CLERKS OFFICE U.S. DIST. COUF AT LYNCHBURG, VA FILED UNITED STATES DISTRICT COURT 8/91/2020 WESTERN DISTRICT OF VIRGINIA JULIA C. DUDLEY. CLERK LYNCHBURG DIVISION BY: s/ CARMEN AMOS DEPUTY CLERK LEAGUE OF WOMEN VOTERS CASE No. 6:20-cv-00024 OF VIRGINIA, ef al., Plaintiffs, v. MEMORANDUM OPINION
VIRGINIA STATE BOARD JUDGE NORMAN K. MOON OF ELECTIONS, ef al., Defendants.
In this action, the League of Women Voters of Virginia and several voters have sued the Virginia State Board of Elections and a number of state officials, arguing that Virginia’s witness signature requirement for absentee ballots is an unconstitutional burden on the right to vote as applied during the COVID-19 pandemic. The parties later reached a partial settlement, and the Court approved their joint request for a consent decree that would enjoin enforcement of the witness signature requirement for Virginia’s primaries on June 23, 2020, for voters who believe they may not safely have a witness present while completing their ballot. Plaintiffs later filed a second motion for a preliminary injunction, and the parties then reached another partial settlement and now jointly seek approval of a second consent decree that would enjoin the witness signature requirement for Virginia’s elections on November 3, 2020, for the same group of at-risk voters. The same reasons that motivated the Court to approve the parties’ previous consent decree carry even more force today, as the pandemic has resurged. Thus, the Court will grant the motion and approve the agreement, which does not address any election beyond the November 3, 2020 elections.
I. BACKGROUND A. Factual Background The Court incorporates the background set forth in its prior opinion, League of Women Voters of Va. v. Va. State Bd. of Elections, No. 6:20-cv-00024 (Dkt. 69), ---F. Supp. ---,2020 WL 2158249 (W.D. Va. May 5, 2020), as supplemented by the more recent significant factual developments included in the record since that point and as set forth below. 1. The COVID-19 Pandemic Over five months ago, Virginia Governor Ralph Northam declared a state of emergency in response to the emergence of COVID-19. Va. Exec. Order No. 2020-51 (March 12, 2020). On May 26, 2020, the Governor extended the state of emergency in response to the continued public
health threat posed by COVID-19, pursuant to an amended version of Executive Order 51. Va. Exec. Order No. 2020-51 Amended (May 26, 2020).This declaration remains in effectuntil it has been rescinded or amended by a subsequent order. Id. Despite the efforts of state and local officials in the Commonwealth, the pandemic rages on. State health guidelines continue to instruct everyone to stay home as much as possible and to continue social distancing practices.1Federal guidelines, issued by the Centers for Disease Control and Prevention (“CDC”), continue to instruct people to “[p]ut 6 feet of distance between yourself
1 See Va. Dep’t of Health, Prevention Tips, https://www.vdh.virginia.gov/coronavirus/prevention-tips/ (last accessed Aug. 21, 2020). and people who don’t live in your household.”2The CDC has also instructedjurisdictions to “offer alternative voting methods that minimize direct contact.”3 There remains no dispute that, based on the Census Bureau’s 2018 Current Population Survey, over twenty-fivepercent of Virginians over the age of eighteen live alone.Dkt. 103(Joint Proposed Findings of Fact and Conclusions of Law) at ¶ 55.4 Nor do the parties dispute that some
at-risk populations are more likely to live alone, such as Virginians over the age of sixty-five. Id. While “[p]eople of every age can and have contracted COVID-19, including severe cases ... geriatric patients are at the greatest risk of severe cases, long-term impairment, and death.” Dkt. 88-1 (Decl. of Dr. Arthur L. Reingold) at ¶ 8. This is also true of those “with immunologic conditions and with other pre-existing conditions, such as hypertension, certain heart conditions, lung diseases . . . diabetes mellitus, obesity, and chronic kidney disease.” Id. When this Court granted the parties’first motion for a partial consent decree and judgment, the Virginia Department of Health had calculated over 20,000 confirmed or probable cases of COVID-19, over 2,700 hospitalizations, and over 700 deaths attributed to the disease in the
Commonwealth alone. League of Women Voters of Va., No. 6:20-cv-00024 (Dkt. 69), 2020 WL 2158249, at *2. In the three months since then, these numbers have now soared above 110,000
2 Centers for Disease Control and Prevention, Coronavirus Disease 2019 (COVID-19), How to Protect Yourself & Others, available at https://www.cdc.gov/coronavirus/2019- ncov/prevent-getting-sick/prevention.html (last accessed Aug. 20, 2020). 3 Centers for Disease Control and Prevention, Recommendations for Election Polling Locations, available at https://www.cdc.gov/coronavirus/2019-ncov/community/election-polling- locations.html (last accessed Aug. 20, 2020). 4 citing 2018 Current Population Survey, U.S. Census Bureau, statistics accessed by using the Census Bureau Current Population Survey Table Creator tool at https://www.census.gov/cps/data/cpstablecreator.html. confirmed or probable cases of COVID-19, over 9,000 hospitalizations, and over 2,400 deaths.5 As before, the data likely undercount the actual number of positive cases because of the Commonwealth’s limitations with regard to testing capacity. Dkt. 88-1 (Decl. of Dr. Arthur L. Reingold) at ¶ 11; Dkt. 103 (Joint Proposed Findings of Fact and Conclusions of Law) at ¶ 49. Indeed, this is true throughout the United States. Dkt. 88-1 (Decl. of Dr. Arthur L. Reingold) at
¶ 11. On July 1, 2020, the Commonwealth began Phase Three of its three-phase re-opening. Va. Exec. Order No. 2020-67 (June 30, 2020). Among other changes, the Governor’s Phase Three Re- opening order removes the capacity restrictions for a number of business sectors, including retail and restaurants,6 and expands the limits on public and private gatherings to permit gatherings of up to 250 people.7 Id. at 9. Phase Three still requires businesses to comply with social distancing practices and face-mask requirements. See, generally, id. On July 28, 2020, the Governor rolled back this re-opening for certain jurisdictions in the Hampton Roads region in response to “disturbing trends in connection with the increasing number of [COVID-19] cases. Va. Exec.
