U.S. v. Village of Port Chester

Full title: UNITED STATES OF AMERICA, CESAR RUÍZ, Plaintiffs, v. VILLAGE OF PORT…

Court: United States District Court, S.D. New York

Date published: Apr 1, 2010

No. 06 Civ. 15173(SCR) (S.D.N.Y. Apr. 1, 2010)

Opinion

06 Civ. 15173 (SCR).

April 1, 2010

OPINION AND ORDER

STEPHEN ROBINSON, District Judge

On January 17, 2008, the Court found that Plaintiffs had demonstrated that the Village of Port Chester's at-large voting system for electing its Board of Trustees violated Section 2 of the Voting Rights Act of 1965. After careful consideration of the parties' proposed remedial plans, the Court issued a summary order on November 6, 2009 concluding that Defendant, the Village, had proposed a legally acceptable remedy and ordered the implementation of at-large elections with cumulative voting. In furtherance of the implementation, the parties were ordered to submit to the Court a Consent Decree detailing the requisite education and outreach program. The Court also lifted the injunction on the Trustee elections, providing that the 2010 elections shall be held in June 2010 on a date agreed to by the parties to give sufficient time for the proper implementation of the new system. This opinion combines the Court's findings in both the liability and remedial phase of the litigation and is the final order in this matter.

I. Procedural Background

The United States of America (the "Government") filed a Complaint on December 15, 2006 against the Village of Port Chester ("Port Chester" or the "Village" or the "Defendant"), alleging a violation of Section 2 of the Voting Rights Act of 1965 as amended, 42 U.S.C. § 1973. The Government claimed that the at-large system used to elect the six members of the Port Chester Board of Trustees denied the Hispanic population of the Village an equal opportunity to participate in the political process and to elect representatives of their choice.

The Government sought a preliminary injunction pursuant to 42 U.S.C. § 1973j(d) to prevent the Village from holding its next election for the Board of Trustees, which was then scheduled for March 20, 2007. The Court issued a preliminary injunction on March 2, 2007, finding: (i) that there would be irreparable harm if the 2007 Trustee election were allowed to proceed under a structural framework that violated the Voting Rights Act; (ii) that the balance of the potential harms weighed in favor of granting an injunction; and (iii) that the Government had demonstrated that it was likely to succeed on the merits of its claim at trial. Accordingly, the Village was enjoined from holding its March 20, 2007 Trustee election pending a trial on the merits in this matter.¹

1 The March 2, 2007 injunction did, however, allow the Village to decide whether it would go forward with its Mayoral election — also scheduled for March 20, 2007 — given that the instant lawsuit does not challenge the Village's system for electing its Mayor. The Village held its Mayoral election as scheduled, and as a result, Gerald Logan, who testified at both the preliminary injunction and trial phases of this proceeding, was replaced as Mayor by Dennis Pilla ("Pilla"), who also testified during the trial phase.

On March 1, 2007, Cesar Ruiz ("Ruiz"; Ruiz and the Government are collectively referred to herein as the "Plaintiffs") filed an Order to Show Cause why he should not be permitted to intervene in this action pursuant to Fed.R.Civ.P. 24. Following an oral argument, this Court granted Ruiz's motion to intervene as a party plaintiff on April 6, 2007.

After settlement negotiations proved unsuccessful, the parties reconvened for a six-day bench trial that concluded on June 5, 2007.² In lieu of oral closing arguments, the parties were granted until July 9, 2007 to submit post-trial briefs in support of their respective positions. Further, an organization called FairVote — which describes itself as having a mission "to advocate for fair representation through voting systems changes"³ — was given permission to submit an amicus curiae brief. The Court concluded that Plaintiffs have established that Port Chester's system for electing its Board of Trustees violates Section 2 of the Voting Rights Act, and directed the parties to submit proposed remedial plans.

2 In the interest of judicial efficiency, this Court accepted all of the testimony and exhibits from the preliminary injunction hearing as if they had been offered and received in the same way at the trial. See Trial Tr. at 2. Because the pagination of the May/June trial transcripts did not resume from where the February hearing transcripts ended, however, this Decision employs different citation formats for the different phases of the proceedings. All references to the transcripts of the May/June 2007 trial will be cited as "Trial Tr. at XX" (pages 1 through 1015), and all references to the transcripts of the February 2007 hearing will be cited as "Hearing Tr. at XX" (pages 1 through 1683).

3 See Br. of Amicus Curiae (docket number 77) at 1.

The Court held hearings on the remedial plans on July 17, 28, 29, 2008 and September 22 and 23, 2008. Port Chester proposed a voting scheme called cumulative voting that would give Hispanics greater opportunities to participate meaningfully in elections while maintaining the at-large system. Plaintiffs presented the districting plan developed in the liability phase as its proposed remedial plan. The Court issued a Summary Order on November 6, 2009 announcing its decision to choose Port Chester's proposed plan, ordered the parties to develop an education and outreach program to ensure a thorough and non-discriminatory implementation of the new system, and lifted the injunction on the Trustee elections provided that the 2010 elections were delayed until June to give enough time to educate the community about cumulative voting.

II. Port Chester's Voting Rights Act Violation

1. Section 2 of the Voting Rights Act

A. Legal Framework

Section 2 of the Voting Rights Act of 1965, as amended, 42 U.S.C. § 1973, reads:

(a) No voting qualification or pre-requisite to voting or standard, practice, or procedure shall be imposed or applied by any State or political subdivision in a manner which results in a denial or abridgement of the right of any citizen of the United States to vote on account of race or color.

(b) A violation of subsection (a) of this section is established if, based on the totality of the 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, that nothing in this section establishes a right to have members of a protected class elected in numbers equal to their proportion in the population.

There is no dispute here that Port Chester's at-large system for electing its Board of Trustees is an electoral practice or procedure that is subject to challenge under this statute. 2. Gingles preconditions and Senate Factors

The Supreme Court construed this statute in its amended version for the first time in an action challenging a multi-member at-large districting scheme. See Thornburg v. Gingles, 478 U.S. 30 (1986). In Gingles, 478 U.S. at 34, the Supreme Court set out three "preconditions" that must be met for a challenge under Section 2 of the Voting Rights Act to be successful:

(1) the minority group must be sufficiently large and geographically compact to constitute a majority in a single-member district;

(2) the minority group must be politically cohesive and vote as a bloc; and

(3) the White majority must vote sufficiently as a bloc to enable it, in the absence of special circumstances, to defeat the minority's preferred candidate.

No specific showing of discriminatory intent is required to prove a Section 2 violation. See id. at 70-73 (Brennan, J. plurality op.); Coleman v. Bd. of Educ. of the City of Mt. Vernon, 990 F. Supp. 221, 227 (S.D.N.Y. 1997) (internal citation omitted); cf. Goosby v. Bd. of the Town of Hempstead, 180 F.3d 476, 498-504 (2d Cir. 1999) (Leval, J. concurring) (hereinafter "Goosby III").

An analysis of the three Gingles factors and whether each has been proven by a preponderance of the evidence is the first step in a two-part analysis of a vote dilution claim on behalf of minority voters. The Supreme Court has found, however, that the satisfactory establishment of the three Gingles preconditions alone is not sufficient for a Section 2 vote dilution claim to succeed. See Johnson v. DeGrandy, 512 U.S. 997, 1011 (1994). Accordingly, this Court must "consider whether, under the totality of the circumstances, the challenged practice impairs the ability of the minority voters to participate equally in the political process." Goosby v. Bd. of the Town of Hempstead, 956 F. Supp. 326, 329 (E.D.N.Y. 1997) (hereinafter "Goosby I") (citing NAACP v. City of Niagara Falls, 65 F.3d 1002, 1007 (2d Cir. 1995)). Various Circuit courts have recognized that "it will only be the very unusual case in which the plaintiffs can establish the existence of the three Gingles factors but still have failed to establish a violation of Section 2 under the totality of the circumstances." Niagara Falls, 65 F.3d at 1019, n. 21; see also Thompson v. Glades County Bd. of County Comm'rs, 493 F.3d 1253, 1261 (11th Cir. 2007); Jenkins v. Red Clay Consol. Sch. Dist. Bd. of Educ., 4 F.3d 1103, 1116 n. 6 (3d Cir. 1993).

Judicial assessment of the totality of the circumstances requires a "searching practical evaluation of the past and present reality." Gingles, 478 U.S. at 45. The key to this inquiry is an examination of the seven principal factors set forth in the Senate Judiciary Committee Report accompanying the 1982 amendments to Section 2 of the Voting Rights Act, the so called "Senate factors." See S. Rep. No. 97-417, 97th Cong. 2nd Sess. 28 (1982) (the "Senate Report"). The additional factors listed in the Senate Report are:

1. the extent of any history of official discrimination in the state or political subdivision that touched the right of members of the minority group to register, vote, or otherwise to participate in the democratic process;

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 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;

4. if there is a candidate slating process, whether the members of the minority group have been denied access to that process;

5. 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, which hinder the ability to participate effectively in the political process;

6. whether political campaigns have been characterized by overt or subtle racial appeals; [and]

7. the extent to which members of the minority group have been elected to public office in the jurisdiction.

In addition, the Senate Report adds two other considerations that may have probative value in vote dilution cases, specifically:

(1) whether there is a significant lack of responsiveness on the part of the elected officials to the particularized needs of the members of the minority group; and

(2) whether the policy underlying the state or political subdivision's use of such voting qualification, prerequisite to voting, or standard, practice or procedure is tenuous.

The list of factors is "neither comprehensive nor exclusive."Gingles, 478 U.S. at 45. Plaintiffs need not prove a majority of these factors, nor even any particular number of them in order to sustain their claims. Instead, "these factors are simply guideposts in a broad-based inquiry in which district judges are expected to roll up their sleeves and examine all aspects of the past and present political environment in which the challenged electoral practice is used."Goosby I, 956 F. Supp. at 331.

B. Findings of fact

1. Overview of the Village of Port Chester

Port Chester is an incorporated village located within the Town of Rye, and is situated in southeastern Westchester County, New York, adjacent to the Connecticut border. According to the 2000 United States Census, Port Chester's population was 27,867, an increase of 12.7 percent from the 1990 Census. From 1990 to 2000, the Hispanic population of the Village grew by 73 percent from 7,446 to 12,884; the Hispanic community now constitutes a plurality of Port Chester's residents. As of the 2000 Census, Port Chester's total population was 46.2 percent Hispanic, 42.8 percent non-Hispanic White, and 6.6 percent non-Hispanic black. Of Port Chester's voting age population ("VAP") of 21,585 in 2000, however, 45.7 percent were non-Hispanic White, 43.4 percent were Hispanic and 6.1 percent were non-Hispanic black. Meanwhile, as of 2000, the Village had a total citizen voting age population ("CVAP") of 13,990, of whom 65.5 percent (9,160) were non-Hispanic White, 21.9 percent (3,070) were Hispanic and 8.9 percent (1,245) were non-Hispanic black. Plaintiffs' expert Dr. Andrew Beveridge ("Dr. Beveridge") estimated that as of July 2006, Port Chester's CVAP totaled 14,259, of which Hispanics constituted 27.5 percent (3,928).

Port Chester is governed by a Mayor and a six-member Board of Trustees, and all of these Village officials are elected pursuant to an at-large voting scheme. The Trustees serve staggered three-year terms, with two Trustee positions open for election each calendar year; the Mayor, who presides over the Board of Trustees, serves a two-year term, and thus must stand for election every other year. Each resident of the Village who is registered to vote may cast up to two votes for Trustee candidates. A voter cannot select the same candidate twice, but a voter may opt to cast just one of his or her two votes and withhold the other, a practice known as "single shot" or "bullet" voting. Village elections for Mayor and Trustees are held "off cycle" — that is, they are not conducted in November alongside other county, state, and national elections, but instead are held in the spring, usually on the third Tuesday in March.

The Village is divided into 16 election districts for the purposes of voting administration. These districts determine at which polling place Port Chester's voters cast their ballots for both Village elections in March and "on-cycle" county, state, and national elections in November.⁴ In addition, it has been the practice of the Republican and Democratic parties in Port Chester to choose "district leaders" for each election precinct. To be clear, however, these precincts in no way correspond to any type of elected representation — voters in each of the Village's election precincts choose from the same slate of candidates in local elections. The Town of Rye, which in addition to the Village of Port Chester also encompasses the incorporated village of Rye Brook as well as the Rye Neck section of Mamaroneck, consecutively numbered all of the election precincts within the Town; those that lie within Port Chester are precincts 5 through 19 and precinct 25.

4 To eliminate any potential confusion, this Court will refer to the 16 numbered election districts as "precincts" for the purposes of this Decision; future references to "districts" will refer to Plaintiffs' proposed political subdivisions for electing representatives to the Board of Trustees.

2. First Gingles precondition: the minority group must be sufficiently large and geographically compact to constitute a majority in a single-member district

a. Criteria for drawing proposed districts

To demonstrate the existence of the first Gingles precondition in an at-large system, the Plaintiffs must be able to draw illustrative single-member districts following traditional districting principles to show that the Hispanic population is sufficiently large and compact so as to constitute a majority in a single-member district. Dr. Beveridge, an expert in the fields of demographics and redistricting, offered two alternative plans — Proposed Plan A ("Plan A") and Proposed Plan A as Modified ("Modified Plan A") — each of which divided the Village into six hypothetical single-member districts that would allow Port Chester to elect the six members of its Board of Trustees using a district-based, rather than an at-large, system. See Gov. Exs. 32 (Plan A) and 33 (Modified Plan A).

To draw the proposed districts in each of the plans, Dr. Beveridge first sought to ensure equality in the total population of each district, and next endeavored to make each district "reasonably compact." See Hearing Tr. at 599. Dr. Beveridge testified that he opted to draw the districts on the basis of total population because in his view "total population is the accepted standard method"; he did not know of any districting process in the United States that has used a method other than total population for drawing district lines.⁵ See Hearing Tr. at 600. Population equality and compactness are "two of the most relevant [re]districting principles" in smaller geographic areas — such as Port Chester — where districting experts need not be concerned about splitting towns and villages when drawing potential district boundaries. Id. Only after these principal criteria were met did Dr. Beveridge seek, to the extent possible, to keep together a portion of the Hispanic community of Port Chester within a single proposed district in a way that did not "pack" or "crack" the Hispanic population.⁶

5 Defendants argue that the Plan A and Modified Plan A create an unconstitutionally extreme deviation in the CVAP of the districts. Defendants contend that the votes of citizens in, for example, District 1 are "devalued" by the proposed districting plans because the number of citizens in District 1 is greater than the number of citizens in District 4, yet the Districts have the same share of political power in the Village since each district would be able to elect one member of the Board of Trustees. While this argument raises important legal questions deserving of full analysis, it is inappropriate to address those questions at this time because the Court has chosen the Village's proposal for cumulative voting over the Plaintiffs' districting plan. See infra III.C. Since the districting plan's legality is not before the Court, it is not compelled to confront Defendant's devaluation concerns. Accordingly, the Court reserves judgment on that question.

