaws of simplicity here with regard to what the General Assembly
was trying to do,” which was to “pick[] out a standard by which they felt they were compliant
with the constitutional requirement for compactness, and that was the floor that was established
13
in Jamerson and Wilkins.” He noted that Dr. McDonald’s test was not peer reviewed. He also
stated: “I think that the elimination of the minority districts is inappropriate,” because there was
a “rather subjective treatment” by Dr. McDonald to draw the alternative districts around the
existing VRA districts.
On cross examination, Dr. Hofeller acknowledged that Alternative Plans 1 and 2 were not
presented as plans that the legislature should adopt. He acknowledged that Dr. McDonald did
not purport to create a bright line test, and that neither Dr. McDonald nor anyone else had
created a bright line score for compactness. He agreed that the constitutions trump “[w]hatever
else the legislature . . . chooses to add to that [which] is within its discretion.”
After the close of the evidence, the court overruled a previously filed motion in limine
regarding Dr. McDonald’s testimony. 8 The circuit court stated that it would consider the
predominance test and Dr. McDonald’s conclusions. The court noted that it disagreed with Dr.
Hofeller’s assertion “that the Supreme Court of Virginia has ever established a constitutionally
required minimum compactness score for measuring the priority given to compactness in
drawing legislative districts.” The court also stated that the standard expressed in Wilkins and
Jamerson “requires that if the evidence offered by both sides of the case would lead reasonable
and objective persons to reach different conclusions, then the legislative determination is ‘fairly
debatable’ and must be upheld.” The circuit court determined that the legislative determination
regarding the compactness of the Challenged Districts was fairly debatable and upheld the
constitutional validity of the Challenged Districts.
8
The ruling of the circuit court on the motion in limine to exclude Dr. McDonald’s
testimony is the subject of an assignment of cross-error filed in this appeal. However, we need
not address that assignment of cross-error in resolving this case.
14
ANALYSIS
The Challengers assert that the circuit court erred in finding that the evidence presented
to it was sufficient to support its conclusion that the constitutionality of the General Assembly’s
redistricting decision regarding each of the Challenged Districts was fairly debatable. In other
words, the Challengers claim that the evidence they presented to the circuit court established that
the General Assembly subordinated the required criterion of compactness to discretionary
criteria, and that such evidence proved that no objective and reasonable person could conclude
that the Challenged Districts were not in violation of the compactness requirement of Article II,
§ 6 of the Constitution of Virginia. We disagree.
Article II, § 6 of the Constitution of Virginia provides that electoral districts are
“established by the General Assembly.” The exercise of that power clearly encompasses the
drawing of legislative districts. As recited above, the districts in the General Assembly’s 2011
Plan were codified in Code § 24.2-303.3 and Code § 24.2-304.03.
Established principles govern our determination whether the General Assembly has
adhered to the Constitution in exercising its legislative power. Redistricting legislation is subject
to “the principles applicable to our review of legislative determinations.” Wilkins, 264 Va. at
462, 571 S.E.2d at 108. The first such principle is that every law enacted by the General
Assembly “is entitled to a strong presumption of validity and will be invalidated by the courts
only if it clearly violates a constitutional provision.” Id. (citation and internal quotation marks
omitted); City of Newport News v. Elizabeth City Cty., 189 Va. 825, 839, 55 S.E.2d 56, 64
(1949).
The party challenging an enactment has the burden of proving that the statute is
unconstitutional, and every reasonable doubt regarding the constitutionality of a legislative
15
enactment must be resolved in favor of its validity. Marshall v. Northern Va. Transp. Authority,
275 Va. 419, 428, 657 S.E.2d 71, 75 (2008); Hess v. Snyder Hunt Corp., 240 Va. 49, 53, 392
S.E.2d 817, 820 (1990). See Blue Cross of Virginia v. Commonwealth, 221 Va. 349, 358-59, 269
S.E.2d 827, 832-33 (1980); see also In re Phillips, 265 Va. 81, 85-86, 574 S.E.2d 270, 272
(2003). Further, in determining the constitutionality of a General Assembly enactment, we must
give the Constitution a liberal construction in favor of sustaining the enactment in question, if
practicable. Marshall, 275 Va. at 428, 657 S.E.2d at 75; Heublein, Inc. v. Department of
Alcoholic Beverage Control, 237 Va. 192, 195, 376 S.E.2d 77, 78 (1989).
