Summary
In Reyes Judge O'Connor found against a group of plaintiffs who brought a similar § 2 challenge to the at-large method of electing members of the Farmers Branch City Council.
Summary of this case from Fabela v. City of Farmers Branch, TexasOpinion
CIVIL ACTION NO. 3:07-CV-900-O.
November 4, 2008
MEMORANDUM OPINION AND ORDER
This is a voting rights case. Plaintiffs Valentine Reyes, Irene Gonzalez, and Gary Garcia (collectively, "Plaintiffs") allege that the at-large election of council members for the city of Farmers Branch dilutes the voting rights of Hispanic citizens in violation of Section 2 of the Voting Rights Act ("VRA"), 42 U.S.C. § 1974, and the Fourteenth Amendment. Doc. No. 1 (3:07-CV-900-O) (N.D. Tex. May 21, 2007) (Plaintiffs' complaint). Plaintiffs seek a permanent injunction prohibiting Farmers Branch from holding elections under the present at-large system in favor of single-member electoral districts or cumulative voting. Id. The parties presented evidence to the Court on May 27-28, 2008 regarding these claims.
Plaintiffs' complaint also alleges violation of Section 5 of the VRA. However, Plaintiffs stated at a pretrial conference that they have abandoned their § 5 claims. Plaintiffs did not present evidence on these claims at trial.
This opinion will cite to the trial transcript using the volume number ("Vol. 1" or "Vol. 2") and at ___, to indicate the page number of the particular volume of the transcript being referenced."
Because resolution of a voting dilution claim requires close analysis of unusually complex factual patterns, and because the decision of such a case has the potential for serious interference with state functions, district courts must explain with particularity their reasoning and the factual conclusions underlying such reasoning. Westwego Citizens for Better Gov't v. City of Westwego, 872 F.2d 1201, 1203 (5th Cir. 1987). In accordance with the requirements of Rule 52 of the Federal Rules of Civil Procedure and Fifth Circuit precedent, the Court now examines the applicable law and sets out its findings of fact and conclusions of law.
I. Legal Framework under the VRA
Section 2 of the Voting Rights Act, 42 U.S.C. § 1973, as amended, provides in subsection (a) that "no voting qualification or prerequisite to voting or standard, practice, or procedure shall be imposed or applied by any State or political subdivision in a manner which results in denial or abridgement of the right of any citizen of the United States to vote on account of race or color . . ." A violation of subsection (a) is established
"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 members of a class of citizens protected by subsection (a) of this section 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, (t)hat nothing in this section establishes a right to have members of a protected class elected in numbers equal to their proportion in the population." Id.
Thus, ultimately, the Court must inquire as to whether minority voters have demonstrated an unequal opportunity to participate in the political process on account of race or ethnicity. See League of United Latin Am. Citizens Council No. 4434 v. Clements, 986 F.2d 728, 742 (5th Cir. 1993) (en banc).
The Supreme Court elaborated on the basic analytical framework established by Section 2 of the VRA in Thornburg v. Gingles, 478 U.S. 30 (1986). Clements, 986 F.2d at 742. In Gingles, the Court set forth a three-part threshold test for analyzing claims that an at-large election scheme dilutes minority voting strength. Id. Under this threshold test, the minority group challenging an at-large election scheme must demonstrate 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 majority votes sufficiently as a bloc to enable it, in the absence of special circumstances, usually to defeat the minority group's preferred candidate. Id. at 742-43. Failure to establish any one of the Gingles factors precludes a finding of vote dilution, because "(t)hese circumstances are necessary preconditions for multimember districts to operate to impair minority voters' ability to elect representatives of their choice." Id. at 743 (quoting Gingles).
Satisfaction of the Gingles factors in the Fifth Circuit does not by itself establish a violation of Section 2. Id. The minority group must further demonstrate that, under the totality of the circumstances, "its members have less opportunity than other members of the electorate to participate in the political process and to elect representatives of their choice." 42 U.S.C. § 1973(b). Unlike the Gingles threshold inquiry, the totality of circumstances inquiry is broad. Clements, 986 F.2d at 747. It requires courts to make a "searching practical evaluation of the past and present reality" of the community and to take a "functional" view of the political process. Gingles, 478 U.S. at 45. The Court is guided by the factors mentioned in the Senate Report accompanying the 1982 amendments to Section 2, and may take into account other relevant factors as well. Clements, 986 F.2d at 747.
The Senate Report accompanying the 1982 amendments to Section 2 of the VRA identifies seven factors that may be relevant to an analysis of the totality of the circumstances. Id. The factors include: (a) the extent of any history of official discrimination in the state or political subdivision that touched the right of the members of the minority group to register, to vote, or otherwise to participate in the democratic process; (b) the extent of racially polarized voting; (c) the extent to which the state or political subdivision has used unusually large election districts, majority vote requirements, anti-single shot provisions, or other voting practices or procedures that may enhance the opportunity for discrimination against the minority group; (d) whether members of the minority group have been denied access to any candidate slating process; (e) the extent to which members of the minority group in the state or political subdivision bear the effects of discrimination in such areas as education, employment, and health; (f) use of racial appeals in campaigns; and (g) the extent to which minority candidates have been elected to public office. Id. at 747-52. This list is not exclusive, and courts may consider other factors relevant to whether a minority group has an unequal opportunity to participate in the political process on account of race or ethnicity. Id. at 753. Additional factors that may be considered include (h) whether there is a significant lack of responsiveness on the part of elected officials to the particularized needs of the members of the minority group, and (i) whether the policy underlying the state or political subdivision's use of the challenged voting qualification, prerequisite to voting, or standard, practice, or procedure is tenuous. Id. at 752-53.
The final determination by this Court must be made by an evaluation of the Gingles threshold factors and then the totality of circumstances. Williams v. City of Dallas, 734 F.Supp. 1317, 1320 (N.D. Tex. 1990). Plaintiffs do not have to show a discriminatory intent to prevail under Section 2 of the VRA. Campos v. City of Baytown, Tex., 80 F.2d 1240, 1242 (5th Cir. 1988). Instead, courts employ a "results test." Id. However, the burden of proof is on plaintiffs, who are required to prove by a preponderance of the evidence that all of the Gingles preconditions are satisfied and that based on the totality of the circumstances the accused election system dilutes the voting strength of the minority group. League of United Latin Am. Citizens No. 4552 (LULAC) v. Roscoe Indep. Sch. Dist., 123 F.3d 843, 846 (5th Cir. 1997).
II. Framework under Fourteenth Amendment
III. Motions to Strike
Rogers v. Lodge458 U.S. 613617Id. Id. See
The basis for each motion to strike, as well as Defendant's objection, is that exhibits and related testimony are precluded by Rule 37 of the Federal Rules of Civil Procedure because information was not disclosed as required by Rule 26. See id.; Vol. 1 at 84-85. Under Rule 37, a party that, without substantial justification, fails to disclose information required by Rule 26(a) or 26(e)(1) is not, unless such failure is harmless, permitted to use as evidence at trial any witness or information not so disclosed. FED. R. CIV. P. 37. Rule 37(c)(1) contains express exceptions that allow admission of evidence when the party's failure to disclose the required information is substantially justified or harmless. Id. A district court considers four factors when determining if such a violation is harmless: (1) the explanation, if any, for the non-disclosing party's failure to comply with the discovery rule; (2) the prejudice to the opposing party; (3) the possibility of curing such prejudice by granting a continuance; and (4) the importance of the evidence and related witnesses' testimony. See Barrett v. Atlantic Richfield Co., 95 F.3d 375, 380 (5th Cir. 1996). The Court has broad discretion in deciding whether a violation of Rule 26(a) is substantially justified or harmless. United States v. $9,041,598.68, 163 F.3d 238, 252 (5th Cir. 1998). The Court will now address these factors for the motions to strike and Defendant's objection.
A. Plaintiffs' Exhibit 7 and Defendant's Exhibit 13
The Court finds that Defendant's Exhibit 13 and Plaintiffs' Exhibit 7 both contain information that was not timely disclosed. See generally Doc. Nos. 38-42 (3:07-CV-900-O) (N.D. Tex. May 22, 2008); Doc. Nos. 46-47 (3:07-CV-900-O) (N.D. Tex. May 25, 2008). However, after considering the four factors outlined above, the Court finds that these violations were harmless.