6 As Dr. Beveridge described, "packing" a minority population would involve forcing as many members of the minority community as possible into a single district to limit their political clout. Conversely, "cracking" a minority community involves spreading the minority community to limit that group's ability to elect a candidate of its choice. See Hearing Tr. at 634-36.

Both of the proposed plans show very limited deviation in the total population among the six proposed districts based on data from the 2000 Census. Given a total Village population of 27,867, an equal division of the population for each district would yield approximately 4,645 individuals in each district. In Plan A, the district with the smallest population is District 5, which contains 4,528 people — 117 less than the ideal — for a deviation of 2.51 percent. See Gov. Ex. 25 at Ex. F. The largest district by population in Plan A is District 3, which contains 4,793 people — 149 more than the ideal — for a total deviation of 3.20 percent. Id. Thus the total population deviation in Plan A measured by the spread between the greatest downward deviation and greatest upward deviation within the districting plan is 5.71 percent, a figure that is comfortably within the bounds of acceptable 10 percent population deviation for state or local legislature districting purposes. See, e.g., Brown v. Thompson, 462 U.S. 835, 842-43 (1983); White v. Regester, 412 U.S. 755 (1973). Under Modified Plan A, the greatest downward deviation is again in District 5, though the difference in this model is only 1.50 percent; similarly, District 3 exhibited the greatest upward deviation in Modified Plan A, but with only a 1.84 percent departure from the ideal figure. See Gov. Ex. 26 at T. 2. Modified Plan A therefore has a total deviation of 3.34 percent, again well within acceptable population deviation parameters. Defendant's expert Dr. Peter Morrison ("Dr. Morrison"), an expert in the fields of demography and drawing and evaluating single-member districts, agreed that the total population balance of both Plan A and Modified Plan A falls within acceptable limits. See Def. Ex. LL at 34.

As to Dr. Beveridge's second criterion, both Plan A and Modified Plan A are reasonably compact. The only challenge that can be construed as relating to the compactness of the proposed districts is the Village's contention that the proposed districts for Trustee representation should have taken into account existing election precincts in the Village. Based on the current election precinct boundaries and the proposed district boundaries in Plan A and Modified Plan A, the district plans would create a system where the population of certain election precincts would be divided among one or more Trustee districts. For example, certain residents of election precinct 14 would be eligible to vote in Trustee District 1, while others would be eligible to vote in District 3, District 4, or District 5. In fact, 10 of the 16 current election precincts would experience these types of cleavages. See Def. Ex. X.

Such a scenario raises administrative and logistical concerns for the Village, as Port Chester would have to ensure that voters are presented with the proper electoral choices when they arrive at their polling places. Assuming that the precinct boundaries remained the same and that the proposed districts were implemented, the Village might, for example, be forced to use different practices and procedures for March elections (which, under the current electoral framework, would implicate the Trustee district boundaries) and November elections (which would not). Dr. Morrison admitted, however, that this issue is a "purely economic consideration" — his concern is that "it would be an expense imposed on the taxpayers of the Village to have to have a different set of geographies used for elections" — though he also acknowledged that he did not analyze any potential expenses or savings that might result. Trial Tr. at 478, 556. Dr. Beveridge did not place great emphasis on election precinct boundaries given that they are simply an administrative mechanism for localities, and given that such precinct lines, in his experience, are commonly redrawn during districting or re-districting processes. See Hearing Tr. at 613. There was no evidence presented at any point in this proceeding to explain why Port Chester's election precincts were drawn the way they were, or why it would be important to preserve those particular boundaries.

On balance, the decision not to give any particular weight to election precinct lines here was sensible; these purely administrative designations do not signify anything of overwhelming import in Port Chester, and do not represent the types of political boundaries that are particularly deserving of deference when crafting proposed district borders. In no way would it have been advisable to place greater emphasis on the maintenance of existing precincts than on the other criteria employed by Dr. Beveridge in crafting Plan A and Modified Plan A.

Finally, though he admits that he used race as part of his districting process, Hearing Tr. at 634, Dr. Beveridge did not, as Defendant suggests, use race as his only criterion, or even as his predominant criterion, in drawing either of the proposed plans. The proposed districts did not result in any impermissible packing or cracking of the Hispanic population of the Village. Indeed, the distribution of Hispanics across the proposed district lines under both Plan A and Modified Plan A results in four districts where Hispanics account for a greater percentage of the total population, VAP, and CVAP in those districts as compared with the Hispanic share of population, VAP, and CVAP in Port Chester as a whole. By way of illustration, the Hispanic community constituted 46.23 percent of the population of the Village based on 2000 data, as well as 43.34 percent of the VAP and 21.87 percent of the CVAP. Under Plan A, the Hispanic population shares of the following four proposed districts exceed these thresholds on all three metrics: District 5 (48.43 percent Hispanic population; 44.31 percent Hispanic VAP 26.00 percent Hispanic CVAP); District 6 (54.37 percent Hispanic population; 51.80 percent Hispanic VAP; 28.13 percent Hispanic CVAP); District 3 (54.45 percent Hispanic population; 51.67 percent Hispanic VAP; 29.95 percent Hispanic CVAP); and District 4 (75.40 percent Hispanic population; 73.83 percent Hispanic VAP; 50.51 percent Hispanic CVAP). Under Modified Plan A, the Hispanic populations of District 6, District 3, and District 4 constitute a greater percentage of the total population, VAP, and CVAP in those districts as compared with the Hispanic population of Port Chester as a whole, while the Hispanic population of District 5 accounts for a greater percentage of CVAP, but slightly smaller percentages of total population and VAP as compared to the Village generally.

In addition to the favorable comparisons between intra-district ethnic compositions and the Village population as a whole, the inter-district distribution of the Village's Hispanic population in both Plan A and Modified Plan A is well-balanced. For example, in Modified Plan A, District 3, District 4, District 5, and District 6 contain 82.05 percent of the total Hispanic CVAP of the Village, with each of those Districts accounting for no less than 19.33 percent and no more than 22.79 percent of the Village-wide share based on 2000 Census data. Similarly, those four districts include 83.94 percent of the Hispanic VAP of Port Chester and 83.47 percent of the total Hispanic population of the Village; while District 4 contains the greatest share of Village-wide Hispanic VAP (28.54 percent) and total Hispanic population (28.10 percent), the remaining three Hispanic-heavy proposed districts contain between 16 and 21 percent of the Village-wide share of Hispanic VAP and total population.

Through cross-examination of Dr. Beveridge, Defendant attempted to demonstrate that the non-Hispanic White population of the Village was impermissibly packed under the proposed districting regime. Neither Dr. Beveridge's methodology nor the resulting data support this contention, and the Court therefore rejects the notion that either Plan A or Modified Plan A packed the non-Hispanic White population of the Village. The mere fact that there are greater concentrations of non-Hispanic Whites in certain areas and greater concentrations of Hispanics in other areas does not indicate any sort of nefarious effort; instead, this merely is a reflection of the reality of residential segregation in Port Chester.⁷ Under Modified Plan A, 72.60 percent of the non-Hispanic White population of the Village is concentrated in District 1, District 2, and District 5, but the greatest proportion in any one district is the 30.18 percent of non-Hispanic Whites living in District 1. Similarly, those three districts also contain 72.18 percent of the non-Hispanic White VAP and 74.67 percent of the non-Hispanic White CVAP of the Village, but none of the districts contains more than a 30 percent share of the non-Hispanic White VAP or CVAP of the Village. These figures simply do not support a finding that the proposed districting plans create impermissible concentrations of the non-Hispanic White population in Port Chester.

7 Dr. Morrison calculated that there was a "moderately high" level of segregation in the Village. Hearing Tr. at 1492. In addition, Plaintiffs' expert Robert Courtney Smith ("Professor Smith"), an expert in the areas of history of discrimination against Hispanics and the socioeconomic disparities of Hispanics in New York, described Port Chester as "a very segregated town."Id. at 372 (expert qualifications); 408 (quotation).

In sum, it is clear that the proposed districts in Plan A and Modified Plan A were drawn in accordance with traditional districting principles of population balancing and compactness, and there is no evidence in the record to indicate that race — of Hispanics or non-Hispanics — was the predominant factor in crafting the proposed district boundaries.

b. Use of 2000 Census data and 2006 estimates

As noted above, Dr. Beveridge relied on data from the 2000 Census in drawing the proposed districts in Plan A and Modified Plan A, and data from the 2000 Census formed the basis of the majority of his expert conclusions. The 2000 figures represent the most recent set of comprehensive Census information for Port Chester — no new complete Census information will be available until sometime in 2011. Various exhibits — such as Gov. Ex. 34 — show the 2000 Census data broken down by the block level, indicating the number of individuals counted for all of the blocks of the Village that appear on that particular map. Defendant attempted to call into question the accuracy and reliability of the Census figures, but the Village's efforts did not produce any clear, concrete, and comprehensive demonstration that the 2000 Census data for Port Chester is in any way significantly unreliable.⁸

8 Defendant presented testimony from Patrick Cleary ("Cleary"), who worked as the Village's principal planner from 1986 through 1990, and has served as a planning consultant to Port Chester for the majority of time since then. Referring to Gov. Ex. 34, Cleary testified that certain blocks on that map (which, according to the 2000 Census, contain some number of residents) are in fact purely industrial areas that have no residential units whatsoever. He further testified that other blocks likely did not contain as many residents as indicated by the 2000 Census data as a result of various development initiatives in the Village. See generally Trial Tr. at 797-813.
During cross-examination, however, Cleary admitted that he uses Census data as a benchmark in his own work as a planner, and it also became evident that Cleary had no idea about the population characteristics of Census blocks in Port Chester other than the few he described during his direct testimony. See id. at 815, 818-19. Cleary's selective examination of certain Census blocks in Port Chester consisted of little more than his personal observations of land uses and development patterns — there was no analysis of population trends, and certainly no sampling of the accuracy of Census data throughout the Village. On the whole, Defendant's contention that the 2000 Census should not be accepted as reliable is unavailing.

As part of his analysis, Dr. Beveridge estimated the demographic changes that he believes have occurred in Port Chester since 2000 based on an extrapolation from the 2000 Census figures and the rate of change in voter registration in the Village. This Court takes Dr. Beveridge's 2006 estimates for what they are: estimates provided by a demographics expert that, while not endowed with the same presumption of reliability as the decennial census, may nevertheless be used by this Court to understand relevant population trends in the Village. Nevertheless, there is no need to rely directly on the 2006 estimates in making any determinative findings of fact or conclusions of law with respect to the ultimate issues in this case.⁹ Accordingly, to the extent that the Court has reviewed and considered the 2006 data, it has only been for background informational purposes. As set forth throughout this Decision and Order, the data on which the Court relies for its assessment of the Gingles preconditions and the Senate factors is the data from the 2000 Census.

9 The Court will not discuss at length Dr. Morrison's effort to discredit Dr. Beveridge's estimates (see Def. Ex. LL. at 25-30), though considerable time was spent on this issue at both the hearing and trial of this matter. The Court was not convinced by Dr. Morrison's analysis in this part of his report because, among other reasons, his charts relied heavily on the accuracy of registration data but did not reflect any of the admitted problems with registration data as an accurate measure of voters in a given area. See section III.B.3 infra.

c. Measuring the effective majority in a single-member district

Within both Plan A and Modified Plan A, the proposed district that is meant to satisfy the first Gingles precondition for the Hispanic community is District 4. The best method to judge whether a particular minority group constitutes an effective majority in a single-member district is to examine the VAP and CVAP data for that district.¹⁰ The Hispanic community comprises 73.83 percent of the VAP in proposed District 4 under the Plan A boundaries, and constitutes 50.51 percent of CVAP according to 2000 Census data. In Modified Plan A, the Hispanic VAP in District 4 is 77.27 percent of the population there, and the Hispanic CVAP makes up 56.27 percent of the district. Even Dr. Morrison conceded that even taking into account possible data errors, Hispanics would constitute a majority of CVAP in District 4 in Modified Plan A. See Hearing Tr. at 1390, 1429, 1430, 1456.

10 Though districts are drawn on the basis of total population, the effectiveness of the minority group is not measured with reference to the total population data. Dr. Morrison indicated in his final expert report that showing that Hispanics represent a majority of CVAP in a single member district "is a typical method for arguing that there is an ability to elect." Def. Ex. LL at 21.

Port Chester challenged the use of VAP and CVAP for measuring whether Hispanics constituted an effective majority in proposed District 4 under any of Plaintiffs' plans by attacking the reliability of the VAP and CVAP figures and by offering alternative approaches to the effective majority question.¹¹ However, Port Chester failed to convince the Court that there are documented discrepancies in VAP and CVAP or that the other methods it proposed, such as using "corrected" Census numbers, voter registration, or voter turnout, were more reliable measures. In particular, the Court finds that voter registration and voter turnout methods have serious shortcomings that render them inappropriate for this analysis. Voter registration overstates the number of eligible voters in a given location because they are only scrutinized and updated from time to time.See Hearing Tr. at 1461-62. Records kept by the United States Election Assistance Commission ("EAC") illustrate this problem; according to EAC voter registration data for New York State from the 2004 general election, a highly implausible 99.3 percent of the CVAP in Westchester County¹² was registered to vote at the time of that election. See Gov. Ex. 92. In smaller counties, the 2004 data revealed complete mathematical impossibilities — according to those figures, 100.3 percent of the CVAP in Orleans County, New York was registered to vote, and 104.0 percent of the CVAP in Sullivan County, New York was on the registration rolls. Id.

11 Dr. Morrison posited that CVAP data for Port Chester may not be fully accurate because there is some evidence that there is over-reporting of citizenship status by Hispanics in the Census in general — that is, Hispanic non-citizens will indicate on Census forms that they are in fact citizens. See Def. Ex. LL at 31. Dr. Morrison did admit, however, that he had no evidence showing any such over-reporting of citizenship status in Port Chester specifically. See Hearing Tr. at 1500.

