Mi Familia Vota v. Fontes
Opinion
1 WO 2 3 4 5 6 IN THE UNITED STATES DISTRICT COURT 7 FOR THE DISTRICT OF ARIZONA 8 9 Mi Familia Vota, et al., No. CV-21-01423-PHX-DWL 10 Plaintiffs, ORDER 11 v. 12 Katie Hobbs, et al., 13 Defendants. 14 15 INTRODUCTION
16 This action involves a challenge to two voting laws that were enacted by the Arizona 17 legislature following the 2020 election. The first is Senate Bill 1485, which provides that 18 voters who do not cast a mail-in ballot in two consecutive election cycles must be removed 19 from Arizona’s permanent early voting list. The second is Senate Bill 1003, which clarifies 20 that the deadline for a voter to attempt to “cure” a missing signature on an early ballot is 21 7:00 PM on election day. (Doc. 1 ¶ 1.)1 22 Plaintiffs in this action are four nonprofit groups, the Democratic Senatorial 23 Campaign Committee, and the Democratic Congressional Campaign Committee. 24 Defendants are Arizona Secretary of State Katie Hobbs, Arizona Attorney General Mark 25 26 1 The parties offer competing shorthand descriptions of the laws at issue. Plaintiffs 27 refer to the two challenged laws as the “Voter Purge Law” and the “Cure Period Law.”
(Doc. 1 ¶ 1.) The State refers to the two challenged laws as the “Periodic Voting 28 Requirement” and the “Poll-Close Deadline.” (Doc. 58 at 3.) The Court will simply refer to the challenged laws as S.B. 1485 and S.B. 1003.
1 Brnovich, the recorders from all 15 Arizona counties, the Republican National Committee, 2 and the National Republican Senatorial Committee. 3 Plaintiffs2 assert three claims. In Count One, Plaintiffs contend that S.B. 1485 and 4 S.B. 1003, “individually and collectively,” violate the First and Fourteenth Amendments 5 because they create an undue burden on the right to vote. (Id. ¶¶ 127-35.) In Count Two, 6 Plaintiffs contend that S.B. 1485 and S.B. 1003, “individually and collectively, violate the 7 Fourteenth and Fifteenth Amendments because they were adopted for the purpose of 8 denying voters of color full and equal access to the political process.” (Id. ¶¶ 136-41.) And 9 in Count Three, Plaintiffs contend that S.B. 1485 and S.B. 1003, “individually and 10 collectively, violate Section 2 of the Voting Rights Act because they were adopted for the 11 purpose of denying voters of color full and equal access to the political process.” (Id. 12 ¶¶ 142-45.) 13 Now pending before the Court is the State’s corrected consolidated motion to 14 dismiss under Rules 12(b)(1) and 12(b)(6) of the Federal Rules of Civil Procedure. (Doc. 15 76.) For the following reasons, the State’s motion is granted in part and denied in part. 16 BACKGROUND 17 I. Factual Background 18 A. Arizona’s History Of Discrimination 19 Plaintiff’s complaint includes a lengthy section entitled “Arizona’s History of 20 Discrimination and Voter Suppression.” (Doc. 1 ¶¶ 97-126.) The State does not, in 21 general, challenge the factual allegations appearing in that portion of the complaint for 22 purposes of the pending motion. Accordingly, the following facts are presumed true for 23 purposes of resolving the State’s motion. 24 … 25 … 26 2 27 Where there is no meaningful difference between the positions of the four nonprofit groups (“Plaintiffs”) and the Democratic Party entities (“Intervenor-Plaintiffs”), the Court 28 refers to them collectively as “Plaintiffs.” Additionally, the Court refers to Secretary Hobbs, Attorney General Brnovich, and the 15 county recorders collectively as “the State.”