The Court finds that the parties failed to timely disclose information because, at least in part, both parties waited for the Court to rule on Defendant's summary judgment motion before deposing experts and preparing for trial. Id. Shortly after the parties designated their experts, Defendant filed its motion for summary judgment. See Doc. 17 (3:07-CV-900-O) (N.D. Tex. Jan. 7, 2008). On April 8, 2008, one month before the original trial date of May 8, 2008, the Court denied this motion. See Doc. 24 (3:07-CV-900-O) (N.D. Tex. Apr. 8, 2008). At that time, none of the parties' experts had been deposed. See Doc. No. 47 (3:07-CV-900-O) (N.D. Tex. May 25, 2008). A few days later, on April 11, 2008, Plaintiffs and Defendant deposed the opposing parties' expert(s). Id. The Court finds that Plaintiffs' Exhibit 7 and Defendant's Exhibit 13 were prepared in reaction to the deposition testimony of the opposing party's expert(s). Accordingly, the relevant information was disclosed to the opposing party after the deadlines set out in Rule 26 and shortly before trial.
Though this explanation is not a substantial justification for the parties' untimely disclosures, the Court finds that the remaining factors to be considered demonstrate that both Plaintiffs' and Defendant's failures to disclose were harmless and that these exhibits and related testimony should be permitted. While both parties generally allege prejudice from untimely disclosures, neither party articulates how any such prejudice arises. See generally Doc. Nos. 38-42 (3:07-CV-900-O) (N.D. Tex. May 22, 2008); Doc. Nos. 46-47 (3:07-CV-900-O) (N.D. Tex. May 25, 2008). Additionally, the harmless nature of this error is demonstrated by the fact that neither party requested a continuance or even suggested that additional time would be helpful. Furthermore, the Court finds that both exhibits contain information that is important to the parties' respective positions. Plaintiffs' Exhibit 7 contains estimates and supporting calculations which might establish that the Hispanic population in Farmers Branch is sufficiently large and geographically compact to constitute a majority in a single member district. Similarly, Defendant's Exhibit 13 contains information which, if found persuasive, would provide a basis for discrediting one of the theories on which Plaintiffs rely to establish that the Hispanic population in Farmers Branch is sufficiently large and geographically compact.
Counsel for Defendant indicated at trial that he would file a response to Plaintiffs' motion to strike. However, a review of the docket indicates no response was filed.
Accordingly, the Court finds that both parties' failure to timely disclose constitutes harmless error and that Plaintiffs' Motion to Strike Defendant's Exhibit 13 and Defendant's Motion to Strike Plaintiffs' Exhibit 7 should be and are hereby DENIED.
B. Plaintiffs' Exhibit 36
Plaintiffs' Exhibit 36 contains calculations relating to Plaintiffs' regression analysis of Farmers Branch election results. This exhibit is being offered to demonstrate that Hispanics in Farmers Branch are politically cohesive, and that the white majority votes sufficiently as a bloc to enable it, in the absence of special circumstances, usually to defeat the Hispanic population's preferred candidate.
For reasons explained below, the Court does not consider Plaintiffs' Exhibit 36 in resolving Plaintiffs' claims. Accordingly, Defendant's Motion to Strike Plaintiffs' Exhibit 36 is hereby DENIED as moot.
C. Defendant's Objection to Plaintiffs' Exhibit 48
At trial, Defendant objected to Plaintiffs' Exhibit 48, which is a list of names of seven people Plaintiffs' expert believes are Hispanic. Vol. 1 at 84-85. Prior to his deposition, Plaintiffs' expert provided Defendant with the names of 48 individuals he believed to be Hispanic. Id. At his deposition on April 11, 2008, Plaintiffs' expert told Defendant's counsel that he had identified a few other people he believed to be Hispanic, but that he didn't have the list with him at that time. Id. In the days before trial, Plaintiffs' expert identified another couple of people, resulting in a total of seven individuals identified after the original 48 people. Id. At trial, Plaintiffs introduced the list of the seven names, drawing an objection from Defendant for failure to supplement. Plaintiffs responded that this exhibit was, in part, a reaction to Defendant's Exhibit 13, which was not timely disclosed. Id. at 85.
The Court finds that Plaintiffs failed to timely disclose at least some of the information contained within their Exhibit 48. See id. at 84-85. However, the Court finds this error was harmless. Defendant did not articulate any specific prejudice that would be caused by this untimely disclosure. In addition, the evidence demonstrates that the seven individuals listed in Exhibit 48 were identified by Plaintiffs' expert through the same methods used to identify the other 48. See Vol. 1 at 37-39; 70-85; 255-272. Therefore, there does not appear to be any risk that Defendant would be unable to cross-examine Plaintiffs' expert or that Defendant would have otherwise been better off had this information been disclosed prior to trial. The Court notes that Defendant knew Plaintiffs' expert had identified additional individuals at the deposition held on April 11, 2008, and did not seek continuance or Court intervention as a result. In addition, the seven names listed in Exhibit 48 are being offered to show that the Hispanic population in Farmers Branch is sufficiently large and geographically compact to constitute a majority in a single member district, and the Court finds the importance of this exhibit and related testimony weighs against excluding this evidence. Accordingly, the Court OVERRULES Defendant's objection to Plaintiffs' Exhibit 48.
IV. Findings of Fact
A. The Parties
Plaintiffs Valentine Reyes, Irene Gonzalez, and Gary Garcia are Hispanic citizens of the city of Farmers Branch, Texas. Defendant is the city of Farmers Branch, which is located near Dallas, Texas. Farmers Branch is governed by a city council composed of five members. Election of city council members is at-large, in that all voters in the city are allowed to vote for candidates running for each of the five city council member positions. Every candidate must run for a particular numbered city council position, and the city holds run-off elections where no one candidate for a particular position receives a majority of votes. Elections are staggered, and, while members of the city council must live in Farmers Branch, there is no requirement that members be elected from different parts of the city.
On May 21, 2007, Plaintiffs Valentine Reyes, Irene Gonzalez, and Gary Garcia filed this suit, alleging that the at-large election of city council members for Farmers Branch dilutes the voting rights of Hispanic citizens, in violation of the Section 2 of the Voting Rights Act ("VRA"), 42 U.S.C. § 1974, and the Fourteenth Amendment. The parties presented evidence regarding these claims at a bench trial held May 27-28, 2008.
B. Fact and Expert Witnesses
At trial, Plaintiffs offered the testimony of Mr. Rendon and Dr. Gambitta. Mr. Rendon is a Hispanic resident of Farmers Branch who recently ran for city counsel, but was not elected. Vol. 1 at 16-18, 20-24, 43. Mr. Rendon testified regarding the ethnic composition of Farmers Branch and his experiences regarding the role Hispanic origin plays in elections in Farmers Branch. Id. at 16-57.
Dr. Gambitta is the Director of the Institute for Law and Public Affairs, as well as an Associate Professor, at the University of Texas at San Antonio. Pls' Ex. 47 at 1. Dr. Gambitta holds B.A., M.A., and Ph.D. degrees in political science, and has worked as a consultant in areas such as survey research design. Id. Dr. Gambitta testified regarding Plaintiffs' ability to establish the Gingles threshold factors, as well as the factors considered by courts in assessing the totality of the circumstances. Vol. 1 at 59-219; Vol. 2 at 77-84.
Defendant offered the testimony of Dr. Rives and Dr. Alford at trial. Dr. Rives is a Senior Lecturer in the Department of Finance at the Fisher College of Business at The Ohio State University. Def's Ex. 1. Dr. Rives holds A.B, M.A., and Ph.D. degrees in Economics, and studied demography and statistics as a postdoctoral research fellow with the Office of Population Research at The Woodrow Wilson School at Princeton University. Id. Dr. Rives has worked as a professor at several universities, as a research fellow with the U.S. Census Bureau, and as a consultant in VRA cases. Id. Dr. Rives testified regarding Plaintiffs' ability to satisfy the first Gingles threshold factor, commonly referred to as Gingles I. Vol. 1 at 220-301.
Dr. Alford is an Associate Professor at Rice University, and holds B.S., M.P.A., M.A., and Ph.D. degrees in political science. Def's Ex. 14. Dr. Alford has worked as a consultant in numerous VRA cases. Id. Dr. Alford testified regarding Plaintiffs' ability to satisfy Gingles I, II and III, as well as regarding the factors considered by the Court in assessing the totality of the circumstances. Vol. 2 at 4-76.