12 This Court recognizes that there is an inherent problem with the data collected in Gov. Ex. 92, because the chart compares the voter registration rolls in November 2004 with the CVAP data as of the 2000 Census. A population increase in these jurisdictions between 2000 and 2004 could account for part of, but certainly not all of, the inflated percentage of Census 2000 CVAP registered to vote in 2004. It is our view, however, that population changes alone cannot account for the significant degree to which the percentage of CVAP on the registration rolls exceeds expected levels; those discrepancies are more likely than not attributable to the types of problems — including deceased voters and voters who moved — highlighted by the Plaintiffs.

This Court also rejects the notion, offered in Dr. Morrison's expert report, that a proper measure of an effective majority must include a consideration of voter turnout. Using Spanish Surname Analysis of voter sign-in sheets in Port Chester, ¹³ Dr. Morrison calculated only between 9.4 and 11.8 percent of all people who cast votes in Village elections between 2001 and 2006 were Hispanic; the greatest percentage of Hispanic voters in any one year — 11.8 percent of turnout — came in the March 2006 elections. See Def. Ex. LL at 23 (Table 5). Based on the borders of proposed District 4 in Plan A, ¹⁴ Dr. Morrison found that in those same elections, only between 21.0 and 28.2 percent of actual voters in the illustrative District were Hispanic, with the greatest percentage of Hispanic voters again occurring in March 2006. According to Dr. Morrison, these figures reveal that even if Hispanics constitute a majority of VAP and CVAP in District 4, they will not amount to an effective majority in that District because they do not turn out in sufficient numbers to elect candidates of their choice without assistance from other demographic groups.

13 Experts for both parties used the Census Bureau List of Spanish Surnames to calculate the number of Hispanic voters in a particular area in various charts they prepared (hereinafter "Spanish Surname Analysis"). Neither party disputes that Spanish Surname Analysis is an accepted methodology, though it is clear that certain individuals who identify as Hispanic will be missed by this approach and certain individuals who do not identify as Hispanic will be included in these counts.

14 Dr. Morrison does not provide any analysis of voter registration or turnout for proposed District 4 in Modified Plan A, the plan that clearly contains a greater proportion of Hispanic population, VAP, and CVAP within the boundaries of District 4.

What this turnout analysis fails to consider is that there may very well be a correlation between the subject matter of this lawsuit — the various circumstances and conditions that contribute to the inability of the Hispanic community to elect candidates of its choice — and the lower turnout by Hispanic citizens in Port Chester. Defendant's expert Dr. Ronald Keith Gaddie ("Dr. Gaddie"), an expert in the fields of elections and voter participation, acknowledged that a district-based system with a majority-minority district would likely increase the number of Hispanic candidates who would run for office (and the number who would win), and that such candidates would likely stimulate increased voter participation — both in terms of registration and turnout — among the Hispanic population. Hearing Tr. at 1289 (qualifications); 1344-48.

It is interesting to note that in the 2007 Mayoral election — held in the wake of this lawsuit, and less than two weeks after the issuance of the preliminary injunction halting the Trustee elections — Hispanic turnout both Village-wide and within the confines of proposed District 4 was the highest it had ever been for a Village election from the years 1995-2007 (15.3 percent Village-wide, and 44.5 percent within proposed District 4). See Def. Ex. LL at Table 5. Though these figures are subject to multiple interpretations, when they are combined with the testimony and other evidence presented in this case, it seems highly likely to this Court that a dramatic change in the electoral structure to give Hispanics a better opportunity to participate would likely result, for myriad reasons, in a marked change in voter turnout. Accordingly, it would be counterintuitive to determine that depressed turnout among Hispanics — a condition that may very well be a direct byproduct of the existing electoral regime — should be a reason to preclude the creation of a new electoral structure in Port Chester.

On balance, the most reliable measure of whether Hispanics constitute an effective majority in proposed District 4 in Plan A and Modified Plan A is the CVAP data for Port Chester. As discussed above, Hispanics constitute a slight CVAP majority in District 4 under Plan A, and an even more substantial majority under Modified Plan A; Plaintiffs, therefore, have made a sufficient showing to satisfy this component of the first Gingles factor.

3. Second Gingles precondition: the minority group must be politically cohesive and vote as a bloc

Plaintiffs' expert, Dr. Lisa Handley ("Dr. Handley"), is an expert in the fields of racially polarized voting, analyzing voting behavior, statistical analysis of voting, and the effect of electoral practices of minority participation and representation. Hearing Tr. at 461. Dr. Handley used three methods of statistical analysis — bivariate ecological regression analysis, ecological inference, and homogeneous precinct analysis — to review election data and to determine how voters cast their ballots in those contests. Hearing Tr. at 480, 500. In her initial report in this case, Dr. Handley used voter registration data to estimate voter behavior — this was the only data that was available to her at the time she prepared the initial expert report. See Gov. Ex. 12. In her subsequent reports, however, Dr. Handley used sign-in data, which reflects actual voter turnout, and therefore provides a more reliable basis for estimating voter preferences. See Gov. Ex. 13. Defendant's expert, Dr. Ronald Weber ("Dr. Weber"), is an expert in the fields of political science, state and local politics, quantitative analysis of voting behavior, and demography. Hearing Tr. at 894. In analyzing election data, Dr. Weber also used bivariate ecological regression analysis and ecological inference methodologies, and used voter sign-in data as the basis for his conclusions. Def. Ex. B; Hearing Tr. at 507.

Both experts analyzed "endogenous" and "exogenous" elections as part of their work for this matter. Endogenous elections are those involving the specific office at issue in the lawsuit (i.e. Port Chester Trustee elections). Hearing Tr. at 470. This Court also treats Mayoral elections in Port Chester as endogenous elections in this case — the Mayoral elections are conducted in precisely the same manner as the Trustee elections with precisely the same set of voters, and in the structure of the Village government, the Mayor presides over meetings of the Board of Trustees and votes along with the Trustees on legislative initiatives. Exogenous elections are contests for positions other than Trustee or Mayor in the Village — for example, county-wide races for judgeships or the office of the district attorney, and state-wide races for attorney general. Hearing Tr. at 471.

To determine whether minority voters vote cohesively, Dr. Handley considers the degree to which those voters support the same candidates, and will look to the gap between the percentage of votes for the minority-preferred candidate and the non-preferred candidate rather than using a particular bright-line threshold for determining cohesion.

For the years 2001-2007, Dr. Handley in her three reports analyzed 16 endogenous elections — 12 of these were the Trustee contests in each of these six years (two positions were up for election each year), and the remaining four were the Mayoral races in 2001, 2003, 2005, and 2007.¹⁵ Dr. Handley testified that "in all of the 16 contests that [she] looked at, Hispanics were cohesive." Trial Tr. at 13. According to Dr. Handley, this was especially true in 2001, when the Trustee election included Ruiz, a Hispanic candidate. Both Dr. Handley and Dr. Weber concluded that virtually 100 percent of Hispanics who voted in that election cast one of their votes for Ruiz, the Hispanic candidate. See Hearing Tr. at 483-84; Def. Ex. B at 29, 37. The data presented in Dr. Handley's reports indicate that Hispanics also voted cohesively in endogenous elections where there was no Hispanic candidate. See Gov. Ex. 13.

15 This Decision concentrates on endogenous elections between 2001 and 2006 because both Dr. Handley and Dr. Weber testified that these contests produced the most reliable data for their analyses. See Hearing Tr. at 499, 506-07 (Dr. Handley explaining how the data for the 1999 and 2000 Trustee elections were unreliable); 1015 (Dr. Weber admitting his concern about the reliability of the data in this case, particularly with respect to the "earlier" data between 1995 and 2000).

Dr. Weber testified that the Hispanic community is not cohesive, in his view, because with the exception of the 2001 Trustee election, the turnout of the Hispanic group is lower than the minimum threshold required to constitute cohesion.¹⁶ Hearing Tr. at 982. Dr. Weber defines the "minimum threshold" for cohesion as 10 percent of the CVAP. Hearing Tr. at 926. The Court rejected this bright-line rule that Dr. Weber conceded was an arbitrary figure that no court has explicitly used and no other expert in the field has adopted.¹⁷ Hearing Tr. at 981. Dr. Weber also conceded that if the Court rejected his 10 percent rule, then the Hispanic voters in Port Chester were cohesive in 13 of 15 Trustee and Mayoral elections and strongly cohesive in 10 of the 15 elections between 2001 and 2006.¹⁸ Hearing Tr. at 977-79.

16 Unlike Dr. Handley, Dr. Weber will only consider a group to be cohesive if 60 percent of that group votes for the same candidate. Given the degree of cohesion of the Hispanic community in Port Chester, however, this requirement was regularly met; accordingly, this Court will not make any findings of fact regarding this 60 percent threshold in this section. But see section III.D.2 infra.

17 When questioned, Dr. Weber conceded that the 10 percent figure was arbitrary. The Court inquired: "But why ten? Why is ten the magic number and not 12 and not 8? . . . Is there any science; is there any statistical data that supports a 10 percent number? Versus a 9 percent, versus an 11 percent number?" Dr. Weber's candid response was "No. No there isn't. No." Hearing Tr. at 980.

18 Dr. Weber did not perform any analysis of the 2007 Mayoral election. Hispanic voters were not cohesive in the remaining two elections, according to Dr. Weber, because they did not reach his 60 percent threshold.

4. Third Gingles precondition: the White majority must vote sufficiently as a bloc to enable it, in the absence of special circumstances, to defeat the minority's preferred candidate

a. Most probative election contests

The parties' experts offered substantially different views of which elections this Court should consider most important in analyzing the third Gingles precondition, though both sides agreed that the 2001 Trustee race in which Ruiz was a candidate was the most significant of the elections they studied. See Hearing Tr. at 469, 473-74; 984. The 2001 Trustee contest was the only endogenous election that involved a Hispanic candidate who was the candidate of choice of the Hispanic community; indeed, it was the only endogenous election from the 2001-2006 time period that involved a Hispanic candidate at all.

Dr. Handley posited that after the 2001 Trustee race, the next most significant elections were three exogenous elections that included Hispanic candidates who were the candidates of choice of the Village's Hispanic community. These included: (i) the 2001 and 2005 races for Westchester County District Attorney, in which Anthony Castro ("Castro"), a candidate of Portuguese ancestry, was defeated both times; and (ii) the 2000 race for Westchester County Family Court Judge, a contest in which Nilda Morales Horowitz ("Judge Morales Horowitz"), became the first person of Hispanic ancestry ever to win election to a countywide office in Westchester. Finally, Dr. Handley added that endogenous races that did not involve Hispanic candidates — so-called "White versus White" endogenous contests — were also of some probative value. Hearing Tr. at 476-77.

In addition to the 2001 Trustee election, Dr. Weber believed that the most important contests for consideration were the 2000 Westchester County Family Court Judge election, the 2005 District Attorney election, and the 2002 race for New York State Attorney General. Hearing Tr. at 901-03. This final example — the only one not listed by Dr. Handley — included a Hispanic candidate, though that candidate was clearly not the candidate of choice of the Hispanic community in Port Chester.

Dr. Handley would have included among her most important elections the exogenous contests for the Port Chester-Rye Brook Board of Education (the "School Board"). Several of these races involved minority candidates, and the electorate for School Board seats includes all of Port Chester and only a small contingent of Rye Brook voters who live outside of the Village. Hearing Tr. at 474-75. Because these elections take place in a single voting precinct, however, it is not possible to perform the same types of statistical analysis for the School Board elections as were performed for all other endogenous and exogenous elections studied in this case. Hearing Tr. at 475. Given these limitations, Dr. Handley could not consider the School Board elections as part of her analysis. Nevertheless, the Government did attempt to use the School Board elections as evidence of non-Hispanic bloc voting by presenting information regarding the outcomes of various School Board elections.

Testimony from various witnesses revealed difficulties with the exogenous elections described by these experts; specifically, in the three elections that involved Castro and Judge Morales Horowitz, there was considerable disagreement about whether voters perceived these candidates to be Hispanic. See, e.g., Hearing Tr. 182-84 (Vega testimony regarding Judge Morales Horowitz campaign); 778-79, 785 (Judge Morales Horowitz testimony regarding her campaign); Gov. Ex. 35. It is clear that these examples only retain the level of relevance attributed to them by Dr. Handley and Dr. Weber if these candidates were thought to be Hispanic. Otherwise, the elections would be no more important than other "White versus White" exogenous races.

After careful consideration, this Court concentrated its examination on the endogenous elections — for Mayor and for various Trustee positions — that were held exclusively within the Village of Port Chester. It is clear that the 2001 Trustee race took on special significance because of Ruiz's candidacy, but the "White versus White" Trustee and Mayoral races also provided important insights into the behavior of Port Chester voters.

While certain exogenous elections involving Hispanic candidates ordinarily would be of significant value — providing insight into how Hispanic voters in Port Chester behaved when presented with the option of a Hispanic candidate — the exogenous elections studied by the respective experts here do not allow for this type of understanding. There is insufficient evidence in the record for this Court to conclude confidently that Westchester voters viewed Castro and Judge Morales Horowitz as Hispanic candidates. Moreover, the inclusion of a Hispanic candidate in the 2002 Attorney General race does not make that contest probative for this Court, precisely because the Hispanic-preferred candidate was not the Hispanic candidate.

In sum, the evidence from all of the endogenous contests studied here was both more convincing and less fraught with factual disputes about, for example, whether candidates were or were not perceived to be Hispanic and how that may or may not have factored into the electoral outcomes. Further, it is clear to this Court that in general, countywide and statewide elections interject a host of different influences and variables into the electoral analysis, not the least of which is that those elections are held "on cycle" in November. In addition, this Court opted not to consider the Government's evidence about the School Board elections in support of the third Gingles precondition; because those elections could not be analyzed statistically, and because fact witnesses offered conflicting, and unverifiable, opinions about which candidates had the support of which communities in those elections, the Court does not find that evidence to be probative on this question. For all of these reasons, the Court concentrated its analysis on the endogenous elections in the Village.

b. Methodologies

Dr. Weber again proposed an arbitrarily assigned percentage for measuring non-Hispanic bloc voting: according to him, 60 percent or more of non-Hispanics have to coalesce or vote for a particular candidate to constitute bloc voting. Hearing Tr. at 926-28. Not only could he point to no court in the United States that has accepted his cohesion requirement for non-minority bloc voting, but he also admits that he knows of no other expert in the field who has adopted or agreed with his non-minority cohesion requirement. Hearing Tr. at 1005. The Court declines to be the first court to endorse a cut-off without any scientific or statistical basis other than it is "simply a number at which [Dr. Weber] feel[s] comfortable." Hearing Tr. at 1001.

