United States Court of Appeals
                          For the Eighth Circuit
                      ___________________________

                              No. 16-4511
                      ___________________________

 Missouri State Conference of the National Association for the Advancement of
       Colored People; Redditt Hudson; F. Willis Johnson; Doris Bailey

                     lllllllllllllllllllll Plaintiffs - Appellees

                                         v.

                     Ferguson-Florissant School District

                    lllllllllllllllllllll Defendant - Appellant

              St. Louis County Board of Election Commissioners

                           lllllllllllllllllllll Defendant

                           ------------------------------

                     Missouri School Boards’ Association

               lllllllllllllllllllllAmicus on Behalf of Appellant(s)
                                     ____________

                  Appeal from United States District Court
                for the Eastern District of Missouri - St. Louis
                                ____________

                        Submitted: December 13, 2017
                             Filed: July 3, 2018
                               ____________

Before SMITH, Chief Judge, ARNOLD and KELLY, Circuit Judges.
                                    ____________

KELLY, Circuit Judge.

       The Missouri State Conference of the National Association for the
Advancement of Colored People and the other named plaintiffs (together, the
NAACP) sued the Ferguson-Florissant School District (FFSD, or the district) for vote
dilution under section 2 of the Voting Rights Act of 1965 (VRA), 52 U.S.C § 10301.
After a bench trial, the district court1 found that (1) the NAACP had proved the
preconditions for a section 2 claim, and (2) the totality of the circumstances indicated
that the district’s black voters had less opportunity to elect their preferred candidate
than other members of the electorate. FFSD appeals both determinations.

                                           I.

       We begin with the underlying facts of the case, the legal framework applicable
to section 2 vote dilution claims, and the proceedings below.

                                       a. Facts

       FFSD is a school district located in northern St. Louis County, Missouri. It
was created after a 1975 desegregation order required the original FFSD to annex two
neighboring school districts “to achieve a meaningful desegregation” within one
unified district. United States v. Missouri, 

, 1366 (8th Cir. 1975) (en
banc). It includes all or part of eleven municipalities, and is governed by a
seven-member school board. Each member is elected “at-large” by the popular vote
of the entire district. Members of the FFSD board serve three-year terms with two or
three seats filled by an election every April.

      1
        The Honorable Rodney W. Sippel, Chief Judge, United States District Court
for the Eastern District of Missouri.

                                          -2-
       Voting in FFSD board elections works as follows. Assume there are three seats
to be filled in a given election. All of the candidates are listed on a single ballot.
Each voter gets three votes, one for each seat, and can cast those three votes for any
three candidates, but cannot vote for any candidate more than once. The top three
vote-getters assume seats on the board. In a two-seat year, the same procedures are
followed with two votes rather than three. The only time this procedure is not
followed is when the election is uncontested (i.e., there are the same number of
candidates as there are available seats). In that circumstance, the candidates simply
assume the positions on the board and the election is cancelled.

                                b. Legal Framework

       Section 2 of the VRA protects against the “denial or abridgement of the right
of any citizen of the United States to vote on account of race or color” in any election
held by a “State or a political subdivision.” Section 2 claims are “established if,
based on the totality of circumstances,” it is shown that members of a racial minority
group “have less opportunity than other members of the electorate to participate in
the political process and to elect representatives of their choice.” One manner in
which a violation may occur is when districts that elect several at-large
representatives “operate to impair blacks’ ability to elect representatives of their
choice.” Thornburg v. Gingles, 

, 42 (1986). Such a circumstance gives
rise to a “claim of vote dilution.” Id.; see also Buckanaga v. Sisseton Indep. Sch.
Dist., No. 54-5, S.D., 

, 471 (8th Cir. 1986) (“The legislative history of
the 1982 amendment to § 2 indicates that it was aimed particularly at discriminatory
at-large election systems which dilute minority voting strength.”).

       There are two steps to proving a section 2 vote dilution claim: (1) satisfying the
so-called “Gingles preconditions,” and (2) showing the violation based on a totality
of the circumstances. In the first step, plaintiffs are required to prove three
preconditions by a preponderance of the evidence:

                                          -3-
       (1) The racial group is sufficiently large and geographically compact to
       constitute a majority in a single-member district; (2) the racial group is
       politically cohesive; and (3) the majority votes sufficiently as a bloc to
       enable it to usually defeat the minority’s preferred candidate.