The Court finds that Mr. Rendon is a fact witness regarding elections in Farmers Branch. In addition, the Court finds that Plaintiffs' expert, Dr. Gambitta, and Defendant's experts, Dr. Rives and Dr. Alford, are qualified, due to their education and experience, to give expert opinions regarding the Gingles factors and the factors considered by the Court in assessing the totality of circumstances.
At trial, Mr. Rendon offered testimony as to what last names indicate Hispanic origin. Vol. 1 at 37. Defendant objected to this testimony, noting that Plaintiffs presented no evidence establishing that Mr. Rendon was qualified to offer an expert opinion on this issue. Id. The Court sustained Defendant's objection during trial and does not consider Mr. Rendon an expert on any subject matter relevant to this lawsuit. Id.
C. Gingles I
To demonstrate a violation of Section 2 of the VRA, the minority group challenging an at-large election scheme must demonstrate that the minority group is sufficiently large and geographically compact to constitute a majority in a single member district. Clements, 986 F.2d at 742. To satisfy this requirement, a plaintiff must demonstrate that it is possible to draw an election district of an appropriate size and shape where the citizen voting age population of the minority group is a majority. Perez v. Pasadena Indep. Sch. Dist., 165 F.3d 368, 372 (5th Cir. 1999). This requirement ensures that, in the absence of the multimember district and at-large voting scheme, the minority group possess the potential to elect representatives of its choice. Clements, 986 F.2d at 743. In other words, if the minority group is not sufficiently large and geographically compact, the multimember form of the district cannot be responsible for the minority voters' inability to elect their candidate of choice. Id.
1. Farmers Branch and 2000 Census data
According to the 2000 Census, Farmers Branch had a total population of 27,505. Pls' Ex. 1 at 1; Def's Ex. 2. Of the total residents of Farmers Branch, 10,241 identified themselves as Hispanic in the 2000 Census, constituting 37.2% of the city's population. Pls' Ex. 1. Of these 10,241 Hispanic residents, 6,467 were of voting age. Pls' Ex. 1 at 2; Def's Ex. 1 at Ex. 1 (Dr. Rives Report).
The pool of eligible voters in Farmers Branch at the time of the 2000 U.S. Census was comprised of 16,008 citizens of voting age. Def's Ex. 1 at 2. Of these 16,008 citizens, 2,500 identified themselves as being of Hispanic origin. Def's Ex. 1, at Exhibit 1. Thus, approximately 15.6% of eligible voters in Farmers Branch were Hispanic at the time of the 2000 Census.
Defendant's experts testified that citizenship and voting age population data were not available for the precise area that constitutes Farmers Branch. However, the Census reports this information at the Census block group level, allowing Dr. Rives to estimate these figures. See Vol. 2 at 29. Defendant's evidence demonstrates that Dr. Rives allocated block groups split by city boundaries to census blocks within the city to determine citizenship and voting age population for Farmers Branch. Def's Ex. 1 at Exhibit 1. Plaintiffs did not present evidence contradicting this method or the resulting estimates.
2. Plaintiffs' Proposed District
In order to satisfy Gingles I, Plaintiffs must propose a district and demonstrate that it is more likely than not that the Hispanic citizen voting age population ("HCVAP") of the proposed district exceeds 50% of the district's total citizen voting age population ("CVAP"). See Perez, 165 F.3d at 372. Plaintiffs' proposed district was created by Dr. Gambitta in association with Plaintiffs' legal counsel with the goal of capturing the highest percentage of Hispanic citizens of voting age. Vol. 1 at 64. The proposed district was drawn to include the eastern portion of voting precinct 1500 and the majority of voting precinct 1502, which is to the southwest of precinct 1500. Id.; Pls' Ex. 8.
The evidence demonstrates that Plaintiffs originally proposed two slightly different demonstration districts, referred to as 1225C and 1226C. However, at trial, the parties focused on district 1226C, and it appears that Plaintiffs are relying solely on this district to meet their burden of proof. Vol. 1 at 238-39. When the Court refers to Plaintiffs' proposed district, if refers to district 1226C.
Plaintiffs argue that several methods of assessing Hispanic voting strength in the proposed district demonstrate that Gingles I is met here. First, Plaintiffs rely the Texas Legislative Council's ("TLC") estimate that Spanish surnamed registered voters constitute 52.5% of the total registered voters in Plaintiffs' proposed district. Vol. 1 at 65. In addition, Plaintiffs rely on an "actual count" of Hispanics and/or Spanish surnamed individuals registered to vote in the proposed district. Vol. 1 at 178; Vol. 2 at 77-78. Plaintiffs also rely on Dr. Gambitta's calculation of HCVAP/CVAP based on Hispanic and non-Hispanic voter registration rates for Dallas County. Vol. 1 at 86-90. Dr. Gambitta used voter registration rates for Dallas County to estimate HCVAP and CVAP for Plaintiffs' proposed district from the number of Hispanic and non-Hispanic registered voters in the proposed district, as determined by Plaintiffs' "actual count." Vol. 1 at 88-90.
a. Texas Legislative Council Estimates
To satisfy Gingles I, Plaintiffs rely in part on an estimate of the number of Spanish surnamed registered voters in Plaintiffs' proposed district made by the Texas Legislative Council. Vol. 1 at 65. The TLC compiles information from Census and voter registration records and has developed a computer system that calculates and/or estimates certain figures generally used by the Texas Legislature and legislative agencies in redistricting. Id. at 68-69.
Here, Plaintiffs provided the TLC with the boundaries of their proposed district, and the TLC system estimated (1) the percentage of the citizen voting age population that was Hispanic based on 2000 Census data, and (2) the percent of registered voters in the proposed district with Spanish surnames based on 2006 voter registration data. Vol. 1 at 64-66; Pls' Ex. 4; Def's Ex. 6. While the Census collects data on Hispanic origin based on participants' identification of their own race and ethnicity, voter registration records do not contain information on ethnicity. Vol. 1 at 69. Spanish surnames are sometimes used as a surrogate for Hispanic origin. See Clements, 986 F.2d at 774, 868-73. The evidence shows that the TLC system designates a voter as a Spanish surnamed registered voter ("SSRV") if the voter's last name is found on the 1990 Spanish surname list developed by the U.S. Census Bureau. Vol. 1 at 145, 243-44.
Based on 2000 Census data, the TLC estimated that the HCVAP within Plaintiffs' proposed district was 41.4% of the CVAP. See Def's Ex. 6. Thus, Plaintiffs are unable to establish a HCVAP majority based solely on 2000 Census data. Accordingly, Plaintiffs looked to the TLC's estimate of SSRVs in the proposed district to establish Gingles I. Vol. 1 at 142-43.
While Dr. Gambitta testified that generally Hispanics under report Hispanic origin on the Census, no evidence was presented on the degree of under reporting such that the Court could conclude that Plaintiffs' proposed district had a HCVAP above 50% of the citizen voter population based on 2000 Census data. See Perez v. Pasadena Indep. Sch. Dist., 958 F.Supp. 1196, 1211 (S.D. Tex. 1997) (where court did not find projections of population based on undercount of Hispanic voters was supported by sufficient evidence to override Census data).
Based on 2006 voter registration records, the TLC estimated that there are 667 SSRVs in Plaintiffs' proposed district out of 1,270 registered voters. Pls' Ex. 4. In other words, the TLC estimated that Spanish surnamed registered voters comprise 52.5% of the registered voters in Plaintiffs' proposed district. Id. Plaintiffs argue that these data indicate that there is a majority HCVAP in Plaintiffs' proposed district. Doc. No. 57 (3:07-CV-900-O) (N.D. Tex. Aug. 25, 2008) (Plaintiffs' Proposed Findings).