Dr. Handley put forth a different analysis of the requirements of Gingles in this regard. She described a functional test that examines whether "the Whites [are] voting for the other candidates to such a degree that the Hispanic preferred candidate is losing." Hearing Tr. 467. It is not necessarily important that the non-Hispanic voters coalesce behind a particular candidate or that a particular percentage of non-Hispanic voters vote for any one candidate — what matters most is that those voters do not cast votes for the Hispanic candidate of choice, and those votes usually result in the defeat of the minority-preferred candidates. Even Dr. Weber agrees with this statement as long as there is reliable data to review. Hearing Tr. at 1004. This Court believes that a more flexible, functional test, like that proposed by Dr. Handley, is appropriate when considering whether there has been non-minority bloc voting.

c. Electoral outcomes

Dr. Handley testified that non-Hispanic voters voted as a bloc to defeat Hispanic candidates of choice in 12 of the 16 endogenous elections she reviewed — a total of 75 percent of the time. Trial Tr. at 13. In all 10 Trustee contests between 2001 and 2005, the Hispanic candidates of choice were different from the non-Hispanic candidates of choice; the candidates of choice of the non-Hispanic voters won 9 of those 10 elections. Gov. Ex. 13 at 1-2. The only year in which a Hispanic candidate of choice ran for a position as Trustee was 2001, the year in which Ruiz was a candidate; notably, Ruiz was the Hispanic community's top candidate of choice in that election according to Dr. Handley's data, but he still was defeated. That election also provided an illustration of Dr. Handley's functional approach to this issue; non-Hispanic votes were not concentrated on any one candidate, but rather were well distributed among the three White candidates who finished ahead of Ruiz in the balloting. See Gov. Ex. 13 at 1. The same pattern held true for the Mayoral elections in Port Chester in 2001, 2003, and 2005. In each of these endogenous contests, the Hispanic candidate of choice differed from the non-Hispanic candidate of choice; in each, the Hispanic candidate of choice was defeated.

In 2006, the only year in the past six in which Hispanic candidates of choice were elected in both Trustee races, it is no coincidence that those candidates were also the candidates of choice of non-Hispanic voters. Similarly, in the 2007 Mayoral election, the Hispanic candidate of choice emerged victorious because non-Hispanic voters did not vote cohesively, and instead split their votes between the two candidates. See Gov. Ex. 46. Of course, the 2007 Mayoral election must be viewed somewhat differently; the election took place just weeks after this Court enjoined the Trustee election that was supposed to take place simultaneously, and the voting rights issues raised by this lawsuit played a prominent role in the campaign itself. See Gov. Exs. 70, 71, 77.

5. Senate Report factors — totality of the circumstances

a. History of official discrimination

Professor Smith testified concerning the history of discrimination against Hispanics in New York State, Westchester County, and Port Chester. Some of that testimony was related to historical events that occurred 30 or 40 years ago — including evidence of New York State's literacy test for voting (abolished in 1966), see Hearing Tr. at 373-74, and a New York City lawsuit from the early 1970s concerning Spanish language assistance at polling places. See Hearing Tr. at 374. Of greater probative value to this Court were some of the more recent examples from Westchester County — including the 1985 Yonkers housing and education discrimination case, see Hearing Tr. at 377, a 2006 State Senate race in Yonkers, see Gov. Ex. 22 at ¶ 22, and the 2005 Consent Decree between the United States and Westchester County pertaining to language assistance at polling sites in the County, see Hearing Tr. at 378-87. Based on these data points and others, Professor Smith offered the general conclusion that there has been discrimination against Latinos in Westchester County.See Hearing Tr. at 376.

Professor Smith analyzed the Consent Decree entered in United States v. Westchester County, 05 Civ. 0650 (CM), as a guidepost for what type of language assistance would be required at polling sites in Port Chester.¹⁹ See Hearing Tr. at 386-87. Based on the county-wide standard established in the Consent Decree — requiring at least one bilingual poll worker²⁰ at each polling site located in an election precinct containing between 100 and 249 Spanish surnamed voters — Professor Smith concluded that Port Chester failed to provide sufficient Spanish language assistance at polling sites in the Village for the Trustee elections held between 2001 and 2006. See Hearing Tr. at 387-90.

19 In the Defendant's Motion to Reconsider, Port Chester brings to the Court's attention that the Consent Decree expired in December 2008 and the Department of Justice did not seek to renew it. As the Court found in its decision to deny the Motion to Reconsider, the expiration of the Consent Decree has no bearing on the Court's finding of a history of discrimination that led to the Consent Decree.

20 The terms "poll worker" and "election inspector" are used interchangeably here, as they were during the Hearing and Trial.

The Port Chester Village Clerk's Office is responsible for conducting Village elections; those responsibilities include posting public notices, creating the official ballots, and assigning four election inspectors (two Democrats, two Republicans) to each of the 16 election precincts in the Village.See Trial Tr. at 79. Joan Marino ("Marino"), Port Chester's Deputy Village Clerk, testified that since she began working there in 1997, it has not been the specific practice of the Village Clerk's office to appoint Spanish-speaking poll workers for election precincts that contain a large number of Spanish speaking voters. See Trial Tr. at 80. Joanne Villanova, who is not a Spanish speaker, testified that she served as an election inspector for Village elections approximately 15 times over a period of approximately 25 years, and only once worked alongside a Spanish-speaking inspector. See Trial Tr. at 924-27.

From 2001-2004, the Town of Rye provided the Village with lists of qualified election inspectors for each major party; these lists, however, did not indicate which inspectors, if any, spoke Spanish, and the Village Clerk's Office made no independent effort to determine which inspectors spoke Spanish. See Trial Tr. at 80-86; Gov. Exs. 49-52. In 2005 and 2006, the lists of qualified election inspectors provided to the Village did denote which poll workers were Spanish speakers — a total of four eligible poll workers were indicated to be Spanish speakers in 2005, and six were listed as Spanish speakers in 2006. See Trial Tr. at 87, 91 and Gov. Ex. 53 (2005 election); Trial Tr. at 92-95 and Gov. Ex. 54 (2006 election). During the 2005 election, only two of the four eligible Spanish-speaking inspectors actually worked at polling places; and only three of the six eligible inspectors worked at polling places during the 2006 election. Trial Tr. at 92, 96. For the 2007 Mayoral election, the Village Clerk's Office obtained a list of bilingual inspectors from the Westchester County Board of Elections for the first time, see Gov. Ex. 56, and assigned 14 Spanish-speaking poll workers to various election precincts, a total higher than in any of Marino's previous years working for the Village. See Trial Tr. at 98-100.

Plaintiffs offered testimony from one Spanish-speaking poll worker — Luz Marina Chavista — who described several situations from her experiences as a poll worker in Port Chester where she observed Hispanic voters being treated differently from White voters. See, e.g., Trial Tr. at 844, 852. In addition, Richard Falanka, the former Village Clerk and Village Manager of Port Chester, testified that there are no Spanish-speaking employees at the Village Hall who would be able to take a complaint from a Spanish-speaking voter at the beginning of the polling day (from 7:00 a.m. until 9:00 a.m.) or at the end of the polling day (from 4:30 p.m. until 9:00 p.m.), though English-speaking employees are available to receive complaints during those times. See Hearing Tr. at 1272-73.

Plaintiffs offered other examples of official discrimination against Hispanics that occurred in Port Chester itself. During his Hearing testimony, Nelson Rodriguez described the events surrounding his 1991 campaign for a seat on the School Board. According to Rodriguez, more than 40 Hispanic voters were turned away from the polls during that election because of poll workers' inability to locate their names on voter lists. See Hearing Tr. at 295-301; Gov. Ex. 9 (affidavits of voters from 1991 School Board election). Rodriguez lost that election by 37 votes, Gov. Ex. 10 at 1, and subsequently challenged the election results by filing an appeal with the New York State Education Department. See Appeal of Nelson Rodriguez, Dec. No. 12,704 (May 26, 1992) (available at http://www.counsel.nysed.gov/Decisions/volume31/d12704.htm). The School Commissioner sustained the appeal and ordered a new election, finding that Rodriguez "amply demonstrated that there were irregularities in the conduct of respondent board's election," and that the "board's failure to locate approximately 39 names is unacceptable." Id. In 1992, the School Board scheduled a special "re-vote" election; Rodriguez ran again, and was defeated by 374 votes. See Gov. Ex. 10 at 2.

Finally, the Government offered in evidence audio-visual recordings of two public hearings held in Port Chester in 2006 regarding the Government's proposed districting plans. Gov. Exs. 101 and 102. The hearings mostly consisted of statements by various citizens of the Village either in support of or in opposition to the districting proposals, and therefore could not be viewed as evidence of official discrimination in the Village. However, there was a noteworthy comment at the first public hearing by Aldo Vitagliano, an attorney who would later be appointed by the Village to serve as special counsel to the newly-formed Voting Rights Commission created to study and evaluate the Government's districting proposals. He suggested that Port Chester's representatives in Congress should introduce an amendment to exempt the Village from the requirements of the Voting Rights Act.

b. Extent of racially polarized voting

Dr. Handley testified that "voting is polarized . . . if Hispanics would have elected a different candidate or set of candidates than Whites, [and] it rises to the level of legal significance if, under these circumstances, the Hispanic preferred candidate usually loses." Hearing Tr. at 467. According to Dr. Handley's analysis, 13 of the 16 endogenous elections were polarized.²¹ Trial Tr. at 13.

21 In the data presented by Dr. Handley in support of her conclusions of polarization, African-American and other minority voters are grouped in with the "non-Hispanic" voting bloc. Thus, to the extent, if any, that African-American voters actually tend to vote in a manner more comparable to the Hispanic citizens of Port Chester than the non-Hispanic White citizens of Port Chester, the polarization data actually understates the separation between the Hispanic and non-Hispanic White communities in the Village.

In many cases, the degree of polarization was significant; in the single-vote elections for Mayor, Hispanic preferred candidates received between 69.6 percent and 96.2 percent of the Hispanic vote in these two-candidate elections, according to Dr. Handley's bivariate ecological regression estimates.²² Gov. Ex. 13 at 2; Gov. Ex. 46 at 1. It therefore follows that the candidates of choice of non-Hispanic voters received little support from Hispanic voters in these elections. In the "vote for two" Trustee elections — where 50 percent support would be the maximum achievable threshold absent "single shot" voting — 10 of 12 Hispanic-preferred candidates received more than 40 percent of Hispanic voter support according to Dr. Handley's bivariate ecological regression estimates.²³ Gov. Ex. 13 at 1-2. Again it follows that the non-Hispanic candidates of choice received little support from Hispanic voters, with percentages often in single digits according to these estimates.

22 Dr. Handley arrived at substantially similar results using ecological inference methodology for the 2001, 2003, and 2007 elections, but her ecological inference estimates for the 2005 Mayoral race ascribed only 52.6 percent of the Hispanic vote to the Hispanic-preferred candidate, as opposed to 96.2 percent support under the bivariate ecological regression method.

23 Nine candidates reached this threshold based on Dr. Handley's ecological inference methodology.

For the same reasons discussed above in connection with the third Gingles factor, this Court does not consider evidence presented about the School Board elections as part of this determination.

c. Electoral practices that enhance opportunities for discrimination

There is no dispute that Port Chester holds its Trustee elections in March, and it is also evident that the six Trustees are elected to staggered three-year terms. See Gov. Ex. 4. Experts on both sides agree that generally, voter turnout is lower in March elections than in November elections, and that this general principle is true of the off-cycle elections in Port Chester for both the Hispanic and the non-Hispanic populations of the Village. Dr. Handley's data demonstrates that Hispanic turnout was markedly greater for November elections in 2004, 2005, and 2006 than it was for March elections those same years.See Gov. Ex. 13 at 6-7; see also Hearing Tr. at 1377 (Dr. Morrison noted that "the participation levels of Hispanics turning out to vote varies widely from a March Trustee election to a November general election"). Meanwhile, Dr. Gaddie agreed that holding local elections off cycle generally results in depressed voter participation, and observed that "every March election has lower turnout than the November elections." Hearing Tr. at 1343-44.

d. Access to the candidate slating process

Candidates for political office in Port Chester are selected through a caucus system organized and administered by the political parties in the Village. At each caucus, a majority of caucus attendees must vote in favor of a particular candidate for that candidate to formally receive the party's nomination for Trustee. Prior to the official party caucus, however, the major political parties invite prospective candidates to interview before the parties' respective nominating committees, which then select their two preferred individuals and forward those names to the parties' caucuses for ratification. Hearing Tr. at 823-25 (describing the Republican Party process); Trial Tr. at 337 (describing the Democratic Party process).

In theory, a candidate who did not win approval from the nominating committee could "storm" the caucus by bringing enough supporters to challenge the nominating committee selections — the formal rule is that the individuals who receive the most support at the caucus become the nominees. For both parties, however, the nod from the nominating committee is the critical step to getting onto the March ballot — no witness could identify a single instance where the nominating committee's selections were defeated by a "storming" of the caucus. See,e.g., Trial Tr. at 163-64 (Pilla was not aware of any challenges at a caucus); 337 (former Village Democratic Party chairman testified that "the people nominated were always approved by the caucus"); Hearing Tr. at 828 (Village Republican Party chairman not aware of any candidate selected by nominating committee who did not become the nominee); 1166 (Rye Town Republican Chairman could not recall a contested caucus). Indeed, no witness could even identify a single bona fide attempt to storm the caucus.

In fact, Dr. Janusz Richards, the Chairman of Port Chester's Republican Committee, initially testified that he believed that candidates were required to interview with the nominating committee in order to receive the Republican nomination. Hearing Tr. at 841-42. Though he ultimately clarified this testimony to make clear that there was no party rule or regulation that required an appearance before the nominating committee, Hearing Tr. at 842, the fact that this political "insider" was not completely clear about the possibility of winning the nomination through the caucus alone makes this Court question whether the "storming" option is known to exist among the general population, much less the Hispanic community. It is worth noting that, as explained further in section III.E.7 below, even when the parties purported to have made outreach efforts to find Hispanic candidates, the evidence is clear that only two Hispanics made it through the nominating committee process and onto the ballot for Port Chester Trustee between 1992 and 2006.

