Benavidez v. Irving Independent School District et al, No. 3:2013cv00087 - Document 91 (N.D. Tex. 2014)

Court Description: MEMORANDUM OPINION AND ORDER. Following a bench trial, the court finds that the 5-2 system for electing trustees of the Irving Independent School District trustees violates § 2 of the Voting Rights Act of 1965, 42 U.S.C. § 1973. Defendants are therefore ordered to submit, within 60 days of the date of this memorandum opinion and order, a plan to remedy the violation. (Ordered by Chief Judge Sidney A Fitzwater on 8/15/2014) (Chief Judge Sidney A Fitzwater)

Download PDF

IN THE UNITED STATES DISTRICT COURT FOR THE NORTHERN DISTRICT OF TEXAS DALLAS DIVISION MANUEL A. BENAVIDEZ, et al., Plaintiffs, VS. IRVING INDEPENDENT SCHOOL DISTRICT, et al., Defendants. § § § § § Civil Action No. 3:13-CV-0087-D § § § § § § MEMORANDUM OPINION AND ORDER This is a suit brought under § 2 of the Voting Rights Act of 1965 ( the Act ), 42 U.S.C. § 1973, challenging the 5-2 system five single-member districts and two at-large positions for electing trustees of the Irving Independent School District ( Irving ISD ). Following a bench trial, and for the reasons that follow,1 the court finds that the Irving ISD 5-2 system for electing trustees violates § 2 of the Act because, under the totality of the circumstances, it denies Hispanic voters an equal opportunity to participate in the electoral process and to elect representatives of their choice. I Plaintiffs Manuel A. Benavidez ( Benavidez ), Juana De Leon, and Daniela De Leon are Hispanic residents of the Irving ISD, which is located in Dallas County, Texas. They 1 The court sets out in this memorandum opinion and order its findings of fact and conclusions of law. See Fed. R. Civ. P. 52(a)(1). bring this suit against Irving ISD and the members of its Board of Trustees, in their official capacities, alleging that the current system for electing members of the Irving ISD Board of Trustees violates § 2 of the Act.2 This is the second § 2 lawsuit that Benavidez has brought challenging the manner in which Irving ISD elects trustees. In the first, Benavidez v. Irving Independent School District, Texas, 690 F.Supp.2d 451 (N.D. Tex. 2010) (Fitzwater, C.J.) ( Benavidez I ), Benavidez challenged Irving ISD s then-system of electing all seven trustees in district-wide elections.3 Under that system, which has since been replaced, all seven trustee candidates ran at-large for specific numbered positions, with approximately one-third of the positions up for election each year. The only pertinent requirement for election was that the candidate reside within the boundaries of the Irving ISD. Any eligible voter residing within the school district could vote for all trustee positions that were up for election in a given year. Because all seven trustee positions were elected at-large, the parties have referred to the system as a pure at-large system, and the court will do the same. Benavidez sought to establish that the pure at-large system denied Hispanic voters an equal opportunity to participate in the electoral process and to elect representatives of their choice. At the time his lawsuit was tried in 2009, the most recent census data were derived 2 The individual defendants sued in their official capacities are Steven Craig Jones, Marilyn Gail Conder Woods Wells, Larry M. Stipes, Valerie D. Jones, Norma C. Gonzales, Randy Randle, and Lee Mosty, who are current or former members of Irving ISD s Board of Trustees. 3 The facts of the first lawsuit are set out extensively in Benavidez I. -2- from the 2000 Census. At trial, Benavidez s expert, David Ely ( Ely ), admitted that he could not draw an illustrative district that had greater than a 50% Hispanic share of the citizen voting age population ( CVAP ) based on 2000 Census data. Ely opined, however, that it was proper to rely on one-year survey data from the 2007 American Community Survey ( 2007 one-year ACS data ), and that the 2007 one-year ACS data proved that there had been a statistically significant increase in the Hispanic share of CVAP in the Irving ISD since 2000. He testified that, based on the 2007 one-year ACS data, he could draw three illustrative districts that had Hispanic CVAP shares ranging from 51.87% to 55.42%. Defendants criticized Benavidez s reliance on one-year data and urged the court to rely on the 2000 Census data instead. In Benavidez I the court held that Benavidez had failed to prove the first prong of Thornburg v. Gingles, 478 U.S. 30 (1986), which requires that a plaintiff prove by a preponderance of the evidence that the minority group is sufficiently large and geographically compact to constitute a majority in a single-member district.4 The court noted that, [i]n the Fifth Circuit, [c]ensus figures are presumed accurate until proven otherwise. Proof of changed figures must be thoroughly documented, have a high degree of accuracy, and be clear, cogent and convincing to override the presumptive correctness of the prior decennial census. Benavidez I, 690 F.Supp.2d at 457 (quoting Valdespino v. Alamo Heights Indep. Sch. Dist., 168 F.3d 848, 853-54 (5th Cir. 1999)) (second alteration in 4 This type of district is often referred to as a majority-minority district. -3- original).5 The court found that, because the 2007 one-year ACS data were not sufficiently reliable to override the presumption that the 2000 Census data were correct,6 and both sides agreed that the 2000 Census data failed to support a finding that the illustrative districts contained a majority Hispanic CVAP share, Benavidez failed to prove the first prong of the Gingles threshold test and did not consider the other elements of his § 2 claim. In rejecting Benavidez s challenge, however, the court offered the following observation: After the results of the 2010 Census are published, Benavidez may be able to obtain the relief he seeks trustees elected from single member districts without the need for another lawsuit. The 2010 Census may confirm Benavidez s contention that a majority Hispanic CVAP district can be drawn. But for now, Benavidez has failed to overcome the strong presumption that the 2000 Census data are correct. These data, when applied to the first Gingles factor, defeat his § 2 claim. Id. at 464. The results of the 2010 Census were subsequently published, and they indicated that the Hispanic share of the total population in the Irving ISD had indeed increased between 2000 and 2010. According to the 2000 Census, Hispanics made up only 35.63% of the 5 See also Fairley v. Hattiesburg, Miss., 584 F.3d 660, 674 (5th Cir. 2009) ( The Census is presumptively correct and typically must be rebutted with clear and convincing evidence. ) (citing NAACP v. Fordice, 252 F.3d 361, 366 (5th Cir. 2001)). 6 The court noted in its prior decision that it only found the 2007 one-year ACS data unreliable as Benavidez had used these data. It observed that one-year ACS data might be reliable when used for myriad purposes, including analysis of large populations and even hypothesis testing of small populations, because hypothesis testing takes into account the larger margins of error. Benavidez I, 690 F.Supp.2d at 464 n.18. Benavidez, however, had relied on 2007 one-year ACS data to analyze a relatively small population and had not performed hypothesis testing to account for the large margins of error observed in that case. -4- district s total population. But according to the 2010 Census, the Hispanic share of the total population had grown to 50.21%. In 2011, following publication of the 2010 Census results, Irving ISD s Board of Trustees studied the possibility of changing the system for electing trustees. In early 2012, the Board adopted a 5-2 system. Under this system, there are five single-member districts and two at-large positions.7 Districts 2-6, inclusive, are single-member districts, and Districts 1 and 7 are elected at-large. In the single-member districts, only voters who reside in a particular district can vote for the trustee who represents that district. To be eligible for election, a candidate for a single-member district must reside within the district. As for each of the two at-large positions, all voters residing within the boundaries of the Irving ISD can vote in the election for that position, and candidates for at-large positions need only reside within the school district rather than a particular district. Because there are five singlemember districts and two at-large positions, the system is a partial at-large or mixed system, as opposed to a pure at-large system, like the one Benavidez challenged in his first lawsuit. As the court will discuss below, a pivotal question in this lawsuit is whether the creation of District 6 is sufficient under the 5-2 system to provide Hispanic voters an equal opportunity to participate in the electoral process and to elect representatives of their choice. Defendants maintain that the creation of District 6 precludes plaintiffs § 2 claim because it 7 Members of the Board of Trustees are elected to staggered three-year terms, and elections are by plurality vote. -5- is a single-member district from which Hispanic voters can elect a representative of their choice. Plaintiffs counter that District 6 is not an effective Hispanic district because it does not contain a majority Hispanic CVAP share. Defendants concede that two five-year ACS special tabulations one based on the years 2007 to 2011, the other based on years 20082012 show that the Hispanic CVAP share of District 6 is 46.75% and 45.23%, respectively, both of which fall below the 50%+ threshold that is necessary to constitute a majority share of Hispanic CVAP. But defendants contend that, based on different data8 and projections of future population growth, the Hispanic CVAP share in District 6 is currently at least 49.88% and could be as high as 60.2%. They posit that, by 2016 (the date of the next election for the District 6 seat on the Board of Trustees), the Hispanic CVAP share will be between 52.18% and 64.91%. Defendants maintain that, given these projections, the preponderance of evidence establishes that District 6 is an effective district for Hispanic representation. And they argue that § 2 does not mandate that Irving ISD draw a district with a majority Hispanic CVAP share even if it is possible to do so so long as one district is an effective district for Hispanic representation. In the history of the Irving ISD Board of Trustees, only two Hispanics have been elected, and no Hispanic candidate has ever been elected in a contested election against a non-Hispanic opponent. When Ruben Franco ( Franco ) was elected in 2000, he defeated Benavidez. When Franco won reelection in 2003, he ran unopposed. In May 2013 the first 8 The court discusses these data extensively infra in § VI(B)(1)-(3). -6- election held under the current 5-2 system Norma Gonzales ( Gonzales ) ran unopposed and was elected. Gonzales resigned from the Board of Trustees approximately four months following her election,9 creating a vacancy that the Board of Trustees filled by soliciting applications from interested residents of District 6. No Hispanic applied for the position, and the Board of Trustees ultimately selected a non-Hispanic candidate to fill the vacancy. There are currently no Hispanic members of the Board of Trustees. II In 1982 Congress substantially revised § 2 of the Voting Rights Act to clarify that a violation requires evidence of discriminatory effects alone, and to make clear that proof of discriminatory intent is not required to establish a violation of Section 2. Benavidez I, 690 F.Supp.2d at 455 (quoting League of United Latin Am. Citizens # 4434 (LULAC) v. Clements, 986 F.2d 728, 741 (5th Cir. 1993) ( LULAC v. Clements I ) (citation and some internal quotation marks omitted)). Section 2(b) now provides that the Act is violated if, based on the totality of circumstances, it is shown that the political processes leading to nomination or election in the State or political subdivision are not equally open to participation by [a class of persons of a certain race or color] in that its members have less opportunity than other members of the electorate to participate in the political process and to elect representatives of their choice. The extent to which members of a protected class have been elected to office in the State or political subdivision is one circumstance which may be considered: Provided, That 9 No evidence was offered at trial concerning the reasons for her resignation, and neither side suggests that these reasons are pertinent to this lawsuit. -7- nothing in this section establishes a right to have members of a protected class elected in numbers equal to their proportion in the population. 