Order 2020-68 Amended (Aug. 3, 2020). These restrictions reduced the permitted size of public and private in-person gatherings down to no more than 50 people and imposed operating-hours limitations on venues that serve alcohol. Id.
5 Virginia Department of Health, COVID-19 & You, available at https://www.vdh.virginia.gov/coronavirus/covid-19-daily-dashboard/ (last accessed Aug. 21, 2020). In a reflection of the rapid pace of the virus’s destruction, I have had to continually revise these numbers upward in the span of time I have spent writing this opinion. 6 Other business sectors, such as entertainment and amusement venues, racetracks and speedways, and fitness and exercise studios, do still have occupancy restrictions in place. See Va. Exec. Order 2020-67 (June 30, 2020) at 4, 7–8. 7 This number is up from 50 people during Phase Two. Va. Exec. Order 2020-65 Amended, at 11 (June 9, 2020). 2. Virginia’s Absentee Voting Scheme Va. Code § 24.2-612 requires that absentee ballots be made available forty-five days prior to the November 3, 2020 elections—Saturday, September 19, 2020. Or, in localities whose general registrar is closed on that day, ballots must be made available on Friday, September 18, 2020. Dkt. 95-1 at ¶ 18; see Dkt. 88-6 (Decl. of Stephanie L. Liles), at ¶ 6.
General registrars maintain lists of voters who have requested and returned absentee ballots, and absentee ballots are only accepted from voters whose names are on that list. Dkt. 64- 3 at ¶ 7. Upon receiving a valid request for an absentee ballot and confirming that the applicant is a registered voter in the jurisdiction, the registrar provides an absentee voter’s package, which includes the following: a sealed envelope containing the ballot and a second envelope containing printed instructions as well as a “Statement of Voter” form, which the absentee voter is required to sign, attesting to their identity, residency status, and that they “marked the ballot(s) in the presence of [a] witness, without assistance or knowledge on the part of anyone as to the manner in which [the voter] marked it” and that they the voter has “not voted and will not vote in this election at any other time or place.” Va. Code § 24.2-706; see Dkt. 84 at ¶ 70.
Pertinent here, Va. Code §§ 24.2-706 and 24.2-707—the provision Plaintiffs challenge as unconstitutional as applied—require that the absentee voter “mark[] the ballot(s) in the presence of [a] witness, without assistance or knowledge on the part of anyone as to the manner in which [the voter] marked it” and that the voter personally attest to the ballot being so witnessed. See Dkt. 84 at ¶ 70. Individuals who make a willfully false material statement or entry on an absentee ballot are at risk of being charged with a Class 5 felony. Va. Code § 24.2-1012. And anyone who falsely signs the name of a different qualified voter on a ballot is at risk of being charged with a Class 4 felony. Va. Code § 24.2-1012. A voter’s failure to obtain a witness signature on their absentee ballot is an omission that is “always material” that renders the ballot invalid. See 1 Va. Admin. Code 20-70-20(B); Dkt. 1 at ¶ 70. However, “[t]he illegibility of a voter’s or witness’s signature . . . shall not be considered an omission or error.” 1 Va. Admin. Code 20-70-20(C)(10); see Dkt. 84 at ¶ 75; Dkt. 91 at ¶ 75. In the May 2020 primary, which required absentee ballots to include a witness signature,
653 ballots were rejected for lack of a witness signature. This is a substantial increase over the 9 ballots rejected for this reason in 2016 and 25 rejected in 2018, Dkt. 96 at 6 (citing Dkt. 88-2 at 17–19). The parties dispute whether this increase resulted from some voters’ fear of obtaining a witness safely or whether it was the result of public confusion over the extent of this Court’s previous consent decree and the parties’ statements regarding it. Compare Dkt. 99 at 12, with Dkt. 88 at 26. Regardless, the Plaintiffs have put forth evidence supporting the claim that, if absentee ballots are rejected at the same rate as in May and historical patterns of absentee ballot usage and turnout hold, nearly 17,000 voters would have their ballots rejected for failure to meet the witness signature requirement. Dkt. 88-2 (Decl. of Dr. John B. Holbein) at 20.
B. Procedural History On April 17, 2020, the League of Women Voters of Virginia and three individual voters— Katherine D. Crowley, Erikka Goff, and Seijra Toogood—(“Plaintiffs”) sued the Virginia State Board of Elections and, in their official capacities, its Chairman (Robert H. Brink), Vice-Chair (John O’Bannon), and Secretary (Jamilah D. LeCruise), as well as the Commissioner of the Virginia Department of Elections (Christopher E. Piper) (“State Defendants”), seeking to enjoin enforcement of Va. Code § 24.2-707(A), which mandates that all absentee ballots be signed by a
witness before submission. Dkt. 1 at ¶¶ 1–2. Given the public health crisis spawned by the COVID- 19 pandemic, Plaintiffs argue that the requirement unduly burdens their right to vote. Plaintiffs filed their second amended complaint on July 17, 2020. Dkt. 84. On April 21, 2020, Plaintiffs moved for a preliminary injunction that: (1) would have prohibited State Defendants from enforcing the aforementioned witness requirement for all Virginia voters for the June 23, 2020 primaries and for any and all subsequent elections in Virginia
until such time as in-person interactions required by compliance with the witness requirement no longer pose a risk to public health and personal safety; (2) would have ordered State Defendants to issue guidance instructing city and county election officials to count otherwise validly cast absentee ballots that are missing a witness signature for Virginia’s June 23 primary elections; and (3) would have ordered State Defendants to conduct a public information campaign informing Virginia voters about the elimination of this requirement, in coordination with city and county election officials. Dkt. 16. Just days after Plaintiffs filed their motion for a preliminary injunction, two groups sought to intervene in this case: a group of three voters, Dkt. 22, and the Republican Party of Virginia
(“RPV”) together with another three voters (Vincent E. Falter, Mildred H. Scott, and Thomas N. Turner, Jr.), Dkt. 28. After the motions to intervene were fully briefed, the Court heard arguments on the motions and denied intervention as to the six voters, but it permitted the RPV to intervene as a defendant in the suit. Dkts. 55, 57. The Court permitted the first group of three voters who were denied intervention in this suit to file amicus briefs in the case and treated all briefs it had filed thus far as amicus briefs. Dkt. 55 at 1 n.1. On April 27, 2020, Plaintiffs and State Defendants filed their “Joint Motion for Entry of Partial Consent Judgment and Decree,” Dkt. 35, which would resolve these parties’ dispute over the application of the witness signature requirement for only the June primary elections. The Court held a second status conference on April 29, 2020, to preliminarily discuss the impact of the joint motion for a partial consent decree on this case. Dkt. 39. On April 30, 2020, the RPV filed a brief opposing the approval of the consent decree. Dkt. 58. The next day, Plaintiffs and State Defendants filed their reply briefs in support of their proposed agreement. Dkts. 62, 64. That week, the Court also accepted three amicus briefs in this case.