Free access — add to your briefcase to read the full text and ask questions with AI
1 WO 2 3 4 5 6 IN THE UNITED STATES DISTRICT COURT 7 FOR THE DISTRICT OF ARIZONA 8 9 Mi Familia Vota, et al., No. CV-21-01423-PHX-DWL 10 Plaintiffs, ORDER 11 v. 12 Katie Hobbs, et al., 13 Defendants. 14 15 INTRODUCTION
16 This action involves a challenge to two voting laws that were enacted by the Arizona 17 legislature following the 2020 election. The first is Senate Bill 1485, which provides that 18 voters who do not cast a mail-in ballot in two consecutive election cycles must be removed 19 from Arizona’s permanent early voting list. The second is Senate Bill 1003, which clarifies 20 that the deadline for a voter to attempt to “cure” a missing signature on an early ballot is 21 7:00 PM on election day. (Doc. 1 ¶ 1.)1 22 Plaintiffs in this action are four nonprofit groups, the Democratic Senatorial 23 Campaign Committee, and the Democratic Congressional Campaign Committee. 24 Defendants are Arizona Secretary of State Katie Hobbs, Arizona Attorney General Mark 25 26 1 The parties offer competing shorthand descriptions of the laws at issue. Plaintiffs 27 refer to the two challenged laws as the “Voter Purge Law” and the “Cure Period Law.”
(Doc. 1 ¶ 1.) The State refers to the two challenged laws as the “Periodic Voting 28 Requirement” and the “Poll-Close Deadline.” (Doc. 58 at 3.) The Court will simply refer to the challenged laws as S.B. 1485 and S.B. 1003.
1 Brnovich, the recorders from all 15 Arizona counties, the Republican National Committee, 2 and the National Republican Senatorial Committee. 3 Plaintiffs2 assert three claims. In Count One, Plaintiffs contend that S.B. 1485 and 4 S.B. 1003, “individually and collectively,” violate the First and Fourteenth Amendments 5 because they create an undue burden on the right to vote. (Id. ¶¶ 127-35.) In Count Two, 6 Plaintiffs contend that S.B. 1485 and S.B. 1003, “individually and collectively, violate the 7 Fourteenth and Fifteenth Amendments because they were adopted for the purpose of 8 denying voters of color full and equal access to the political process.” (Id. ¶¶ 136-41.) And 9 in Count Three, Plaintiffs contend that S.B. 1485 and S.B. 1003, “individually and 10 collectively, violate Section 2 of the Voting Rights Act because they were adopted for the 11 purpose of denying voters of color full and equal access to the political process.” (Id. 12 ¶¶ 142-45.) 13 Now pending before the Court is the State’s corrected consolidated motion to 14 dismiss under Rules 12(b)(1) and 12(b)(6) of the Federal Rules of Civil Procedure. (Doc. 15 76.) For the following reasons, the State’s motion is granted in part and denied in part. 16 BACKGROUND 17 I. Factual Background 18 A. Arizona’s History Of Discrimination 19 Plaintiff’s complaint includes a lengthy section entitled “Arizona’s History of 20 Discrimination and Voter Suppression.” (Doc. 1 ¶¶ 97-126.) The State does not, in 21 general, challenge the factual allegations appearing in that portion of the complaint for 22 purposes of the pending motion. Accordingly, the following facts are presumed true for 23 purposes of resolving the State’s motion. 24 … 25 … 26 2 27 Where there is no meaningful difference between the positions of the four nonprofit groups (“Plaintiffs”) and the Democratic Party entities (“Intervenor-Plaintiffs”), the Court 28 refers to them collectively as “Plaintiffs.” Additionally, the Court refers to Secretary Hobbs, Attorney General Brnovich, and the 15 county recorders collectively as “the State.”