In response, Defendant argues that the TLC computer system overestimated the number of SSRVs in Plaintiffs' proposed district. Doc. No. 56 (3:07-CV-900-O) (N.D. Tex. Aug. 25, 2008) (Defendant's Proposed Findings). Dr. Rives testified that the TLC system contains data for entire precincts, and that when only a portion of a precinct is considered, the system must determine how many voters to allocate to a particular portion of the precinct. Vol. 1 at 240-241. Dr. Rives stated that the system allocates voters as though they were distributed evenly throughout a precinct. Id. Dr. Rives further testified that the TLC system overestimated SSRVs in Plaintiffs' proposed district because SSRVs are not evenly spread out in precinct 1500, as assumed by the TLC system. Vol. 1 at 242-249. Dr. Rives stated that the proposed district contains a portion of precinct 1500 with relatively fewer SSRVs than the rest of the precinct, resulting in over allocation of SSRVs to the proposed district. Id. at 247-249. Dr. Rives explained that the uneven distribution of SSRVs in precinct 1500 may be due to the fact that there is a higher proportion of apartment housing in the part of precinct 1500 in Plaintiffs' proposed district, while the rest of the precinct contains mainly single family homes. Id. Dr. Rives stated that Public Use Microdata Sample information ("PUMS data") for the area including Farmers Branch indicate that Hispanic owner-occupied units are roughly 2.8 times more likely to be headed by a Hispanic citizen than Hispanic renter-occupied units. Id. at 252. Dr. Rives noted that this was not a precise calculation, as PUMS data is collected for an area that includes but is much larger than Farmers Branch. Id. However, Dr. Rives testified that the differences in citizenship might be the cause of the uneven distribution of registered voters with Spanish surnames in precinct 1500 and resulting over allocation of SSRVs to the proposed district. Id. at 246-253. Dr. Rives concluded that the TLC overestimated the number of SSRVs in Plaintiffs' proposed district. Id. at 254-55.
In assessing the TLC evidence, the Court finds it necessary to first address the probativeness of the TLC's use of Spanish surname data. Case law demonstrates a preference for reliance on Census data in voting rights cases, although the Fifth Circuit has noted that situations occur when Census data are not sufficiently probative, and that use of non-census data in such cases may be appropriate. See Valdespino, et al., v. Alamo Heights Indep. Sch. Dist., 168 F.3d 848, 852 (5th Cir. 1999); Westwego, 906 F.2d at 1045, n. 3. In Rodriguez v. Bexar County, Tex., 385 F.3d 853 (5th Cir. 2004), the Fifth Circuit stated that without a strict showing of probativeness, Spanish-surname data are disfavored, and census data based upon self-identification provide the proper basis for analyzing claims that the votes of Hispanics have been diluted in violation of Section 2 of the VRA. Id. at 867, n. 18.
Based on the evidence presented at trial, the Court finds that the SSRV data utilized by the TLC are sufficiently probative to be considered by the Court. First, this case was filed in 2007, many years after the 2000 Census data were collected. Courts in the Fifth Circuit have noted that Census data are less probative the further away from the Census cycle one gets. See Perez v. Pasadena Indep. Sch. Dist., 958 F.Supp. 1196, 1212-13 (S.D. Tex. 1997) (where court noted that case was tried past midpoint of Census cycle, and therefore Census data were inaccurate to some extent). In addition, the evidence shows that the Spanish surname list used by the TLC in identifying SSRVs has been published and heavily tested by the U.S. Census Bureau. Vol. 1 at 172-73. This testing indicates that there is a relationship between Hispanic origin and the Spanish surnames listed on the 1990 Spanish surname list. Vol. 1 at 262-63. Although courts have criticized the tendency of this type of data to misidentify Hispanic persons as non-Hispanic and vice versa, the Court finds that it can consider this type of data when, as here, Census data are out-dated and therefore less likely to be accurate. See Rodriguez, 385 F.3d at 867. Accordingly, the Court finds the TLC's estimate of SSRVs in Plaintiffs' proposed district have probative value on the issue of whether Plaintiffs have satisfied Gingles I.
The Court notes that Dr. Gambitta testified that SSRV data, in connection with the Defendant's "actual count" of SSRVs, are not probative. Vol. 1 at 162-63. This issue will be addressed infra.
Though the Court finds that the SSRV data are probative, the Court does not find the TLC's estimate based on this data demonstrates that it is more likely than not that Plaintiffs' proposed district contains a majority HCVAP, as required by Gingles. Defendant has produced evidence which calls into question the reliability of Plaintiffs' estimate, which exceeds the minority majority required by Gingles I by only a few percent. The Court credits Dr. Rives testimony that splitting precincts can result in an inaccurate estimate of SSRVs, and notes that this testimony was corroborated by a document from the TLC that warned about the accuracy of applying its methods to small areas. See Def's Ex. 7 (TLC document stating that election data for a small geographic unit such as a census block yields a very low confidence level, and that, as with other statistical data, confidence levels increase as the blocks are aggregated into larger units such as counties). The Court finds that this evidence indicates that the TLC estimates are to some degree unreliable when applied to Plaintiffs' proposed district, a relatively small area.
The Court also credits Dr. Rives testimony that the TLC system overestimated the number SSRVs in Plaintiffs' proposed district because the proposed district contains a portion of precinct 1500 with a lower concentration of SSRVs than the rest of the precinct. The Court finds persuasive Dr. Rives explanation that SSRVs are not distributed evenly throughout precinct 1500 because the part of precinct 1500 in Plaintiffs' proposed district contains mainly apartments, as opposed to single family homes. Accordingly, the Court finds that the TLC estimate suffers from reliability issues, and does not demonstrate that it is more likely than not that Plaintiffs' proposed district contains a Hispanic citizen voting age majority.
As in Brewer v. Ham, 876 F.2d 448, 451-52 (5th Cir. 1989), the most Plaintiffs have shown with the TLC estimate is that perhaps there is a slim minority majority district. See Brewer, 876 F.2d at 451-52 (holding district court did not err in finding proposed district with total minority population of 55.91% did not meet Gingles I, as expert testified that voting age population could drop below 50% level and that at most Plaintiffs had shown perhaps there was a majority minority voting age population). While Plaintiffs are not required to demonstrate Hispanics would enjoy an overwhelming majority, just that they would possess the potential to elect the candidate of their choice, Plaintiffs must demonstrate this potential by a preponderance of the evidence. See Westwego, 946 F.2d at 1117. The Court finds that, considering the TLC estimate, it is just as likely that Plaintiffs' proposed district has below or equal to 50% HCVAP.
Additionally, the evidence presented by the parties demonstrates that it is possible to directly count, rather than estimate, the number of SSRVs in Plaintiffs' proposed district. The Court finds that an actual count of SSRVs within the proposed district, if accurate and reliable, offers more direct and probative evidence of Hispanic voting strength within the proposed district than an estimate. Accordingly, the Court will now evaluate whether Plaintiffs' "actual count" of SSRVs within Plaintiffs' proposed district demonstrates that it is more likely than not that the proposed district contains a Hispanic citizen voting age majority.
b. "Actual Count" of Registered Voters
Both parties argue that an "actual count" of Spanish surname registered voters in the proposed district supports their respective positions. Doc. Nos. 56, 57 (3:07-CV-900-O) (N.D. Tex. Aug. 25, 2008). The evidence demonstrates that, in an attempt to refute Plaintiffs' TLC data, Dr. Rives obtained a list of voters, registered as of July 31, 2007, from Dallas County Elections. Id. at 242-44. This list included the address of each voter, as well as the Texas Secretary of State's designation of certain voters as SSRVs, coded based on the 1990 Spanish surname list developed by the U.S. Census Bureau. Id. at 243. Dr. Rives then "geocoded" the list, determining whether each registered voter lived within the boundaries of Plaintiffs' proposed district based on the voter's address, and counted the number of SSRVs within the district. Id. at 243-44. Dr. Rives testified that, based on this work, there are 600 SSRVs in the proposed district out of 1291 registered voters. Id. at 245. In other words, Dr. Rives testified that Spanish surname registered voters account for approximately 46.5% of registered voters in Plaintiffs' proposed district. Dr. Rives concluded that there is not a HCVAP majority in the proposed district. Vol. 1 at 273.
The Court notes that while Defendant counted SSRVs, Plaintiffs counted SSRVs as well as individuals they believe to be Hispanic based on factors other than surname, as will be discussed infra.