Again, this system greatly favors those with existing political ties or other institutional support. Members of the Hispanic community have few positions of leadership within the major political parties in Port Chester, even at the entry-level district leader position that can often be a steppingstone to public office. See Hearing Tr. at 172. With the exception of a brief "renegade" effort led by Ruiz to seat Hispanic-preferred district leaders in the Democratic Party in 2004, very few Hispanics have served at even this entry level leadership position in either party. See Hearing Tr. at 56-57; 67-69 (Democratic Party); 832, 1169 (Republican Party).

e. Discrimination in other areas that hinders the ability of Hispanics to participate effectively in the political process

Professor Smith and Dr. Morrison acknowledged that Hispanics in Port Chester have lower levels of educational attainment on average and lower incomes on average than non-Hispanics. See Def. Ex. LL at 13 (Morrison Report); Gov. Ex. 22 at ¶ 24 (Smith Declaration). In addition, Hispanics in Port Chester were more likely than non-Hispanics to live in overcrowded housing, to rent their homes, and to have lived in their homes for less than five years. See Def. Ex. LL at 13; Gov. Ex. 22 at ¶ 25. According to Professor Smith, educational disparities in Port Chester are "stark" — approximately 55 percent of Hispanic men and 48 percent of Hispanic women aged 25 or older had attained less than a high school education, while only 14 percent of White men and 16 percent of White women were limited to this level of education. Gov. Ex. 22 at ¶ 25. A total of 17 percent of Port Chester Hispanics lived below the poverty line in 1999, while the same was true of only 0.6 percent of the White population of the Village. Id. at ¶ 24. These economic disparities persist despite the fact that Hispanics have "comparable or higher rates of participation in the labor force compared to other groups." Id. at ¶ 25. Figures from the 2000 Census reveal that 75 percent of Hispanic men participate in the labor force, as compared with 71 percent of White men; meanwhile, 53 percent of Hispanic women participate in the labor force, as compared with 56 percent of White women. Id. In sum, it is clear that Hispanics and Whites in Port Chester "have significant differences in socioeconomic status." Hearing Tr. at 399.

Professor Smith testified that "lower socioeconomic status leads to lower levels of political participation." Hearing Tr. at 399 (Smith). Though he agreed that the fact that the Hispanic community is on average younger and more recently arrived in the United States than the non-Hispanic citizens of Port Chester could contribute to lower Hispanic voter turnout²⁴, Hearing Tr. at 445-48, there was no testimony to suggest that the presence of these factors negates the effects of socioeconomic status. Meanwhile, Dr. Gaddie agreed that socioeconomic status is the foundational influence on political participation. Hearing Tr. at 1342. Further, Dr. Gaddie noted that empirical studies have repeatedly shown that individuals who score lower on socioeconomic status criteria are less prone to participate in politics. Hearing Tr. at 1342.

24 To explain lower rates of Hispanic political participation in Port Chester, Dr. Morrison attempted to place great emphasis on the fact that the Hispanic community in the Village contains many recent arrivals and is generally a more transient population. See Trial Tr. at 427. Dr. Morrison, however, has not conducted any particular studies to determine how transient the Hispanic community actually is in Port Chester, let alone how this would impact the rate of political participation of the Hispanic population. See Trial Tr. at 409-11.

Dr. Gaddie also testified, however, that socioeconomic status factors such as age, wealth, education and literacy alone are not enough to predict rates of political participation. Hearing Tr. at 1292-94. He offered the proposition that in addition to the socioeconomic status factors that contribute to one's "civics skills set," it is important to take into account mobilization efforts. Hearing Tr. at 1294. Political mobilization, he suggested, is not determined by socioeconomic status, but rather by the degree of in-person campaigning and other get-out-the-vote efforts in communities of lower socioeconomic status. See Hearing Tr. at 1302-06. Even Dr. Handley acknowledges that factors other than socioeconomic status must contribute to our understanding of participation rates in Port Chester, given that participation fluctuates greatly between March and November elections, even within the same calendar year. See Gov. Ex. 13 at 7.

f. Racial appeals in political campaigns

None of the evidence offered by the Government at the Hearing phase of these proceedings provided a clear indication that political campaigns in Port Chester have been marred by racial appeals. The Plaintiff attempted to demonstrate through the testimony of Dr. Maria Munoz Kantha ("Dr. Kantha") that the 2005 contest for Westchester County District Attorney between Janet DiFiore and Anthony Castro was characterized by subtle racial appeals in the form of a campaign flyer. See Gov. Ex. 11 (original campaign flyer); Hearing Tr. at 1218-24. Dr. Kantha's testimony made clear that there were some in the Hispanic community who viewed Gov. Ex. 11 as a racial appeal — as illustrated by Gov. Ex. 103, a number of individuals called a press conference to voice their displeasure with the flyer. See Hearing Tr. at 1124-28. This Court viewed Gov. Ex. 11 as nothing more than a piece of partisan political propaganda in the midst of a hard-fought campaign. The Court also heard testimony at the Hearing from Mr. John Reavis and Mrs. Doris J. Bailey-Reavis about racial epithets that were spoken or written at two points during Mr. Reavis's 1996 campaigning for a seat on the School Board. See Hearing Tr. at 333-34 (Mr. Reavis); 353-54 (Mrs. Bailey-Reavis).

At the trial, however, the Court received extensive testimony about a flyer — admitted in evidence as Gov. Ex. 63 — that was used as part of the 2007 Mayoral election in the Village. Without question, this flyer must be considered a racial appeal. Bart Didden ("Didden"), who was slated to be a Republican candidate for Trustee in the March 2007 election before the election was enjoined, developed a first draft of this flyer approximately two or three weeks before the March 20 election²⁵; thus, the flyer was first created after the conclusion of the hearing phase of this proceeding, and perhaps even after this Court issued the preliminary injunction on March 2, 2007. See Trial Tr. at 247-49.

25 According to Didden, two other individuals collaborated on the final version of the flyer: John Crane, a current member of the Board of Trustees, and Dominic Bencivenga, who was then a member of the School Board. Both Crane and Bencivenga testified that they were only minimally involved in the creation of the flyer. Bencivenga stated that he did not have anything to do with the drafting of the flyer, though he did make certain comments and recommendations about the content of the flyer when he viewed it in draft form. Trial Tr. at 360, 364-66. Crane, at the very least, saw the flyers before they were mailed. Trial Tr. at 364-67. Neither Crane nor Bencivenga did anything to stop Didden from sending out the flyer, even though they both recognized that the document was "troubling" and/or "racist, sexist and disgusting."

Didden called the flyer a "hard-hitting, issues oriented" piece that was designed to convince voters not to vote for Pilla, who was at the time the Democratic candidate for Mayor of Port Chester. See Trial Tr. at 278. The flyer also includes personal attacks on two Hispanic leaders in the Village — Ruiz, and Blanca Lopez, who was Pilla's campaign manager. Ruiz is described in the flyer as a "hot dog vendor-turned-professional-consultant Ceaser (sic) Ruiz," while Lopez's name appears many times in many different contexts. For example, the flyer states that "what Blanca cares about is only Hispanic," and accuses Lopez of being "not only a double agent" but "a super secret triple agent," apparently because Lopez submitted a declaration in support of the Government's motion for a preliminary injunction, and therefore "testified" "against Port Chester."

Pilla is attacked because of his apparent support of issues of importance to Hispanics in the Village. The flyer declares that Lopez is pushing for more affordable housing, more subsidized housing, and more Section 8 housing, and warns that "she is going to get if (sic) because Lopez and Pilla are in bed together on the Village affordable housing sub committee, the wolf is in the house thanks to Pilla!" Further, the flyer proclaims that "Blanca say's (sic) jump, fetch, beg or bark and Pilla does it. The Hispanics are running the show already."

The flyer also criticizes Pilla for his position with respect to this lawsuit, suggesting that Pilla changed his views on the lawsuit at the behest of Lopez; specifically, the flyer claims "flip flop Pilla sells out on command of campaign manager/Hispanic leader Blanca Lopez." Pilla allegedly "abandon's (sic) the Village by reneging on his commitment to fight splitting the Village up into districts and pitting neighbor against neighbor," and "is selling you and me out to the Department of Justice." Language in the flyer also mischaracterizes the lawsuit as an attempt to portray the residents of the Village as racists, urging recipients not to "elect carpet baggers (sic)" but rather to "elect people who care about our history, heritage and what our kids will be told about us in the future, are we to be known as racists or law abiding free Americans."

According to Didden he mailed the flyer to approximately 1,000 households in the Village, asserting that he did so because of his "civic responsibility to the community that I live in." Trial Tr. at 278-79. Didden, however, did not sign the flyer or otherwise indicate that he was its primary author, and when he mailed the material he did so at a mailbox in Greenwich, Connecticut, because he "did not want to be observed in front of the Port Chester post office with a thousand envelopes putting them in the post box because that could lead to someone suspecting that I had something to do with the mailing." Trial Tr. at 280-81.

Various Village officials testified that they believed the flyer was a racist document. In response to a question from the Court, Domenick Cicatelli, currently a Trustee and, in March 2007, the Republican candidate for Mayor, testified that the flyer appears to be a racial appeal, and called the flyer "troubling." Trial Tr. at 761. Trustee Crane testified that the flyer was "racist, sexist, disgusting," and "highly inflammatory." Trial Tr. at 680. Bencivenga, in response to a question from the Court, indicated that he believed the flyer could fairly be characterized as racist. Trial Tr. at 384.

g. Election of Hispanics to public office in the jurisdiction

At the time of the liability phase of this case, no Hispanic candidate had ever been elected to public office in Port Chester — not Mayor, not to the Board of Trustees, and not to the School Board. On May 19, 2009, Blanca Lopez was elected to the Port Chester School Board, making her the first Hispanic to be elected to a jurisdiction sharing most of its precincts with the Village of Port Chester. This election was not included in the analysis because the parties' experts determined that it did not have sufficient probative value. See, supra II.B.4.a.

While the lack of statistical data from the School Board elections made it difficult to consider the results of those elections as part of the racial polarization analysis, this Court does consider the outcomes of those elections as additional evidence of this Senate factor. Ms. Lopez's victory is set against a bleak backdrop: between 1991 and 2006, three Hispanic candidates ran for the School Board a total of four times²⁶, and all were defeated. See Gov. Ex. 10.²⁷ Indeed, before Ms. Lopez's election only one member of the Hispanic community has ever been elected to any federal, state, county, or local office for any jurisdiction in which Port Chester is located — Judge Morales Horowitz, who, as discussed above, was elected Family Court Judge in 2000. See Hearing Tr. at 782.

26 Nelson Rodriguez ran in 1991 and ran again in 1992 as part of the state-mandated re-vote for the 1991 election. Rodriguez's 1988 School Board candidacy was also unsuccessful. Hearing Tr. at 278

27 In addition, though this lawsuit does not name African-Americans as a minority group experiencing a violation of the Voting Rights Act in Port Chester, it is worth noting that no African-American has ever been elected Mayor or to the Board of Trustees, and no African-American has ever been elected to the School Board, despite the fact that African-American candidates ran for the School Board eight times between 1991 and 2006.

In all of the Trustee elections studied by both sides in this case up to and including 2007, only two Hispanics have ever been on the ballot — Jose Santos ran as a Republican in 1992 and Cesar Ruiz ran as a Democrat in 2001 — and both finished last in their respective fields. See Gov. Ex. 4 at 13 (Santos results) and 22 (Ruiz results). There is no indication from the evidence in this case that a Hispanic candidate has ever run for Mayor in Port Chester.

The Village attempted to elicit testimony concerning various theories for why no Hispanics have been elected, both through expert witnesses and from Village residents who offered their views as to which Hispanic candidates actually garnered the support of the Hispanic community and why. This Court has already addressed the testimony from the various experts, and further concludes that the speculative testimony from the Village's other witnesses regarding the preferences of Hispanic voters is of no particular relevance to the issues presented in this case. Defendant also endeavored to show that both major political parties in Port Chester made concerted efforts to encourage Hispanic candidates to run for office. See,e.g., Hearing Tr. at 811-14 (Republican Party efforts); 1088 (Democratic Party efforts). It was clear that at least some of the recruiting was conducted at least in part in response to the Justice Department's investigation here; more importantly, few Hispanic candidates ultimately were put forward by the parties, despite these purported outreach strategies.

h. Additional factors in the Senate Report

i. Lack of responsiveness to the particularized needs of Hispanics

Plaintiffs have not attempted to make an issue of Port Chester's lack of responsiveness to the particularized needs of members of the Hispanic community.

ii. Tenuousness of the challenged voting practice or procedure

Plaintiffs put forward no evidence to suggest that the policy rationales underlying Port Chester's voting system are tenuous. Port Chester has had an at-large system of elections in place since 1868, more than a century before the Hispanic population became a plurality. The Village has offered evidence that it holds local elections in March to insulate them from the vagaries of the national election cycle and, in part, to bring the Village in line with other New York State localities.

C. Conclusions of Law

1. First Gingles precondition: the minority group must be sufficiently large and geographically compact to constitute a majority in a single-member district

To satisfy the first Gingles precondition, Plaintiffs must prove that the minority group "is sufficiently large and geographically compact to constitute a majority in a single-member district." Gingles, 478 U.S. at 50. In sum, unless minority voters possess the potential to elect representatives in the absence of an at-large voting system, they cannot claim that their voting rights have been implicated by that system. See id. at 50, n. 17. This requirement is designed to ensure that the minority population in the subject area will have a real opportunity to elect candidates of its choice.

As a threshold matter, although there was some anecdotal evidence presented that the 2000 Census might not have placed every voter in the exact block of their residence, this Court recognizes that Census data is presumptively accurate. See Valdespino v. Alamo Heights Indep. Sch. Dist., 168 F.3d 848, 853-54 (5th Cir. 1999); Johnson v. DeSoto County Bd. of Comm'rs, 204 F.3d 1335, 1341 (11th Cir. 2000) ("the presumption is that Census figures are continually accurate"). The Valdespino court determined that "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." Valdespino, 168 F.3d at 854. Defendants have not come close to meeting that burden here through Cleary's testimony, and this Court accepts the 2000 Census data as reliably accurate, though not perfect, in this case.