1 1. Voting Discrimination 2 The United States acquired present-day Arizona from Mexico under the 1848 Treaty 3 of Guadalupe Hidalgo and the 1853 Gadsden Purchase. (Doc. 1 ¶ 98.) Indigenous nations 4 had no authority over the transfer of their lands, and by the 1880s, Native Americans were 5 largely confined to reservations, often through violent means. (Id. ¶¶ 98-99.) In 1928, the 6 Arizona Supreme Court held that Native Americans were ineligible to vote in Arizona, 7 despite a 1924 federal law that declared Native Americans to be citizens of the United 8 States and their state of residence. (Id. ¶ 102.) 9 In 1909, Arizona’s territorial legislature adopted an English language literacy test 10 as a prerequisite to voter registration. (Id. ¶ 100.) When Congress passed a law the next 11 year that authorized Arizona to draft a state constitution as a prelude to statehood, the law 12 prohibited Arizona from using the literacy test as an eligibility requirement to vote on the 13 proposed constitution. (Id.) Once Arizona achieved statehood in 1912, the legislature re- 14 imposed an English literacy test for voting, which was not repealed until 1972. (Id. ¶ 101.) 15 Although the Arizona Supreme Court recognized Native Americans’ right to vote 16 in 1948, Arizona’s literacy test disenfranchised 80-90% of Native Americans that year and 17 still disenfranchised about half of Native Americans by the 1960s. (Id. ¶¶ 102-03.) County 18 officials also used the literacy test to prevent eligible Latino and black citizens from voting. 19 (Id.) 20 In 1964, Arizona Republicans strategically challenged voters’ right to vote, 21 particularly the right of voters of color, at the polls. (Id. ¶ 105.) 22 In 1970, the Arizona legislature purged the voter rolls and required all citizens to 23 re-register to vote. (Id. ¶ 106.) Many Latino voters did not realize they needed to re- 24 register, and in the 1970 election Democrat Raul Castro narrowly lost the governor’s race 25 despite receiving 90% of the Latino vote. (Id.) 26 In 1975, Congress amended the Voting Rights Act and made all Arizona 27 jurisdictions subject to “preclearance.” (Id. ¶ 107.) In the 1980s and 1990s, the United 28 States Department of Justice issued 17 preclearance objections to proposed changes in
1 Arizona election procedures, concluding that the changes had the purpose or effect of 2 discriminating against Arizona’s Native American or Latino voters. (Id. ¶ 108.) 3 During the presidential primary election in 2016, voters in Maricopa County—a 4 county in which more than 45% of residents are people of color—waited in line for up to 5 five hours to vote after county officials cut polling locations by 85% as compared to the 6 2008 presidential primary. (Id. ¶ 109.) “In Phoenix, where a majority of voters are people 7 of color, there was one polling location for every 108,000 residents, while in Cave 8 Creek/Carefree, a predominantly white community, there was one polling location for 9 every 8,500 residents and in Peoria, also predominantly white, there was one polling 10 location for every 54,000 residents.” (Id. ¶ 110.) 11 In 2012, the official Spanish-language pamphlet in Maricopa County stated that the 12 November 6 election would be held on November 8. (Id. ¶ 111.) Four years later, in 2016, 13 Spanish-language ballots in Maricopa County incorrectly translated a ballot proposition. 14 (Id.) 15 2. Racial Discrimination 16 Arizona’s educational system was formally segregated by race for decades. (Id. 17 ¶ 113.) Arizona mandated English-only education in public schools as early as 1919, and 18 in 2000, Arizona banned bilingual education. (Id. ¶ 114.) Arizona also has a history of 19 failing to adequately fund its English Language Learning Program. (Id.) After World War 20 II, Phoenix placed Latino veterans in housing units separated from white Arizonans. (Id. 21 ¶ 116.) Latinos were not permitted to use the same theaters, swimming pools, parks, or 22 restaurants as whites. (Id.) 23 In Arizona, 34% of Native Americans, 19% of black people, and 19% of Hispanics 24 live below the poverty line, compared to only 9% of white people. (Id. ¶ 117.) The 25 unemployment rate is higher for Latinos, Native Americans, and black people than white 26 people. (Id.) White Arizonans are more likely than Latino, Native American, and black 27 Arizonans to graduate high school and are nearly three times more likely to have a 28 bachelor’s degree than Latino and Native American Arizonans. (Id. ¶ 115.) In 2017, home