In response, Plaintiffs asked Dr. Gambitta to review the list of 1291 registered voters identified by Defendant as residing within Plaintiffs' proposed district. Vol. 1 at 70-76; Pls' Ex. 5. Dr. Gambitta initially identified 48 names he believed should have been included on Defendant's list. Vol. 1 at 70-76; Pls' Ex. 6. The evidence shows that Dr. Gambitta considered various factors in identifying these 48 individuals. Dr. Gambitta identified ten individuals with last names he believes are misspelled, but when properly spelled, are found on the Spanish surname list. Vol. 1 at 70-76. For example, Dr. Gambitta identified Gloria Aaleman, believing her properly spelled last name to be Aleman, which is on the Spanish surname list. Id. Dr. Gambitta also identified eight women with hyphenated last names, where at least one of the last names is on the Spanish surname list. Id. at 76. In addition, Dr. Gambitta identified individuals with last names he believes to be Hispanic, even though the last names are not on the Spanish surname list. See Pls' Ex. 5; Pls' Ex. 6 (identifying four individuals with last name Garpa); Vol. 1 at 258-61 (where Dr. Rives testifies that Garpa is not on the Spanish Surname list). Dr. Gambitta also identified individuals without last names on the Spanish surname list or otherwise deemed Hispanic by Dr. Gambitta if the voter's first and/or middle name was Hispanic. For example, Dr. Gambitta identified Cindy Mezo Hogan, Victoria Fuentes Miller, and Sylvia Hernandez Marsh as people that should have been on Defendant's list. Vol. 1 at 76. Dr. Gambitta noted that many middle names of women are actually maiden names, and therefore a Hispanic middle name may indicate Hispanic origin. Id. Dr. Gambitta also identified a few individuals without Hispanic first, middle, or last names where Dr. Gambitta believed other indicators of Hispanic origin were present. For example, Dr. Gambita testified that he identified Monica Marsh because voter registration records indicate that Sylvia Hernandez Marsh lived at the same address as Monica Marsh. Id. at 165-166. Dr. Gambita noted that Sylvia Marsh had a Hispanic middle name and was old enough to be Monica's mother. Id. at 168. Dr. Gambitta testified that Monica could be Sylvia's step-daughter or not a daughter at all, but stated that he included her because there was some probability of Monica being Hispanic. Id. at 167; see also Pls' Ex. 7 (listing Veronica Elizabeth Reed as a SSRV/Hispanic voter because voter Margie Ann Rumbo Reed is old enough to be Veronica's mother and lives at the same address).
After Dr. Gambitta constructed his initial list of 48 people he believed were left off of Defendant's list of SSRVs, Plaintiffs' counsel asked Mr. Rendon to visit the homes of these individuals to confirm their Hispanic origin. Id. at 77-78. Plaintiffs had contemplated hiring a professional survey firm, but knew Mr. Rendon would be going door-to-door while campaigning, and hoped to save money by having Mr. Rendon conduct the survey. Id. at 78. Mr. Rendon testified that he went to the homes of the 48 individuals. Id. at 38. Some individuals identified by Dr. Gambitta were not home when Mr. Rendon visited. Id. at 39-40. In those instances, Mr. Rendon asked neighbors regarding the individuals' ethnicity. Id. at 39-40.
Mr. Rendon testified that, based on his door-to-door survey, only one woman Dr. Gambitta identified, Cindy Meza Hogan, was not Hispanic. Id. at 38. Dr. Gambitta testified that Mr. Rendon also observed that two individuals were listed several times by Dr. Gambitta due to various misspellings of their names within voter registration records. Id. at 79; see also Pls' Ex. 6 (Dacia Tinagera, Dacia Tinajera, Dacia Tinjajera, Juan Tinajeka, Juan Tinajera, and Juan Alfredo Tinajera were initially identified by Dr. Gambitta as six separate people on the list of 48 individuals, though each Dacia T. and Juan T. have the same birthday and address).
Dr. Gambitta testified that, after removing duplicates and Ms. Meza, Mr. Rendon confirmed that there were 644 SSRVs/Hispanics in the proposed district. Id. at 80. In the days before trial, Dr. Gambita identified an additional seven individuals he believes should be included in Plaintiffs' "actual count." Id. Dr. Gambitta also identified three non-Hispanic names that appear to be duplicates. See id. at 74-76 (identifying 2 Nathan Hales with the same birthday and address, 2 Manda Soaps with the same address, and a Michael Warren Neibert and Michael Warren Neidert with the same address). Dr. Gambitta concluded that there are 651 Hispanic registered voters (600 SSRVs identified by Defendant plus those individuals identified by Dr. Gambitta) out of 1285 total registered voters (1291 voters identified by Defendant minus duplicates identified by Dr. Gambitta) in the proposed district. Vol. 2 at 77-78. In other words, Dr. Gambitta testified that 50.7% of registered voters in the proposed district are SSRVs/Hispanic. Dr. Gambitta testified that this demonstrates Plaintiffs' proposed district meets Gingles I.
Dr. Gambitta testified that there are 651 individuals that are SSRVs or Hispanic voters in Plaintiffs' proposed district. Vol. 2 at 77-78. However, Dr. Gambitta's testimony and Plaintiffs' exhibits suggest that Dr. Gambitta identified only 50 individuals in addition to the 600 SSRVs identified by Defendant. See Pls' Ex. 6 (listing Cindy Meza Hogan, Dacia Tinagera, Dacia Tinajera, Dacia Tinjajera, Juan Tinajeka, Juan Tinajera, and Juan Alfredo Tinajera, with each Dacia T. and each Juan T. having the same address and birthday, therefore causing an over count by 5 individuals); Vol. 1 at 74-80 (demonstrating that Dr. Gambitta initially identified 48 individuals, four of which were later determined to be duplicates and one not Hispanic. Dr. Gambitta later identified an additional seven individuals).
Dr. Gambitta testified that six individuals should be subtracted from the total number of registered voters identified by Defendant. See Vol. 2 at 77-78. However, Dr. Gambitta's testimony and Plaintiffs' exhibits suggest that seven people should be subtracted from the list of total registered voters in Plaintiffs' proposed district. See Pls' Ex. 6; Vol. 1 at 74-76.
Using the figures in footnotes 9 and 10, it appears that Dr. Gambitta has testified that 50.6% of registered voters would be SSRVs/Hispanic (650 SSRVs/1284 registered voters).
Defendant questions whether Dr. Gambitta's methods and results are reliable and could be replicated. Dr. Rives criticized Dr. Gambitta for not adhering to the Spanish surname list in identifying people to include as SSRVs. Vol. 1 at 257. Dr. Rives conceded that the eight individuals with hyphenated names should have been included pursuant to the Spanish surname list. Id. at 256-57. However, Dr. Rives testified that Dr. Gambitta erred by listing individuals without last names found on the Spanish surname list without also identifying and removing people with Spanish surnames who are not Hispanic. Id. at 259-60, 265-66. Dr. Rives testified that Dr. Gambitta's method of altering the Spanish surname list has not been tested, and has an unknown error rate, unlike the Spanish surname list developed by the Census Bureau and used by the State of Texas. Id. at 271. Dr. Rives also stated that he does not know what survey methodology Mr. Rendon used when going door-to-door to determine if the individuals on Plaintiffs' list were Hispanic, and that there is no way of judging the survey's reliability. Id. at 299. Dr. Rives also presented evidence regarding the 2000 Spanish Surname list developed by the U.S. Census Bureau. Id. at 256. The 2000 Spanish Surname list gives the probability that a certain surname will be Hispanic, whereas the 1990 Spanish Surname list assumes that names on the list are always Hispanic and does not contain probabilities that individual surnames will be Hispanic. Id. at 256-58. Dr. Rives calculated that, based on the 2000 Spanish surname list, one would expect 582.23 Spanish surnamed voters to reside in Plaintiffs' proposed district. Def's Ex. 13; Vol. 1 at 286. Dr. Rives concluded that Dr. Gambitta's methods are not reliable and that Plaintiffs' proposed district does not have a majority SSRV or HCVAP. Id. at 299.
Considering this concession, Defendant's expert admitted that 47.5% of registered voters would be SSRVs (608 SSRVs/1285 registered voters).
If there were 582.23 Hispanics registered to vote in the proposed district, 45.3% of registered voters would be Hispanic.
The Court does not find that Plaintiffs' "actual count" of SSRVs/Hispanic voters demonstrates that it is more likely than not that Plaintiffs' proposed district contains a majority HCVAP, as required by Gingles. The Court does not credit Dr. Gambitta's testimony that there are 651 registered Hispanic voters out of 1285 registered voters in Plaintiffs' proposed district, and finds that the methods used in arriving at this conclusion that are not reliable or capable of replication. The Court credits Dr. Rives testimony criticizing the reliability of Dr. Gambitta's "actual count."