First, the size and shape of the illustrative districts contained in Plaintiffs' Plan A and Modified Plan A comport with traditional districting principles of population equality and compactness. See Shaw v. Reno, 509 U.S. 630, 651 (1993) (quotingUnited Jewish Orgs. v. Carey, 430 U.S. 144, 168 (1977)). In addition, while respect for existing political boundaries is also a valued traditional districting method, see Miller v. Johnson, 515 U.S. 900, 919 (1995), election precincts are not such important political boundaries that they should negate a districting proposal, particularly where, as here, other key districting principles are obeyed. The Court finds no fault with Dr. Beveridge's decision to disregard the precinct boundaries when drawing the proposed districts. Finally, based on the testimony of the parties' respective experts, this Court is firmly convinced that race was not the "predominant, overriding factor explaining" Dr. Beveridge's Modified Plan A. See id, 515 U.S. at 920.

While traditional districting principles typically require the use of total population in drawing district boundaries, in determining whether the minority group at issue has a sufficient majority in an illustrative district to satisfy the first Gingles precondition, courts look to the VAP, and in particular to the CVAP, as the relevant population in the district. See, e.g.,Rodriguez v. Pataki, 308 F. Supp. 2d 346, 378 n. 38 (S.D.N.Y. 2004) (three-judge panel) (citing Valdespino, 168 F.3d at 851-53;Negron v. City of Miami Beach, 113 F.3d 1563, 1569 (11th Cir. 1997); and France v. Pataki, 71 F. Supp. 2d 317, 326 (S.D.N.Y. 1999)).

Plaintiffs have proven that Hispanics comprise 56.27 percent of the CVAP in proposed District 4 under Modified Plan A, which clearly represents a majority of CVAP in that area. See Goosby I, 956 F. Supp. at 348 (finding that Plaintiffs satisfied the firstGingles precondition with a proposed district where African-Americans comprised 52.57 percent of the VAP in the district). Though the Supreme Court has held in dicta that it is theoretically possible for a minority group to lack "real electoral opportunity" in a district even if that group constitutes a majority of CVAP in that district, see League of United Latin Am. Citizens (LULAC) v. Perry, 548 U.S. 399, 428 (2006), the weight of authority indicates that a CVAP majority will typically constitute an effective majority for the purposes of the first Gingles precondition. Indeed, even the Perry Court indicated that a 57.5 percent CVAP majority did possess electoral opportunity protected by Section 2 of the Voting Rights Act. Id.

Though the Village argued that the proportion of Hispanic registrants and the turnout rate among Hispanics in proposed District 4 are such that the Hispanic population would not constitute an effective voting majority, there are significant shortcomings in both lines of reasoning. As to the turnout issue, we agree with the view expressed by the Ninth Circuit about the proper consideration of voter turnout in a Section 2 analysis: "if low voter turnout could defeat a Section 2 claim, excluded minority voters would find themselves in a vicious cycle: their exclusion from the political process would increase apathy, which in turn would undermine their ability to bring a legal challenge to the discriminatory practices, which would perpetuate low voter turnout, and so on." United States v. Blaine County, 363 F.3d 897, 911 (9th Cir. 2004). A similar logic applies to voter registration — if, as expected, the elimination of a Section 2 violation will increase opportunities for Hispanics in Port Chester to participate in the political process of the Village, it seems likely that such participation will extend down to the simplest level of participation: registering to vote.

Thus, as to the first Gingles precondition, this Court finds that Plaintiffs have demonstrated adequately through Modified Plan A that Hispanics in Port Chester are sufficiently large in number and geographically compact to constitute an effective majority in a single-member district in the Village. Accordingly, Plaintiffs have satisfied the first Gingles factor.

2. Second Gingles precondition: the minority group must be politically cohesive and vote as a bloc

The second Gingles precondition requires Plaintiffs to demonstrate that Hispanics in Port Chester are politically cohesive. See Gingles, 478 U.S. at 51. According to the Gingles Court, "if the minority group is not politically cohesive, it cannot be said that the selection of a multimember electoral structure thwarts distinctive minority group interests." Id. Plaintiffs proved during these proceedings that Hispanic voters in Port Chester voted cohesively in all 16 election contests in the Village between 2001 and 2007. The methods employed by Dr. Handley to reach these conclusions have been accepted by numerous courts in voting rights cases. See Gingles, 478 U.S. at 52-53 (accepting bivariate ecological regression analysis); Rodriguez, 308 F. Supp. 2d at 388 (accepting ecological inference methodology).

Further, this Court declines to adopt Dr. Weber's position that the Hispanic community in Port Chester cannot be considered cohesive unless 10 percent of the Hispanic CVAP votes in a given election. This Court does not believe that there should be any arbitrarily fixed percentage for CVAP participation in order to find cohesion; such a bright-line threshold for minority CVAP turnout is not helpful or appropriate here. See Blaine County , 363 F.3d at (9th Cir. 2004); Uno v. City of Holyoke, 72 F.3d 973, 987 (1st Cir. 1995). Dr. Weber conceded that if this Court were to reject his turnout requirements, Hispanics in Port Chester were cohesive in 13 of 15 elections between 2001 and 2006. Accordingly, this Court concludes that Plaintiffs have proven that Hispanics in Port Chester vote cohesively, and therefore that Plaintiffs have fulfilled the second Gingles precondition.

3. Third Gingles precondition: the White majority must vote sufficiently as a bloc to enable it, in the absence of special circumstances, to defeat the minority's preferred candidate

To satisfy the third Gingles precondition, Plaintiffs must demonstrate 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. Proving this third point enables the minority group to show that "submergence in a White multimember district impedes its ability to elect its chosen representatives." Id. Further, the requirement that the White majority be repeatedly successful "distinguishes structural dilution from the mere loss of an occasional election." Id.

In assessing which elections should be afforded the greatest probative value, the Court was guided by two Second Circuit pronouncements on this question. First, it is clear that, in this Circuit, at least, district courts must consider "White versus White" elections as part of a Section 2 analysis. Niagara Falls, 65 F.3d at 1015-17. In addition, "exogenous elections — those not involving the particular office at issue — are less probative than elections involving the specific office that is the subject of the litigation." Goosby III, 180 F.3d at 497 (quoting Clark v. Calhoun County, Miss., 88 F.3d 1393, 1397 (5th Cir. 1996)); see Niagara Falls, 65 F.3d at 1015 n. 16. In light of these decisions, we believe that the proper focus in this case was on endogenous elections in the Village, even though all but one of these elections were "White versus White" contests.

Defendant argues that any elections that occurred after spring 2001 cannot form the basis of a Section 2 violation here, because "if a Section 2 violation exists it must be one that could have been reasonably evident to the Village contemporaneous with the release of the decennial census data. . . . The law cannot require the Village to create districts because sometime in the future, (but before the next census) population changes might lead to an allegation that its at-large system dilutes minority voting rights." Def. Post-Trial Mem. of L. at 13. The Village does not cite any case law in support of the proposition that this Court should not consider the most recent elections in the jurisdiction as part of its Section 2 analysis. A brief look atGoosby I reveals that other district courts in this Circuit have not adhered to Defendant's position. The original Complaint in Goosby I was filed in 1988, and a bench trial was held in July 1996. See Goosby I, 956 F. Supp. at 329. By Defendant's logic, the Goosby I court should not have considered any Town of Hempstead elections that took place after spring 1991, yet the record is replete with evidence from two elections held in 1993. See, e.g., id. at 334. The argument that post-2001 elections should not be considered here is without merit.

Courts have employed methods that are very similar to Dr. Handley's functional approach to assess whether Whites vote as a bloc to defeat Hispanic-preferred candidates. See Gingles, 478 U.S. at 53. That is, the critical point is whether White voters are voting for other candidates to such a degree that Hispanic-preferred candidates are consistently defeated. See,e.g., Jenkins v. Red Clay Consol. Sch. Dist. Bd. of Educ., 4 F.3d 1103, 1123 (3d Cir. 1993) ("the correct question is . . . whether, as a practical matter, the usual result of the bloc voting that exists is the defeat of the minority-preferred candidate"). The motivations of the White voters under such a framework are irrelevant — indeed, Plaintiffs did not introduce evidence sufficient to prove that the non-Hispanic community in Port Chester voted the way it did because of any sort of racial bias. Contrary to Dr. Weber's view, however, the degree of White voter cohesion is also irrelevant, and the Court declines to adopt Dr. Weber's 60 percent cohesion requirement in this matter.

The evidence here is clear that in 12 of the 16 elections this Court views as most probative in this case, the candidates of choice of Hispanic voters in Port Chester were defeated by the candidates of choice of non-Hispanic voters. Defendant is correct to point out that three of the four elections in which Hispanic candidates were not defeated are among the most recent contests in the Village — the two Trustee races in 2006 and the 2007 Mayoral election. Nevertheless, it is well-settled that "in 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." Gingles, 478 U.S. at 57. Moreover, the March 2007 election — which took place in the after this Court issued the preliminary injunction, and in which this lawsuit became a central campaign issue — was characterized by the type of "special circumstances" that make the results of this election somewhat of an outlier in the overall analysis.²⁸ In addition, Defendant seeks to use data from elections held between 1995 and 2000 to rebut Dr. Handley's conclusions; however, this Court concurs with the views of the parties' experts that the data from those elections is unreliable, and therefore should not form the basis of any legal conclusions here.

28 Though the Second Circuit has not specifically defined the contours of the "special circumstances" doctrine, the Court thinks the Ninth Circuit's reasoning on this point is instructive. That court found that "to invoke the special circumstances doctrine regarding an election that occurred after a Section 2 lawsuit is filed, plaintiffs must show that a particular election was surrounded by unusual circumstances. Those unusual circumstances must demonstrate that the election was not representative of the typical way in which the electoral process functions. The focus is voter behavior, not voter motivation." Ruiz v. City of Santa Maria, 160 F.3d 543, 557 (9th Cir. 1998).

Defendant attempts to explain differences in voter behavior by claiming that partisan politics, and not racial polarization, is the cause of these electoral outcomes. The Second Circuit has counseled, however, that arguments concerning the causes of racially polarized outcomes are to be considered as part of the totality of the circumstances analysis, and not as part of theGingles determination. See Goosby III, 180 F.3d at 493 ("the best reading of the several opinions in Gingles, however, is one that treats causation as irrelevant in the inquiry into the threeGingles preconditions, see Gingles, 478 U.S. at 62 (Brennan, J., plurality op.) but relevant in the totality of circumstances inquiry").

In sum, it is clear to this Court that Hispanic voters and non-Hispanic voters in Port Chester prefer different candidates, and that non-Hispanic voters generally vote as a bloc to defeat Hispanic-preferred candidates. Accordingly, Plaintiffs have succeeded in proving the third Gingles precondition, and therefore have established all three of the required Gingles preconditions in this case. We now turn to the totality of the circumstances analysis.

4. Senate Report factors — totality of the circumstances

As outlined above, even though this Court has found that Plaintiffs have satisfied all three Gingles preconditions, it is also necessary to consider the totality of the circumstances before finding a Section 2 violation. See Johnson v. DeGrandy, 512 U.S. 997, 1011 (1994); Goosby III, 180 F.3d at 492. As discussed in further detail below, it is this Court's view that Plaintiffs have proved that all seven of the Senate factors are present in Port Chester; accordingly, the totality of the circumstances clearly indicates that Defendant's method of electing the members of its Board of Trustees violates Section 2 of the Voting Rights Act.

a. History of official discrimination

This Court is persuaded that there is some history of official discrimination in Port Chester that continues to touch the rights of Hispanics to participate in the political process. Other courts in this district have cited the "regrettable history of discrimination in employment, housing and education in the Westchester County area," New Rochelle Voter Defense Fund, 308 F. Supp. 2d at 159, as part of a Voting Rights Act analysis. While this Court finds the New York State and Westchester County examples of discrimination to be relevant, we were far more influenced by the examples from Port Chester itself, including the lack of Spanish-language voter assistance in the Village and the 1991 and 1992 School Board elections. At the very least, it is apparent that the Village Clerk's Office has failed to take proactive steps to address the needs of the Hispanic population in Port Chester, despite the rapid growth of the Hispanic community. While the evidence in support of this conclusion is not overwhelming, this Court does believe that, on balance, the first Senate factor supports a finding in favor of Plaintiff.

b. Extent of racially polarized voting

The Gingles Court cited this as one of the two most important Senate factors, see Gingles, 478 U.S. at 48 n. 15, and this Court considers it significant that this factor strongly bolsters Plaintiffs' position in this matter. According to Gingles, 478 U.S. at 53, "racial polarization exists where there is a consistent relationship between [the] race of the voter and the way in which the voter votes." As discussed above in connection with the second and third Gingles factors, the evidence presented in this case demonstrates that such a consistent relationship clearly exists in Port Chester: Hispanic voters vote cohesively, and the non-Hispanic community tends to vote as a bloc, generally resulting in the defeat of the Hispanic preferred candidates. Defendant argued at various points that the outcomes of the Village elections could be viewed as a factor of partisan political preferences rather than racial polarization of the electorate. That there is some correlation between political party and the voting preferences of Hispanics in Port Chester, however, does not contradict the conclusion that voting in the Village is polarized along racial lines. See Goosby I, 956 F. Supp. 355. Senate factor two clearly suggests that judgment for Plaintiffs is appropriate here.

c. Electoral practices that enhance opportunities for discrimination

Port Chester's practice of holding local elections "off-cycle" in March and staggering its Trustee elections combines to enhance the opportunity for discrimination against the Hispanic voting population. There is little question that the difference between holding an election "off-cycle" in March as opposed to holding it in November alongside major state and national elections can have a significant impact on voter behavior. See NAACP v. Hampton County Election Comm., 470 U.S. 166, 178 (1985) (noting that in the jurisdiction at issue, "an election in March is likely to draw significantly fewer voters than an election held simultaneously with a general election in November"). The lower turnout rates for March elections in Port Chester is at least partly the result of a structural flaw in the system, and is indicative of the Section 2 violation here; holding local elections at a time when only the most engaged and politically astute citizens — those citizens who feel the most enfranchised — are likely to vote will almost certainly result in the diminished influence of groups who feel generally excluded from the political fabric of the community.