42 U.S.C. § 1973(b). In Gingles the Supreme Court first considered the 1982 amended version of § 2, setting out the current framework for analyzing § 2 cases. To prevail on a § 2 claim, a plaintiff must first prove that (1) the minority group is sufficiently large and geographically compact to constitute a majority in a single member district, (2) the minority group is politically cohesive, and (3) the white majority votes sufficiently as a bloc to enable it in the absence of special circumstances, such as the minority candidate running unopposed usually to defeat the minority s preferred candidate. Benavidez I, 690 F.Supp.2d at 455 (quoting Gingles, 478 U.S. at 50-51). Failure to establish any one of the Gingles factors precludes a finding of vote dilution, because these circumstances are necessary preconditions for multimember districts to operate to impair minority voters ability to elect representatives of their choice. Id. (quoting LULAC v. Clements I, 986 F.2d at 743) (brackets and internal quotation marks omitted). [The Fifth Circuit] has interpreted the Gingles factors as a bright line test. Id. at 456 (quoting Valdespino, 168 F.3d at 852) (brackets in original). Each factor must be proved. Failure to establish any one of these threshold requirements is fatal. Id. (quoting Valdespino, 168 F.3d at 852) (citations, brackets, and some internal quotation marks omitted). If a plaintiff meets the threshold Gingles test, the court must then engage in a broader totality of the circumstances inquiry, considering whether the minority group has demonstrated that under the totality of the circumstances, its -8- members have less opportunity than other members of the electorate to participate in the political process and to elect representatives of their choice. Id. at 456 n.7 (quoting LULAC v. Clements I, 986 F.2d at 747) (internal quotation marks omitted). In conducting this broad inquiry, a court must be flexible in its totality inquiry and guided by factors drawn from the Senate Judiciary Committee report on the 1982 amendments to the Voting Rights Act and reference Zimmer v. McKeithen, 485 F.2d 1297 (5th Cir. 1973). Fabela v. City of Farmers Branch, Tex., 2012 WL 3135545, at *3 (N.D. Tex. Aug. 2, 2012) (Fitzwater, C.J.) (quoting Teague v. Attala Cnty., Miss., 92 F.3d 283, 292 (5th Cir. 1996)). [T]he determination whether [an] at-large election scheme violates section 2 depends upon a searching practical evaluation of the past and present reality and on a functional view of the political process. Westwego Citizens for Better Gov t v. City of Westwego, 946 F.2d 1109, 1120 (5th Cir. 1991) ( Westwego III ) (quoting Gingles, 478 U.S. at 45; Westwego Citizens for Better Gov t v. City of Westwego, 872 F.2d 1201, 1204 (5th Cir. 1989) ( Westwego I )). The factors to be considered when evaluating the totality of the circumstances include: [1] the history of voting-related discrimination in the State or political subdivision; [2] the extent to which voting in the elections of the 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[;] [4] the extent to which minority group members bear the effects -9- of past discrimination in areas such as education, employment, and health, which hinder their ability to participate effectively in the political process; [5] the use of overt or subtle racial appeals in political campaigns; [6] the extent to which members of the minority group have been elected to public office in the jurisdiction[;] [7] evidence demonstrating that elected officials are unresponsive to the particularized needs of the members of the minority group[;] [8] [evidence] that the policy underlying the State s or the political subdivision s use of the contested practice or structure is tenuous[;] [9] whether the number of districts in which the minority group forms an effective majority is roughly proportional to its share of the population in the relevant area[.] Fairley v. Hattiesburg, Miss., 584 F.3d 660, 672-73 (5th Cir. 2009) (quoting League of United Latin Am. Citizens (LULAC) v. Perry, 548 U.S. 399, 426 (2006) ( LULAC v. Perry ). The existence of racially polarized voting and the extent to which minority group members have been elected to public office are the most important factors to be considered. Fabela, 2012 WL 3135545, at *3 (citing Gingles, 478 U.S. at 48 n.15). In analyzing the totality of circumstances, the court is required to effect a flexible, fact-intensive inquiry predicated on an intensely local appraisal of the design and impact of the contested electoral mechanisms[.] NAACP v. Fordice, 252 F.3d 361, 367 (5th Cir. 2001) (quoting Magnolia Bar Ass n, Inc. v. Lee, 994 F.2d 1143, 1147 (5th Cir. 1993)). Multimember districts and at-large election schemes . . . are not per se violative of - 10 - minority voters rights. Minority voters who contend that the multimember form of districting violates § 2, must prove that the use of a multimember electoral structure operates to minimize or cancel out their ability to elect their preferred candidates. Benavidez I, 690 F.Supp.2d at 456 (citing Gingles, 478 U.S. at 48). A plaintiff must prove a § 2 violation by a preponderance of the evidence. Id. (citing League of United Latin Am. Citizens # 4552 (LULAC) v. Roscoe Indep. Sch. Dist., 123 F.3d 843, 846 (5th Cir. 1997)). III Before analyzing the merits of plaintiffs § 2 claim, the court will address an objection that defendants made before and during trial to two of plaintiffs exhibits: Ps. Exs. 53 and 56. Essentially, defendants maintain that plaintiffs are attempting at the eleventh hour to prove a § 2 violation on the basis that two Hispanic CVAP majority districts can be created for electing trustees of the Irving ISD.10 A Ely prepared an expert report approximately one year before trial, and he signed it on June 20, 2013. His report analyzed a hypothetical districting scheme for electing Irving ISD trustees from seven single-member districts. Ely opined that there are multiple configurations in which it is possible to create a single member district with 1/7th of the population of the Irving ISD in which Hispanics comprise a majority of eligible voters, but 10 Defendants maintain that plaintiffs failed to timely disclose the information contained in these exhibits and that this failure prejudiced their trial strategy because they were not given enough time to develop rebuttal expert testimony. - 11 - that there is no such district in the existing plan. Ps. Ex. 33 at ¶ 25. Ely did not opine that it was possible to create two single-member districts with a majority of Hispanic CVAP. Defendants expert, Norfleet Rives, Ph.D. ( Dr. Rives ), testified during his March 10, 2014 deposition that defendants possessed newly released 2012 ACS data. On June 20, 2014 both sides submitted their pretrial disclosures. Defendants pretrial disclosures included twenty exhibits that reflected an analysis of the newly released data. Of these twenty exhibits, seven were modified exhibits to Dr. Rives s expert report. When Ely received these exhibits, he analyzed them and eventually created two new exhibits Ps. Exs. 53 and 56 using the newly released data and applying the methodology that Dr. Rives used in his expert report.11 Ps. Ex. 53 is a map of seven single-member districts, and Exhibit 56 is a table with data corresponding to the map. When plaintiffs attempted to introduce the exhibits at trial, defendants objected, arguing that they were untimely disclosed, in violation of Fed. R. Civ. P. 26(a)(3) and 37(c)(1).12 Initially, it appeared that plaintiffs intended to offer testimony and exhibits (including Ps. Exs. 53 and 56) regarding the possibility of creating two districts with a Hispanic CVAP 11 Ely modified these data in certain ways (e.g., by redrawing the boundaries of the districts), but the modifications are not material to the court s decision. 12 Defendants moved to strike these exhibits before trial. The court concluded that defendants motion to strike should be treated procedurally as an objection to the introduction of exhibits at trial. Defendants timely objected to the exhibits when they were offered at trial, relying on the grounds and rationale contained in their pretrial motion to strike. The court gave them a running objection to the exhibits and related testimony, and it stated that it would carry the objections throughout the trial and then rule on the objection, if necessary, as part of its decision on the merits. - 12 - majority solely for the purpose of rebutting Dr. Rives s methodology and the newly released ACS data when calculating the Hispanic CVAP share for District 6. Ely testified at trial that, if the court were to accept the reliability of Dr. Rives s methodology and the newly released ACS data when calculating the Hispanic CVAP for District 6, then plaintiffs would actually be able to draw not one, but two, single-member districts with a majority share of Hispanic CVAP. But Ely made it clear during his testimony that he did not consider Dr. Rives s data and methodology to be reliable. For example, Ely testified: I don t believe that that analysis is reliable. And for that reason I didn t initially draw a two district plan. I drew one district plan according to the analysis that I believed was appropriate. However, since the defendants and the defendants expert persisted with the idea that all this other analysis was appropriate then I drew an additional plan that was consistent with that analysis, with two districts. Tr. 1:186 (paragraph break omitted). Ely created Ps. Exs. 53 and 56 only to show that, if Dr. Rives s approach were correct, there are actually two single-member districts that could be drawn with a majority Hispanic CVAP. For example, Ely testified: If the court chooses to accept the growth rates that were projected by [Dr.] Rives, which I have no expectation that the court will choose to accept, then I wanted to be in a position of saying, well, in that case we can have two. Id. at 1:182 (bracketed material added). Just before trial, however, plaintiffs appeared to pivot concerning their purpose for and intended use of Ely s testimony and of evidence (including Ps. Exs. 53 and 56) showing the potential for two Hispanic CVAP majority districts. On July 14, 2014 they filed their - 13 - amended proposed findings of fact and conclusions of law, in which they asserted that two Hispanic districts can be drawn. See Ps. Am. Prop. Findings of Fact No. 23 ( Plaintiffs have shown that Latinos are sufficiently geographically compact and numerous in the IISD that at least two districts can be drawn which have Latino majorities of eligible voters. ); but see Ps. Am. Prop. Findings of Fact and Concl. of Law No. 67 ( Specifically, Plaintiffs have shown that they can create at least one district in the IISD that is geographically compact and that has a Latino majority of eligible voters. (emphasis added)). And at trial, they did not only rely on Ely s evidence to rebut defendants evidence. During closing argument, although plaintiffs counsel relied on Ely s testimony concerning two districts to refute defendants evidence concerning the Hispanic CVAP composition of District 6, see Tr. 3:125 ( And if we accept defense s argument that District 6 . . . is growing into majority Hispanic CVAP district, both of David Ely s districts that he drew in this case would be well above 50 percent ), plaintiffs counsel also argued that, [a]s to Gingles prong 1, David Ely demonstrated that two majority Hispanic districts in the seven single-member district system can be drawn in accordance with traditional districting principles, id. at 3:124. The court therefore concludes that plaintiffs are attempting to rely on Ely s evidence, including Ps. Exs. 53 and 56, both to refute defendants proof and to argue that the evidence supports the creation of two Hispanic districts. B Despite plaintiffs dual intended use of Ps. Exs. 53 and 56, the court overrules defendants objection to their introduction in evidence. Rule 37 states that, [i]f a party fails - 14 - to provide information . . . as required by Rule 26(a) or (e), the party is not allowed to use that information . . . to supply evidence . . . at a trial, unless the failure was