On May 4, 2020, the Court heard oral argument from Plaintiffs, State Defendants and Intervenor-Defendant RPV on the merits of both the pending motion for a preliminary injunction, Dkt. 16, and the motion for approval of the consent decree, Dkt. 35. The next day, the Court issued a Memorandum Opinion and Order approving the Plaintiffs and State Defendants’ partial consent decree. Dkt. 68. Pursuant to that consent decree, Plaintiffs then withdrew their first motion for a preliminary injunction. Dkt. 73. On July 24, 2020, the Plaintiffs filed a second motion for a preliminary injunction. Dkt. 87. This second motion sought substantially the same relief as Plaintiffs’ first motion for a preliminary injunction—the key difference being that such relief would apply to the November 3, 2020
elections rather than the already completed June 23, 2020 primaries. The Court ordered expedited briefing on the motion. Dkt. 94. Twelve days after the filing of the second motion for a preliminary injunction, Plaintiffs and State Defendants reached another a partial settlement and filed a joint motion to approve a partial consent judgment and decree, which would enjoin the witness signature requirement for only the November 3, 2020 general election. Dkt. 95. This was supplemented on August 20, 2020, to clarify that the parties intended the term “general election” to encompass all elections held in Virginia on November 3, 2020—including the twenty-seven special elections being held that day and any ballot referenda or other measures appearing on ballots.8 Dkt. 108-1 at ¶ 1 (stating that “in referring to the ‘November 3, 2020 General Election’ and the ‘November Election’ [in the proposed partial consent decree, the Plaintiffs and State Defendants] intended to include all elections and measures for which Virginiavoters are able to cast ballots with an election date of November 3, 2020, including any special elections, ballot referenda, or other measures
appearing on Virginia voters’ ballots that are set to that Election Date.”). On August 14, the Court heard oral argument from Plaintiffs, State Defendants, and the RPVon the merits of the motion to approve the proposed consent judgment and decree as well as the second motion for a preliminary injunction. On August 18, 2020, the Court received joint proposed findings of fact and conclusions of law from the RPV and the Plaintiffs9 as well as individual findings from the RPV, the Plaintiffs, and the State Defendants. Dkts. 103–106. The State Defendants’ proposed findings expressly endorse and agree with Plaintiffs’ proposed individual findings, with the exception of ¶ 24. Dkt. 106.
C. Proposed Partial Consent Decree On August 5, 2020, Plaintiffs and the State Defendants filed their “Joint Motion for Entry of Partial Consent Judgment and Decree,” Dkt.95, which would resolve these parties’ dispute over the application of the witness signature requirement for only the upcoming November 3, 2020 elections. The proposed partial consent judgment and decree provides in part: Defendants shall issue updated instructions to include with all absentee ballots as provided in Va. Code. § 24.2-706—or issue guidance instructing all relevant city 8 The Court will use the term “November 3, 2020 elections” to refer to all elections and measures for which Virginia voters are able to cast ballots on that date. 9 The Court appreciates the parties’ substantial efforts to reach agreement wherever possible on aspects of the factual background of this case. The Court has found the parties’ joint proposed findings of fact as well as individual proposed findings very useful in its disposition of this motion. and county election officials to modify or amend the printed instructions accompanying each absentee ballot— to inform voters that any absentee ballot cast in the November Election without a witness signature will not be rejected on that basis and specifically inform voters in bold print that they may disregard the witness signature line on the absentee ballot envelope if they believe they may not safely have a witness present while completing their ballot.
Dkt. 95-1 at 8. The Commissioner of the Virginia Department of Elections would also be required under this agreement to undertake “additional reasonable steps to inform the public that the witness requirement will not be enforced for those absentee voters who believe they may not safely have a witness present while completing their ballot, and issue guidance instructing all relevant city and county election officials to do the same.” Id. at 8–9. Plaintiffs would then withdraw their motion for a preliminary injunction and waive any entitlement to attorneys’ fees and costs that have accrued up to the date the agreement is approved by the Court. Id. at 9 To be clear, this agreement is limited in duration and scope to elections and ballot measures for a single date: November 3, 2020. As with this Court’s prior consent decree, this subsequent proposed consent decree does not address any future election. Dkt. 95 at 7. II. LEGAL STANDARD Consent decrees have a dual nature, carrying elements “of both judgment and contract.” Szaller v. Am. Nat. Red Cross, 293 F.3d 148, 152 (4th Cir. 2002). As such, the Fourth Circuit has instructed courts to scrutinize their terms in order to first determine whether such agreements are “fair, adequate, and reasonable” and also to ensure that in approving and issuing a consent judgment, that a court does not stamp its imprimatur upon an agreement that is “illegal, a product of collusion, or against the public interest.” United States v. North Carolina, 180 F.3d 574, 581 (4th Cir. 1999); see Ohio Valley Envt’l Coal. v. Pocahontas Land Corp., No. 2:15-cv-15515, 2017 WL 988115, at *2 (S.D. W. Va. Mar. 14, 2017). In all this, courts should “be guided by the general principle that settlements are encouraged . . . [but] should not blindly accept the terms of a proposed settlement.” United