The Supreme Court has recognized that staggered elections may enhance the discriminatory effect of certain voting systems. See,e.g., Lockhart v. United States, 460 U.S. 125, 143 (1983). Particularly given that many of Port Chester's Trustee elections have been close in terms of number of votes received, it is substantially less likely that White bloc voting could defeat all Hispanic-preferred candidates if all six trustees were chosen at one time. There has been no evidence to suggest that the Village adopted either of these practices with the intention of discriminating against Hispanic citizens, but as noted above, intent is not the touchstone of a Section 2 violation. What is important here is that off-cycle and staggered Trustee elections contribute to the Hispanic community's difficulty in electing its candidates of choice and "enhance the opportunity for discrimination" against Hispanics. Thus, this Senate factor points toward judgment for the Plaintiffs.

d. Access to the candidate slating process

While the candidate selection process of Port Chester's two major political parties formally allows for candidates to have open access to the ballot through the party caucus system, the reality of local politics in this community is that virtually binding decisions are made at closed meetings of the parties' respective nominating committees, which allow limited access to outsiders or upstart candidates.

The Second Circuit has held that a system that provides only a theoretical avenue for minority or other upstart candidates to get their names on the ballot while for all practical purposes making it extremely difficult for such candidates to have a meaningful opportunity to participate does in fact contribute to a violation of Section 2 of the Voting Rights Act. See Goosby III, 180 F.3d at 496 (describing Town of Hempstead process where Republican Committee members "theoretically are empowered to choose a slate of candidates for the Town Board, [but where] the actual selection process has been much different").

The candidate slating process employed by Port Chester's political parties to select their candidates for Trustee positions effectively limits access to those who are invited to interview before the parties' nominating committees, a situation which makes it all the more difficult for Hispanic citizens in the Village to elect their candidates of choice. Accordingly, this Court concludes that Senate factor four supports judgment for the Plaintiffs.

e. Discrimination in other areas that hinders the ability of Hispanics to participate effectively in the political process

While experts from both sides agreed that factors other than the socioeconomic disparities between Hispanics and non-Hispanics in Port Chester contribute to the differences in political participation rates in the Village, experts also agreed that there are substantial differences in education and income between Hispanics and non-Hispanics.

The Supreme Court has recognized that "political participation by minorities tends to be depressed where minority group members suffer effects of prior discrimination such as inferior education, poor employment opportunities, and low incomes."Gingles, 478 U.S. at 69. The Senate Report itself specifies that: "the courts have recognized that disproportionate educational, employment, income level and living conditions arising from past discrimination tend to depress minority political participation. Where these conditions are shown, and where the level of [minority] participation in politics is depressed, plaintiffs need not prove any further causal nexus between their disparate socioeconomic status and the depressed level of political participation. S. Rep. No. 97-417 at 29 n. 114. Various circuit courts, including the Second Circuit, have followed this line of reasoning, finding that plaintiffs are not required to prove a causal connection between socioeconomic factors and depressed political participation. See Niagara Falls, 65 F.3d at 1021 ("according to the Senate Report, voting rights plaintiffs need not establish [a causal] nexus where both disparate socioeconomic conditions and depressed political participation are shown to exist"); Teague v. Attala County, 92 F.3d 283, 294 (5th Cir. 1996); United States v. Marengo County Comm., 731 F.2d 1546, 1569 (11th Cir. 1984). Instead, the burden falls to Defendant to show that the cause is something else. Marengo County Comm., 731 F.2d at 1569.

Port Chester has not offered any persuasive evidence to suggest that socioeconomic factors do not contribute to differing rates of political participation of Hispanics in Port Chester. For this Senate factor to support a Section 2 violation, it is not necessary to find that socioeconomic status alone led to disparate levels of political participation; indeed, such a finding would be nearly impossible given the number of factors that contribute to any one individual's decisions about political participation. That said, it is the view of this Court that the effects of Hispanics' socioeconomic status, when combined with the structure of elections in Port Chester, limit the opportunities of the Hispanic community to participate in the political process and to elect candidates of choice. Senate factor five, therefore, also supports a finding of a Section 2 violation here.

f. Racial appeals in political campaigns

There can be no question that the most recent election for Mayor of Port Chester was marred by a racial appeal. Though Plaintiffs did not present any compelling evidence of racial appeals prior to 2007, the fact that such a blatant racial message — one which several witnesses conceded was racist — emerged in the midst of the ongoing proceedings in this case is troubling to this Court. The district court in Goosby I found that far more subtle racial appeals than this one contributed to the Section 2 violation in the Town of Hempstead. See Goosby I, 956 F. Supp. at 343, 353. Thus, in light of the evidence presented at trial, it is clear that this Senate factor weighs in favor of a ruling for Plaintiffs.

g. Election of Hispanics to public office in the jurisdiction

The Gingles Court also cited this as one of the two most important Senate factors. See Gingles, 478 U.S. at 48 n. 15. The evidence here is undisputed that no member of the Hispanic community in Port Chester has ever been elected to the Board of Trustees. With the exception of the recent election of Blanca Lopez to the School Board, an election the Court has not considered as probative in this case, no other Hispanics have been elected to public office in the Village. In short, there cannot be any more compelling case in support of Senate factor seven. See Goosby I, 956 F. Supp. at 343-44 (finding that Senate factor seven supported a Section 2 violation where only one African-American had ever been elected to the Town Board). Without question, this critical Senate factor supports a finding of a Section 2 violation.

h. Additional factors in the Senate Report

The Senate Report makes clear that the issue of a political subdivision's responsiveness has little probative value, particularly where the plaintiff has not made it an issue in the case — "defendants' proof of some responsiveness would not negate plaintiffs' showing by other, more objective factors enumerated here that minority voters nevertheless were shut out of equal access to the political process." See Marengo County Comm., 731 F.2d at 1572 (citing S. Rep. No. 97-417 at 29 n. 116). Thus, while Defendant at various points attempted to demonstrate that the Village is in fact sensitive to the needs of the Hispanic community, that alone is not enough to overcome this Court's findings with regard to the other Senate factors. Further, the current Port Chester system is not a marked departure from past practices in the Village, nor is it necessarily a significant departure from the structure employed by other localities in New York State. Plaintiff does not appear to contest these conclusions.

5. Conclusion

Having conducted the thorough and careful analysis required by the statute, this Court finds that Plaintiffs have demonstrated that the Village of Port Chester's at-large system for electing its Board of Trustees violates Section 2 of the Voting Rights Act. Plaintiffs have proven the existence of all three Gingles preconditions, and have shown clearly that under the totality of the circumstances, the at-large election system for electing members of the Board of Trustees prevents Hispanic voters from participating equally in the political process in the Village.

Defendant argued throughout the course of this case that, given time and assuming the continued growth of the Hispanic population of the Village, the Hispanic community could come to dominate the political landscape in Port Chester even under the current at-large system. This Court, however, is not charged with projecting what might happen years, or decades, from now; rather, we are faced with the current political reality in the Village, and based on the evidence presented, the Village is currently in violation of Section 2 of the Voting Rights Act.

III. Implementing a Remedial Plan

Both parties gave oral argument on their proposed remedies at conferences on July 17, 28, 29, 2008 and September 22 and 23, 2008. In October 2009, both parties submitted letters to the Court requesting a swift resolution to the case before the next Village Trustee elections on March 16, 2010. Port Chester's letter stated that an indication of the Court's decision could enable the elections to proceed if it was received on or before November 9, 2009, the last date by law to publish notices regarding the March election or specifying boundaries of election districts. See N.Y.S. Elec. Law § 15-104(3)(a), § 15-110(5).

The Court issued a summary order informing the parties of its decision to adopt Port Chester's proposal for cumulative voting. It also lifted the injunction on the Trustee elections, but required the 2010 election to be delayed until June 2010 to ensure enough time to properly implement the new system.

A. The parties' proposed remedial plans

1. Port Chester proposes a cumulative voting system

The Village of Port Chester proposes an at-large, cumulative voting scheme with the elimination of staggered terms. Each voter would be allotted the same number of votes as there are seats up for election and would be free to allocate them however he or she chooses. Voters may choose to "plump" all their votes on one candidate — the strategy of choice for minority communities who want to indicate a strong preference for a particular candidate. Defendant also acknowledges the need for an education program to help voters, and in particular Port Chester's Hispanic population, understand how cumulative voting works and what their strategic options are under the system.

2. Plaintiffs propose six single-member districts

Plaintiffs propose a districting plan that divides Port Chester into six single-member districts with one majority-minority Hispanic district. The plan is identical to the one drawn by Dr. Beveridge during the liability phase to demonstrate the Gingles factors. The plan's six districts have roughly equal populations, with deviations of about 3.34 percent. Dr. Beveridge has also testified that his plan satisfies a number of different measures of compactness. The majority-minority district has 56.27% Hispanic Citizen Voting Age Population (CVAP) according to the 2000 Census, and 70.35% Hispanic CVAP according to 2006 estimates of Port Chester's population. There are four districts in total in which the Hispanic share of the CVAP is greater than in the Village as a whole. However, the majority-minority district is the only district in which Hispanics would be able to elect a representative of their choice completely on their own (i.e. without cross-over voting).

B. Legal Standard for Choosing a Proposed Plan

The Court must give the defendant jurisdiction the first opportunity to suggest a legally acceptable remedial plan, based on the theory that the judiciary should not intrude on legislative policy any more than necessary. White v. Weiser, 412 U.S. 783, 794-95 (1973); Upham v. Seamon, 456 U.S. 37, 41 (1982);Cottier v. City of Martin, 445 F.3d 1113, 1123 (8th Cir. 2006). The Court must also defer to the choice of the governing legislative body so long as the choice is consistent with federal statutes and the Constitution. Whitcomb v. Chavis, 403 U.S. 124, 160-61 (1971); White, 412 U.S. at 797. The degree of deference is quite strong. A district court may not substitute its own remedial plan for defendant's legally acceptable one, even if it believes another plan would be better. See, e.g., Upham, 456 U.S. at 42.

Courts have explicitly recognized this deference applies to claims under Section 2 of the Voting Rights Act, as well as to one-person, one-vote cases. See, e.g., Growe v. Emison, 507 U.S. 25, 34 (1993); Cottier, 445 F.3d at 1123; Cane v. Worcester County, 35 F.3d 921, 927-28 (4th Cir. 1994); Harper v. City of Chicago Heights, 223 F.3d 593, 601-602 (7th Cir. 2000); McGhee v. Granville County, 860 F.2d 110, 115 (4th Cir. 1988). Courts have also expressly applied deference to the defendant jurisdiction's remedial plan in cases involving local legislative bodies.Harper, 223 F.3d at 601-02 (citing White v. Weiser). Therefore, if the Village's proposal is a legally acceptable remedy, the Court must accept it regardless of any alternative remedies proposed by Plaintiffs.

C. The Court adopts Port Chester's proposed remedy because it is legally acceptable

1. Cumulative voting is lawful as a remedy under the Voting Rights Act and New York Law

There is no case law that rejects cumulative voting as a lawful remedy under the Voting Rights Act. Recently, a district court in the Northern District of Ohio did exactly what Port Chester is asking of the Court in this case: it accepted the defendant's proposal for limited voting instead of the plaintiffs' districting plan to remedy a Section 2 violation. United States v. Euclid City School Bd. ("Euclid III"), 632 F. Supp. 2d 740.²⁹ Federal courts have repeatedly mentioned cumulative voting as a remedial option in Voting Rights Act cases. Holder v. Hall, 512 U.S. 874, 897-99, 908-13 (1994) (Thomas, J., concurring in the judgment) ("Nothing in our present understanding of the Voting Rights Act places a principled limit on the authority of federal courts that would prevent them from instituting a system of cumulative voting as a remedy under Section 2"); Branch v. Smith, 538 U.S. 254, 309-10 (2003) (O'Connor, J., concurring) ("a court could design an at-large election plan that awards seats on a cumulative basis, or by some other method that would result in a plan that satisfies the Voting Rights Act"); LULAC v. Clements, 986 F.2d 728, 814-15 (5th Cir. 1990) ("[S]tate policy choices may require the district court to carefully consider remedies such as cumulative voting" and other remedies), rev'd on other grounds, 999 F.2d 831 (5th Cir. 1993) (en banc); United States v. Marengo County Comm'n, 731 1546, 1560 (11th Cir. 1984); Dillard v. Town of Louisville, 730 F. Supp. 1546, 1548 n. 8 (M.D. Ala. 1990);Dillard v. Chilton County Bd. of Educ., 699 F. Supp. 870, 875 (M.D. Ala. 1988), aff'd, 868 F.2d 1274 (11th Cir. 1989); Euclid III, 632 F.Supp.2d at 752 n. 11 (N.D. Ohio 2009). Cumulative voting has also been mentioned as an option in New York voting rights cases. Lopez Torres v. New York State Bd. of Elec., 411 F.Supp.2d 212 (E.D.N.Y. 2006),rev'd on other grounds, 552 U.S. 196 (2008).

29 The court in Euclid III chose defendant's alternative proposal for limited voting over cumulative voting for nuanced reasons that were specific to the jurisdiction, such as a law requiring staggered elections and the prevalence of limited voting in the state. Euclid III, 632 F.Supp. at 746.

Plaintiffs would like the Court to believe that cumulative voting has been consistently rejected as a remedy to a Section 2 violation. This is a misstatement of the case law. None of the cases cited by Plaintiffs are rejecting cumulative voting as a concept, ³⁰ and a number of them go out of their way to clarify that the decision should not be taken as a condemnation of cumulative voting, see, e.g., Cane, 35 F.3d at 928-29 (4th Cir. 1994); Harper, 223 F.3d at 601 (7th Cir. 2000). Instead, the circuit courts found either that the district court improperly imposed its own remedy without first finding that defendant's plan was not legally acceptable, see Harper, 223 F.3d at 601, or the district court's plan did not adequately take into account the preferences of the defendant, Cane, 35 F.3d at 928-29. In others, cumulative voting was deemed inappropriate in judicial elections for reasons unique to the judiciary, see, e.g., Nipper v. Smith, 39 F.3d 1494 (11th Cir. 1994) (en banc). In this case, the Court will adopt defendant's proposal for cumulative voting as legally acceptable, rather than "conjure up such an election scheme [on its own] and impose it" on the Village, Dillard, 376 F.3d at 1268.