1 ownership by people of color was significantly lower than by white Arizonans. (Id. ¶ 118.) 2 Also in 2017, Latinos, Native Americans, and black people in Arizona ranked below white 3 people in “relative healthiness.” (Id. ¶ 119.) Native Americans and black Arizonans are 4 more likely than white Arizonans to die before the age of 65. (Id.) Latinos, Native 5 Americans, and black people are overrepresented in Arizona jails. (Id.) 6 3. Voting Consequences 7 Although the voter turnout rate among Native Americans nationwide is anywhere 8 from one to ten percentage points lower than that of other groups, turnout among Arizona’s 9 Native American population is even lower—15 to 20 percentage points lower than the 10 statewide voter turnout in 2020 and 2016, respectively. (Id. ¶ 122.) 11 In 2012, 40.4% of Arizona’s Latino citizens and 46% of Arizona’s black citizens 12 voted, as compared to 62.4% of Arizona’s non-Hispanic white citizens. (Id. ¶ 123.) In 13 2016, nearly 75% of white citizens were registered to vote and more than 68% voted. (Id. 14 ¶ 124.) That same year, only 57% of Latino citizens were registered to vote and only 47.4% 15 voted. (Id.) Additionally, only 50.9% of black citizens voted in 2016. (Id.) In both the 16 2012 and 2016 elections, the presidential candidate preferred by minority Arizonans did 17 not win the state’s general election. (Id. ¶ 125.) 18 Between 1990 and 2020, Arizona’s Latino population nearly doubled from 18.8% 19 of the general population to 30.7%. (Id. ¶ 50.) Arizona’s black population increased 20 during the same time span from 3% to 6.2%, and Arizona’s Asian population increased 21 from 1.4% to 4.9%. (Id.) 22 In the 2020 general election, 59% of voters of color voted for the Biden-Harris 23 ticket, including 61% of Latinos and 80% of voters in the Navajo Nation and Hopi 24 Reservation. (Id. ¶ 49.) 25 B. Arizona’s Early Voting System 26 Since 1991, all eligible Arizona voters have been able to participate in early voting. 27 (Id. ¶ 42.) Today, a voter who wishes to vote early may request a mailed ballot, so long as
1 the request is received no later than 5:00 PM on the eleventh day before the election. (Id. 2 ¶ 43.) 3 In 2007, Arizona created the permanent early voting list (“PEVL”). (Id. ¶ 42.) 4 Voters who join the PEVL are automatically sent an early ballot no later than the first day 5 of the 27-day early voting period. (Id. ¶ 43.) Voters may return their early ballots by mail 6 postage-free; in person at a polling place, voting center, or election official’s office; or by 7 putting the ballot in a dropbox if a dropbox is provided by the voter’s county. (Id.) 8 Arizona law has long required counties to follow certain procedures to verify voters’ 9 addresses and desire to remain on the PEVL and to ensure that voters whose registrations 10 are moved to “inactive” status are removed from the PEVL. (Doc. 55 ¶ 46.) Before each 11 election, county recorders must mail a notice to all voters on the PEVL that, among other 12 things, provides the voter a means to update his or her address or decline to receive a mail 13 ballot for the upcoming election. (Id.) If that notice is returned as undeliverable, and the 14 voter does not respond to a follow-up notice within 35 days, the voter’s registration status 15 will be changed from active to inactive. (Id.) 16 Around 60% of Arizona voters voted by mail in the 2010 general election. (Doc. 1 17 ¶ 44.) This figure rose to 66% in 2012 and 80% in 2016. (Id.) In 2020, 2.5 million 18 Arizonans voted by mail or early in-person, which was 88% of all voters and over 900,000 19 more early votes than were cast in 2016. (Doc. 1 ¶ 45; Doc. 55 ¶ 4.) Between 2012 and 20 2020, most of the voters on the PEVL voted, and they did so at a higher rate than voters 21 who were not on the PEVL. (Doc. 1 ¶ 70.) As of 2020, at least 75% of all registered voters 22 are on the PEVL. (Doc. 55 ¶ 47.) In contrast, around 21% of voters nationally cast their 23 ballot by mail. (Id.) 24 In 2020, voters of color were disproportionately likely to be new to voting and new 25 to voting by mail. (Doc. 1 ¶ 46.) When comparing the population of new early voters to 26 the population of all registered voters, voters of color were a larger percentage of new early 27 voters than they were of all registered voters. (Id.)