substantially justified or is harmless. Rule 37(c)(1). Assuming arguendo that plaintiffs did not timely disclose the information on which Ps. Exs. 53 and 56 are based and that their failure to disclose the information was not substantially justified, the court finds that this failure is harmless. As the court explains below, see infra § VI(B)(3), it does not find that Dr. Rives s data and methodology are sufficiently reliable to support his opinion that District 6 currently has a majority Hispanic CVAP share. Nor does the court find that Ely s opinions based on these same data data that Ely himself rejects, see Tr. 1:182, 186 are reliable. Plaintiffs failure to timely disclose Ps. Exs. 53 and 56 is therefore harmless because the court is not relying on the exhibits or the related testimony to reject Dr. Rives s data and methodology or to find that two majority Hispanic CVAP single-member districts can be drawn. IV The court now turns to the merits of plaintiffs § 2 claim, considering whether they have met their burden of establishing each of the Gingles factors.13 13 Defendants appear to question whether the Gingles preconditions apply at all, or apply only in a modified form, because plaintiffs are challenging a partial at-large system rather than a pure at-large system of electing trustees. Defendants have cited no authority, and the court is aware of none, standing for the proposition that the Gingles preconditions do not apply, or should be modified, simply because the challenged system includes both single-member districts and at-large positions. Moreover, courts in this circuit have applied the Gingles factors to partial at-large or mixed systems without questioning whether the framework applies. See, e.g., St. Bernard Citizens for Better Gov t v. St. Bernard Parish Sch. - 15 - A Under the first prong of Gingles, plaintiffs must prove that the Hispanic population in the Irving ISD is sufficiently large and geographically compact to constitute a majority in a single member district. LULAC v. Clements I, 986 F.2d at 742. In Bartlett v. Strickland, 556 U.S. 1 (2009) (plurality opinion), the Supreme Court considered the minimum-size question, concluding that the majority-minority rule relies on an objective, numerical test: Do minorities make up more than 50 percent of the voting-age population in the relevant geographic area? Id. at 18. In this circuit, to satisfy this test, plaintiffs must prove that there is a single potential single-member district (the demonstrative or illustrative district) in which a majority of the CVAP is Hispanic. Reyes v. City of Farmers Branch, Tex., 586 F.3d 1019, 1023 (5th Cir. 2009). Plaintiffs rely on Ely s expert testimony to satisfy this element of Gingles.14 Ely used data from the 2010 Census and a five-year data set covering years 2007-2011 ( 2007-2011 Bd., 2002 WL 2022589 (E.D. La. Aug. 26, 2002) (applying Gingles factors to 5-2 system and finding § 2 violation); Williams v. City of Dallas, 734 F. Supp. 1317 (N.D. Tex. 1990) (Buchmeyer, J.) (applying Gingles factors to 8-3 system and finding § 2 violation), aff d, 38 F.3d 569 (5th Cir. 1994) (unpublished table decision); League of United Latin Am. Citizens, Council No. 4836 (LULAC) v. Midland Indep. Sch. Dist., 648 F. Supp. 596 (W.D. Tex. 1986) (applying Gingles factors to 4-3 plan and 2-5 plan and finding that both violate § 2), aff d, 829 F.2d 546 (5th Cir. 1987) (en banc) (per curiam). To the extent defendants maintain that the Gingles preconditions do not apply in this case, or that they apply but must be modified in the context of this § 2 challenge, the court disagrees. 14 As is commonly the case in § 2 litigation, [plaintiffs ] claim turns on the expert witness factual testimony. Benavidez I, 690 F.Supp.2d at 457 n.9 (quoting the statement in LULAC v. Clements I, 986 F.2d at 736, that [a]s with all cases under the Voting Rights Act, this one is driven by the facts. ). - 16 - five-year ACS data ) to draw three illustrative districts (Districts A, B, and C). Each district s Hispanic CVAP share is greater than 50%. Ely testified that each illustrative district satisfies the first Gingles prong and that any of the three could be the demonstration district. Defendants neither challenge the reliability of the data on which Ely based these calculations nor materially dispute the validity of the methodology Ely applied to the data.15 Rather, they argue that the illustrative districts are not geographically compact. They maintain that the illustrative districts are bizarrely shaped and ignore traditional districting principles because they reach out to include small, isolated pockets of Hispanic citizens, or, in some instances, to encompass pockets of non-citizens.16 Plaintiffs respond that defendants position is too heavily predicated on geometric compactness, and that each illustrative district is geographically compact within the meaning of Gingles first prong. 15 There are references in the record to two criticisms of Ely s methodology neither of which is material to the court s decision. Dr. Rives criticized Ely for applying an unnecessarily complicated rounding adjustment when he created the districts that are reflected in Ps. Exs. 53 and 56. These exhibits and the underlying methodology that includes this rounding adjustment pertain to Ely s opinion that, under Dr. Rives s approach, two single-member districts with a majority Hispanic CVAP share can be drawn. See supra § III. As the court explains infra in § VI(B)(3), the court finds Dr. Rives s data and methodology to be unreliable and does not rely on them in reaching any of its factual findings regarding the Gingles factors or the totality of circumstances. Defendants do not argue that this rounding-adjustment criticism applies to the methodology Ely used to draw Districts A, B, and C. Dr. Rives also criticized how Ely interpreted data regarding Spanish surname registered voters ( SSRV data ). But Dr. Rives concluded that, even though Ely and Dr. Rives interpreted the SSRV data differently, these differences turned out to be negligible, and it was a nonissue. Tr. 2:179; see also id. at 3:22. 16 For purposes of this decision, it is irrelevant whether a non-citizen was present in the Irving ISD legally or illegally. - 17 - B Because defendants do not challenge Ely s conclusion that each illustrative district contains greater than 50% Hispanic CVAP, the court will focus on their compactness challenge. Plaintiffs rely on Ely s expert testimony to establish that the illustrative districts are geographically compact. At trial, Ely admitted that District A is geographically very awkward and very ugly. Tr. 1:187, 188. But he opined that the district is compact nevertheless because people within the district are not separated by huge distances, each piece of the district is an entire Census block group17 (i.e., he did not divide Census block groups to draw the boundaries), and the district is connected by major transportation corridors. He also testified that the district did not cut back and forth to pick up isolated pockets of population, and that there are no fingers that pick up remote areas. Ely testified in less equivocal terms about the geographic compactness of Districts B and C, stating that both were geographically compact without calling them awkward or ugly. He noted that the greatest distance between any two residences within District B was no more than seven miles and no more than five miles within District C, and that both districts were connected by major transportation corridors. He also explained that District C was modeled largely on District B, and that he modified District C only slightly to reduce the distances that were included within the district and to simplify the district s boundaries. Finally, he opined 17 A block group is a unit of measurement that the Census uses in its population statistics. See, e.g., Fabela, 2012 WL 3135545, at *5 (differentiating Census blocks from block groups and Census tracts). - 18 - that election officials presented with the map for any of his proposed districts would have no difficulty determining what the boundaries were and administering an election. Defendants rely on Dr. Rives s expert testimony to rebut Ely s testimony. Dr. Rives testified that, on visual inspection of the three illustrative districts, they did not appear compact to him. He testified that he used redistricting software to measure the compactness of the districts, and that the illustrative districts did not score well on this basis. He also noted that the total perimeter of each illustrative district was relatively large and that each district had [a] lot of edges to traverse. Id. at 2:160. On cross-examination, however, Dr. Rives acknowledged that his analysis was limited to evaluating the shape of each district, and that he did not perform any analysis of functional measures of compactness, such as determining whether residents of the district attended the same schools or churches. See id. at 3:16. Ely persuasively responded to Dr. Rives s criticisms of his conclusion concerning geographic compactness. He opined that the redistricting software on which Dr. Rives relied uses numeric models that only test the geometric shape of each district, and that they do not analyze functional compactness. According to Ely, measuring geographic compactness exclusively in terms of geometric shape is potentially misleading. He gave an example of a district drawn as a perfect square, but where the population is divided into four isolated segments and placed in the four corners of the district. Ely persuasively pointed out that Dr. Rives s compactness measures would suggest that this hypothetical district is compact merely because of the shape of its outside edges, when, in reality, the district would be - 19 - extremely non-compact if great distances separated the four corners of the district. See id. at 1:201. Defendants did not adequately address Ely s rebuttal of Dr. Rives s criticism of the compactness of Ely s illustrative districts. Geometric shape is a significant factor that courts can and must consider in a Gingles compactness inquiry. Sensley v. Albritton, 385 F.3d 591, 596 (5th Cir. 2004). But the compactness inquiry is less about measuring aesthetics and more about analyzing the functional realities of the minority population. See id. at 596 (noting that geometric shape necessarily directly relates to . . . population dispersal of the minority community in question ); Clark v. Calhoun Cnty., Miss., 21 F.3d 92, 95 (5th Cir. 1994) ( Clark I ) ( The first Gingles precondition does not require some aesthetic ideal of compactness, but simply that the [minority] population be sufficiently compact to constitute a majority in a single-member district. ); Rodriguez v. Harris Cnty., Tex., 964 F.Supp.2d 686, 738 (S.D. Tex. 2013) ( Geographical symmetry or attractiveness is a desirable consideration for districting, but only to the extent it facilitates the political process. (citation omitted)). While no precise rule has emerged governing § 2 compactness, the inquiry should take into account traditional districting principles such as maintaining communities of interest and traditional boundaries. LULAC v. Perry, 548 U.S. at 433 (citations and internal quotation marks omitted). For example, a district would not be sufficiently compact if it was so spread out that there was no sense of community, that is, if its members and its representative could not effectively and efficiently stay in touch with each other; or if it was so convoluted that there was no sense of community, that is, if its members and its representative could not - 20 - easily tell who actually lived within the district. Dillard v. Baldwin Cnty. Bd. of Educ., 686 F. Supp. 1459, 1466 (M.D. Ala. 1988); see also Clark I, 21 F.3d at 96 (noting that courts have concluded that the first Gingles precondition is not satisfied if the proposed district does not retain a natural sense of community such that it can be effectively represented ). The court finds that at least illustrative Districts B and C are geographically compact.18 Dr. Rives s testimony does not materially undermine Ely s conclusions that Districts B and C are geographically compact. Although Dr. Rives testified that both districts did not score well on the basis of his compactness measures, he made clear that his analysis was limited to assessing only the geometric shape of the districts. And geometric shape is not the controlling