The Court notes that, when questioned about his methods, Dr. Gambitta testified that his knowledge of what names are Hispanic was "just in his head" and based on "common sense" and "general observation" and that the methodology he used was not written down or otherwise published. Vol. 1 at 178-184. This demonstrates that there was no published method by which Dr. Gambitta identified voters to include in his "actual count." While Defendant relied upon coding of individuals by the State of Texas pursuant to the Spanish surname list, Plaintiffs' expert did not follow any specific or uniform method of assessing which registered voters should be considered SSRVs.
Plaintiffs argue that Defendant's criticisms of Dr. Gambitta's work is unfounded, as Mr. Rendon confirmed that almost everyone identified by Dr. Gambitta is Hispanic. Dr. Gambitta testified that his methods were confirmed to be accurate within approximately 2% by Mr. Rendon's door-to-door survey, as Mr. Rendon confirmed that 47 of the 48 individuals Dr. Gambitta listed were Hispanic. Vol. 1 at 170. However, the Court finds that even if each of the 50 or 51 individuals identified by Dr. Gambitta are Hispanic, Dr. Gambitta's methods still have reliability issues because Dr. Gambitta identified individuals without last names found on the Spanish surname list without also identifying and removing people with Spanish surnames who are not Hispanic. See Vol. 1 at 262-66.
The Court credits Dr. Rives' testimony regarding this issue. Dr. Rives testified that Spanish surname and Hispanic origin are not the same thing, and that a person can have a Spanish surname but not be Hispanic. Vol. 1 at 262. It is called commission errors when non-Hispanic individuals are included in a SSRV count. Id. Dr. Rives also testified that a person can have a surname that is not on the Spanish surname list, but be Hispanic. Id. It is called omission error when Hispanic individuals are not included in a SSRV count. Id. Dr. Rives testified that commission and omission errors are not insignificant, and that Dr. Gambitta has addressed omission, but not commission errors. Id. at 263. In other words, Dr. Gambitta only addressed people who may have been left off Defendant's list, but did not address people on the list who may not belong there. See id. The Court also credits Dr. Rives' testimony that Dr. Gambitta added his own criteria which are not part of the 1990 Spanish surname list, and that such methods have not been published or otherwise written down, and have an unknown error rate.
The Court, however, does not find that evidence regarding the 2000 Spanish surname list, including Defendant's Exhibit 13 and Dr. Rives testimony that one would expect 582.23 Spanish surnamed voters in the proposed district, weighs in Defendant's favor. See Vol. 1 at 287. The Court credits testimony that it is much more common for the 1990 Spanish surname list to be used in assessing Hispanic voting strength, which is the list used by the State of Texas and Texas Legislative Counsel. Vol. 2 at 53-54.
The Court finds that Dr. Gambitta's consideration of omission but not commission errors likely resulted in an over count of Hispanic voters. Though Dr. Gambitta assumed all 600 SSRVs identified by Defendant were Hispanic, both Dr. Rives and Dr. Gambitta testified that some of the 600 people coded as SSRVs by the state of Texas are probably not Hispanic. Id. at 173-174. For example, Dr. Gambitta testified that Garcia is a Spanish surname, but that only 95% of people with this surname will actually be of Hispanic origin, such as where a non-Hispanic woman marries a Hispanic man. Id. at 172. Dr. Gambitta concedes that there is overestimate built into the list, and that he did not do any testing to see how his adjustment would affect the overestimate. Id.
The Court's finding that Plaintiffs' "actual count" is likely an over count is supported by Dr. Gambitta's identification of women with Hispanic middle names and non-Hispanic surnames, and failure to remove women with Spanish surnames and non-Hispanic middle names from consideration. For example, Sylvia Hernandez Marsh and Victoria Fuentes Miller were identified by Dr. Gambitta as people left off Defendant's list. However, Dr. Gambitta did not remove Katherine Yonick Cantu and Vicky Leach Agraz from Defendant's list of SSRVs. See Pls' Ex. 5; Pls' Ex. 6. The Court notes that Plaintiffs did not have Mr. Rendon or anyone else confirm that the 600 SSRVs identified by Defendant were Hispanic. This demonstrates that Dr. Gambitta's methods account for omission but not commission error, and are likely to have resulted in an inflated "actual count."
The Court notes that during trial, Dr. Gambitta stated, upon questioning by Defendant's counsel, that Katherine Yonick Cantu and Vicky Leach Agraz should be removed from his count of SSRVs/Hispanic voters. See Vol. 1 at 175-176. However, Dr. Gambitta's ultimate conclusion that 651 registered Hispanic voters reside in the proposed district included these two women. Considering the figures in footnotes 9 and 10, and Dr. Gambitta's concession at trial that Katherine Yonick Cantu and Vicky Leach Agraz should be removed from the count, Dr. Gambitta's testimony appears to be that 50.5% of the total number of registered voters in Plaintiffs' proposed district would be SSRVs and/or Hispanic voters.
Having found reliability issues with respect to Plaintiffs' "actual count," the Court does not find this data demonstrates that it is more likely than not that Plaintiffs' proposed district contains a majority HCVAP, as required by Gingles. While Plaintiffs have once again shown there is perhaps a slim minority majority in Plaintiffs' proposed district, the Court finds that, based on the evidence, it is as likely that the Hispanic voting population is equal to or below 50% of the citizen voting age population of the district. See Brewer, 876 F.2d at 451-52.
Plaintiffs, in their post-trial submission, assert that Dr. Gambitta properly identified Spanish surnamed registered voters because in Ovalle v. State, 13 S.W.3d 774 (Tex.Crim.App. 2000), the Texas Court of Criminal Appeals considered an individual with a misspelled last name, as well as three women without Spanish surnames with Hispanic first names, to be Hispanic. See Doc. No. 57 (3:07-CV-900-O) (N.D. Tex. Aug. 25, 2008); Ovalle v. State, 13 S.W.3d at 780, n. 22. In Ovalle, the Texas Criminal Court of Appeals considered an appeal from a defendant convicted of capital murder and sentenced to death. Ovalle, 13 S.W.3d at 776. Appellant alleged that his sentence should be reversed because, among other things, persons of Hispanic origin had been systematically excluded from grand juries in Navarro County in violation of the Fourteenth Amendment. Id. at 777. The Court denied appellant's Fourteenth Amendment claim, finding that nineteen grand jurors had served that were probably Hispanic. Id. at 783.
In determining that nineteen likely Hispanic individuals served on grand juries in Navarro County, the court counted as Hispanic 10 individuals (one individual who served twice) with surnames considered Hispanic pursuant to Word and Perkin's list of the 639 most commonly heavily Hispanic surnames, including one name that appeared to be misspelled. Id. at 783, n. 31. In addition, the court counted four individuals with surnames that were considered heavily Hispanic, although the names were not common. Id. It is unclear what source contained this information, although one of the sources may have been the 1990 Spanish surname list. Id. It appears that the court also considered an additional source listing probabilities of names being Hispanic in Texas, as opposed to other states. Id. Finally, the court also considered Hispanic three women with Hispanic first names, but with non-Hispanic surnames. The court stated that it was "common knowledge" that these first names were Hispanic, and that this common knowledge was confirmed by various websites. Id.
The Court does not find that Plaintiffs' post-trial citation of Ovalle changes the Court's finding that Dr. Gambitta's "actual count" fails to demonstrate that Plaintiffs' proposed district contains an HCVAP majority. The Court finds that Dr. Gambitta's methods of assessing SSRVs/Hispanic voters contained error, likely resulting in an inflated "actual count" by addressing omission but not commission errors. In contrast, the Ovalle court appears to have considered commission error. The court noted that it would not include "Commiato" as a Hispanic grand juror because the surname violates the Buechley rule against double letters. Ovalle, 13 S.W.3d at 783, n. 31. In addition, the court did not consider Avera and Honeo as Hispanic grand jurors because these surnames are only rarely Hispanic. Id. There is no evidence that Dr. Gambitta relied on such rules or methods. See Vol. 1 at 178-184 (where Dr. Gambitta testified that his knowledge of names which are Hispanic was "just in his head" and based on "common sense" and "general observation" and that the methodology he used was not written down or otherwise published).