Bone Shirt v. Hazeltime, 

, 1018 (8th Cir. 2006) (cleaned up) (quoting
League of United Latin Am. Citizens v. Perry (LULAC), 

, 425 (2006)).
Satisfying these three preconditions “carries a plaintiff a long way towards showing
a Section 2 violation,” but does not suffice. Harvell v. Blytheville Sch. Dist. No. 5,

, 1390 (8th Cir. 1995) (en banc).

      In the second step, “the statutory test directs us to consider the ‘totality of the
circumstances’ to determine whether members of a racial group have less opportunity
than do other members of the electorate.” 

–26. The
Supreme Court has further instructed that, in applying this standard, we are to
consider a list of factors that were included in the Senate Report on the 1982
amendments to the VRA. 

These factors are:

      (1) “the history of voting-related discrimination in the State or political
      subdivision;”
      (2) “the extent to which voting in the elections of that State or political
      subdivision is racially polarized;”
      (3) “the extent to which the State or political subdivision has used voting
      practices or procedures that tend to enhance the opportunity for discrimination
      against the minority group, such as unusually large election districts, majority
      vote requirements, and prohibitions against bullet voting;”
      (4) “the exclusion of members of the minority group from candidate slating
      processes;”
      (5) “the extent to which minority group members bear the effects of past
      discrimination in areas such as education, employment, and health, which
      hinder their ability to participate effectively in the political process;”
      (6) “the use of overt or subtle racial appeals in political campaigns;”
      (7) “the extent to which members of the minority group have been elected to
      public office in the jurisdiction;”

                                          -4-
      (8) the extent to which “elected officials are unresponsive to the particularized
      needs of the members of the minority group;” and
      (9) whether “the policy underlying the State’s or political subdivision’s use of
      the contested practice or structure is tenuous.”

–45; see also 

; Bone 

–22. The Senate Report also “stresses . . . this list of typical factors is neither
comprehensive nor exclusive.” 

.

                                c. Proceedings Below

       The NAACP sued FFSD and the St. Louis County Board of Elections, alleging
that FFSD’s at-large elections violated section 2 of the VRA because the
elections—along with the historical and socioeconomic realities present in the
district—denied black residents a meaningful opportunity to elect representatives of
their choice. The NAACP proposed several remedies.

       The district court conducted a six-day bench trial. After the trial, it issued a
lengthy decision containing both extensive factual findings and thorough legal
analysis. The district court concluded that the NAACP satisfied the Gingles
preconditions and proved based on the totality of the circumstances that FFSD’s
election system provided black voters with less opportunity to elect candidates of
their choice compared to white voters.

                                           II.

       On appeal, FFSD contests the district court’s findings on both the Gingles
preconditions and totality-of-the-circumstances vote dilution. “We review the district
court’s factual findings for clear error, including the district court’s factual
determination of whether the Section 2 requirements are satisfied.” Bone 

(citing 

). “Legal questions and mixed questions

                                           -5-
of law and fact are . . . reviewed de novo.” 

71 F.3d at 1386); see
also 

(“Where ‘the ultimate finding of dilution’ is based on
‘a misreading of the governing law,’ however, there is reversible error.” (quoting
Johnson v. De Grandy, 

, 1022 (1994))).

                                         III.

      FFSD contests the district court’s conclusions regarding the first and third
Gingles preconditions. We address each argument in turn.

                         a. The First Gingles Precondition

       The first Gingles precondition requires that “the racial group is sufficiently
large and geographically compact to constitute a majority in a single-member
district.” 

(cleaned up). The district court relied on the 2010
Decennial Census to conclude that whites continued to outnumber blacks within
FFSD. The district court also held that, even if black voters comprised a bare
majority of the voting age population in the district, that would not prevent a vote-
dilution claim from going forward if the district’s black voters had been historically
discriminated against and disenfranchised. On appeal, FFSD contests both
determinations. FFSD also argues that the district court failed to consider whether
a revised election system would remedy vote dilution.

                               i. Use of Census Data

       FFSD argues the district court erred in relying on the 2010 Decennial Census
data instead of data it submitted from the American Community Survey (ACS). As
an initial matter—and, as the district court noted—“[t]he census is presumed accurate
until proven otherwise.” See McNeil v. Springfield Park Dist., 

, 946
(7th Cir. 1988); cf. 