consideration in analyzing the first Gingles precondition. Ely presented unrefuted testimony that the greatest distance between two residences within District B is no more than seven miles and no more than five miles in District C; that both districts are connected by major transportation corridors; and that election officials would be able to determine district boundaries and administer an election plan if presented with a map of the proposed district. Defendants have not adduced any evidence that the districts break apart communities of interest or otherwise violate traditional districting principles. They assert that the districts are drawn to pick up isolated pockets of non-citizens, but they have not presented any evidence that these pockets of non-citizens represent distinct communities, with interests that are different from those of the other people included in the district. In fact, 18 Because the court finds that Districts B and C satisfy the first Gingles precondition, the court need not decide whether District A is geographically compact. - 21 - the trial evidence suggests that many of these non-citizens are Hispanic and are presumably members of the same communities. And there is no indication from the trial record that either district is so bizarrely shaped as to suggest that race was the predominant factor in drawing the boundaries.19 V The court now considers the second and third prongs of Gingles, which require that plaintiffs prove that Hispanic voters are politically cohesive and that the white majority votes sufficiently as a bloc to enable it in the absence of special circumstances, such as the minority candidate running unopposed usually to defeat the minority s preferred candidate. Gingles, 478 U.S. at 51 (citations omitted). A Racially polarized voting, i.e., where there is a consistent relationship between [the] race of the voter and the way in which the voter votes, is relevant to a vote dilution claim. Fabela, 2012 WL 3135545, at *8 (quoting Gingles, 478 U.S. at 53 n.21 & 56) (some internal quotation marks omitted). It tends to prove that the minority group members constitute a politically cohesive unit, under the second Gingles prong, and that they are unable to elect representatives of their choice because, under the third Gingles prong, the majority group is 19 In Shaw v. Reno, 509 U.S. 630 (1993), the Supreme Court held that a voting scheme so extremely irregular on its face that it rationally can be viewed only as an effort to segregate the races for purposes of voting triggers strict scrutiny under the Equal Protection Clause of the Constitution. Id. at 642. Here, as in Clark I, the illustrative districts are not nearly as bizarre as the district under consideration in Shaw. See Clark I, 21 F.3d at 95. - 22 - similarly politically cohesive and votes sufficiently as a bloc usually to defeat [their] preferred candidates. Gingles, 478 U.S. at 56; see also, e.g., Westwego I, 872 F.2d at 1207 ( Evidence of racially polarized voting is the linchpin of a section 2 vote dilution claim and is relevant to establishing [the second and third elements] . . . in Gingles[.] (citations omitted)). Because . . . the extent of bloc voting necessary to demonstrate that a minority s ability to elect its preferred representatives is impaired varies according to several factual circumstances, the degree of bloc voting which constitutes the threshold of legal significance will vary from district to district. Gingles, 478 U.S. at 55-56. A showing that a significant number of minority group members usually vote for the same candidates is one way of proving the political cohesiveness necessary to a vote dilution claim, and, consequently, establishes minority bloc voting within the context of § 2. And, in general, a white bloc vote that normally will defeat the combined strength of minority support plus white crossover votes rises to the level of legally significant white bloc voting. The amount of white bloc voting that can generally minimize or cancel, [minority] voters ability to elect representatives of their choice, however, will vary from district to district according to a number of factors, including the nature of the allegedly dilutive electoral mechanism; the presence or absence of other potentially dilutive electoral devices, such as majority vote requirements, designated posts, and prohibitions against bullet voting; the percentage of registered voters in the district who are members of the minority group; the size of the district; and, in multimember districts, the number of seats open and the number of candidates in the field. Id. at 56 (citations omitted). - 23 - B Plaintiffs expert, Richard Engstrom, Ph.D. ( Dr. Engstrom ), presented data regarding eight Irving ISD elections occurring in 2006, 2008, 2010, 2012, and 2013. These elections are the most recent in which voters were presented with a choice between a Hispanic and a non-Hispanic candidate. In each election, the Hispanic-preferred candidate lost. After receiving data from the Dallas County Elections Department regarding the names of voters who cast ballots, Ely used Spanish surnames to identify whether the votes were cast by Hispanics or non-Hispanics, and he provided this information to Dr. Engstrom. Dr. Engstrom then conducted a statistical analysis called ecological inference ( EI ), which is accessible through R software ( EI in R ), to estimate the percentage of Hispanic and non-Hispanic voters who voted for the Hispanic candidate in each election. His report offers the point estimate, which is the best estimate according to the statistical procedure, Tr. 2:59, and the confidence interval,20 which identifies the range of estimates within which we can be 95 percent confident, statistically, that the true value of a group s support for a candidate falls, Ps. Ex. 42 at 5; see also Tr. 2:59-60. The results, using EI in R, are as follows: 20 It is standard in the area of political science for confidence intervals to be set at 95%. - 24 - Election Percent of Hispanic Voters Place 3, 2006: Benavidez Place 4, 2006: Carranza Place 2, 2008: Chac Place 5, 2010: Portillo Place 3, 2012: FernandezMott Place 4, 2012: Gonzales District 5, 2013: Benavidez 95% Confidence Interval 95% Confidence Interval 7.5% 68.2% - 99.2% 5.3% - 11.1% 11.9% 50.6% - 98.5% 9.1% - 17.3% 15.0% 55.6% - 100.0% 9.7% - 21.5% 13.6% 57.3% - 99.9% 10.0% - 17.7% 32.7% 59.9% - 100.0% 28.3% - 37.7% 58.0% Point Estimate 95% Confidence Interval 9.2% - 15.0% 90.1% Point Estimate 95% Confidence Interval 38.5% - 98.3% 87.8% Point Estimate 95% Confidence Interval 11.7% 91.5% Point Estimate 95% Confidence Interval 6.5% - 15.6% 90.3% Point Estimate 95% Confidence Interval 59.1% - 99.5% 93.2% Point Estimate 95% Confidence Interval 10.9% 81.4% Point Estimate Place 1, 2008: Ponce 93.0% Point Estimate Percent of NonHispanic Voters 9.0% 30.7% - 87.4% 2.9% - 17.4% Ps. Ex. 44. Based on this analysis, Dr. Engstrom opined that voting in Irving ISD trustee elections is racially polarized. He observed that, in seven of the eight elections, the Hispanic candidate - 25 - received more than 80% of the Hispanic votes (and in the eighth election, the Hispanic candidate received nearly 60% of the Hispanic votes). He also observed that, in six of the eight elections, the lowest level of the confidence interval was still above 50%. Dr. Engstrom opined that the analysis shows that (1) Hispanic voters preferred the Hispanic candidate in each of the eight elections; (2) non-Hispanic voters preferred the non-Hispanic candidate in each of the eight elections; and (3) the lack of non-Hispanic support for the Hispanic candidate effectively functioned in every instance as a veto over the election of the Hispanic candidate. Defendants presented expert testimony from John Alford, Ph.D. ( Dr. Alford ), in an attempt to refute Dr. Engstrom s opinions. Dr. Alford did not perform an alternative analysis (i.e., an analysis using a different methodology or different data), nor did he criticize Dr. Engstrom s data or methodology as being unreliable. In fact, he testified that he performed an analysis to verify the results of Dr. Engstrom s study and found no substantive differences between his results and Dr. Engstrom s results. See Tr. 3:65. Instead, Dr. Alford attempted to undermine Dr. Engstrom s opinions by arguing that Dr. Engstrom s analysis largely pertained to the seven elections held between 2006 and 2012 elections that occurred before Irving ISD switched from the pure at-large system challenged in Benavidez I to the 5-2 system challenged here. Dr. Alford opined on this basis that, as a result, Dr. Engstrom s analysis was not significantly probative as to whether there is racially polarized voting under the current system. - 26 - C The court finds that plaintiffs have proved racial bloc voting through statistical evidence from eight elections,21 thus satisfying the second and third prong of Gingles.22 See Westwego III, 946 F.2d at 1118 ( Usually, plaintiffs in a vote dilution case will attempt to establish both the second and third Gingles factors with statistical evidence of racial polarization of the electorate. ). Plaintiffs have proved that Hispanic voters vote along racial lines. Id. at 1122. The point estimates, which are undisputedly the best estimates in the data, for Hispanic votes in favor of a Hispanic candidate establish a pattern of Hispanic bloc voting: 93.0% in the Place 3, 2006 election; 81.4% in the Place 4, 2006 election; 93.2% in the Place 1, 2008 election; 90.3% in the Place 2, 2008 election; 91.5% in the Place 5, 2010 election; 87.8% in the Place 3, 2012 election; 90.1% in the Place 4, 2012 election; and 58.0% in the District 5, 2013 election. The point estimates for seven of the eight elections establish overwhelming 21 It is sufficient under the circumstances for plaintiffs to have provided statistical evidence regarding eight elections, given that these eight are the most recent involving a choice between a Hispanic candidate and a non-Hispanic candidate. See Campos v. City of Baytown, Tex., 840 F.2d 1240, 1245 (5th Cir. 1988) (stating that the number of elections that must be studied in order to determine whether voting is polarized will vary according to pertinent circumstances, and holding that [t]he district court was warranted in its focus on those [five] races that had a minority member as a candidate (quoting Gingles, 478 U.S. at 57 n.25) (alteration omitted)); see also Westwego I, 872 F.2d at 1208 n.7 ( [T]he evidence most probative of racially polarized voting must be drawn from elections including both black and white candidates. ). 