Here, we are only concerned with the compactness requirement stated in Article II, § 6 of
the Constitution of Virginia. Article II, § 6 of the Constitution of Virginia provides that
legislative districts “shall be composed of contiguous and compact territory.” The Challengers
allege that the Challenged Districts were clearly drawn in violation of the compactness
requirement in Article II, § 6.
Compactness is a somewhat abstract concept. Determining compactness does not admit
to a bright line approach in determining whether a legislative district is, in fact, compact. As
attested to by the Challengers’ expert, Dr. McDonald, social scientists have developed at least 50
different methods of measuring compactness. The proliferation of measures does not provide
clarity. It does exactly the opposite. In the social science community, there is no universal
definition or consensus as to what constitutes the best measure for compact districts.
In addition to statistical compactness measures, the record demonstrates that
policymakers and citizens define compactness in a variety of ways. Delegate Jones testified to
his spatial and political assessment of compactness, which included consideration of geographic
size, ease of travel and representation, and communities of interest. Also, in vetoing the initial
16
redistricting bill (H.B. 5001), Governor McDonnell linked compactness not only to statistical
compactness scores but also to the number of localities split between districts and to “a plain
visual examination of the districts.”
In Wilkins and Jamerson, we have held that courts must “give proper deference to the
wide discretion accorded the General Assembly in its value judgment of the relative degree of
compactness required when reconciling the multiple concerns of [redistricting].” Jamerson, 244
Va. at 517, 423 S.E.2d at 186. We have stated that redistricting is, in a sense, political. Id. at
510, 423 S.E.2d at 182. Thus, giving the General Assembly the opportunity to determine the
relative level of compactness in the first instance places the decision, of what degree of
compactness best serves local needs, with the institution best informed to make that decision.
Although whether a voting district is “compact” is justiciable, it is a question of “fact”
triggering the “fairly debatable” standard. See Jamerson, 244 Va. at 509-10, 423 S.E.2d at 182.
When the constitutionality of a statute depends on facts, the determination of
those facts by the legislature can be set aside if it is clearly erroneous, arbitrary, or
wholly unwarranted. If the evidence offered in support of the facts in issue would
lead objective and reasonable persons to reach different conclusions, the
legislative determination is considered fairly debatable and such a determination
must be upheld by the courts.
Wilkins, 264 Va. at 462, 571 S.E.2d at 108 (emphasis added).
Specifically,
if the validity of the legislature’s reconciliation of various criteria is fairly
debatable and not clearly erroneous, arbitrary, or wholly unwarranted, neither the
court below nor this Court can conclude that the resulting electoral district fails to
comply with the compactness and contiguous requirements of Article II, § 6.
Id. at 463, 571 S.E.2d at 108.
Accordingly, our decisions in Jamerson and Wilkins do not vest the judiciary with
authority to establish a standard of compactness in the first instance; rather, our role is to “review
17
legislative determinations of fact” regarding compactness to ensure that they are not “clearly
erroneous, arbitrary, or wholly unwarranted.” Jamerson, 244 Va. at 510, 423 S.E.2d at 182.
Thus, our precedent requires our courts to give the General Assembly deference and wide
discretion in determining compactness of legislative districts. However, the judiciary has the
authority and obligation to intervene when an abuse of that discretion is shown by a grave,
palpable, and unreasonable deviation from the principles fixed by the Constitution. Id. at 510,
423 S.E.2d at 182. In considering the validity of a redistricting statute, we have said that “when
fair-minded men[,] from an examination of the [redistricting,] can entertain no reasonable doubt
that there is a grave, unnecessary, and unreasonable [error], the Constitution has been violated
and it is the duty of the court to so declare.” Id. (citation, alteration and internal quotation marks
omitted).
In addition, as an appellate court, we also “must give proper deference” to a circuit
court’s resolution of disputed facts, if supported by credible evidence. We must consider the
evidence and all reasonable inferences that can be drawn therefrom in the light most favorable to
the parties who prevailed before the circuit court. Jamerson, 244 Va. at 517, 423 S.E.2d at 186.