& n.1 (citing figures from the 1980

                                         -6-
census). FFSD asserts that it successfully overcame this presumption based on the
ACS and the testimony of its expert. The ACS is a sample survey that seeks to
provide up-to-date estimates because the Decennial Census (as is implied by its
name) is only conducted every ten years. The 2011–2013 version of the ACS
suggested that the black voting age population within FFSD was approximately
50.3%.2 The district’s expert, Dr. Jonathan Rodden, also created projections of the
black voting age population in FFSD based on historical trends within the district.
These projections were based on the ACS. According to FFSD, the ACS and Dr.
Rodden’s projections established that black voters made up a majority of the district.

       The district court found, however, that changes in the ACS did not show, with
any reliable degree of certainty, that the population within the district had shifted
since the 2010 Census. Further, the district court noted that “[t]he Census Bureau
itself cautions against using ACS estimates rather than the Decennial Census
complete count to determine the population of a given geographic area,” and cited,
as an example, the fact that the ACS had projected that the overall population of St.
Louis would grow throughout the 2000s, only to be disproved when the actual data
for the 2010 Census were collected. In light of these methodological limitations, the
district court determined that the most reliable data available were the 2010 Census
data. We detect no clear error in that determination.

                            ii. Bare Numerical Majority

      The district court also concluded that, “even if [it] were to find that African
Americans constitute a majority of the District’s [voting age population], that would
not be the end of the analysis.” The district court explained that “[r]acial minorities

      2
       This number comes from the district court’s “corrected” estimate of “any-part
black” voting age population that extrapolates from the ACS, which only provides an
estimate of the “single-race black” population and does not give an estimate of the
percentage of “any-part black” population, unlike the Decennial Census itself.

                                         -7-
do not suddenly lose the broad protections of the VRA at the moment that they
surpass 50% of a jurisdiction’s [voting age population].” FFSD disagrees, asserting
that only a numerical minority enjoys the protections of the VRA.

       As Gingles notes, under the VRA, the term “minority” does not refer to a
purely numerical fact. Rather, section 2(a) protects the voting rights of “any citizen
who is a member of a protected class of racial or language 

; see also Salas v. Sw. Tex. Jr. Coll. Dist., 

, 1547 (5th Cir. 1992)
(“The plain text of the statute, as affirmed by case law, makes clear that the Act is
concerned with protecting the minority in its capacity as a national racial or language
group.”); 

(“The Voting Rights Act was passed in 1965 to effectuate the
guarantees of the Fifteenth Amendment.”). And the Supreme Court has affirmatively
stated that “it may be possible for a citizen voting-age majority to lack real electoral
opportunity,” and thereby require the protection of section 2. 

.

       Nonetheless, FFSD argues for a rule that a racial minority cannot prevail on a
section 2 claim when it constitutes a bare numerical majority within the district. To
support this per se rule, the district relies on Smith v. Brunswick Cty. Bd. of
Supervisors, 

(4th Cir. 1993). Even if Smith can be read as adopting
a bright-line rule, however, the weight of authority is to the contrary. The Second,
Fifth, Eleventh, and D.C. Circuits all have declined to adopt the rule the district
advocates. See Pope v. Cty. of Albany, 

, 575 n.8 (2d Cir. 2012) (“[T]he
law allows plaintiffs to challenge legislatively created bare majority-minority districts
on the ground that they do not present the ‘real electoral opportunity’ protected by
Section 2.” (quoting 

)); Kingman Park Civic Ass’n v.
Williams, 

, 1041 (D.C. Cir. 2003) (“Vote dilution claims must be
assessed in light of the demographic and political context, and it is conceivable that
minority voters might have less opportunity to elect representatives of their choice
even where they remain an absolute majority in a contested voting district.” (cleaned

                                          -8-
up)); 

(“Unimpeachable authority from our circuit has rejected
any per se rule that a racial minority that is a majority of a political subdivision
cannot experience vote dilution.” (quoting Monroe v. City of Woodville, 

, 1333 (5th Cir, 1989))); Meek v. Metro. Dade Cty., 

, 1546 (11th
Cir. 1990) (reaffirming, post circuit reorganization, Zimmer v. McKeithen, 

, 1300 (5th Cir. 1973), which had held “infirm” the conclusion “that an at-large
scheme cannot work a dilution of black voting strength where blacks . . . comprise a
majority of the total population of the parish”).

       The section 2 inquiry is one that “requires an ‘intensely local appraisal’ of the
challenged district.” 