22 See Monroe v. City of Woodville, Miss., 897 F.2d 763, 764 (5th Cir. 1990) (per curiam) (stating that [s]tatistical proof of political cohesion is likely to be the most persuasive form of evidence, although other evidence may also establish this phenomenon ). - 27 - support by Hispanic voters for the Hispanic candidates. Cf. Gingles, 478 U.S. at 59 ( [B]lack voters support for black candidates was overwhelming in almost every election. In all but 5 of 16 primary elections, black support for black candidates ranged between 71% and 92%; and in the general elections, black support for black Democratic candidates ranged between 87% and 96%. ); Clark v. Calhoun Cnty., Miss., 88 F.3d 1393, 1397 (5th Cir. 1996) ( Clark II ) (holding that district court s finding that racially polarized voting exists is beyond question, and pointing to evidence that, inter alia, the black candidate received an estimated 71.6% of the black vote but only 7.8% of the white vote ). Although the point estimate for the 2013 election for the District 5 seat (58.0%) is lower than the rest,23 a pattern of racial bloc voting that extends over a period of time is more probative of a claim that a district experiences legally significant polarization than are the results of a single election. Gingles, 478 U.S. at 57 ( [I]n a district where elections are shown usually to be polarized, the fact that racially polarized voting is not present in one or a few individual elections does not necessarily negate the conclusion that the district experiences legally significant bloc voting. ); see also Teague, 92 F.3d at 288 ( [T]he results of a couple of elections do not discount the presence of racial bloc voting[;] . . . a showing that bloc voting is not absolute does not preclude a finding of racial polarization[;] . . . to be legally significant, racial polarization need only be the usual pattern for a significant proportion 23 In that election, the Hispanic candidate received only 9.0% of the non-Hispanic votes based on the point estimate, indicating that non-Hispanics still effectively vetoed the candidate preferred by a majority of Hispanic voters. - 28 - of voters. (quoting Gingles, 478 U.S. at 57) (citations, alterations, and some internal quotation marks omitted)). Moreover, Dr. Alford admitted on cross-examination that the data suggested a pattern of racially polarized voting. Tr. 3:65 ( And here s where I think Dr. Engstrom and I sort of agree but maybe don t quite get to the same point: I think we agree as to that pattern [of racially polarized voting]. ). Dr. Alford simply disagreed on the ultimate question whether, under the totality of circumstances, the data proved the 5-2 system caused vote dilution, in violation of § 2. Plaintiffs have also established a pattern of non-Hispanic bloc voting. The point estimates for non-Hispanic votes in favor of Hispanic candidates in these eight elections never exceeded 33%, and in most years did not exceed even 15%: 10.9% in the Place 3, 2006 election; 11.7% in the Place 4, 2006 election; 7.5% in the Place 1, 2008 election; 11.9% in the Place 2, 2008 election; 15.0% in the Place 5, 2010 election; 13.6% in the Place 3, 2012 election; 32.7% in the Place 4, 2012 election; and 9.0% in the District 5, 2013 election. Cf. Gingles, 478 U.S. at 59 (recognizing that district court had found a substantial majority of white voters would rarely, if ever, vote for a black candidate because, inter alia, [i]n the primary elections, white support for black candidates ranged between 8% and 50%, and in the general elections it ranged between 28% and 49% ); Campos v. City of Baytown, Tex., 840 F.2d 1240, 1249 (5th Cir. 1988) (holding that the evidence was sufficient to show that the whites vote sufficiently as a bloc to overcome the minority votes plus the white crossover vote and noting three election results that are illustrative: Mario Delgado received approximately 83% of the minority votes but only 37% of the Anglo vote and lost - 29 - the election. . . . Tony Campos received 63% of the minority vote but only 29% of the white vote and lost the election. . . . Ruby Hardy received approximately 78% of the minority vote but only 3% of the white vote and lost the election. ). Moreover, none of the points within the confidence intervals reached majority level the highest confidence interval is 37.7% for the 2012 election for Place 4. Thus the point estimates and the confidence intervals for these eight elections strongly support the finding that non-Hispanic voters generally do not provide crossover votes to Hispanic-preferred candidates in Irving ISD trustee elections. Accordingly, the court finds that plaintiffs have met their burden of satisfying the second and third preconditions of Gingles. VI Because plaintiffs have proved the threshold Gingles preconditions, the court now decides whether they have established that, under the totality of the circumstances, they do not possess the same opportunities to participate in the political process and elect representatives of their choice enjoyed by other voters. League of United Latin Am. Citizens, Council No. 4434 (LULAC) v. Clements, 999 F.2d 831, 849 (5th Cir. 1993) (en banc) ( LULAC v. Clements II ).24 24 Defendants appear to contend that, even if the Gingles factors apply to plaintiffs challenge to the Irving ISD 5-2 system, the fact that plaintiffs have established the Gingles preconditions does not of itself prove a violation of § 2. Plaintiffs do not dispute the premise that they must not only satisfy the Gingles preconditions but must also prove that, under the totality of the circumstances, Hispanic voters do not possess the same opportunities to participate in the political process and elect representatives of their choice. And it is well settled that the Gingles factors are necessary, but not sufficient, conditions for establishing a § 2 violation. See LULAC v. Clements II, 999 F.2d at 849 ( Satisfaction of these three - 30 - A Plaintiffs assert that the following factors establish a violation of § 2 under the totality of circumstances: (1) Irving ISD s trustee elections are racially polarized; (2) no Hispanic candidate has ever been elected to Irving ISD s Board of Trustees in a contested election against a non-Hispanic opponent; (3) there is a history of official discrimination in the City of Irving, as recognized by the court in Benavidez v. City of Irving, Texas, 638 F.Supp.2d 709 (N.D. Tex. 2009) (Solis, J.) ( Benavidez v. City of Irving ); (4) Irving ISD used a place system when its electoral system was a pure at-large system and continues to use a place system for the two at-large seats under the current 5-2 plan, and a place system tends to preconditions is necessary, but not sufficient to establish liability under § 2. (citations omitted)). Plaintiffs did rely at trial on the Fifth Circuit s statement in Clark I that it will be only the very unusual case in which the plaintiffs can establish the existence of the three Gingles factors but still . . . fail[] to establish a violation of § 2 under the totality of circumstances. Clark I, 21 F.3d at 97 (quoting Jenkins v. Red Clay Consol. Sch. Dist. Bd. of Educ., 4 F.3d 1103, 1135 (3d Cir. 1993)). Clark I, however, does not suggest that the totality of the circumstances is an empty formalism or that clearing the Gingles hurdles preordains liability. To the contrary, this final inquiry can be powerful indeed. At the same time, it is more than an intuitive call of the trial judge; the trial court must anchor its judgment in evidence. Clark II, 88 F.3d at 1396-97. In fact, when this court quoted the same observation from Clark I in its decision in Fabela, it followed with the finding that the Fabela plaintiffs ha[d] proved a voting rights violation under the totality of the circumstances. Fabela, 2012 WL 3135545, at *13. Plaintiffs must therefore prove by a preponderance of the evidence that, under the 5-2 system, Hispanic voters have less opportunity than other members of the electorate to participate in the political process and to elect members of their choice. - 31 - enhance the likelihood of vote dilution; (5) Hispanics residing in the Irving ISD exhibit lower educational attainment, lower income, and higher poverty rates factors that hinder their ability to participate in the political process; and (6) the policy underlying the 5-2 system is tenuous. Defendants do not specifically contest each factor on which plaintiffs rely. Instead, they assert on the following grounds that plaintiffs have not established a § 2 violation: (1) District 6 under the current 5-2 system is a district that ensures effective Hispanic representation; (2) the current 5-2 system has only been in effect since early 2012, and, as a result, there is insufficient evidence of vote dilution under the challenged system to find a § 2 violation; (3) the Irving ISD Board of Trustees voluntarily chose to convert from a pure at-large system to the current 5-2 system after the results of the 2010 Census were published, and the U.S. Department of Justice precleared the proposed plan; (4) in the only election in which a Hispanic candidate ran under the 5-2 plan, she won; and (5) the only way that plaintiffs are able to draw demonstration districts is by packing the districts with noncitizens, which has the effect of diluting votes of Hispanic citizens who live outside the demonstration districts. B Because defendants focus less on plaintiffs grounds for contending that they have met their burden of proof and more on reasons for asserting that the current 5-2 system survives § 2 scrutiny under the totality of the circumstances, the court will first analyze defendants contentions in § VI(B)-(E) and then in § VI(F) consider plaintiffs evidence and reasoning. - 32 - 1 Defendants primary argument is that the presence of District 6 under the current 5-2 system ensures effective Hispanic representation and thus precludes a finding of § 2 liability. Plaintiffs challenge this position in two ways. First, they argue that the data and methodology on which defendants rely to prove that District 6 is an effective district for Hispanic representation are unreliable. According to plaintiffs, the best available data and methodology prove that the share of Hispanic CVAP in District 6 is below 50%, and there is no evidence of sufficient crossover voting from non-Hispanic voters to ensure that Hispanic voters can consistently elect representatives of their choice. Second, plaintiffs posit in the alternative that, if the court were to find that the data and methodology on which defendants rely to prove that District 6 is an effective Hispanic district are sufficiently reliable, then that same data and methodology actually compel a finding that plaintiffs can draw two single-member districts with a majority Hispanic CVAP share. See supra at § III. According to plaintiffs reasoning, the current 5-2 system violates § 2 either because District 6 is not an effective Hispanic district or, if it is, because a second majority Hispanic CVAP district can and should be drawn. 2 Defendants rely on Dr. Rives s expert testimony to support their argument that District 6 is an effective district for Hispanic representation. Defendants acknowledge that two special tabulations the 2007-2011 five-year ACS data and another five-year ACS data set covering the years 2008-2012 ( 2008-2012 five-year ACS data ) indicate that the share of - 33 - Hispanic CVAP is 46.75% or 45.23%, respectively. But Dr. Rives testified that, because the most recent data in these five-year sets are approximately two years old, it is appropriate to consider the upward trend in the Hispanic population within the Irving ISD. Accordingly, he took one-year ACS data from 2005, 2006, 2007, 2008, 2009, 2010, 2011, and 2012 to calculate projections for 2014 and 2016. See Tr. 2:197-98. Based on this combination of one-year data, Dr. Rives derived two projections: one based on the one-year data files from 2005 to 2012, and one based on the one-year data files from 2005 to 2011. He created this second projection, which excluded the 2012 one-year ACS data, because, in his view, the 2012 data file must contain an error, since it reflects a disproportionately large drop in the percentage of Hispanic voting age population ( VAP ), and such a drop cannot be reconciled with the overall trend in the data. For each projection, he distinguished two measures: a computed share and a trended share. The distinction between computed share and trended share, which is explained in Dr. Rives s expert report, see Ds. Ex. 43 at 4-5 & n.1; Ds. Ex. 17 (Note 3), is immaterial to the court s decision. Dr. Rives s estimates, which pertain to District 6 in the current 5-2 system, are as follows: - 34 - District 6 2014 Projection without ACS 2012 2014 Projection with ACS 2012 2016 Projection without ACS 2012 2016 Projection with ACS 2012 Hispanic CVAP 6,228 5,673 6,710 6,038 Non-Hispanic CVAP 5,002 5,682 4,748 5,533 Hispanic Share of CVAP (Computed Share) 55.46% 49.96% 58.56% 52.18% Hispanic Share of CVAP (Trended Share) 60.23% 49.88% 64.91% 52.94% Ds. Ex. 17. Defendants rely on these estimates to argue that District 6 is currently an effective district for Hispanic representation because they show that the share of Hispanic CVAP is at least 49.88% (2014 Projection with ACS 2012 data, trended share) and more likely higher than 50% (based on the other estimates). Plaintiffs introduced Ely s expert testimony to refute the reliability of Dr. Rives s data and methodology. According to Ely, Dr. Rives s approach is unreliable for several reasons: (1) he applied a uniform district-wide rate of growth to individual sub-populations without determining whether the component sub-populations were changing over time in a way similar to the district-wide rate of growth; (2) he combined a series of one-year data files to perform his linear regression instead of using the more reliable pooled five-year data files, - 35 - each of which possesses a higher indicia of reliability; (3) his model projects a trend out to 2014 and 2016 based on data that, at most, extend only to 2012 (the most recent one-year ACS data), which means that there is relatively little overlap between the period in which the trend is established and the period of the projection; and (4) Dr. Rives s decision to exclude the 2012 one-year ACS data because they are an outlier is ad hoc and not supported by a reliable statistical methodology.25 Ely also explained two fundamental differences between the data and methodology he used for his analysis, discussed above in § IV(A), and the data and methodology used by Dr. Rives: (1) Ely s approach did not rely on any projections about future growth, and (2) Ely relied exclusively on five-year pooled data (and did not rely on one-year data for any of his analysis). 