1 Finally, Plaintiffs allege, and the State does not dispute for purposes of the motion 2 to dismiss, that “[i]n the nearly three decades it has existed, there have been no examples 3 of widespread election fraud in connection with Arizona’s vote-by-mail system.” (Id. 4 ¶ 47.) 5 C. The 2020 Election And The Legislative Response 6 In the 2020 general election, over 3.4 million Arizonans voted, which was nearly 7 80% of registered Arizona voters. (Id. ¶ 48.) 59% of Arizona’s total voting age population 8 voted, which was nearly 10% higher than in 2016. (Id.) 9 The 80% turnout in 2020 was about 5% higher than turnout in 2012 and 2016, and 10 turnout increased most noticeably in areas heavily populated by people of color. (Doc. 55 11 ¶ 3.) Several precincts in South Phoenix, home to large numbers of black and Latino 12 residents, saw increases of about 10%, and precincts in Arizona’s Native American 13 reservations saw increases of about 12-13%. (Id.) However, minority turnout remained 14 low compared to statewide turnout—for instance, turnout on the reservations was around 15 65%. (Id.) 16 Plaintiffs allege that Arizona’s governor, local and national election officials, and 17 the state and federal judiciary rejected the notion that the 2020 election was affected by 18 widespread voter fraud. (Doc. 1 ¶¶ 51-57.) Plaintiffs further allege that election audits 19 produced no evidence of fraud and were alleged to be motivated by conspiracy theories 20 (id. ¶¶ 57-63); that multiple lawsuits alleging fraud were dismissed for lack of evidence 21 (Doc. 55 ¶¶ 66-67); that no Arizona legislator has identified any instance of voter fraud in 22 connection with a mail-in ballot (Doc. 1 ¶ 66); and that “there is no evidence of widespread 23 voter fraud that undermined the integrity of Arizona’s 2020 election” (id.). 24 After the 2020 election, legislators across the country introduced more than 400 bills 25 (a dramatic increase from prior years) that would make it harder to register to vote, remain 26 on the voter rolls, or cast a ballot. (Id. ¶ 65.) The Arizona legislature proposed S.B. 1485 27 and S.B. 1003 (collectively, “the challenged legislation”). During debates, legislators 28 opposed to the challenged legislation argued it would result in fewer citizens of color
1 voting. (Id. ¶ 67.) Democratic Representative Reginald Bolding, who is black, stated that 2 S.B. 1485 would make it harder for “independent voters, seniors, Native Americans, Black, 3 brown and low income people to vote.” (Doc. 55 ¶ 114.) Republican Representative Travis 4 Grantham replied: “I feel personally that motives were [attributed to] members, including 5 myself with regards to colored people, Black people, whatever people this individual wants 6 to single out and their ability to vote . . . . I think he should be sat down and he shouldn’t 7 be allowed to speak.” (Id.) 8 In March 2021, Republican Representative John Kavanagh made a statement 9 regarding the challenged legislation. Although only a portion of the statement appears in 10 the body of the complaint, a citation to a news article that purports to contain the entire 11 statement appears in an accompanying footnote. (Doc. 1 ¶ 67 & n.17.) As discussed in 12 more detail below, the State argues in its motion papers that the context arising from the 13 fuller statement is important when evaluating the statement (Doc. 76 at 14) and Plaintiffs 14 do not dispute the accuracy of the State’s description of the fuller statement—instead, 15 Plaintiffs only dispute the relevance of the fuller statement at this stage of the proceedings 16 (Doc 99 at 11-12). In any event, because the news article containing Representative 17 Kavanagh’s fuller statement is incorporated by reference into the complaint (by virtue of 18 Plaintiffs’ citation to it), the fuller version of the statement is properly part of the record 19 for purposes of the current dispute.3 The fuller version, as derived from the cited news 20 article, is as follows: 21 Democrats value as many people as possible voting, and they’re willing to risk fraud. Republicans are more concerned about fraud, so we don’t mind 22 putting security measures in that won’t let everybody vote—but everybody 23 shouldn’t be voting. 24 ***
25 Not everybody wants to vote, and if somebody is uninterested in voting, that probably means that they’re totally uninformed on the issues. Quantity is 26 important, but we have to look at the quality of votes, as well. 27 3 United States v. Ritchie, 342 F.3d 903, 908 (9th Cir. 2003) (“A court may . . .