States v. North Carolina, 180 F.3d at 581. In considering a proposed consent decree, the Court must assess “the strength of the plaintiff’s case,” id. (citing Flinn v. FMC Corp., 528 F.2d 1169, 1173 (4th Cir. 1975)), and [w]hile this assessment does not require the court to conduct “a trial or a rehearsal of the trial,” the court must take the necessary steps to ensure that it is able to reach “an informed, just and reasoned decision.” Id. (internal quotation marks omitted.) In particular, the “court should consider the extent of discovery that has taken place, the stage of the proceedings, the want of collusion in the settlement and the experience of plaintiffs’ counsel who negotiated the settlement.” Id. (quoting Carson v. American Brands, Inc., 606 F.2d 420, 430 (4th Cir. 1979) (en banc) (Winter, J., dissenting), adopted by Carson v. Am. Brands, Inc., 654 F.2d 300, 301 (4th Cir. 1981) (en banc) (per curiam)); see United States v. E.I. du Pont de Nemours & Co., No. 5:16-cv-00082, 2017 WL 3220449, at *11 (W.D. Va. July 28, 2017). As in this case, “when a settlement has been negotiated by a specially equipped agency, the presumption in favor of settlement is particularly strong.” Md. Dept. of Env’t v. GenOn Ash Mgmt., LLC, Nos. 10-cv-0826, 11-cv-1209, 12-cv-3755, 2013 WL 2637475, at *1 (D. Md. June 11, 2013). III. ANALYSIS The proposed settlement agreement now before the Court is nearly identical to that which this Court approved on May 5, 2020. The key difference, of course, is that it applies to the November 3, 2020 elections as opposed to the June 23, 2020 primaries. Importantly, just last week both the First Circuit and the Supreme Court denied the Republican National Committee and Rhode Island Republican Party’s (intervenor-defendants in that case) application to stay a consent decree between Rhode Island and voting rights groups issued by District of Rhode Island. Republican Nat. Comm. v. Common Cause RI, No. 20A28, 2020 WL 4680151, at *1 (U.S. Aug. 13, 2020). As here, the Rhode Island consent decree enjoined the state’s witness signature requirement for absentee voting. The Supreme Court held that “here the state election officials support the challenged decree, and no state official has expressed opposition. Under [these] circumstances, [the political party] applicants lack a cognizable interest in the State’s ability to ‘enforce its duly enacted’ laws.” Id.(internal citations omitted). Further, the Court stated how the status quo in Rhode Island during the pandemic was set in that state’s previous election, which had suspended the witness signature
requirement, and it indicated that changing it could confuse the “many voters” who could be under the belief that it was also suspended for later elections. Id. For these reasons, the Supreme Court distinguished its prior recent Order granting the stay of a district court’s order that had enjoined a witness signature requirement in Alabama,Merrill v. People First of Ala., No. 19A1063, 2020 WL 3604049 (U.S. July 2, 2020). The same reasons underlying the District of Rhode Island’s approval of the consent decree support this Court’s approval of Virginia’s proposed partial consent decree in this case. There is little, if any, meaningful difference between the pending motion to approve the proposed consent decree currently before this Court and the Rhode Island consent decree recently before Supreme
Court, First Circuit, and District of Rhode Island. Consequently, the reasoning in those decisions has provided significant guidance to the Court in deciding this motion. The Court has scrutinized the terms of the proposed partial consent decree, Dkt. 95-1, and all other submissions in the record. The briefing and oral argument in support of and against the proposed partial consent decree were ably presented by counsel. As the following analysis explains, the Court has determined that, based on the strength of Plaintiffs’ case and the quality of the counsel representing the parties to the agreement,as well as theotherfactors for consideration, the proposed consent decree is “fair, adequate, and reasonable.” Further, the Court finds that that this agreement is in the public interest, and that it is not unlawful or the product of collusion between the drafting parties. The Court will approve the proposed consent decree.
A. Fair, Adequate, and Reasonable In analyzing whether an agreement is fair, adequate, and reasonable, the Court must examine the strength of the plaintiff’s case. United States v. North Carolina, 180 F.3d at 581 (citing Flinn, 528 F.2d at 1172–73). The Fourth Circuit has further directed courts to consider “the extent of discovery that has taken place, the stage of the proceedings, the want of collusion in the settlement and the experience of plaintiffs’ counsel who adopted the settlement.” United States v. North Carolina, 180 F.3d at 581. This Court has already determined that the respective counsel for the parties who proposed the settlement provided fair and adequate legal counsel in representing Plaintiffs and State Defendants in their settlement negotiations. League of Women Voters of Va., No. 6:20-cv-00024
(Dkt. 69), 2020 WL 2158249, at *5. The case is now in a more advanced stage of proceedings than when the Court approved the previous consent decree. Indeed, Plaintiffs sought discovery, including on election data for the 2016, 2018, and 2020 elections, Dkt. 96 at 6, and responsive pleadings have been entered in the case, Dkts. 91, 93. These factors point in favor of the fairness, adequacy, and reasonableness of the consent decree. The Court also finds that the limited nature of the settlement agreement’s relief weighs in favor of a finding that it is “fair, adequate, and reasonable.” See United States v. North Carolina, 180 F.3d at 581. Like the prior consent decree, the proposed consent decree is limited in duration and scope to elections and ballot measures for a single date: November 3, 2020.