The Challengers had the initial burden to show “probative evidence of
unreasonableness,” and the legislature could respond to that showing with evidence that its
actions were reasonable. See Board of Supervisors v. Jackson, 221 Va. 328, 333, 269 S.E.2d
381, 384-85 (1980). At that point, “[i]f such evidence of reasonableness is sufficient to make the
issue fairly debatable, the legislative action must be sustained.” Id. at 333, 269 S.E.2d at 385.
Thus, the key question that must be answered in resolving this appeal is simply whether the
legislature’s determination, that the Challenged Districts are in fact compact, is fairly debatable.
If so, the circuit court’s judgment must be affirmed.
18
In support of their claim, the Challengers presented Dr. McDonald’s expert testimony.
He testified about a predominance test that he devised for this litigation. Applying his test, Dr.
McDonald concluded that the composite district compactness scores (using the Reock and
Polsby-Popper measures) of each of the Challenged Districts had been “degraded” by more than
50 percent from the compactness score of an alternative district drawn to be as compact as
possible, without consideration of any discretionary criteria. Based upon the results of his test,
he opined that the General Assembly subordinated the constitutional requirement that the
districts be compact to other discretionary considerations in drawing the Challenged Districts.
The Challengers rely upon that opinion as their evidence that the Challenged Districts are clearly
in violation of the compactness requirement expressed in Article II, § 6.
The Challengers’ theory that Dr. McDonald’s testimony clearly proves there was a
violation of the compactness requirement in Article II, § 6 faces several challenges. Notably, it
is based upon the measure of a requirement, predominance, that is not found in our Constitution,
and it does not account for the discretionary factors the legislature balances when drawing
districts; it relies on measurements (statistical scores) that do not have a universally accepted
meaning; and the test devised by Dr. McDonald is novel and unproven.
First, Dr. McDonald’s approach assumes that evaluating compactness involves
considering whether discretionary factors predominated over other factors. This assumption has
no support in the text of the Constitution nor from precedents in Virginia law. To the contrary,
our Court has previously stated that “any purpose that may underlie the design of an electoral
district, while relevant to challenges under other portions of the Constitution of Virginia . . . is
not determinative in a challenge based on Article II, § 6.” Wilkins, 264 Va. at 462, 571 S.E.2d at
108. Instead, “[d]eterminations of contiguity and compactness, as we said in Jamerson, are
19
limited to consideration of the district from a spatial perspective, taking into consideration the
other facts that a legislative body must balance in designing a district.” Wilkins, 264 Va. at 462,
571 S.E.2d at 108. The proper inquiry is whether the resultant district is, in fact, compact. In
making that determination, our precedent has recognized that “traditional [discretionary]
redistricting elements not contained in the statute . . . are also legitimate legislative
considerations.” Id. at 464, 571 S.E.2d at 109. Where there is evidence that the General
Assembly considered the constitutional requirement of compactness in reconciling the different
demands upon it in drawing legislative districts, deference is given to the General Assembly in
its value judgment as to the relative degree of compactness required. Jamerson, 244 Va. at 517,
423 S.E.2d at 186.
It is also important to note that, contrary to a core premise in Dr. McDonald’s test, the
Constitution of Virginia does not require districts to be as compact as possible. 9 Article II, § 6
simply requires that districts “shall be . . . compact.” See, e.g., Wilkins, 264 Va. at 466, 571
S.E.2d at 110 (observing that “nothing in this record indicates that the District is repugnant to the
constitutional principles of compact and contiguous electoral districts,” even though the district
was “far from the most compact district”). As the Board observed, the compactness requirement
is not based on the same type of objective comparative inquiry as the equal population
requirement.
9
Unlike Virginia’s general requirement that districts be “compact,” other states provide
specific standards for compactness, either by controlling the process of redistricting or the maps
drawn as a result. For example, in Arizona, an independent redistricting commission begins the
mapping process by creating “districts of equal population in a grid-like pattern across the state,”
and then making adjustments “as necessary to accommodate” enumerated goals, including that
districts “shall be geographically compact and contiguous to the extent practicable.” Ariz. Const.
art. IV, pt. 2, § 1. Colorado requires that “[e]ach district shall be as compact in area as possible
and the aggregate linear distance of all district boundaries shall be as short as possible.” Colo.