(quoting 

). We
agree with the majority of our sister circuits that such an appraisal should take into
account the history of discrimination and disenfranchisement in the district, and the
way that those historical problems may still affect the district’s political landscape.
The district’s proposed per se rule would mean that any section 2 claim would be
defeated the moment black voters made up a bare numerical majority of the district,
regardless of whether minority voters in that district still face actual impediments and
disadvantages. Such a rule does not comport with the VRA’s substantive requirement
that racial minorities have equal opportunity “to participate in the political process
and to elect representatives of their choice.” 52 U.S.C. § 10301(b); see also 

. In short, minority voters do not lose VRA protection simply because
they represent a bare numerical majority within the district.

                         iii. History of Disenfranchisement

      FFSD argues that the district court erred in using statewide data about felony
disenfranchisement, voter registration, and home ownership to conclude that the
voting age population statistics likely overstated the influence that black voters had
within the district. Although neither party presented FFSD-specific data as to felony
disenfranchisement or voter registration, NAACP submitted evidence about how

                                          -9-
these issues disproportionately affect black Missourians. Particularly as to felony
disenfranchisement, the district court used the statewide data as a starting point, and
then credited and relied on expert testimony about how the high criminal conviction
rate of black FFSD residents likely led to a high rate of disenfranchisement in the
district. The court also noted that home ownership, for which FFSD-specific data
were presented, “is a strong predictor of voting in local elections.” Whenever
possible, of course, parties must provide the court with the most location-specific data
available. But we discern no clear error in this case because the district court still
conducted an “intensely local appraisal,” even if part of the appraisal included
statewide statistics.

                             iv. Remedial Consideration

       The Supreme Court has described the Gingles preconditions as “threshold
conditions” that must be established before liability can be assessed. And, only after
“a Section 2 violation is found” should a district court turn to the task of “developing
a constitutional remedy.” Bone 

. Nevertheless, FFSD takes
the position that the first Gingles precondition cannot be satisfied if the single-
member districts that the NAACP proposed would give black voters no greater
chance to elect a candidate of their choice than the current at-large system. But as we
have said, “at the initial stage of the Gingles precondition analysis, the plaintiffs are
only required to produce a potentially viable and stable solution.” 

(emphasis in original). Of course, the district court must still determine that the
proposed solution demonstrates that the minority group is “sufficiently large and
geographically compact to constitute a majority of a single member district,” or, in
other words, that it is “potentially viable and stable.” However, at this stage of the
proceedings, NAACP is not required to proffer the best option for remedying the
asserted violation. See 

district court correctly noted, the Gingles
preconditions are designed to establish liability, and not a remedy.”); see also

n.17 (explaining that the purpose for this requirement is that,

                                          -10-
“[u]nless minority voters possess the potential to elect representatives in the absence
of the challenged structure or practice, they cannot claim to have been injured by that
structure or practice.”). The first precondition “seeks to establish whether a workable
solution is possible,” Bone Shirt at 1019, and is met when “the minority group [is]
able to demonstrate that it is sufficiently large and geographically compact to
constitute a majority in a single member district.” 

. To the
extent FFSD complains that the proper analysis was lacking in this case, we further
note that the district court has since approved a remedial plan—a plan that FFSD does
not challenge. Under these circumstances, we discern no clear error in the district
court’s finding that the first Gingles precondition was satisfied.

                        b. The Third Gingles Precondition

        The third Gingles precondition requires that “the majority votes sufficiently as
a bloc to enable it usually to defeat the minority’s preferred candidate.” 

(cleaned up). First, FFSD argues, the district court erred in concluding
that it was possible for black voter preferences to be “subsumed” by white majority
bloc voting because the racial groups are close to numerical parity within the district.
Second, it asserts that the district court failed to consider certain special
circumstances and narrow margins of victory that might weigh in its favor. And,
third, FFSD claims that the district court put too much weight on the special
circumstances that it found attendant to the 2014 and 2015 elections.3




      3
       FFSD asserts that these errors were legal—which would require de novo
review—but their true disagreement is with the district court’s factual findings and
not the legal standard that it applied.

                                         -11-
                                i. White Bloc Voting

       The district court made extensive findings in support of its conclusion that
white voters in FFSD tended to vote cohesively and that their preferences were
sufficiently uniform and distinct from those of black voters (i.e., they voted as a
“bloc”) such that they were able “usually to defeat the minority’s preferred
candidate.” FFSD argues, however, that the district’s demographics make it
“impossible” for white bloc voting to defeat a black-preferred candidate because
black voters have slight numeric superiority in the district. To the extent this
argument is merely a reframing of FFSD’s position concerning the racial makeup of
the voting age population in the district, we again find it unpersuasive.