Here, the evidence suggests that Dr. Gambitta counted as Hispanic any and all individuals for which there was an arguable basis for suggesting Hispanic origin, but that Dr. Gambitta did not review the 600 SSRVs identified by Defendant to determine if there was any basis for concluding these individuals were not Hispanic. Accordingly, the Court finds that Dr. Gambitta's methods were flawed, and that this is not cured by Plaintiffs' after-the-fact citation to Ovalle. Plaintiffs bear the burden of proving by a preponderance of the evidence that the HCVAP in the proposed district exceeds 50% in this case. The Court considers the potential inaccuracy in Plaintiffs' "actual count," and finds that Plaintiffs have failed to demonstrate HCVAP in the proposed district more likely than not exceeds 50%.
The Court notes that Plaintiffs' expert Dr. Gambitta testified that the Spanish surname list has a high degree of probativness in a macro sense, but not in the micro sense, such as with respect to Farmers Branch. Vol. 1 at 163. Dr. Gambitta states that the probability of a certain name on the Spanish surname list being Hispanic depends upon the community to which it is being applied. Id. Dr. Gambitta points out that a Barron in Idaho is probably not Hispanic, but a Barron in the County of Dallas has a higher probability of being Hispanic. Id. at 162-63.
It appears that Dr. Gambitta was attempting to make the point that there is a higher probability that a person with a Spanish surname from Farmers Branch, Texas will be Hispanic than a person with the same surname residing in other communities. The Court does not find, as urged by Defendant, that Dr. Gambitta was conceding that the basis for his expert opinion is not probative enough to be considered by the Court. See id. at 164; Doc. No. 56 (3:07-CV-900-O) (N.D. Tex. Aug. 25, 2008).
Dr. Gambitta appears to suggest that because Farmers Branch has a large Hispanic population, there is a higher probability that a Spanish surname registered voter in the proposed district is Hispanic and that all 600 SSRVs identified by Defendant can be assumed to be Hispanic. In other words, Dr. Gambitta argues that Plaintiffs' "actual count" is not flawed because Dr. Gambitta failed to account for commission error. The Court disagrees. Both Dr. Rives and Dr. Gambitta have testified that it is likely that not all SSRVs identified by Defendant are Hispanic. Id. at 173-76, 272. Dr. Gambitta conceded at trial that individuals such as Katherine Yonick Cantu and Vicky Leach Agraz should be removed from his count of SSRVs/Hispanic voters. See Vol. 1 at 175-176. In addition, even though Plaintiffs have produced evidence demonstrating that Farmers Branch has a relatively high percentage of Hispanics, the evidence also suggests that Farmers Branch does not have a large Hispanic citizen voting age population. See Pls' Ex. 1 (indicating that 37.2% of the city's population was Hispanic at the time of the 2000 Census); Def's Exhibit 1 (demonstrating that approximately 15.6% of eligible voters in Farmers Branch were Hispanic at the time of the 2000 Census). The Court does not find that Dr. Gambitta can rely on the unsubstantiated assumption that all 600 SSRVs identified by Defendant were Hispanic, and erred in considering omission but not commission errors. Accordingly, the Court finds that Plaintiffs' "actual count" does not demonstrate that it is more likely than not that Plaintiffs' proposed district contains a majority HCVAP, as required by Gingles. The Court now considers Plaintiffs' argument that Hispanic and non-Hispanic voter registration rates demonstrate that the proposed district contains a majority HCVAP.
c. Voter Registration Rates
Plaintiffs argue that the lower rate at which Hispanics register to vote, compared with non-Hispanics, demonstrates that there is HCVAP majority in the proposed district. Doc. No. 57 (3:07-CV-900-O) (N.D. Tex. Aug. 25, 2008). Plaintiffs point out that the pool of eligible voters is necessarily larger than the number of registered voters. Id. Dr. Gambitta produced evidence that only one of every 1.79 Hispanic citizen of voting age in Dallas County registered to vote, while one out of every 1.279 non-Hispanic citizen of voting age in Dallas County registered to vote. Vol. 1 at 87. Dr. Gambitta testified that if there are 600 Hispanics registered to vote in the proposed district out of 1291 total registered voters, 54.8 percent or 55 percent of the eligible voters in the proposed district are Hispanic. Id. at 88.
As discussed in Plaintiffs' Exhibit 7, in arriving at this factor, Dr. Gambitta looked at 2000 Census data for Dallas County and subtracted the HCVAP (182,740.23) from the total CVAP (1,296,030) for Dallas County, and determined that there are 1,113,289.77 non-Hispanic citizens of voting age population in Dallas County overall. Dr. Gambitta then looked at the SSRV data for Dallas County in 2000, using Spanish surname as a surrogate for Hispanic origin. In 2000, there were 102,061 SSRVs and 869,953 non-Spanish surnamed registered voters. From this, Dr. Gambitta concluded that 78.14% of non-Hispanic citizens of voting age registered to vote in Dallas County in 2000, while 55.85% of the Hispanic/Spanish-surnamed individuals registered to vote. Id. Based on these percentages, Dr. Gambitta concluded that there is one non-Hispanic registered voter for each 1.279 non-Hispanic citizen of voting age, while there is one Hispanic registered voter for each 1.790 Hispanic citizen of voting age. See Pls' Ex. 7.
Defendant responds that Dr. Gambitta's attempt to estimate HCVAP/CVAP is based on an unreasonable assumption. Doc. No. 56 (3:07-CV-900-O) (N.D. Tex. Aug. 25, 2008). Specifically, Defendant criticizes Dr. Gambitta's assumption that the factors he calculated based on statistics for Dallas County apply to Farmers Branch or, more specifically, to Plaintiffs' proposed district. Id. Defendant argues that, while these factors may apply generally to Dallas County, the evidence demonstrates that, unlike Dallas County as a whole, there is not much difference between the registration rates of Hispanics and non-Hispanics in the area that comprises Plaintiffs' proposed district. Id. Defendant's expert Dr. Alford testified that a higher percent of Hispanic citizens register to vote in the area comprising Plaintiffs' proposed district. Vol. 2 at 25-28. Dr. Alford notes that, based on 2000 Census data and 2000 voter registration records, the TLC estimated that the HCVAP in Plaintiffs' proposed district was 41.4% of the CVAP of the district, and that SSRVs are 40.1% of registered voters in the district, a difference of just 1.3%. Id.; see also Def's Ex. 6. Dr. Alford testified that this registration rate is not much different than the rate of registration observed in non-Hispanics in the area. Vol. 2 at 26-27. Based on this, Dr. Alford concluded that if Hispanic registered voters are less than a majority, the rates at which Hispanics and non-Hispanics register to vote fail to demonstrate that HCVAP exceeds 50% the CVAP of Plaintiffs' proposed district. Id. at 27-28.
Dr. Gambitta relied on the ratios he calculated in describing the relationship between eligible and registered voters, while Dr. Alford relied on percentages in discussing this relationship. To make it easier to compare Dr. Gambitta and Dr. Alford's opinions, the Court performed its own calculations, set out below, based on the data in evidence. As Defendant's Exhibit 6 demonstrates, the TLC estimated that there were 1895 citizens of voting age in the proposed district at the time of the 2000 Census, 41.4% of which were Hispanic. Def's Ex. 6. Thus, 784.53 citizens of voting age were Hispanic, while 1110.47 were non-Hispanic. Additionally, the TLC estimated that in 2000, there were 1,296 registered voters in Plaintiffs' proposed district, 40.1% of which were SSRVs. Def's Ex. 6. Thus, 519.7 SSRVs and 776.3 non-SSRVs in the proposed district are estimated to have been registered in 2000. Presuming SSRVs are Hispanic, 69.9% of the non-Hispanic CVAP in the proposed district registered to vote in 2000 (776.3 non-SSRVs out of 1110.47 non-Hispanic citizens of voting age), while 66.2% of the HCVAP registered to vote in the proposed district (519.7 SSRVs out of 784.53 Hispanic citizens of voting age). Thus, based on the TLC data, there is one non-Hispanic registered voter for each 1.43 non-Hispanic citizen of voting age, while there is one Hispanic registered voter for each 1.5 Hispanic citizen of voting age. These factors are much more similar than those calculated by Dr. Gambitta, who found that there is one non-Hispanic registered voter for each 1.279 non-Hispanic citizen of voting age, while there is one Hispanic registered voter for each 1.790 Hispanic citizen of voting age. See Pls' Ex. 7; Def's Ex. 6.