The RPV argues that the Plaintiffs and State Defendants’ failure to address other portions of Virginia law that require witness signatures, such as the Commonwealth’s emergency absentee ballot scheme, somehow makes the agreement before the Court less fair, reasonable, or adequate. But, if anything, this fact reflects the limited nature of the agreement. Virginia’s emergency absentee ballot scheme under Va. Code 24.2-705 was not challenged in Plaintiffs’ complaint. Moreover, Virginia’s emergency absentee ballot scheme permits a voter’s designated representative to request an absentee ballot up until 2:00 PM on the day before the election—this
involves different risks than when the absentee voter requests the ballot on their own. Virginia law requires a witness to sign the request form that the voter completes and signs in order to obtain the emergency ballot, and it requires the designated representative who obtained the ballot to witness the voter voting the ballot. The benefit of the witness signature requirement in this context is clearly more significant, and this is reflected by the verification procedures in place for identifying the designated representative who obtains the ballot for the voter and witnesses them vote the ballot. Va. Code § 24.2-705 (“The designated representative shall complete a statement, subject to felony penalties for making false statements pursuant to § 24.2-1016, that (i) he is the designated representative of
the requesting voter; (ii) he personally delivered the ballot to the voter who applied for it; (iii) in his presence, the voter marked the ballot, the ballot was placed in the envelope provided, the envelope was sealed, and the statement on its reverse side was signed by the requesting voter; and (iv) the ballot was returned, under seal, to the general registrar at the registrar's office.”). That the parties would not, pursuant to a settlement, seek to enjoin a law that was not challenged by the Plaintiffs should be expected. Contrary to the RPV’s contention, the failure to do so does not mean the proposed consent decree is any less fair, adequate, or reasonable. It is also significant to the Court that Plaintiffs continue to plead a probable violation of federal law. See Kasper v. Bd. of Election Comm’rs of the City of Chi., 814 F.2d 332, 342 (7th Cir. 1987); League of Women Voters of Va., No. 6:20-cv-00024 (Dkt. 69), 2020 WL 2158249, at *6. The Supreme Court has made clear that it has not identified any “litmus test for measuring the severity of a burden that a state law imposes on a political party, an individual voter, or a discrete class of voters.” Crawford v. Marion Cty. Election Bd., 553, U.S. 181, 191 (2008). Rather, “however slight that burden may appear” the reviewing court must find that it is “justified by
relevant and legitimate state interests ‘sufficiently weighty to justify the limitation.’” Id. (quoting Norman v. Reed, 502 U.S. 279, 288–89 (1992)); McLaughlin v. N.C. Bd. of Elections, 65 F.3d 1215, 1221 n.6 (4th Cir. 1995) (“We believe that a regulation which imposes only moderate burdens could well fail the Anderson balancing test when the interests that it serves are minor, notwithstanding that the regulation is rational.”). The Anderson-Burdick analysis set forth in the Court’s previous Memorandum Opinion, League of Women Voters of Va., No. 6:20-cv-00024 (Dkt. 69), 2020 WL 2158249, at *7–9, looks much the same on review of the current factual record. The witness signature requirement still creates a burden on a substantial and discrete class of Virginia’s electorate while the pandemic rages on,10 and this burden is not justified by the value of that requirement as an anti-fraud measure
in face of the risk posed by COVID-19. Since this Court issued its previous Memorandum Opinion, Plaintiffs have filed declarations from a number of voters who detail the health risks they are undertaking by attempting to meet the witness signature requirement. Each of the individual Plaintiffs, who all live alone, have attested that they do not feel safe interacting with a witness in order to fulfill the requirement.
10 What is more, the record indicates that these conditions are expected to continue for at least the rest of the year. Dkt. 103 (Joint Proposed Findings of Fact and Conclusions of Law) at ¶ 13 (citing e.g., League of Women Voters of Va. v. Va. State Bd. of Elections, Dkt. 6-20-cv-00024, 2020 WL 2158249, at *6 (W.D. Va. May 5, 2020)); Dkt. 88-1 at ¶¶ 14–19). And so, despite their desire to vote in upcoming elections, they feel they must refrain from voting. See Dkt. 84 at ¶¶ 10–19; Dkt. 88-3 (Decl. of Carol D. Peterson) at ¶¶ 5–6; Dkt. 88-7 (Decl. of Seijra Toogood) at ¶¶ 7–8; Dkt. 88-8 (Decl. of Katherine Crowley) at ¶ 6; Dkt. 88-9 (Decl. of Gayle Hardy) at ¶ 7; Dkt. 88-10 (Decl. of Tracy Safran) at ¶¶ 5–6. Plaintiff League of Women Voters of Virginia (“the League”) has “nearly 2,000 members across Virginia, some of whom live
alone and are vulnerable to becoming severely ill from COVID-19 because they are elderly and/or have underlying health conditions.” Dkt. 84 at ¶ 21. The League states that some amount of these members will be unable to comply with the witness requirement on absentee voting “without placing their own health at risk.” Id. at ¶ 21. Plaintiffs have also submitted two new expert declarations:11 a second declaration from epidemiologist Dr. Arthur Reingold, Dkt. 88-1—whose two declarations remain the only expert medical testimony in the record—and a declaration by Dr. John Holbein, a professor at the University of Virginia who studies voter participation, Dkt. 88-2. The testimony in Dr. Reingold’s second declaration builds on his first declaration, addressing relevant developments in COVID-19
research; this includes testimony that “SARS-CoV-2 is also aerosolized, such that tiny droplets containing the virus remain in the air and can be inhaled by others who come into contact with that air and SARS-CoV-2 can also be transmitted in that fashion,” Dkt. 88-1 at ¶ 10. Dr. Holbein’s testimony provides an expert opinion on academic, governmental, legal, and media sources to estimate the number of voters who might be estimated to submit a ballot without a witness signature and the effect of the witness signature on Virginia voter participation during the pandemic. Dkt. 88-2 at ¶¶ 1–2. In his testimony, Dr. Holbein ultimately concludes that “Virginia’s
11 Despite contesting the impact of aspects of their testimony, the parties do not contest their expert qualifications. witness requirement will decrease voter turnout, especially so among older, disabled, disadvantaged, and minority citizens. During a pandemic, this requirement effectively neuters the gains that vote-by-mail systems afford and disenfranchises voters.” Id. at ¶ 2. The Court reiterates that “[i]n ordinary times, Virginia’s witness signature requirement may not be a significant burden on the right to vote.” League of Women Voters of Va., No. 6:20-
cv-00024 (Dkt. 69), 2020 WL 2158249, at *8. However, the witness requirement remains “both too restrictive and not restrictive enough to effectively prevent voter fraud.” Id. (quoting N.C. State Conference of NAACP v. McCrory, 831 F.3d 204, 235 (4th Cir. 2016); see Common Cause RI v. Gorbea, No. 1:20-cv-00318, 2020 WL 4365608, at *4 (D.R.I. July 30, 2020) (approving consent decree that enjoined Rhode Island’s witness signature requirement and stating that “the Court would have found that the mail-ballot witness or notary requirement, as applied during the COVID-19 pandemic, is violative of the First and Fourteenth Amendments to the United States Constitution because it places an unconstitutional burden on the right to vote”); Thomas v. Andino, No. 3:20-cv-01730, --- F. 3d ----, 2020 WL 2617329, at *21 (D.S.C. May 25, 2020) (enjoining