       Reviewing the district court’s analysis as a whole, we see no clear error. The
district court looked at all 12 of the contested4 elections held between 2000 and 2015,
with a particular focus on the more recent elections between 2011 and 2015. It
assessed the expert testimony about those elections from both Dr. Rodden, the
district’s expert, and Dr. Richard Engstrom, the NAACP’s expert. This testimony
aided the district court in determining which candidates were black-preferred, which
candidates were supported by the white bloc, and whether white bloc voting has
successfully defeated most black-preferred candidates. After analyzing the experts’
methodological approaches and the completeness of their analyses, the district court
credited Dr. Engstrom’s testimony because his analysis was “accurate, complete, and
reliable.” It relied on some of Dr. Rodden’s findings and testimony as well, but noted
that they had “limitations” and that the NAACP expert’s methods were “superior to
the other methods proposed.”


      4
       The 2005, 2007, 2008, and 2010 elections were uncontested and the
candidates therefore took the available seats without an election. The uncontested
election of black-preferred candidates “reveal[s] little about either minority cohesion
or white bloc voting.” Uno v. City of Holyoke, 

, 988 (1st Cir. 1995).

                                         -12-
       The district conducted a thorough analysis of the expert testimony, considering
each election individually. It concluded that most of the black candidates over both
periods of time had been defeated. There was no clear error in the district court’s
factual findings, which were supported by careful reasoning and analysis, regarding
the prevalence of white bloc voting and its effect on election results.5

                      ii. Unconsidered Special Circumstances

        The third Gingles precondition also requires that the district court analyze the
“special circumstances” that attend elections to make sure that there are no non-racial
factors at play that would appear to either defeat or demonstrate a section 2 

, 54, 57. In Gingles, the Court discussed special circumstances in the
context of explaining why “the success of a minority candidate in a particular election
does not necessarily prove that the district did not experience polarized voting in that
election.” 

The Court explained, “special circumstances, such as the
absence of an opponent, incumbency, or the utilization of bullet voting, may explain
minority electoral success in a polarized contest.” Id.6 It further cautioned that “[t]his
list of special circumstances is illustrative not exclusive.” 

n.26.

       FFSD argues that the district court failed to consider special circumstances
relating to the 2011 and 2013 elections. In the 2011 election, both of the black-
preferred candidates were defeated. In 2013, the sole black-preferred candidate was
defeated. FFSD argues that these were only narrow defeats, that black voters did not

      5
       We also note that the district court concluded that “even under [FFSD]’s
approaches,” plaintiffs had satisfied the third Gingles precondition. Because we
identify no clear error in the district court’s fact-findings, we need not address this
conclusion.
      6
        Bullet voting is when a voter casts their ballot with one vote for the preferred
candidate, but leaves the other votes blank in an effort not to dilute the strength of
their vote for the preferred candidate.

                                          -13-
allocate their votes efficiently, and that there were other special circumstances. The
other special circumstance that FFSD focuses on was a controversial vote by the
FFSD school board, just prior to the 2011 election, wherein the board voted
unanimously to give the former superintendent of the district (and his wife) a lifetime
health insurance policy. FFSD argues that it was this board vote, and not racial bloc
voting, that was responsible for the election results in 2011 and 2013.

       As an initial matter, it is not clear how much weight should be placed on the
narrowness of the black-preferred candidates’ defeats. Narrow losses could indicate
no meaningful vote dilution at all. Or such losses may signal particularly targeted and
effective vote dilution efforts. Of course, “the ultimate right of § 2 is equality of
opportunity, not a guarantee of electoral success for minority-preferred candidates of
whatever race.” Johnson v. De Grandy, 

, 1014 n.11 (1994). But here,
the narrowness of the losses does not change the fact that these candidates were
defeated and black voters were, therefore, not represented by their preferred candidate
on the board. We find no clear error in how the district court factored the close
defeats into its analysis. And regarding the alleged inefficient allocation of votes,
black voters are not required to vote as a “monolith” in order to make out a vote
dilution claim. See Sanchez v. Colorado, 

, 1319 (10th Cir. 1996); cf.
Gomez v. City of Watsonville, 

, 1416 (9th Cir. 1988) (admonishing the
district court for speculating about how the minority voting population could have
been more cohesive).