The Court finds that Plaintiffs have failed to demonstrate that Plaintiffs' proposed district has a majority HCVAP due to the relative rates of Hispanic and non-Hispanic voter registration. The Court credits Dr. Alford's testimony questioning the validity of the assumption that registration rates for Dallas County apply to Plaintiffs' proposed district. Defendant produced evidence that is more specific to the area that constitutes Plaintiffs' proposed district that demonstrates that the rates between Hispanic and non-Hispanic voter registration are very similar. The Court notes that Dr. Gambitta agreed that Dallas County and Farmers Branch are "certainly very different," and also testified that he believed the TLC estimates were accurate. Vol. 1 at 189, 196. Additionally, the Court notes that Dr. Gambitta testified that "Hispanics in Farmers Branch are registered to vote slightly less in percentages than the Anglo population." Id. at 110-11 (emphasis added); see also Pls' Ex. 39 (containing calculations for the number of Hispanic council members expected in Farmers Branch, indicating that the HCVAP of Farmers Branch is 16.4% and that Hispanics registered to vote comprise 16.1% of the Farmers Branch population). Considering this evidence, the Court credits Dr. Alford's conclusion that if Hispanic registered voters are less than a majority of the proposed district, the rates of voter registration do not demonstrate that HCVAP exceeds 50% of the CVAP of Plaintiffs' proposed district.
The Court notes that it previously found the TLC estimate was insufficient evidence for Plaintiffs to rely on to demonstrate that the proposed district has a majority HCVAP. See supra. However, the Court finds that the TLC estimate can be relied upon by Defendant here to call into question Dr. Gambitta's assumption that Dallas County registration rates apply to Plaintiffs' proposed district. The Court notes that Plaintiffs were using the TLC estimate to demonstrate the number of SSRV voters in the proposed district even though this number could be counted directly, whereas the number of Hispanic citizens eligible to vote can not be directly counted and must be estimated from Census data. Additionally, there has been no evidence presented that the TLC system's assumption that voters are dispersed evenly across a precinct would affect calculation of relative rates of voter registration. Accordingly, the Court finds that the TLC estimate was appropriately relied upon by Dr. Alford and Defendant.
The Court notes that Dr. Gambitta and Plaintiffs' attorneys repeatedly stated that the eligible pool of voters must be larger than the number registered, and therefore HCVAP would be larger than the number SSRVs in the proposed district. The Court finds that while this statement is true, this does not demonstrate that Plaintiffs' proposed district has a HCVAP majority, as it is also true that there is a larger pool of non-Hispanic eligible voters than non-Hispanics registered to vote. Plaintiffs have failed to show that it is more likely than not that Hispanics register to vote at a lower rate than non-Hispanics in the proposed district such that the pool of eligible Hispanic voters exceeds that of non-Hispanics, and the Court is unable to conclude that it is more likely than not that Plaintiff's proposed district has a HCVAP majority. The Court notes that it previously found that Plaintiffs' "actual count" did not demonstrate that it is more likely than not that the proposed district has a HCVAP over 50% of CVAP.
The Court's finding is supported by evidence indicating that individuals of unknown gender were not considered SSRVs in 2000, but were included in 2006 and 2007 counts of SSRVs. Id. at 196-97. This indicates that Dr. Gambitta's estimate of HCVAP in the proposed district in 2007, calculated based on the ratio of HCVAP to SSRVs in the year 2000, may be somewhat inflated, as the assessment of SSRVs in 2007 includes an additional category of people (individuals for which gender information is not known). The Court notes that this is not necessarily an insignificant issue, as Defendant's Exhibit 12 indicates that there was no gender information for approximately 126 of the 600 SSRVs identified by Defendant. See Def's Ex. 12.
Accordingly, the Court finds that Plaintiffs have failed to demonstrate that it is more likely than not that Plaintiffs' proposed district has a majority HCVAP due to the relative rates of Hispanic and non-Hispanic voter registration. In addition, the Court finds that the TLC estimate and Plaintiffs' "actual count" also fail to demonstrate that it is more likely than not that the proposed district has a majority HCVAP. While the evidence before the Court shows that the HCVAP population is somewhere near 50%, the Court finds it is just as likely that the HCVAP of the proposed district is below or equal to 50% of the citizen voting age population as above 50% HCVAP/CVAP.
Having concluded that Plaintiffs have failed to meet one of the threshold factors set out in Thornburg v. Gingles, 478 U.S. 30 (1986), the Court does not find it necessary to make findings with respect to Gingles I and II or regarding the totality of circumstances. Clements, 986 F.2d at 743. The Court also does not reach Defendant's argument that Plaintiffs' proposed district cannot be a Gingles I district because, while its total population is appropriate, the citizen population is below that of other districts such that the district violates the one-person, one-vote standard. Accordingly, the Court proceeds to make findings regarding Plaintiffs' Fourteenth Amendment claims.
D. Fourteenth Amendment
Plaintiffs allege that the at-large election of council members of the city of Farmers Branch dilutes the voting rights of Hispanic citizens in violation of the Fourteenth Amendment. However, the Court finds insufficient evidence of intentional discrimination in the adoption or maintenance of the election system in Farmers Branch. Plaintiffs have failed to show that the multimember form of Farmers Branch elections are responsible for Hispanics inability to elect the candidate of their choice. See Clements, 986 F.2d at 743.
It is unclear whether a plaintiff challenging an electoral system like the system in Farmers Branch can establish a constitutional vote dilution claim where a Section 2 VRA claim has failed. See Johnson v. DeSoto County Bd. Of Comm., 204 F.3d 1335, 1344-45 (11th Cir. 2000) (noting that it found no case in which a circuit court has concluded that an at-large or multi-member-district electoral system, although not in violation of section 2, unconstitutionally dilutes minority voting strength). Assuming it is possible, the Court finds Plaintiffs have failed to demonstrate Defendant enacted or maintains its electoral system with a discriminatory intent, and that the multimember form of the district is responsible for the minority voters' inability to elect their candidate of choice.
V. Conclusions of Law
A. Voting Rights Act
To demonstrate a violation of Section 2 of the VRA, the minority group challenging an at-large election scheme must demonstrate that the minority group is sufficiently large and geographically compact to constitute a majority in a single member district. Clements, 986 F.2d at 742. To satisfy this requirement, a plaintiff must demonstrate that it is possible to draw an election district of an appropriate size and shape where the citizen voting age population of the minority group is a majority. Perez, 165 F.3d at 372.
Here, Plaintiffs have failed to demonstrate that it is possible to draw a district that is sufficiently large and geographically compact. The TLC's estimate of SSRVs in the proposed district is unreliable and likely an overestimate, and is not sufficient evidence for the Court to find that Plaintiffs' proposed district contains a majority HCVAP. Similarly, Plaintiffs' "actual count" of SSRVs and Hispanic voters does not provide a basis for the Court to find more likely than not that the proposed district has a majority HCVAP. Plaintiffs' "actual count" has reliability issues caused by Dr. Gambitta's failure to consider commission error and by his use of methods that have not been published and cannot be replicated. In addition, Plaintiffs have failed to demonstrate that it is more likely than not that Plaintiffs' proposed district has a majority HCVAP due to the relative rates of Hispanic and non-Hispanic voter registration. Defendant produced evidence that is more specific to the area that constitutes Plaintiffs' proposed district that demonstrates that the rates of Hispanic and non-Hispanic voter registration are very similar. Accordingly, the Court is unable to conclude that it is more likely than not that Plaintiffs' proposed district contains a majority of Hispanic citizens of voting age.
B. Fourteenth Amendment
On the present record, the Court does not find that Plaintiffs' have proved a discriminatory intent in the adoption or maintenance of the election system in Farmers Branch. The Court notes that Plaintiffs have failed to show that the multimember form of Farmers Branch elections are responsible for Hispanics inability to elect the candidate of their choice. Accordingly, the Court is unable to find for Plaintiffs' on their Fourteenth Amendment claims.
V. Conclusions
For the forgoing reasons, the Court finds that Plaintiffs have failed to meet their burden with respect to their claims under Section 2 of the Voting Rights Act and the Fourteenth Amendment. The Court dismisses Plaintiffs' claims with prejudice and will issue judgment accordingly.