28 consider . . . documents incorporated by reference in the complaint . . . without converting the motion to dismiss into a motion for summary judgment.”).
1 Timothy Bella, A GOP Lawmaker Says the ‘Quality’ of a Vote Matters. Critics Say That’s 2 ‘Straight out of Jim Crow’, Wash. Post (Mar. 13, 2021) 3 https://www.washingtonpost.com/politics/2021/03/13/arizona-quality-votes-kavanagh/. 4 D. S.B. 1485 5 Under S.B. 1485, “[a] voter’s failure to vote an early ballot once received does not 6 constitute grounds to remove the voter from the [PEVL], except that a county recorder shall 7 remove a voter from the [PEVL] if . . . [t]he voter fails to vote using an early ballot in all 8 of the following elections for two consecutive election cycles: (a) A regular primary and 9 regular general election for which there was a federal race on the ballot[; and] (b) A city or 10 town candidate primary or first election and a city or town candidate second, general or 11 runoff election.” A.R.S. § 16-544(K)(2). Once a voter has failed to vote using an early 12 ballot in the specified elections for two consecutive election cycles, “the county recorder 13 or other officer in charge of elections shall send a notice . . . [that] shall inform the voter 14 that if the voter wishes to remain on the [PEVL], the voter shall do both of the following 15 with the notice received: 1. Confirm in writing the voter’s desire to remain on the active 16 early voting list[; and] 2. Return the completed notice to the county recorder or other officer 17 in charge of elections within ninety days after the notice is sent to the voter.” Id. § 16- 18 544(L). “If a voter receives a notice as prescribed by subsection L of this section and the 19 voter fails to respond within the ninety-day period, the county recorder or other officer in 20 charge of elections shall remove the voter’s name from the active early voting list.” Id. 21 § 16-544(M). 22 A voter who is removed from the PEVL may still vote in person. (Doc. 1 ¶ 83.) 23 However, voting in person will not prevent the voter from being removed from the PEVL. 24 (Doc. 55 ¶ 92.) 25 S.B. 1485 requires counties to implement a new system to track and mail additional 26 notices, which will impose further and new costs upon the counties. (Id. ¶ 104.) Between 27 125,000 and 150,000 voters will be removed from the PEVL after S.B. 1485 takes effect. 28 (Doc. 1 ¶ 76.) If it had been enacted in 2019 (and in effect in 2020), 126,000 voters who
1 ultimately voted would have been removed from the PEVL. (Id.) If enacted in 2015 or 2 2017 (before the 2016 and 2018 elections, respectively), more than 200,000 voters would 3 have been removed from the PEVL. (Id.) 4 White Arizonans account for 71% of all registered voters but only 54% of removed 5 voters will be white. (Id. ¶ 77.) Black and Hispanic Americans are more likely to be 6 intermittent or rare voters as compared to white voters, and low-income Americans are 7 similarly more likely to be intermittent or rare voters. (Id. ¶ 75.) Latinos are 19% of 8 registered voters but will account for 33% of removed voters, blacks are 4% of registered 9 voters but will account for 5% of removed voters, and Native Americans are 0.9% of 10 registered voters but will account for 1.3% of removed voters. (Id. ¶ 77.) Of the eight 11 legislative districts in Arizona with the highest number of voters likely to be removed from 12 the PEVL, seven are majority-minority districts. (Id.) 13 Voters who lack residential mail service will find it challenging to receive and return 14 the removal notice in the 90-day window they are given. (Id. ¶ 79.) Voters who need a 15 translated notice, who are disproportionately voters of color, will struggle to understand 16 and respond to the notice. (Id. ¶ 80.) Voters who have limited funds to purchase postage, 17 limited mobility, or otherwise have limited access to a post or election office will struggle 18 to respond to removal notices in writing. (Id. ¶ 81.) Elderly voters and voters experiencing 19 poverty are more likely to be removed because they are more likely to vote intermittently. 20 (Doc. 55 ¶ 96.) Young voters, such as college students, move frequently and may not 21 receive removal notices. (Id. ¶ 98.) Finally, the removal process is often based on error- 22 ridden lists and there is a substantial risk of manipulation. (Doc. 1 ¶ 74.) 23 Although a voter who is removed from the PEVL may still vote in person, young 24 voters are more likely to be outside the county in which they would be permitted to vote in 25 person by the time they realize they will not be receiving a mailed ballot. (Doc. 55 ¶ 98.) 26 And even if the voter is within the correct county, Arizona has had the “most widespread 27 reduction” in polling places of any state over the last decade—the state now has 320 fewer 28 polling places than it did in 2012. (Id. ¶ 48.) At least one in five people over the age of 70