30 Cf. Cousin v. Sundquist, 145 F.3d 818, 829-30 (6th Cir. 1998) (the dicta expresses discomfort with the use of cumulative voting as a remedial measure in Section 2 violation cases but in particular with respect to judicial elections for reasons unique to the judiciary; the holding, however, does not rely on this at all since the court in that case found no Section 2 violation).

Cumulative voting is also not prohibited by New York law. New York's Constitution gives local legislatures full authority to adopt all laws "not inconsistent with" the state constitution or statutes concerning its own "affairs or government." CONSTITUTION OF STATE OF NEW YORK, Art. IX, § 2( 3)(c). This includes the "membership and composition" of a village's legislative body. Id. The same authority is codified in state statutes. See MUNICIPAL HOME RULE LAW, Art. 2, § 10(1)(i)-(ii). In the Northern District of Ohio, the district court rejected the argument that because Ohio law is silent on the issue of cumulative or limited voting, the court should not give deference to the defendant's plan for cumulative or limited voting. Euclid III, 632 F. Supp. 2d at 750 n. 9. Here too, the Court does not find that cumulative voting is prohibited by New York law just because the law is silent on the issue. The Court also does not find that the absence of cumulative voting in other New York villages means that Port Chester should get less deference, as Plaintiffs suggest. See Memorandum of Law of the U.S. in Support of Plaintiffs' Joint Proposed Remedial Plan, at 18.

2. Port Chester's cumulative voting plan would cleanse the Section 2 violation

For the Village's plan to cleanse the Section 2 violation, it must afford Hispanics in Port Chester an "equal opportunity to participate in the political processes and to elect candidates of their choice." Thornburg v. Gingles, 478 U.S. 30, 44; see also Hall v. Virginia, 385 F.3d 421 (4th Cir. 2004); 42 U.S.C. § 1973(b) (2000). This does not mean that Port Chester is obligated to guarantee electoral success for Hispanics, but rather the plan must provide a genuine opportunity "to exercise an electoral power that is commensurate with its population."LULAC v. Perry, 548 U.S. 399, 428 (2006); see also Johnson v. DeGrandy, 512 U.S. 997, 1014 n. 11 (1994) ("[T]he ultimate right of § 2 is equality of opportunity, not a guarantee of electoral success for minority-preferred candidates of whatever race").

The defendant's plan to remedy a Section 2 violation should also not create a new Section 2 violation. See, e.g., United States v. City of Euclid ("Euclid II"), 523 F. Supp. 2d 641, 644 (N.D. Ohio 2007). Moreover, the defendant's plan must also meet the constitutional requirements of one-person, one vote and the prohibition on the improper use of race in districting. See Goosby v. Town Bd. of Town of Hempstead, 981 F. Supp. 751, 755-56 (E.D.N.Y. 1997), aff'd, 180 F.3d 476 (2d Cir. 1999).

a. Port Chester's plan gives Hispanics a genuine opportunity to elect a representative of their choice

Courts evaluate whether cumulative voting will actually give minorities the opportunity to elect candidates of their choosing using a commonly-accepted and reliable political science concept called the "threshold of exclusion." See, e.g., Cottier v. City of Martin, 475 F. Supp. 2d 932, 937 (D.S.D. 2007); Steven J. Mulroy, The Way Out: A Legal Standard for Imposing Alternative Electoral Systems as Voting Rights Remedies, 33 HARV. L. REV. 333, 337 (1998). The threshold of exclusion "is the percentage of the vote that will guarantee the winning of a seat even under the most unfavorable circumstances." Cottier, 475 F. Supp. 2d at 937 (quoting Dillard v. Chilton County Bd. of Educ., 699 F. Supp. 870, 874 (M.D.Ala. 1988), aff'd, 868 F.2d 1274 (11th Cir. 1989). The threshold of exclusion is calculated according to the following formula: 1/(1 + number of seats available). Mulroy, at 1880.

The threshold of exclusion takes into account the following "worst case scenario:" (1) the majority sponsors as many candidates as there are seats to be filled (in this case, six candidates); and (2) the majority spreads its votes evenly among its candidates, with no support for the minority-preferred candidate. The threshold of exclusion also assumes that the minority population will allocate all of their votes to the minority-preferred candidate, often called "plumping" votes. If the minority population exceeds the threshold of exclusion and "plumps" their votes, they are virtually guaranteed to elect their preferred candidate even under the worst case scenario.Dillard, 699 F. Supp. at 874; see also Richard Engstrom, Report on Cumulative Voting for United States v. Village of Port Chester, Feb. 7, 2008, at ¶¶ 12-14 [hereinafter Engstrom Report].

Because the minority population achieves electoral success by plumping their votes, the cohesiveness of the minority voting bloc is very important. See Mulroy, at 1908. The Court finds that it is highly likely that Hispanics in Port Chester will plump their votes behind a candidate of choice because of the degree to which Hispanics voted cohesively in other elections. See supra II.B.3. For example, virtually 100 percent of Hispanics who voted cast a vote for Hispanic Trustee candidate Ruiz, a Plaintiff in this case. Id. Given this level of cohesiveness, it is reasonable to expect that the Hispanic population would continue to vote as a bloc and would therefore be able to take advantage of their voting power under a cumulative voting plan.

Currently, Port Chester has a six-member Board of Trustees, and each Trustee serves staggered three-year terms such that two positions are open for election each calendar year. Defendants propose eliminating the staggered terms so that six seats are up for election at each election. The threshold of exclusion would be 1/1+6 or 14.3 percent. In Port Chester, the Hispanic percentage of the CVAP according to the 2000 Census was 21.9 percent. Dr. Beveridge estimates that in 2006, the Hispanic CVAP will be 27.5 percent. Both of these figures are well above the threshold of exclusion: the 2000 count is 153.1 percent and the 2006 estimate is 192.3 percent above. Thus, Hispanics would have a genuine opportunity to elect one representative of their choice under Defendant's plan. Furthermore, the 2006 estimates suggest that Hispanics are also close to being able to guarantee the election two representatives of their choice using plumping. Because Hispanic CVAP is much greater than the threshold of exclusion, Hispanics still have an opportunity to elect their preferred candidate even if not all Hispanics plump their votes behind a single candidate of choice. See Remedy Hearing Tr., July 17, 2008, at 28.

In addition, both Plaintiffs' and Defendant's experts recognize that the opportunity to elect a candidate of choice tends to dramatically increase voter registration and turnout in the minority community. See Engstrom Report, at ¶ 25 (citing Beverage, Trial Tr., Feb. 16, 2007, at 606-07, 633; Handley, Trial Tr., Feb 15, 2007, at 466; Gaddie, Trial Tr., Feb. 22, 2007, at 1327). Courts have acknowledged that voter turnout is depressed in a discriminatory system and therefore may not be a reliable measure for turnout under a non-discriminatory plan.See, e.g., Solomon v. Liberty County ("Solomon I"), 865 F.2d 1566, 1574 (11th Cir. 1988); Harvell v. Blytheville School Dist. No. 5, 71 F.3d 1382, 1388 (8th Cir. 1995) ("[L]ow voter turnout has often been considered the result of the minority's inability to effectively participate in the political process"); cf. Rodriguez v. Pataki, 308 F. Supp. 2d 346, 401 (S.D.N.Y. 2004) (accepting the theory that minority voter turnout increases with genuine opportunity to elect a minority-preferred candidate, but finding that plaintiffs may not rely without evidence on this "warming effect" alone to create the requisite majority-minority district). Therefore, the Court can expect that turnout will likely increase as Hispanics in Port Chester realize their opportunity to elect their preferred representative.

b. Port Chester's plan does not create a new Section 2 violation so long as it is accompanied by a sufficient educational program and other conditions to be embodied in a consent decree

Because cumulative voting is not a common form of voting in this country, it is not automatically understood by voters. Also, the very rules of cumulative voting that enable minority populations to elect representatives of their choice are relatively complex and require voter education. See Lani Guinier, No Two Seats, the Elusive Quest for Political Equality, 77 VA. L. REV. 1413, 1471 n. 211 (1991). Particularly when a cumulative voting plan is proposed in a jurisdiction where vote dilution is due in part to historical discrimination in education and socio-economic factors, it must contain a plan to educate voters on the new process or else it is counterproductive to correcting the Section 2 violation. See Euclid III, 632 F. Supp. 2d at 756-57, 757 n. 15.

In this case, Port Chester's plan offers Hispanics a genuine opportunity to elect a representative of their choice only if they understand cumulative voting and how to take advantage of their electoral power. During the liability phase of this case, the Court found that Hispanics in Port Chester are at a distinct socio-economic and educational disadvantage compared to non-Hispanics. See supra II.B.5.e. The Court also found that Port Chester has historically failed to provide bilingual poll workers or election materials in Spanish to enable Hispanic voters to participate. See supra II.B.5.a. Against the background of these disparities and historical discrimination against Hispanics, it is imperative that the Village adequately address the barriers that might keep Hispanics from participating in the new system.

Defendant does propose an educational program to help voters understand cumulative voting rules, as well as to educate voters on their strategic options and how cumulative voting allows them to express the intensity of their preferences. In particular, Hispanic voters would need to grasp their power to plump votes in order to elect a candidate of their choice. See Engstrom Report, at ¶¶ 37-38. At the remedy hearing, Dr. Engstrom, the Village's expert witness, emphasized the need for special education when implementing a new, alternative voting system. See Remedy Hearing Tr., July 17, 2008, at 66-68 (Engstrom's suggestion that voters receive pamphlets, practice voting on test ballots, and learn what their options are under cumulative voting); see also Malroy, at 1893.

However, Port Chester's education plan does not contain enough details to reassure the Court that there will be a thorough effort to educate Hispanic voters. A federal court may make modifications to a defendant jurisdiction's plan, but the court is "limited to those necessary to cure any constitutional or statutory defect." Branch v. Smith, 538 U.S. 254, 309-10 (2003) (O'Connor, J., concurring) (citing Upham). In this case, the Court finds it is necessary to modify the Defendant's plan to eliminate any possibility of perpetuating the Section 2 violation that may result if Hispanic voters do not fully understand cumulative voting. Therefore, as a condition of accepting Port Chester's cumulative voting plan, the Court ordered both parties to determine the necessary conditions for the non-discriminatory implementation of cumulative voting, with a specific focus on the education program and election day support for Spanish-speakers. The parties detailed these conditions in a Consent Decree that the Court reviewed and approved on December 22, 2009. The Court further approved an addendum to the Consent Decree on February 23, 2010.

c. Port Chester's plan does not violate one-person, one-vote or use of race improperly in districting

In addition to satisfying Section 2, the defendant's remedial plan must also comply with the Fourteenth Amendment's one-person, one-vote requirement. See Abrams v. Johnson, 521 U.S. 74, 98 (1997). District courts have consistently found that cumulative voting complies with one-person, one-vote because the entire population is contained in one district and each voter is given the same number of votes. See, e.g., Cottier v. City of Martin, 475 F. Supp. 2d 932, 939 (D.S.D. 2007); McCoy v. Chicago Heights, 6 F. Supp. 2d 973, 984 (N.D.Ill. 1998), rev'd sub nom. on other grounds by Harper v. City of Chicago Heights, 223 F.3d 593 (7th Cir. 2000); Cane v. Worcester County, 847 F. Supp. 369, 374 n. 8 (D.Md. 1994), rev'd on other grounds, 35 F.3d 921 (4th Cir. 1994). Thus, Port Chester's cumulative voting plan does not violate one-person, one-vote.

Since Shaw v. Reno, 509 U.S. 630 (1993), the Supreme Court has recognized that the Equal Protection Clause of the Fourteenth Amendment limits the use of race in districting. If race is the predominant factor motivating the districting plan such that the legislature subordinated traditional race-neutral districting principles, the Court should apply strict scrutiny. Under strict scrutiny, the challenged plan will only survive if it is narrowly tailored to serve a compelling state interest. Id.; Miller v. Johnson, 525 U.S. 900, 916 (1995).

Plaintiffs do not argue that the Village's plan is a Shaw violation, nor does the plan involve any consideration of race since every voter is treated exactly the same. In fact, cumulative voting and other alternative voting schemes have received focus precisely because they avoid the Shaw problem that plagued drawing single-member districts. See, e.g., Jason Kirksey, et al., "Shaw v. Reno and the New Election Systems: The Cumulative Voting Alternative," Voting Rights Rev. 10 (Spring 1995). Therefore, the Court finds that cumulative voting in this case does not improperly use race.

3. The Court has no obligation to consider whether districting would be better if the defendant's plan is legally acceptable

As explained above, the Court is required to defer to the defendant's remedial plan and evaluate only whether it is legally acceptable. If the defendant's plan has a statutory or constitutional infirmity, the Court must fashion a remedy that complies with Section 2 and also "to the greatest extent possible give[s] effect to the legislative policy judgments underlying the current electoral scheme or the legally unacceptable one offered by the legislative body." Cane v. Worcester County, 35 F.3d 921, 928 (4th Cir. 1994). Since the only criteria for judging the sufficiency of Port Chester's plan is statutory and constitutional acceptableness, the Court need not consider whether Plaintiffs' remedial plan is better. Therefore, Plaintiffs' assertions that single-member districts are preferable remedies in Section 2 violation cases are not relevant to this determination. Nor is it relevant that Plaintiffs have proposed a districting plan that would itself pass constitutional muster. Had Port Chester proposed or supported the districting plan, the Court would examine it for legal acceptableness. However, Port Chester has clearly stated its preference for cumulative voting in a multi-member district and since the Court has found that plan to be legally acceptable, the inquiry must end there.

IV. Conclusion

The Court conducted a careful analysis of Port Chester's at-large voting scheme for electing its Board of Trustees and determined that the system violated Section 2 of the Voting Rights Act. Plaintiffs have proven the existence of all threeGingles preconditions, and have shown clearly that under the totality of the circumstances, the at-large election system for electing members of the Board of Trustees prevents Hispanic voters from participating equally in the political process in the Village.

Having found a Section 2 violation, the Court evaluated the parties' proposed remedial plans. Following the high level of deference accorded to the defendant jurisdiction, the Court adopted Port Chester's proposal for cumulative voting because it was deemed legally acceptable under the Voting Rights Act, the Constitution, and New York law. The Court has approved the parties' agreed upon Consent Decree detailing the education and outreach program that will ensure the effective and non-discriminatory implementation of the new system. To give sufficient time for implementation, the Court also orders Port Chester to hold its 2010 Trustee elections in June 2010.

It is So Ordered

Dated: White Plains, New York

April 1, 2010