Const. art. V, § 47.
20
Second, Dr. McDonald’s approach depended solely on calculations of relative
compactness using several common compactness score measures. It is undisputed that there is
no accepted bright-line test or score in the social sciences for when a district can no longer be
considered “compact.” Dr. McDonald acknowledged that there is “no universal definition of
what constitutes the best measure for compact districts,” and that “[i]n the social science
community, there’s no threshold for when a district is compact or noncompact.” Our decisions
in Jamerson and Wilkins do not purport to adopt a bright-line standard; indeed, those opinions do
not even recount the compactness scores of the districts upheld therein. However, the record
indicates that the districts challenged in this case all have Reock scores exceeding the lowest
score for a district upheld in Jamerson and within 0.01 of the lowest score upheld in Wilkins.
Nine of the Challenged Districts have Polsby-Popper scores greater than or equal to the lowest
score for districts upheld in Jamerson and Wilkins. Two of the Challenged Districts (HD 72 and
SD 28) have Polsby-Popper scores which are 0.02 below the lowest score for a district upheld in
those cases. However, Dr. McDonald agreed that a 0.02 difference is not “a large difference”
and generally is not meaningful. Thus, compactness scores for the Challenged Districts were at
least approximately the same or greater than the scores of districts previously found to be
constitutionally compact.
Third, Dr. McDonald’s predominance test was admittedly novel and created for this case.
He conceded that the predominance test had not previously been used or published, and he was
not aware of any other expert, court opinion or state using the “predominance test” that he used
to develop degradation percentages regarding compactness of the legislative districts. Dr.
McDonald admitted that the entire question the new test purports to answer—whether other
redistricting criteria predominated over compactness—“is not a question that I’ve ever seen in
21
any legal or scholarly writings that are out there.” Consequently, the predominance test has no
supporting basis in social science scholarship or redistricting case law.
As noted above, the standard applicable at trial is highly deferential to the government,
even after a challenger meets his prima facie burden. 10 The government need only present
“some evidence of reasonableness,” and the evidence need not establish the government’s
position by a “preponderance of the evidence;” it is enough if “the evidence offered in support of
the opposing views would lead objective and reasonable persons to reach different conclusions.”
Ames v. Town of Painter, 239 Va. 343, 348, 389 S.E.2d 702, 704 (1990) (citation and internal
quotation marks omitted).
The Board and the House countered the Challengers’ case with extensive evidence that
the Challenged Districts were reasonable, including testimony and statements from the
individuals who crafted the 2011 Plan and experts who disagreed with and challenged Dr.
McDonald’s opinions regarding compactness. For purposes of this appeal, that evidence is
viewed in the light most favorable to the Board and the House, the prevailing parties at trial.
The evidence produced by the House and the Board showed that compactness was
considered by the General Assembly during the redistricting. Both the House and the Senate
adopted criteria for the 2011 redistricting that included “compact and contiguous” districts under
10
The Challengers assert that the circuit court found their prima facie burden of proof
was met based upon the court’s denial of the House and Board’s motions to strike. However, at
that stage of the trial, the Challengers enjoyed “all reasonable inferences drawn from the
evidence in the light most favorable” to them. Izadpanah v. Boeing Joint Venture, 243 Va. 81,
82, 412 S.E.2d 708, 709 (1992). The presumption in the Challengers’ favor, however, does not
carry beyond that procedural stage, and they bore the burden at trial with no benefit of the doubt.
Thus, prevailing on the motions to strike did not mean that the Challengers carried any burden
relevant to the court’s decision after trial on the merits. See Kiddell v. Labowitz, 284 Va. 611,
621 n.3, 733 S.E.2d 622, 627 n.3 (2012) (explaining that denial of a motion to strike was not a
finding that the non-moving party met its substantive burden).
22
Article II, § 6 and consistent with Wilkins and Jamerson. Several of the floor speeches and other
exhibits introduced into the record mentioned compactness. Morgan and Delegate Jones testified
that they checked the compactness scores of districts as they were creating the 2011 Plan against
scores of the districts approved in Wilkins and Jamerson. At trial, the defense presented
testimony from two experts about the 2011 redistricting and compactness—Dr. Hood regarding
the Senate districts, and Dr. Hofeller regarding the House districts. Both experts agreed that all
of the Challenged Districts were sufficiently compact.