1 do not drive. (Doc. 1 ¶ 83.) People of color are less likely to have reliable access to a 2 vehicle and are more likely to have a job with inflexible scheduling, making it difficult to 3 travel to vote in person. (Id.) Voters who live on the state’s tribal lands are likely to face 4 transportation accessibility issues and long distances to travel to a polling place. (Doc. 55 5 ¶ 97.) 6 E. S.B. 1003 7 Under Arizona law, early vote ballots must include a signed affidavit. (Doc. 1 8 ¶ 85). S.B. 1003 provides that if an early ballot arrives without a signed affidavit, the voter 9 must sign the affidavit by 7:00 PM on election day for the ballot to be counted. (Id. ¶ 86.) 10 In contrast, if an early ballot arrives with a signature that does not appear to match voter 11 registration records, a 2019 law allows the voter to cure the mismatched signature up to 12 five days after a federal election. (Id. ¶ 85.) 13 Before the 2020 election, Defendant Hobbs wanted to issue guidance in Arizona’s 14 Election Procedures Manual (“EPM”) that would allow missing and mismatched signatures 15 to be cured on the same timeline—within five days of the election. (Id. ¶ 85 n.21.) 16 Defendant Hobbs has stated that there is no reason to distinguish between unsigned and 17 mismatched ballots and that Arizona could easily allow unsigned ballots to be cured during 18 a five-day post-election cure period because Arizona already allows such a cure period for 19 other voter identification issues. (Id. ¶ 96.) Arizona is the only state in the country that 20 imposes an inconsistent cure period for voters. (Doc. 55 ¶ 55.) Defendant Brnovich 21 objected to the proposed guidance, 4 so Defendant Hobbs issued revised guidance that 22 missing signatures would have to be cured by 7:00 PM on election day. (Doc. 1 ¶ 85 n.21.) 23 Even if an unsigned ballot is received well before election day, S.B. 1003 requires 24 only that election officials make “reasonable efforts” to contact the voter and allow the 25 voter to sign the ballot before 7:00 PM on election day. (Id. ¶ 88.) The law does not 26 specify what reasonable efforts must be made. (Id.) Nor is there a provision in the law 27 4 28 Arizona law requires the Attorney General to approve the draft EPM before it goes into effect. (Doc. 55 ¶ 54.)
1 that provides alternatives for voters who cannot reasonably travel to the place where their 2 ballots are held. (Id. ¶ 89.) 3 Native American voters already struggle to access polling places and election 4 offices, which would slow the process by which they receive notice of signature 5 deficiencies. (Id. ¶ 91.) Black and Latino voters often live in neighborhoods with limited 6 access to public transportation and few election offices, and people of color are nearly twice 7 as likely to lack access to a car as white Americans. (Id. ¶ 92.) In Maricopa County, voters 8 in neighborhoods with high concentrations of black and Latino voters would have to travel 9 two hours by public transportation—each way—to provide a missing signature, or the 10 equivalent of between $7 and $19.50 in taxi fare. (Id.) In Yuma County, neighborhoods 11 with high concentrations of black and Latino voters are between 25 and 90 minutes away 12 from election offices by car, or the equivalent of a $75 taxi ride. (Id.) No provision of S.B. 13 1003 requires that notice be given in a language spoken by the voter, and voters with a 14 language barrier are more likely to violate the signature requirement, less likely to be aware 15 of the need to cure deficiencies, and more likely to fail to successfully communicate with 16 election officials. (Id. ¶ 93.) Disabled voters often struggle to access public transportation, 17 and voters who cannot physically provide a valid signature risk having their ballots 18 discarded without sufficient time to cure the ballot. (Id. ¶ 94.) 19 F. Plaintiffs’ Organizing Efforts 20 Plaintiff Living United for Change in Arizona (“LUCHA”) is a nonprofit 21 organization that organizes voter registration drives and educates voters by, for example, 22 encouraging them to register for Arizona’s PEVL. (Id. ¶ 9.) Up to 80% of LUCHA’s 23 93,000 members are early voters, and LUCHA’s membership includes people of color, 24 students, the elderly, and the economically disadvantaged. (Id. ¶ 11.) 25 As a result of the challenged legislation, LUCHA must “divert money, personnel, 26 time and resources away from other activities” to ensure that voters, “particularly of color 27 and those who are low income, can navigate the restrictions” created by the challenged 28 legislation. (Id. ¶ 10.) These diversions have already occurred and will continue to occur.