3 The court finds that Dr. Rives s methodology and data are too unreliable to support a finding that District 6 contains a majority Hispanic CVAP share. It is undisputed that, based on two sets of pooled five-year data the 2007-2011 five-year ACS data and the 20082012 five-year ACS data the share of Hispanic CVAP in District 6 is less than 50%. Although Dr. Rives testified that his calculations differed slightly from Ely s when analyzing 25 Ely also criticized Dr. Rives s original methodology, as described in his initial expert report, for taking a linear estimate for population growth and applying it in a non-linear way. But Dr. Rives testified that he adjusted his methodology to account for this criticism after Ely s deposition was taken, and he clarified that his final projections were not subject to the same criticism. Ely appeared to acknowledge this revision during trial and did not argue that Dr. Rives s final projections were flawed for this particular reason. Accordingly, the court does not rely on this criticism as a basis for its decision. - 36 - these two five-year special tabulations, he nevertheless concluded that, based on these data, the share of Hispanic CVAP is less than 50%. See Tr. 2:127, 129. Defendants acknowledge that both the 2007-2011 five-year ACS data and 2008-2012 five-year ACS data support the finding that the share of Hispanic CVAP in District 6 is less than 50%, but they argue that the court should rely instead on Dr. Rives s approach, which combines multiple one-year data files and depends on projections into the future. But defendants have not sufficiently responded to Ely s criticisms of this approach. In particular, they have failed to explain why a methodology that relies on the combination of separate one-year data files, which exhibits higher variability year-to-year (as reflected by the 2012 ACS data, which Dr. Rives urges the court to ignore), and uncertain growth projections into the future is more reliable than a methodology that is based entirely on pooled five-year data and does not rely on any future growth projections. In addition, Dr. Rives s own testimony was somewhat equivocal on whether District 6 currently contains a majority Hispanic CVAP share. When asked whether he believed that the share of Hispanic CVAP in District 6 is above 50% today (2014), he answered: Yes. Based upon . . . this analysis, just looking at the momentum of Hispanic and non-Hispanic growth for the school district as a whole and Hispanic citizen-voting-age population growing faster over the last eight or nine years . . . than nonHispanic and and assuming . . . that that trend were to continue for a few more years, and that what we re observing at the school district level is is somehow relevant to a district that s, what, one fifth of the size of the of the school district, then with those two stipulations there they re above 50 [percent], and I I think that s probably where they are. I say and I want to underscore probably, because I m not betting - 37 - on these percentages, but they certainly in 2016 they re well above. Tr. 2:155 (emphasis, ellipses, and bracketed material added). He also testified: [T]his is not a case where I said [the Hispanic CVAP share] definitely is and the percentage is X. I didn t say that. I said . . . the trend is up, whether you looked at the numbers or whether you put a line of regression through them, the trend is up and you re you know, you re going from where were we, 2013 well, let s see, we were back I guess around 2009. You re going to 2014 and 2016, so you ve got a little bit of time there where other things can happen. I said . . . if the trend were to continue and if the trend we observe at the school district level has some applicability to . . . the trustee district[,] then one has to wonder whether a district that was 40 in the high 40 percent already in the 2011 file might actually be majority Hispanic, what, two or three years later and and then again in 2016. . . . I m not saying it is. But I m saying one has to wonder because . . . it s sort of compelling to say, well, the trend is up, it s a fairly strong trend, the non-Hispanic [trend] is going the other way, and so the . . . sort of conditions look good for majority Hispanic number today and then a couple years down the road. Id. at 2:182 (emphasis, ellipses, and bracketed material added). As his answers implicitly concede, Dr. Rives s conclusions are based on two significant assumptions: (1) that the trend observed in the past will continue in the future, and (2) that the rates of growth observed for Hispanic and non-Hispanic populations district-wide are comparable to the rates of growth for Hispanic and non-Hispanic populations within the single-member districts of the Irving ISD. And Dr. Rives admitted that, based on the pooled five-year data on which Ely relied - 38 - for his assessment, see supra § IV(A), there is no evidence that District 6 currently contains a majority Hispanic CVAP share. See id. at 2:127, 129. In many respects, the court s conclusion on this issue reflects the rationale of Benavidez I. There, the presumptively correct 2000 Census data indicated that Benavidez could not draw a demonstration district with a share of Hispanic CVAP greater than 50%. Benavidez I, 690 F.Supp.2d at 457 ( By Ely s own admission, none of the illustrative districts on which Benavidez relies has greater than a 50% Hispanic share of CVAP according to 2000 Census data. ). Benavidez attempted to overcome this problem by relying on 2007 one-year ACS data, and the court held that he had failed to demonstrate that the oneyear data were more reliable than the older 2000 Census data. Id. at 459 ( Considering the large margins of error for the 2007 estimates of the Hispanic CVAP, the court finds that Benavidez has failed to prove that the 2007 ACS one-year data are sufficiently reliable to overcome the presumption that the 2000 Census is correct. ).26 The court noted: Ely relies on the uncertain 2007 estimates in calculating the growth rate to project the 2008 population size, making his estimated growth rate and the 2008 population estimates correspondingly unreliable. Id. Here, the roles are reversed, but the rationale of Benavidez I applies. The more reliable data the 2007-2011 five-year ACS data and the 2008-2012 five-year ACS 26 It is somewhat ironic that defendants now rely on one-year data and future growth projections given that they criticized the plaintiff s reliance on the same in Benavidez I. See Benavidez I, 690 F.Supp.2d at 459. - 39 - data indicate that the share of Hispanic CVAP in District 6 is less than 50%. Defendants attempt to overcome this problem by relying on a series of one-year data and future growth projections to prove that the share of Hispanic CVAP is greater than 50%. But similar to the plaintiff in Benavidez I, defendants have not demonstrated that estimates based on a series of one-year data and uncertain growth projections are more reliable than those derived from two sets of pooled five-year data. That Dr. Rives s data are more recent does not necessarily make them more reliable. See id. at 457-58 (explaining why three- and five-year pooled estimates are generally more reliable than one-year estimates under ACS s own guides and concluding that 2000 Census data were more reliable than 2007 one-year ACS data despite their older age). In sum, the court finds that the best available data and methodology indicate that the share of Hispanic CVAP in District 6 is less than 50%.27 4 That the share of Hispanic CVAP in District 6 is less than 50% is not necessarily dispositive of the question whether District 6 is an effective district for Hispanic representation.28 That question depends on the totality of circumstances. In theory, Hispanic 27 Neither the 2007-2011 five-year ACS data nor the 2008-2012 five-year ACS data enjoy the presumption of correctness that the 2000 Census data received in Benavidez I. But this difference is immaterial to the court s conclusion. Here, the two five-year data sets are sufficiently reliable to justify the court s conclusion without resort to any presumptions about their correctness. And defendants have not explained why a series of one-year data is more reliable than the pooled five-year data. 28 In other words, that plaintiffs can prove that a single-member district with a majority Hispanic CVAP share can be drawn, and that defendants have not drawn such a district, does not of itself prove that the 5-2 system causes vote dilution, in violation of § 2. - 40 - voters in District 6 might have an equal opportunity to participate in the political process and elect representatives of their choice if, for example, there were substantial crossover voting from non-Hispanic voters. Cf. Bartlett, 556 U.S. at 24 ( States can and in proper cases should defend against alleged § 2 violations by pointing to crossover voting patterns and to effective crossover districts. Those can be evidence, for example, of diminished bloc voting under the third Gingles factor or of equal political opportunity under the § 2 totality-of-the-circumstances analysis. ). But the trial evidence does not support such a finding. And, as discussed supra in § V(B)-(C), plaintiffs have proved that Irving ISD trustee elections are racially polarized, with a low degree of crossover voting from nonHispanic voters. The only other affirmative evidence that District 6 is an effective district for Hispanic representation is the undisputed fact that Gonzales, a Hispanic, was elected in the first (and only) election for the District 6 seat that has taken place under the 5-2 system. The court finds that Gonzales s election is a special circumstance, and that it does not materially undermine plaintiffs showing that District 6 is not an effective district for Hispanic representation.29 29 Defendants argue that, when Gonzales resigned from her position, no Hispanic candidate applied for appointment to replace her. They suggested at trial that the lack of a Hispanic trustee member should be attributed to the failure of Hispanics to run for office in the first place rather than to illegal vote dilution, and that this factor cuts in their favor under the totality of circumstances. The court disagrees. It is reasonable to infer from the evidence that no Hispanic applied for the position based on the belief that his or her service could be short-lived, because it would be necessary to run for election at the next election, and Hispanics are historically defeated when they run against non-Hispanics. According to the - 41 - The Supreme Court has cautioned that special circumstances . . . may explain minority electoral success in [an otherwise] polarized contest, and that such aberrational victories do not necessarily disprove racial vote dilution. Rodriguez v. Bexar Cnty., Tex., 385 F.3d 853, 864 (5th Cir. 2004) (quoting Gingles, 478 U.S. at 57).30 As explained in Gingles, the special circumstances analysis was designed to prevent defendant jurisdictions from arguing that a minority candidate s occasional victory in an otherwise racially polarized electorate defeats a vote dilution claim. Id. (citing Gingles, 478 U.S. at 57). Here, the court finds that Gonzales s election is a special circumstance for two reasons: (1) she ran unopposed, and (2) her election occurred in the post-litigation context, because this lawsuit had already been filed.31 Dr. Engstrom testified that Gonzales s election was a special circumstance because she ran unopposed. He observed that, since her election, no other Hispanic candidate has run unopposed. And he explained that her election does not undermine his opinions concerning the presence of racially polarized voting because the election of an unopposed candidate does not present voters with a choice between a Hispanic candidate and a non-Hispanic candidate. trial testimony of Jerry Christian, a former Irving ISD trustee, there is deep-seated racial prejudice in the Irving ISD community, and this prejudice is reflected in board elections. See Tr. 1:72-73. Accordingly, defendants reliance on the failure of a Hispanic to apply for Gonzales s vacated position does not favor defendants and arguably favors plaintiffs. 