witness signature requirement that did not include any witness verification measures because the burdens it imposed during the pandemic “far outweigh the imprecise, and (as admitted by SCEC Defendants) ineffective, state interests of combating voter fraud and protecting voting integrity”). The RPV argues that people, even those who are at a greater risk of developing severe or life-threatening complications from COVID-19, have learned over the past several months how to manage their lives safely. But for those who are at a risk of developing severe or life-threatening COVID-19, every potential exposure is a risk—even when taking appropriate precautions, such as wearing a face mask, frequently washing one’s hands, and complying with social distancing guidelines. Such risks may be necessary to obtain food and other necessities, but the burden one might be forced to accept to feed oneself differs in kind from the burden that the First and Fourteenth Amendments tolerate on the right to vote. Moreover, even with the available arsenal of conceivable precautions one could take to reduce risk of contracting the virus, many would be dissuaded from exercising their vote both because of the risk of illness and the efforts involved in mitigating that risk—especially those who
are elderly, immunocompromised, or otherwise at grave risk from the virus. See Dkt. 88-2 (Decl. of Dr. John B. Holbein) at 20. That substantial burden on the right to vote has not been justified by countervailing, demonstrated interests in the witness requirement. Indeed, the evidence in the record reflects that that the virus can be transmitted with more “ease” than was believed when this Court entered the first consent decree in this case. In addition to droplet transmission when an infected individual sneezes, coughs, and the like, Dkt. 103 (Joint Proposed Findings of Fact and Conclusions of Law) at ¶ 5 (citing CDC, How It Spreads, https://www.cdc.gov/coronavirus/2019-ncov/prevent-getting-sick/how-covid-spreads.html (last updated June 16, 2020)),and the parties agree that the evidence in the record demonstrates that the
virus can be “aerosolized,” meaning that contagious particles of the virus can remain airborne for periods of time waiting to be inhaled by an unwitting host. Dkt. 103 (Joint Proposed Findings of Fact and Conclusions of Law) at ¶ 6 (citing Dkt. 88-1 at ¶ 10); Dkt. 106(State Defendants Proposed Findings of Fact and Conclusions of Law). The First Circuit, in denying an application to stay the District of Rhode Island’s approval of a consent decree enjoining a similar witness signature requirement than the one at issue here, put the issue well: Could a determined and resourceful voter intent on voting manage to work around these impediments? Certainly. But it is also certain that the burdens are much more unusual and substantial than those that voters are generally expected to bear. Taking an unusual and in fact unnecessary chance with your life is a heavy burden to bear simply to vote. See Common Cause RI. v. Gorbea, No. 20-1753 , 2020 WL 4579367, at *2 (1st Cir. Aug. 7, 2020), aff’dNo. 20A28, 2020 WL 4680151 (U.S. Aug. 13, 2020). As discussed at length in this Court’s previous Memorandum Opinion, the witness signature requirement is not restrictive enough in that the record does not demonstrate that it is especially effective in preventing voter fraud on its own. League of Women Voters of Va., No. 6:20-cv-00024 (Dkt. 69), 2020 WL 2158249, at *9. The only new evidence that RPV has brought to contest this finding is the declaration of a former North Carolina election official who did not testify as to any particular experience with enforcing Virginia’s election scheme. Moreover, the
witness signature requirement in North Carolina is meaningfully distinct from Virginia’s in two ways. The first is that, for all elections before or after the year 2020, North Carolina law requires that an absentee voter must complete their ballot in the presence of two witnesses. The second, is that those witnesses are required “tosign the application and certificate as witnesses and to indicate those persons’ addresses. Failure to list a ZIP code does not invalidate the application and certificate.” N.C. Gen. Stat. § 163-231(a)(5). The ballot must then either be notarized or “the two persons in whose presence the voter marked that voter’s ballots to certify that the voter is the registered voter submitting the marked ballots.” N.C. Gen. Stat. § 163-231(a)(6). In response to the pandemic, for any election held in the year 2020, North Carolina has reduced the two-witness requirement down to one witness and, further, requires the witness to include on the ballot their
printed name and address. 2020 N.C. Sess. Laws 2020-17 (H.B. 1169) § 1.(a). In comparison, Virginia merely requires a single witness to mark the absentee ballot envelope with an illegible signature. No address required. Compare Va. Code § 24.2-707, with 1 Va. Admin. Code 20-70- 20(C)(10). Ms. Strach’s testimony states that witness signature requirements “add a level of formality to the process of absentee voting that I think is important to the integrity of the election.” Dkt. 99- 1 (Decl. of Kimberly W. Strach) at ¶ 20. Further, she states that even those requirements that do not require a legible signature or any witness verification measures [like Virginia’s] can be useful because state boards of elections have “little time to investigate any election irregularities and the
more tools at their disposal to do so, the better.” Id. at ¶ 21. The risk to some voters’ health is not outweighed by an additional “level of formality,” particularly when considering the constellation of other safeguards in place to prevent absentee voter fraud. See supra Subsection I.A.2. Because of the compromises the parties reached during settlement, experience of the counsel in this case, and the clear strength of the Plaintiffs’ case against the State Defendants, the Court concludes that the proposed partial settlement agreement is fair, adequate, and reasonable. See United States v. North Carolina, 180 F.3d at 581.
B. Proposed Consent Decree’s Relationship to the Public Interest The RPV argues that the public interest is served “by allowing for state control of its election mechanics by democratically accountable officials.” That may well be preferable, but when the Court is confronted with aproposed settlement that would remedy a probable, impending violation of federal law—a settlement reached and advanced by the Attorney General for the Commonwealth, an elected statewide official and thechief law enforcement officer of Virginia12— a federal court cannot abdicate its separate institutional role and simply set such an agreement asidein the hope that an electoral solution to the same problem will arrive in time.
12 SeeVa. House of Delegates v. Bethune-Hill, 139 S. Ct. 1945, 1951 (2019) (“Authority and responsibility for representing the State’s interests in civil litigation, Virginia law prescribes, rest exclusively with the State’s Attorney General”) (citing Va. Code § 2.2-507(A)). Just as the previous settlement did, the entry of this consent decree wouldpermit the State Defendants to avoid the likely unconstitutional application of a state law—at least with regard to theNovember3, 2020 elections. It is neitherunlawful,nor against the public interest. See Common Cause RI v. Gorbea, No. 1:20-cv-00318, 2020 WL 4365608, at *1, 5 (D.R.I. July 30, 2020); Common Cause Ind. v. Marion Cty. Election Bd., No. 1:17-cv-01388, 2018 WL 3770134, at *2
(S.D. Ind. Aug. 9, 2018) (holding that intervening party the State of Indiana’s objection to a consent decree governing early in-person voting that “was necessary to remedy a probable violation of federal law” was not compelling because it only argued that “it is not in the public interest for a federal court to enter, enforce, and monitor a consent decree that dictates the operation of state-run elections”). For these reasons, the Court concludes that the proposed consent decree is not illegal nor against the public interest.SeeUnited States v. North Carolina, 180 F.3d at 581.