       As to the controversy surrounding the 2011 health insurance vote, the district
court was thorough in its discussion of the elections that followed that vote. And its
analysis of the multiple factors that were relevant in those elections is persuasive.
Incumbents who cast the purportedly unpopular health insurance vote came up for
reelection in 2011, 2012, and 2013, due to the staggered election system. In 2011,
only one of the black-preferred candidates running for election was an incumbent
(who had voted for the health insurance policy), yet both black-preferred candidates

                                         -14-
were defeated, receiving nearly identical percentages of the vote.7 And the black
incumbent would have been classified as black-preferred under either Dr. Engstrom’s
or Dr. Rodden’s estimates. In 2012, there was only one black-preferred candidate.
She received near-unanimous support from black voters in the district, but
nonetheless lost to a white incumbent (who had voted for the health insurance policy).
In 2013, there was one black-preferred incumbent and one white incumbent seeking
reelection (both of whom had voted for the health insurance policy), but the white
incumbent was successful, while the black-preferred incumbent was defeated. This
evidence shows that this health insurance vote—regardless of its perceived popularity
or unpopularity—likely did not affect incumbent reelection across racial lines.
Therefore, any error on the part of the district court for not directly addressing this
issue in its analysis was harmless.

                       iii. Considered Special Circumstances

       FFSD next objects to the special circumstances that the district court did
consider, which relate to the 2014 and 2015 elections. In 2014, there were three
black-preferred candidates who ran together as a slate called “Grade A for Change.”
These candidates were motivated to run, in part, due to the board’s treatment of Dr.
Art McCoy, the first black superintendent of the district, who had recently resigned.
This issue also served to motivate black voters. In 2015, the board election followed
on the heels of the death of Michael Brown, which had led to large demonstrations,
political unrest, and national attention on the Ferguson area.8 The district court

      7
       A white incumbent (who had voted for the health insurance policy) was also
defeated in 2011, but no evidence about other possible reasons for that loss was
presented to the district court.
      8
       Additionally, the one black-preferred candidate who was ultimately successful
explicitly encouraged voters to engage in “bullet voting.” Bullet voting is one of the
special circumstances specifically noted by the Court in 

; see


                                         -15-
concluded that special circumstances were present in the 2014 and 2015 elections, but
were not overwhelming, and only “slightly” discounted the election of one black-
preferred candidate in each of those elections. It is apparent from the record that the
district court considered the evidence in context and did not clearly err in deciding
how much weight to give those special circumstances.

      In sum, we find no clear error in the district court’s determination that the third
Gingles precondition was satisfied.

                                          IV.

       Once the three preconditions have been satisfied, plaintiffs must still show that
the “totality of the circumstances” demonstrates a section 2 violation. More
particularly, NAACP “ultimately must prove that the totality of the circumstances
indicates minority voters had ‘less opportunity than other members of the electorate
to participate in the political process and elect representatives of their choice.” Bone

(quoting section 2 of the VRA). FFSD challenges the manner
in which the district court considered several of the nine non-exhaustive factors in
reaching its ultimate conclusion that FFSD’s voting system dilutes the votes of black
voters. We address each contested factor in turn.




also Ruiz v. City of Santa Maria, 

, 555 (9th Cir. 1998) (“The ability of
minority voters to bullet vote does not remedy a vote dilution injury.”).


                                          -16-
                                     a. Factor 7

       Two factors—Factors 29 and 7—“predominate the totality-of-the-circumstances
analysis.” 

see 

; see also 

n.15. As to Factor 7—“the extent to which members of the minority group have been
elected to public office in the jurisdiction,” Bone 

—the district
court concluded that “African American electoral successes have generally been
minimal in FFSD, and their overall success rates continue to be disproportionately
low compared to those of white candidates.” FFSD makes three arguments as to why
this conclusion is wrong.

       FFSD argues that the district court “unduly narrowed” its analysis under this
factor by failing to consider non-racial reasons for electoral defeat of black-preferred
candidates, and alleges that the court “failed to consider the circumstances
surrounding electoral defeat” of black-preferred candidates. To some degree, these
arguments echo FFSD’s arguments under the third Gingles precondition, which we
concluded lacked merit. And we note that FFSD does not contest the finding that
success rates of black candidates continue to be lower than for white candidates.