30 The Fifth Circuit has clarified that a special circumstances analysis cannot be used to explain away the consistent success of Hispanic candidates in a number of races over multiple election cycles. Rodriguez, 385 F.3d at 864. Here, the court applies the special circumstances analysis to the election of one candidate in a single election. 31 Plaintiffs filed suit in January 2013. Gonzales was elected in May 2013. - 42 - Dr. Engstrom also testified that Gonzales s election was a special circumstance because it occurred during the post-litigation context, and litigation can potentially affect the candidate pool. Defendants did not present any evidence rebutting Dr. Engstrom s testimony on this issue. The court therefore finds that plaintiffs have proved under the totality of circumstances that, under the current 5-2 plan for electing Irving ISD trustees, District 6 is not an effective district for Hispanic representation. The court concludes that the creation of District 6 does not preclude plaintiffs § 2 claim, and it proceeds to evaluate the remaining factors under the totality of circumstances. C Defendants contend that the current 5-2 system has only been in effect since early 2012 and, as a result, there is insufficient evidence of vote dilution under the challenged system to establish a § 2 violation. The court disagrees. Although the recency of the 5-2 system is a relevant consideration, this fact does not completely displace or outweigh evidence of vote dilution in the pre-2012 context. There is no evidence that the 5-2 system materially changes the past realities of voting within the Irving ISD e.g., whether Hispanics vote cohesively, or whether non-Hispanics vote as a bloc. Cf. Westwego III, 946 F.2d at 1120 (noting that determination whether electoral system violates § 2 depends, inter alia, on an evaluation of the past and present reality and on a functional view of the political - 43 - process (quoting Gingles, 478 U.S. at 45) (internal quotation marks omitted)).32 D Defendants argue that the Board of Trustees voluntarily chose to convert from a pure at-large system to the current 5-2 system, and that the U.S. Department of Justice precleared the proposed plan. They maintain that, because the plan was precleared, the 5-2 system does not cause vote dilution, in violation of § 2. The court disagrees. Section 5 [of the Act] . . . forbids voting changes with any discriminatory purpose as well as voting changes that diminish the ability of citizens, on account of race, color, or language minority status, to elect their preferred candidates of choice. Shelby Cnty., Ala. v. Holder, ___ U.S. ___, 133 S. Ct. 2612, 2621 (2013) (quoting 42 U.S.C. §§ 1973c(b)-(d)).33 It requires covered jurisdictions to obtain preclearance for any proposed change in its voting plans. Id. Here, it is undisputed that defendants applied for and received preclearance from the U.S. Department of Justice before implementing the 5-2 system. Contrary to defendants suggestion, the Supreme Court has repeatedly recognized that § 2 and § 5 have different aims with different requirements, and that a change that is permissible under § 5 may in fact violate § 2. See Georgia v. Ashcroft, 539 U.S. 461, 477 32 At times throughout trial, defendants appeared to suggest that any evidence about voting patterns before February 2012 was irrelevant or lacked probative value simply because it pertained to a different electoral system. Defendants have cited no authority, and the court is aware of none, that stands for such a sweeping proposition. 33 Defendants obtained preclearance before the Supreme Court decided Shelby County. Because it is immaterial to the court s decision, the court will analyze defendants argument as if § 5 were still in effect and fully enforceable against Irving ISD. - 44 - (2003) ( Thus, a plan that merely preserves current minority voting strength is entitled to § 5 preclearance. Indeed, a voting change with a discriminatory but nonretrogressive purpose or effect does not violate § 5. (citations omitted)); Reno v. Bossier Parish Sch. Bd., 528 U.S. 320, 375 (2000) ( As we have repeatedly noted, in vote-dilution cases § 5 prevents nothing but backsliding, and preclearance under § 5 affirms nothing but the absence of backsliding. ); Reno v. Bossier Parish Sch. Bd., 520 U.S. 471, 480 (1997) ( But § 5, we have held, is designed to combat only those effects that are retrogressive. ). As the Court observed in City of Lockhart v. United States, 460 U.S. 125 (1983), even a plan that could have a discriminatory effect under some circumstances, id. at 135, is still entitled to preclearance under § 5 if it does not increase the degree of discrimination against [the minority population], id. at 134 (emphasis added). As these cases make clear, the fact that a plan lacks a retrogressive effect (and is thus entitled to preclearance) does not necessarily imply that the resulting electoral system is permissible under § 2, which prohibits vote dilution. Accordingly, although the fact that a plan has been precleared under § 5 can be considered when evaluating the totality of the circumstances, the court finds from the evidence presented here that the fact that defendants received preclearance for the 5-2 system does not weigh strongly in defendants favor. E Defendants posit that plaintiffs cannot establish § 2 liability because the demonstration districts on which they rely are packed with non-citizens, and, as a result, dilute the votes - 45 - of Hispanic citizens who live outside the demonstration districts.34 In a proposed conclusion of law, they cite Concerned Citizens for Equality v. McDonald, 63 F.3d 413, 417 (5th Cir. 34 There are references throughout the record to an alleged denominator problem with plaintiffs numbers. For example, in a proposed finding of fact, defendants state: Although the other four districts contain comparable numbers of adult Hispanic citizens, they do not approach majority Hispanic CVAP status because they do not contain as many non-citizens whose presence in the district would reduce the denominator. Ds. Prop. Findings of Fact No. 48; see also Tr: 2:13, 2:92-93, 2:150-51, 3:46-47 & 3:148. Although defendants do not clearly explain the implication of their argument, the court understands the objection as follows and finds, for the reasons stated in § VI(F), that it does not preclude a finding of § 2 liability. Under traditional districting principles, a sovereign must divide the total population of the jurisdiction by the number of districts it has chosen to create and then allocate the resulting shares across the districts in roughly equal fashion. For example, in an n-district system, where n represents the number of districts to be created, each district should have roughly 1/nth of the total population. Calculating the share of Hispanic CVAP in a district is a separate inquiry. To calculate the share (i.e., the percentage) of Hispanic CVAP in a district, one divides the number of Hispanic citizens of voting age by the sum of the number of Hispanic citizens of voting age plus the number of non-Hispanic citizens who are of voting age. (Non-citizens are not part of the equation i.e., they are not in either the numerator or the denominator.) One way of increasing the share of Hispanic CVAP in a district is by including more non-citizens in the district s total population. This effect occurs because each non-citizen will count for purposes of satisfying the 1/nth rule discussed above, but will not count in the calculation of Hispanic CVAP share. As a result, the inclusion of each additional non-citizen in a district effectively bumps out a non-Hispanic citizen of voting age, thereby increasing the ratio of Hispanic CVAP to non-Hispanic CVAP in the district. In other words, because the district s total population remains constant at 1/nth, each non-citizen who is added to the district effectively replaces a non-Hispanic citizen of voting age, thereby increasing the share of Hispanic CVAP. Thus defendants argument that adding non-citizens to the district reduces the denominator is somewhat imprecise. Because non-citizens are not part of the equation for calculating the share of Hispanic CVAP, plaintiffs have not reduced the denominator. What defendants really appear to mean is that, by adding non-citizens to the district s total population, plaintiffs have indirectly caused the number of non-Hispanic citizens of voting age to decrease. - 46 - 1995), for the proposition that a plaintiff may not circumvent the first prong of Gingles by artificially reducing the number of minority group members necessary to constitute a majority in a district through assuming a larger number of districts, and hence a smaller district population, than currently exists. Ds. Prop. Concl. of Law No. 23. And they argue, by analogy, that the reasoning of McDonald precludes attempted circumvention of Gingles by packing the district with non-citizens so that fewer citizens are required to establish a citizen-voting-age population majority. Id. The court disagrees. McDonald does not support defendants position. In McDonald the plaintiffs challenged an electoral system with four single-member districts. McDonald, 63 F.3d at 415. They were unable to draw a majority-minority district under the four-district model, so they proposed a hypothetical five-district system in which there was one majority-minority district. Id. The Fifth Circuit held that the plaintiffs had failed to satisfy the first prong of Gingles. It reasoned that a § 2 plaintiff cannot simply satisfy this precondition by proposing an alternative electoral system that would increase the total size of the governmental body (i.e., the total number of elected seats to be filled). If this were the rule, a § 2 plaintiff could circumvent the first prong by invok[ing] hypothetical mutations and transfigurations of the existing political structure. Id. at 417. Here, plaintiffs theory does not run afoul of McDonald. Under the existing 5-2 system, there are already seven trustees. Plaintiffs theory of the case is that, given seven single-member districts, it is possible to draw one such district with a majority Hispanic CVAP share. They do not rely on an alternative electoral system with more than seven trustee positions, and, as a result, they have not invoked a - 47 - hypothetical mutation or transfiguration of the existing political structure of the Irving ISD. Defendants reliance on the reasoning of McDonald is also unavailing. The presence of a large number of non-citizens in the demonstration district does not render that district invalid. Courts have consistently held that, when there is evidence of a relatively large non-citizen population, plaintiffs must establish the first prong of Gingles by proving that a district can be drawn with a majority-minority CVAP share. See Reyes, 586 F.3d at 1023 ( Indeed, several sister Circuits have joined the Fifth in requiring voting rights plaintiffs to prove that the minority citizen voting-age population comprises a majority. ); see also id. at 1023 n.12 (collecting cases). Indeed, the reason that courts have insisted on the use of CVAP rather than VAP is because, absent this requirement, a plaintiff might be able to satisfy the first Gingles prong by relying on VAP alone, even though the minority group actually lacks the potential to elect a representative of its choice because there are insufficient eligible minority voters in the jurisdiction. See Campos v. City of Houston, 113 F.3d 544, 548 (5th Cir. 1997); Brewer v. Ham, 876 F.2d 448, 452 (5th Cir. 1989) ( It would be a Pyrrhic victory for a court to create a single-member district in which a minority population dominant in absolute, but not in voting age numbers, continued to be defeated at the polls. (quoting Overton v. City of Austin, 871 F.2d 529, 542 (5th Cir. 1989) (Jones, J., concurring))). None of these cases, however, suggests that a plaintiff whose demonstration district includes a high number of non-citizens necessarily fails to satisfy the first prong of Gingles or that it otherwise precludes a finding of § 2 liability. See Fabela, 2012 WL - 48 - 3135545, at *6 n.13 (rejecting substantially similar argument and noting that defendants do not cite any binding decision that holds that a plaintiff cannot satisfy the first prong of Gingles by including a particular number of non-citizens in the demonstration district. ). And so long as the demonstration district is geographically compact and does not exhibit a population deviation that would violate the one person, one vote requirement of the Equal Protection Clause,35 the mere fact that the demonstration district has been drawn to include more non-citizens than an extant district under the challenged electoral system does not render the demonstration district invalid for purposes of satisfying the first Gingles prong or establishing § 2 liability.36 The court finds that the record does not support the premise that the only way of drawing a single-member district with a majority Hispanic CVAP share would cause vote dilution in another single-member district or a violation of the one-person, one-vote principle. Defendants suggestion to the contrary is not supported by the trial 35 See Connor v. Finch, 431 U.S. 407 (1977); Reynolds v. Sims, 377 U.S. 533 (1964); Fairley, 584 F.3d at 675 ( If a population deviance exceeds 10%, it constitutes a prima facie case of invidious discrimination that requires the municipality to prove a legitimate reason for the discrepancy. ). 