C. Product of Collusion This Court begins with the presumption that settlement negotiations were conducted in good faith and that the agreement that resulted was not the product of fraud or collusion. See McCurley v. Flower Foods, Inc., No. 5:16-cv-00194, 2018 WL 6650138, at *2 (D.S.C. Sept. 10,2018);Funkhouser v. City of Portsmouth, No.2:13-cv-520, 2015 WL 12765639, at *3 (E.D. Va. May 14, 2015). Moreover, once any parties begin working toward a settlement, especially one that requires approval by a court, the relationship necessarily begins to align in a way that differs from an activelycontested,adversarial proceeding. That alignment only increases once the parties have actually reached a settlement and present it to a court for approval. But that
is not the same as fraud or collusion. The Court recognizes that the parties have, again, swiftly come to an agreement addressing an upcoming election in the days following the Plaintiffs’ second motion for a preliminary injunction. But the RPV has not brought forth any evidence that would permit the Court to overcome its presumption of good faith and fair dealing in the settlement negotiations between the Plaintiffs’ and State Defendants—instead, it relies on speculation and circumstantial evidence. For example, the RPV points to how the State Defendants have “so far failed to defend the
absentee ballot witness statute, and in fact they are selectively enforcing some witness signature requirements and not others.” Dkt. 99 at 41. The RPV also points to how the Plaintiffs and State Defendants have “almost immediately decided to enter into the Consent Decree” on a second occasion. Id. But such speculation is no substitute for evidence of collusion. See Common Cause RI, No. 1:20-cv-00318, 2020 WL 4365608, at *4 (D.R.I. July 30, 2020) (concluding that a consent decree enjoining a witness signature requirement during the COVID-19 pandemic was not collusive where “no evidence of collusion among the parties has been presented to this Court; in fact, the parties have represented that they engaged in good-faith negotiations in the crafting of the Consent Decree’s terms” despite the Republican Party intervenors’ arguments that the suit had
lacked the requisite adversarial vigor). It is no surprise to this Court that the parties were able to reach a settlement quickly for a second time, given this Court’s articulation that the Plaintiffs had pleaded a probable violation of federal law as to the witness signature requirement’s application to the June 2020 primaries. COVID-19 has resurged in the Commonwealth—if it ever truly receded—and this Court’s finding of such a probable violation presumably sent a signal to the State Defendants as to their chances of defending the requirement in court. The Court will not demand that litigants have blinders on to their chances of success. Rather, one would expect them to always be attuned to the likelihood of winning on the merits and, in turn, the possibility or even necessity of settlement. That a defendant in such a position would settle quickly is hardly evidence of collusion. Further, contrary to the RPV’s contention, the Plaintiffs and State Defendants’ failure to include other witness signature requirements that appear in Virginia’s election scheme within the scope of the consent decree also provides no evidence of collusion. If anything, the more limited
nature of the agreement demonstrates cuts the other way entirely. The RPV also argues that the parties do not have standing to seek their consent decree because there is no case or controversy between them, as required by the Constitution. This is because, as the RPV contends, Plaintiffs are “aligned” with the State Defendants. Dkt. 99 at 42. The RPV made the same argument—based on the same caselaw—in its opposition to the prior consent decree, Dkt. 58 at 13–14, and the Court rejected the argument. The RPV cites no intervening developments that would change the Court’s reasoning as it was set forth in the previous opinion. League of Women Voters of Va., No. 6:20-cv-00024 (Dkt. 69), 2020 WL 2158249, at *11. As then, the Court finds the cases that the RPV cited to be inapposite to the facts
presented in this case. For example, in Moore v. Charlotte-Mecklenburg Board of Education, 402 U.S. 47, 47–48 (1971), the Supreme Court stated that “at the hearing both parties argued to the three-judge court that the anti-busing law was constitutional and urged that order of the district court adopting the [plan] should be set aside.” Because the Court was “confronted with the anomaly that both litigants desire precisely the same result, namely a holding that the anti-busing statute is constitutional,” it held that there was no case or controversy as required under Article III of the U.S. Constitution. Id. In contrast, here, the parties have reached a settlement and have jointly requested a consent decree. They are, of course, aligned in their desire for the Court to approve the consent decree, but that is true of any motion for a consent decree. The RPV has not provided any persuasive reason why the Supreme Court’s order in Moore should undermine the framework for assessing a consent decree as set forth in United States v. North Carolina, 180 F.3d 574 (4th Cir. 1999). The fact remains that the State Defendants have not betrayed any indication that they will not enforce the witness signature requirement under color of state law if this consent decree is not
entered or Plaintiffs’ motion for a preliminary injunction is not granted. And, if the requirement were enforced, Plaintiffs argue that they would be harmed. That is the case or controversy before this Court, and the Court cannot identify anything in the record that demonstrates that this case or controversy has ceased to exist. Indeed, the litigation as to the remainder of Plaintiffs’ sought relief is ongoing, and the Plaintiffs and State Defendants still have legally cognizable interests at stake. See Genesis Healthcare Corp. v. Symczyk, 569 U.S. 66, 71 (2013) (“Article III, § 2, of the Constitution limits the jurisdiction of federal courts to ‘Cases’ and ‘Controversies,’ which restricts the authority of federal courts to resolving the legal rights of litigants in actual controversies . . . . In order to invoke federal-court jurisdiction, a plaintiff must demonstrate that he possesses a
legally cognizable interest, or personal stake, in the outcome of the action.” (internal quotations and citations omitted)). The RPV has not put forth sufficient evidence to overcome the presumption of good faith and fair dealing. But, even without such a presumption, the Court would be unable to find any basis for the conclusion that the parties colluded in bringing this suit and reaching their settlement. D. Burdens on Third-Party Obligations, Rights, or Duties As with the prior consent decree, the proposed consent decree does not bind the RPV to take or not to take any action. Local No. 93, Intern. Ass’n of Firefighters, AFL-CIO C.L.C. v. City of Cleveland, 478 U.S. 501, 529–30 (1986). It does not purport to resolve or otherwise extinguish any claims that the RPV, or any other party, might have. Id. Rather, just as before, this consent decree is limited to one of several verification requirements in place for absentee ballots in Virginia. It involves none of the RPV’s statutory functions or rights in providing for the nomination of its candidates. Va. Code § 24.2-508. And, as the Supreme Court stated just last week, a political party “lack[s] a cognizable interest in the State’s ability to ‘enforce its duly enacted’ laws.” Repub. Nat’l Comm., No. 20A28, 2020 WL 4680151,
at *1 (U.S. Aug. 13, 2020) (quoting Abbott v. Perez, 138 S. Ct. 2305, 2324 n.17 (2018)). The RPV, again, raises its concern that the consent agreement would endanger the security of the election. In support of its argument, it points to a declaration of a former North Carolina election official and argues that her testimony demonstrates that Virginia voters will be “disenfranchised” if the Court approves the consent decree. Dkt. 99 at 35 (quoting Dkt. 99-1, Decl. of Kimberly Strach, at ¶ 13). This quote is taken out of context—the declaration says no such thing. In full, it states, Voters in North Carolina’s Congressional District 9 were disenfranchised by [a fraudulent] absentee ballot scheme [in 2018] because some voters did not cast their own votes, and residents of the district went without congressional representation for a period of time while the [fraud] investigation concluded and before the new election was completed.