       What FFSD also asserts, however, is that the district court failed to consider
election-specific evidence relevant to the three “Grade A for Change” candidates in
2014: that it was only the winning candidate of the three who was highly qualified,
presented herself as a professional, and “campaigned extremely hard.” These reasons,
not race-based ones, explain the defeat of two black-preferred candidates in that
election, according to the district. However, the facts of the 2014 election do not
support this argument. In that election, the three black-preferred candidates received

      9
       Factor 2, “the extent to which voting in the elections of that State or political
subdivision is racially polarized,” is not at issue on appeal. See 

–45.


                                         -17-
similar levels of support. According to estimates of both parties’ experts, the winning
black-preferred candidate received between one and three percent more black votes
than the unsuccessful black-preferred candidates. Due to the significant margins of
error, there was no way to conclude with any level of certainty that the winning black-
preferred candidates received any more votes from black FFSD residents than did the
unsuccessful black-preferred candidates. In short, in spite of the arguments about
candidate and campaign quality, all of the black-preferred candidates received
effectively equal shares of the vote in 2014. As a result, any error in failing to
consider the additional information regarding the campaigns of successful and
unsuccessful candidates would not have affected the overall totality-of-the-
circumstances analysis.

       The core question posed in Factor 7 is whether black candidates have
historically been successful in the district, not whether individual black candidates
were more attractive candidates or could have run better campaigns. See Bone 

(discussing the number—in that case “not one”—of successful
Native-American candidates under Factor 7); 

(noting “only
minimal electoral success under the present scheme,” as represented by “only [one]
minority victory in nine attempts”). The district court considered the evidence FFSD
presented, but discounted its usefulness in determining whether NAACP had proven
a section 2 violation. Under a totality-of-the-circumstances analysis, the district
court’s assessment of FFSD’s evidence was not clear error.

       Next, FFSD argues that the district court “failed to give recent elections more
weight.” As discussed, the district court “slightly” discounted the 2014 and 2015
elections. And FFSD asserts that the district court failed to accord adequate weight
to the 2016 election in particular, wherein one black representative and one white




                                         -18-
representative were elected. In its quantitative analysis,10 the district court discussed
the relative numerical successes of black candidates and white candidates over both
a longer window—from 2000 to 2016—and a short window—from 2011 to
2016—and found that white candidates were far more successful across both. And,
in discussing the more recent elections, the district court explained that “despite the
growth of the African American population in the District since 2000, minority
electoral success has not improved, with near equal rates of success in the last six
contested elections as there w[ere] between 2000 and 2015.” The district court
considered all the relevant elections appropriately,11 and accurately noted no change
in more recent elections relative to the historical pattern. This was not clear error.

        Finally, FFSD argues that the district court should have given greater weight
to the fact that, after the 2016 election, three of the seven members of the FFSD board
were black, which “place[d] African American representation at near proportionate
levels with their population.” The district court observed that section 2 does not
provide a “safe harbor” based solely on current proportionality. It stated, “the totality


      10
         FFSD argues that it was wrong to engage in this analysis quantitatively,
because the focus should have been on the totality of the circumstances. However,
the district court’s quantitative assessment was not incompatible with a totality-of-
the-circumstances analysis. Quantitative analysis can help demonstrate trends across
a number of elections and the district court assessed each factor individually, before
then engaging in a balancing test, which took all of the factors and circumstances into
account.
      11
        The district court first conducted its quantitative analysis without the 2016
election, but then incorporated it, explaining that “[i]ncluding the results of the 2016
election . . . does not substantially change the results.” Elsewhere in its lengthy
opinion, the district court explained that it was reopening evidence to take judicial
notice of the results of the 2016 election, but that it was “unable to draw significant
legal conclusions” based on that election because it had not been included in the
testimony or analysis provided by the experts for either party.


                                          -19-
of the circumstances and a searching and practical evaluation of the facts suggest that
the more recent successes of African American candidates are not representative of
the community’s ability to be elected to the Board.” See 

(explaining that, while “proportional representation is an important factor to consider
in evaluating the validity of an electoral process,” it “is not . . . the statutory
touchstone and does not provide an absolute safe harbor in which a defendant can
seek refuge from the totality of the circumstances” (citing De 

–18)).12 Simply put, proportionality alone does not indicate that a racial minority
has achieved the ability to effectively participate in the political process. This is one
reason why a totality-of-the-circumstances inquiry is required. It was not erroneous
for the district court to note the appearance of proportionality but nevertheless to
conclude that a section 2 violation was on-going.