Both experts also cast doubt upon the validity of Dr. McDonald’s predominance test. Dr.
Hofeller testified that the predominance test was not a proper and useful way to assess
compliance with the compactness requirement. He faulted the test’s lack of foundation, noting
that it lacked peer review, exposure, supporting research, and use during an actual redistricting.
Dr. Hofeller also testified that the low level of overlap between the enacted and alternative
districts meant the new test compared apples to oranges. He questioned the complexity added by
the new test and whether it could be applied within Virginia’s compressed redistricting
timeframe. Lastly, he opined that excluding VRA districts from its application cast doubt on the
validity of the predominance test, because there “needs to be one redistricting standard for all the
districts.”
Dr. Hood opined that any compactness test should be able to be applied to all the districts
in a plan, and noted that the Challengers’ new test admittedly could not be applied to
approximately 50 percent of the Senate districts, because those districts were either governed by
the VRA or bordered VRA districts. Dr. Hood also opined that the new test’s averaging of
23
different compactness scores for a district was not a typical or recommended practice. Indeed,
Dr. McDonald admitted that he could not recall having averaged a district’s compactness scores
before, and admitted that averaging could mask results on particular scores.
The circuit court accepted all the evidence from both sides and, after weighing it,
concluded that reasonable minds would disagree as to whether the Challenged Districts are
sufficiently compact. Taken together, the evidence presented by the Challengers and that
presented on behalf of the House and the Board was sufficient to establish that the constitutional
validity of the Challenged Districts under the compactness requirement was fairly debatable.
The circuit court was not required to accept the opinion of Dr. McDonald as being true or
accurate. Also, even if the circuit court accepted the correctness of Dr. McDonald’s opinion that
his test measured the effect of discretionary criteria upon compactness, the circuit court was not
bound to accept the Challengers’ theory that proof that the General Assembly subordinated the
criterion of compactness to discretionary criteria per se resulted in a clear violation of the
compactness requirement expressed in Article II, § 6.
The test that Dr. McDonald based his opinion upon was created for the purpose of this
litigation and had never been used by anyone before to measure “predominance” and
“degradation.” It is based upon measures of compactness and creates a rule regarding it although
the creator of the test concedes that there is no consensus on how to measure compactness and
there is no bright line demonstrating when a district is compact and when it is not. Further, it is
undisputed that the compactness of the Challenged Districts is approximately equivalent to or
better than that of districts upheld in the past, as measured by widely used compactness scores.
Moreover, the Challengers offered no direct evidence that the legislature ignored compactness,
24
or any other evidence, other than Dr. McDonald’s opinion based upon statistical compactness
scores, as a basis for their assertion that the Challenged Districts violated the compactness
requirement.
Further, there was evidence that the legislature attempted to reconcile discretionary
considerations with the acknowledged requirement in Article II, § 6 that the districts be compact.
The legislature’s efforts to ensure that the compactness scores of the districts in the 2011 Plan
remained largely at or above the scores of the districts approved by this Court in Jamerson and
Wilkins, while not dispositive, are evidence that could lead reasonable people to find that the
Challenged Districts were constitutionally compact. Even if the districts in the 2011 Plan were
not as compact as they could have been, as we have explained above, that is not what our
Constitution requires. Our Constitution speaks to the result of the redistricting process, and
mandates that districts be compact in the end. It does not attempt to curtail the legislative
process that creates the end result. Nor does it require that compactness be given priority over
other considerations, much less establish a standard to determine whether the legislature gave
proper priority to compactness. Thus, there is evidence to support the ruling that the
determination of the General Assembly regarding compactness of the Challenged Districts is
fairly debatable, and not clearly erroneous, arbitrary, or wholly unwarranted, and we must
uphold the legislature’s decision to draw the Challenged Districts as it did.
CONCLUSION
The circuit court did not err in concluding that evidence was presented at trial that would
“lead reasonable and objective people to differ” regarding the compactness of the Challenged
Districts, and declaring the constitutional validity of the Challenged Districts under the fairly
25
debatable standard applied to determinations made by the legislature. Accordingly, we will
affirm the judgment of the circuit court.
Affirmed.
26