Dkt. 99-1 at ¶ 13. As they contend, “Ms. Strach’s affidavit clearly warns this Court that eliminating the witness signature requirement will hamper the ability of investigators to detect fraud, disenfranchise voters as a whole, and have a disparate impact on vulnerable populations.” Dkt. 99 at 36. The Court remains unconvinced. Ms. Strach’s testimony states that the witness signature requirements “add a level of formality to the process of absentee voting that I think is important to the integrity of the election.” Id. at ¶ 20. Further, she states that even those requirements that do not require a legible signature or any witness verification measures can be useful because state boards of elections have “little time to investigate any election irregularities and the more tools at their disposal to do so, the better.” Id. at ¶ 21. The Court’s reasoning as to whether the consent decree addresses a probable violation of federal law applies with equal force here: the strength of the witness signature requirement’s marginal utility as an anti-fraud measure is likely outweighed
by the burden it places on the right to vote for a substantial and discrete class of Virginia’s electorate. That a former North Carolina election official who lacks familiarity with enforcing a Virginia election regulation has testified that its utility is limited to its role as another tool in the arsenal and, specifically, one that adds a layer of “formality” to the absentee voting process does little to tip the scales. The value of her testimony in this case drops even further when weighed against the opinion of the Virginia State Board of Elections—the body that actually has experience in enforcing Virginia’s absentee ballot scheme and investigating Virginia election fraud. That body is before the Court now, requesting the approval of this proposed consent decree—indeed, as a
party to the suit, it crafted this settlement with the Plaintiffs. The Court finds more persuasive that body’s assessment of the value of this requirement as compared to testimony from a former election official of a different state, who has experience enforcing a different absentee ballot scheme with a different witness signature requirement.13 See Common Cause RI, 2020 WL 4579367, at *3 (1st Cir. Aug. 7, 2020). Indeed, the previous consent decree has set a new status
13 As already discussed, North Carolina’s standard witness signature requirement (for elections held outside the year 2020) requires two witness signatures, the printed address of those witnesses, and either notarization of the absentee ballot or a certification from the witnesses that the voter is the registered voter submitting the marked ballots. N.C. Gen. Stat. § 163-231(a)(6). quo in Virginia—one in which the witness signature requirement is not enforced during the pandemic. Cf. Republican Nat. Comm., 20A28, 2020 WL 4680151, at *1 (U.S. Aug. 13, 2020). Every indication before the Court is that the June primary was conducted without the witness signature requirement and without any corresponding increase in voter confusion or election fraud—the RPV has not provided any evidence to the contrary and no state official or
entity has come forth to intervene or file an amicus brief expressing otherwise. To change the status quo set by the June election by setting aside the consent decree would likely do more to confuse and deter voters. Id. (“Second, Rhode Island just conducted an election without any attestation requirement, in which 150,000 mail-in ballots were requested. So, the status quo (indeed the only experience) for most recent voters is that no witnesses are required. Instructions omitting the two-witness or notary requirement have been on the state's website since at least mid-July.”); Republican Nat. Comm., 20A28, 2020 WL 4680151, at *1 (U.S. Aug. 13, 2020) (“The status quo is one in which the challenged requirement has not been in effect, given the rules used in Rhode Island’s last election, and many Rhode Island voters may well hold that belief.”); see also
Democratic Nat’l Comm. v. Bostelmann, 140 S. Ct. 1205, 1207 (2020) (“This Court has repeatedly emphasized that lower federal courts should ordinarily not alter the election rules on the eve of an election.”) (citing Purcell v. Gonzalez, 549 U.S. 1 (2006)). As before, the Court has given the RPV, as an intervening party, its chance to be heard both on brief and at this Court’s Fairness Hearing. It has ably advanced its objections to the proposed consent decree. But it is not entitled to block the other parties to this suit from settling any claims in this action. See Local No. 93, 478 U.S. at 529; Sierra Club v. North Dakota, 868 F.3d 1062, 1067 (9th Cir. 2017). In any event, this Court has concluded that the proposed consent decree is fair, adequate and reasonable, not illegal or the product of collusion, and no argument or evidence presented by the RPV warrants a contrary conclusion. CONCLUSION For the foregoing reasons, the Court will approve the proposed settlement agreement, Dkt. 95-1. It finds that the agreement is fair, adequate, and reasonable and, further, it is not the product of collusion, illegal, or against the public interest. Lastly, the agreement does not purport to extinguish the claims of any third party, nor does it impose any obligations or duties on any third party. The Clerk of the Court is directed to send a copy of this Memorandum Opinion to all counsel of record. Entered this _2ist day of August, 2020. wen (aarttan fen’ _ SENIOR Nit HO STATES DISTRICT JUDGE