36 At trial, defendants appeared at times to recognize this fact. During closing argument, for example, defendants counsel stated: Noncitizens are going to be present in any plan the city draws or the plaintiffs draw. We don t contend it s illegal to draw plans that have large numbers of . . . noncitizens if there is a natural concentration of them. But it does have the effect of causing imbalance in the weight of the vote. Tr. 3:148 (ellipsis added). In addition, Dr. Rives testified that District 6 under the current 5-2 system has a relatively high number of non-citizens compared to the other districts, id. at 3:12, and that any plan drawn to create a district with a high share of Hispanic CVAP would inevitably include a high number of non-citizens because of the particular demographics of the Irving ISD, id. at 3:13-14. - 49 - evidence. F The court now turns to the grounds on which plaintiffs rely, and it finds that they have proved a § 2 violation under the totality of circumstances. It is well-established that the existence of racially polarized voting and the extent to which minority group members have been elected to public office are the most important factors to be considered in a totality determination. Fabela, 2012 WL 3135545, at *13 (citing Gingles, 478 U.S. at 48 n.15). Both of these factors favor plaintiffs. Plaintiffs have proved that the elections for Irving ISD trustees are moderately to highly racially polarized, because Hispanic candidates receive support from an estimated 58.0% to 93.2% of Hispanic voters compared to only 7.5% to 32.7% of non-Hispanic voters. The parties have stipulated that [n]o Hispanic has ever been elected to the Irving ISD Board in a contested election against a non-Hispanic opponent. Pretrial Order ¶ 9 (stipulation of fact). Plaintiffs have also presented the testimony of Dr. Engstrom that Irving ISD s 5-2 system uses a place system for the two at-large positions. According to Dr. Engstrom, a place system is one in which candidates run for a particular position, and each voter can only vote once for each open position. A place system makes it more likely that majority voters can control access to the seats up for election, because minority voters will not be able to use single-shot or bullet voting, a practice where minority voters vote only for their mostpreferred candidate and withhold their other votes, thereby increasing the relative significance of the votes they actually cast. Because voters only get one vote each for the - 50 - two at-large seats under the 5-2 system, minority voters are precluded from using single-shot voting because they are not given other votes to withhold. Courts have consistently recognized that single-shot voting increases the likelihood that minority candidates are elected, see, e.g., Gingles, 478 U.S. at 38 n.5, and that mechanisms that prevent single-shot voting are an enhancing factor within the meaning of Senate Factor 3, see, e.g., Westwego III, 946 F.2d at 1113 n.3; accord Salas v. Sw. Tex. Jr. Coll. Dist., 964 F.2d 1542, 1544 n.1 (5th Cir. 1992) ( [A numbered-post system] prevents the use of bullet, or single shot, voting. ) (quoting Campos, 840 F.2d at 1242 n.1). This factor weighs in favor of plaintiffs, because a place system can enhance the difficulty faced by Hispanics in seeking access to the political process in the Irving ISD. See Gingles, 478 U.S. at 45. Plaintiffs also presented testimony from Jerry Christian ( Christian ), a non-Hispanic, who served as an Irving ISD trustee from 2004 to 2013. Christian testified that, during his tenure as a trustee, he observed significant growth in the Hispanic population in the Irving ISD, and that some people seemed uncomfortable with this demographic change. He testified that he believed there was deep-rooted racial prejudice in the community, and that such prejudice was reflected in school board elections. See Tr. 1:72-73. And he stated that, at one point, he doubted whether a Hispanic could ever get elected to the Board of Trustees. Id. at 1:75. Although this uncontroverted evidence does not fit neatly into any particular factor,37 it supports the court s finding that voting in Irving ISD trustee elections is racially 37 Because the list of factors enumerated in the Senate Report is not exclusive, the court is free to consider this evidence under the totality of circumstances. See Gingles, 478 U.S. - 51 - polarized, and adds probative context to the stipulated fact that no Hispanic candidate has ever been elected as an Irving ISD trustee in a contested election against a non-Hispanic opponent. Plaintiffs have also proved that Hispanics residing in the Irving ISD exhibit lower educational attainment, lower income, and higher poverty rates. Ely testified and it is uncontroverted that ACS data bear out these trends for the Hispanic population in the Irving ISD, see Ps. Ex. 48, and that these characteristics can be evidence of lingering effects of discrimination. These lingering effects can hinder the ability of Hispanics to participate effectively in the political process. See, e.g., Fairley, 584 F.3d at 672-673. The court finds that this factor favors plaintiffs, although the court does not give it special or controlling emphasis.38 Plaintiffs proffer other arguments for a finding of a voting rights violation under the at 45 ( The [Senate] Report stresses, however, that this list of typical factors is neither comprehensive nor exclusive. ); LULAC v. Clements I, 986 F.2d at 753 ( The list of relevant factors in the Senate Report is not exclusive. ); Westwego III, 946 F.2d at 1120 (noting that other factors may be relevant ). 38 Plaintiffs are not required to prove a causal nexus between socioeconomic status and depressed participation in the political process, but they must prove that participation in the political process is in fact depressed among minority citizens. See LULAC v. Clements II, 999 F.2d at 867 ( As this statement discloses, the Senate Report, while not insisting upon a causal nexus between socioeconomic status and depressed participation, clearly did not dispense with proof that participation in the political process is in fact depressed among minority citizens. ). Here, there is uncontroverted evidence in the trial record that Hispanics participation in the political process is depressed in the Irving ISD. See, e.g., Tr. 2:80. - 52 - totality of the circumstances, but the court need not address them in detail.39 The two most important the existence of racially polarized voting and the extent to which minority group members have been elected to public office clearly support plaintiffs. And the other evidence the court has considered in § VI(F) demonstrates that plaintiffs have satisfied their burden of proof. Plaintiffs have proved that, under the totality of the circumstances, Hispanic voters in the Irving ISD have less opportunity than other members of the electorate to participate in the political process and to elect representatives of their choice. VII The court holds that the 5-2 system of electing members of Irving ISD s Board of Trustees violates § 2 of the Act.40 Defendants are therefore ordered to submit, within 60 days of the date of this memorandum opinion and order, a plan to remedy the violation. See Westwego III, 946 F.2d at 1124 ( [I]t is appropriate to give affected political subdivisions at 39 Although plaintiffs urge the court to consider Benavidez v. City of Irving, 638 F.Supp.2d 709, the court declines to rely on it as establishing that there is a history of official discrimination against Hispanics in the Irving ISD. The facts and circumstances of that case were different, and plaintiffs have not produced sufficient evidence of official discrimination to find that this factor favors plaintiffs in this case. 40 Defendants argued at trial that their failure to adopt the maximally-inclusive electoral system is not sufficient to establish § 2 liability. Nothing in this memorandum opinion and order should be interpreted to suggest otherwise. The question in this case is not whether the 5-2 system is maximally inclusive. See Johnson v. De Grandy, 512 U.S. 997, 1016 (1994) ( But reading the first Gingles condition in effect to define dilution as a failure to maximize in the face of bloc voting (plus some other incidents of societal bias to be expected where bloc voting occurs) causes its own dangers, and they are not to be courted. ). The question is whether, under the totality of the circumstances, the 5-2 system denies Hispanics an equal opportunity to participate in the political process and to elect representatives of their choice. - 53 - all levels of government the first opportunity to devise remedies for violations of the Voting Rights Act. ); see also E. Jefferson Coal. for Leadership & Dev. v. Parish of Jefferson, 703 F. Supp. 28, 29 (E.D. La. 1989) ( Only when the legislative plan is uncorrectably invalid should the court consider a scheme submitted by a private litigant or formulate its own plan. ), aff d, 926 F.2d 487 (5th Cir. 1991). In imposing this obligation on defendants, the court notes the following. First, the proposed remedy need not include the creation of one of plaintiffs proposed illustrative districts. See Clark I, 21 F.3d at 95 (citing Westwego III, 946 F.2d at 1124). Second, the court is not suggesting that defendants must abandon the 5-2 system. [A]t-large election schemes . . . are not per se violative of minority voters rights. Benavidez I, 690 F.Supp.2d at 456 (quoting Gingles, 478 U.S. at 48). And plaintiffs did not prove that it is impossible to draw an effective Hispanic district while retaining a 5-2 system; they only proved that District 6 is insufficient as currently drawn. But it is noteworthy that defendants own data indicate that the Hispanic population in the Irving ISD has continued to grow since the 2010 Census, suggesting that, in time, a 7-0 system composed of seven single-member districts and no at-large positions may be necessary to ensure compliance with § 2. Third, although defendants are not legally required to create a single-member district that has a share of Hispanic CVAP that is greater than 50%, if they do not, they must be able to demonstrate on the basis of sufficiently reliable data that there is sufficient crossover voting from non-Hispanic voters or some other set of circumstances that will ensure effective - 54 - Hispanic representation under the proposed plan.41 Fourth, in light of the court s finding that District 6 is inadequate as presently configured, the court will not approve a proposed remedy that only adjusts immaterially the boundaries of District 6. Plaintiffs may file objections to the proposal within 30 days of the date it is filed with the clerk of court. SO ORDERED. August 15, 2014. _________________________________ SIDNEY A. FITZWATER CHIEF JUDGE 41 As explained in this decision, the evidence presented in this trial would not support such a finding. - 55 -

Some case metadata and case summaries were written with the help of AI, which can produce inaccuracies. You should read the full case before relying on it for legal research purposes.

This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.