                                b. Factors 1, 3, and 5

       FFSD also challenges the district court’s factual findings under Factor 1 (“the
history of voting-related discrimination in the State or political subdivision”), Factor
3 (“the extent to which the State or political subdivision has used voting practices or
procedures that tend to enhance the opportunity for discrimination against the


      12
        We also note that, in most cases, the proportionality inquiry is focused on the
number of minority-majority districts being created and proportionality of
opportunity, not proportionality of representation. See De 

n.11 (explaining that “‘Proportionality’ as the term is used here links the number of
majority-minority voting districts to minority members’ share of the relevant
population,” which is consistent with the VRA’s requirement that “the ultimate right
of § 2 is equality of opportunity, not a guarantee of electoral success for
minority-preferred candidates of whatever race.”); see also African American Voting
Rights Legal Def. Fund, Inc. v. Villa, 

, 1353 (8th Cir. 1995) (comparing
proportionality “between the percentage of blacks in the voting age population . . .
and the percentage of safe black wards”).


                                          -20-
minority group, such as unusually large districts, majority vote requirements, and
prohibitions against bullet voting”), and Factor 5 (“the extent to which minority group
members bear the effects of past discrimination in areas such as education,
employment, and health, which hinder their ability to participate effectively in the
political process”). See 

–45. In finding that each of these
factors weighed in favor of the NAACP, the district court extensively recounted the
history of official discrimination that had occurred throughout Missouri, Metropolitan
St. Louis, and FFSD. The court examined specific jurisdictions within FFSD, and
relied on the continued effects of discrimination in the area. FFSD argues, however,
that this was clearly erroneous because it did not involve the requisite “intensely local
appraisal.”

       We see no clear error in the district court’s exhaustive factual findings, even
where some are based on statewide data or expert testimony applying general data to
the district. The district court relied on the testimony of Drs. Kimball and Gordon to
conclude that “there is a history of officially sanctioned discrimination in the region
and the District, and that history is not just a distant memory.” It further found that
“[g]iven the extent to which African Americans in FFSD continue to experience the
effect of discrimination, their ability to participate in the political process is
impacted.” And, the district court noted, “[t]here is also some evidence that staggered
terms in FFSD enhance the opportunity for discrimination, particularly because they
are combined with at-large voting.” See 

(“[S]taggered
terms[] and at-large structure . . . tend to suppress minority voters’ influence.”). The
evidence amply supports these conclusions.

                                  c. Factors 4 and 8

      The district court determined that Factor 4 (“the exclusion of members of the
minority group from candidate slating processes”) “weigh[ed] very slightly in favor



                                          -21-
of Plaintiffs” and that Factor 8 (whether “elected official are unresponsive to the
particularized needs of the members of the minority group”) should be accorded
“neutral weight.” FFSD, however, argues that both factors should have weighed in
its favor.

       As to Factor 4, the district court assessed whether slating processes were
available in the district, applying Eighth Circuit precedent that defines a slating group
as a group that “consists of a small number of individuals who select candidates to
run as a bloc to fill seats which are up for election.” Clay v. Bd. of Educ. of St. Louis,

, 1362 n.11 (8th Cir. 1996). It determined that there were two local
slating organizations and that, while the organizations had endorsement processes
that were open to everyone, “African American candidates have less success than
white candidates in getting endorsed by these organizations.” The district court
accorded this factor “very slight[]” weight because there was little evidence that
access was explicitly denied to black candidates, but there was undisputed evidence
that white candidates were more likely to be endorsed. Even if the district court
erred, however, any such error would not have affected the result, as this factor was
given such low weight, and no one factor is outcome-determinative.

       Similarly, in assessing Factor 8, the district court was faced with conflicting
evidence. FFSD had made some efforts to meet the particularized needs of the black
community, but there were noted instances of unresponsiveness. The district court’s
determination that “the Board is making efforts to respond” but “ha[d], at times, been
unresponsive” was not clearly erroneous, nor was its ultimate conclusion that this
factor should be accorded neutral weight.




                                          -22-
                                        V.

      The district court found a section 2 violation after engaging in the requisite
precondition analysis and conducting a thorough totality-of-the-circumstances
balancing. We affirm.
                      ______________________________




                                       -23-