1 (Id.) LUCHA must now train volunteers about the challenged legislation, create voter 2 education campaigns to combat misinformation about the challenged legislation, and 3 dedicate more support to ballot-curing efforts. (Id.) 4 Plaintiff League of Conservation Voters (“Chispa AZ”), a network of around 20,000 5 members and volunteers, aims to increase political participation among Latinos and low- 6 income communities of color in Arizona. (Id. ¶ 12.) Chispa AZ performs similar services 7 as LUCHA and those services will be diverted by the challenged legislation, thus requiring 8 more resources overall, in a similar way. (Id. ¶¶ 12-13.) 9 Plaintiff Mi Familia Vota (“MFV”) is a national nonprofit organization dedicated to 10 uniting Latino, immigrant, and allied communities to promote social and economic justice. 11 (Id. ¶ 15.) MFV has operations in six states, including Arizona, where it is headquartered. 12 (Id.) MFV has 14,000 members in Arizona. (Id.) MFV performs similar services as 13 LUCHA and those services will be diverted by the challenged legislation, thus requiring 14 more resources overall, in a similar way. (Id. ¶¶ 16-17.) During a voter registration drive 15 performed before the 2020 election, more than 90% of individuals registered by MFV also 16 signed up for the PEVL. (Id. ¶ 16.) 17 Plaintiff Arizona Coalition for Change (“AZC4C”) performs similar services as 18 LUCHA and those services will be diverted by the challenged legislation, thus requiring 19 more resources overall, in a similar way. (Id. ¶¶ 19-20.) 20 II. Relevant Procedural Background 21 On August 17, 2021, Plaintiffs filed the complaint. (Doc. 1.) 22 On September 2, 2021, the Republican Party committees moved to intervene. (Doc. 23 28.) 24 On September 24, 2021, the Democratic Party committees moved to intervene. 25 (Doc. 50.) 26 On October 4, 2021, the Court granted the motions to intervene. (Doc. 53.) That 27 same day, Intervenor-Plaintiffs filed a complaint. (Doc. 55.)
1 On November 24, 2021, the State filed the motion now pending before the Court— 2 a corrected, consolidated motion to dismiss both complaints under Rules 12(b)(1) and 3 12(b)(6). (Doc. 76.) 4 On November 26, 2021, Intervenor-Defendants filed a joinder in the State’s motion. 5 (Doc. 77.) 6 On November 30, 2021, the United States filed a statement of interest. (Doc. 78.) 7 On December 12, 2021, the State filed a notice regarding the Ninth Circuit’s 8 decision in Arizona Democratic Party v. Hobbs, 18 F.4th 1179 (9th Cir. 2021). (Doc. 83.) 9 On January 4, 2022, Plaintiffs and Intervenor-Plaintiffs filed responses to the State’s 10 motion to dismiss. (Docs. 99, 100.) 11 On February 16, 2022, the State filed a consolidated reply in support of its motion 12 to dismiss. (Doc. 118.) 13 On March 14, 2022, Plaintiffs filed a notice that they would not be seeking a 14 preliminary injunction. (Doc. 123.) 15 On May 16, 2022, the Court issued a tentative ruling. (Doc. 144.) 16 On June 7, 2022, the Court heard oral argument. (Doc. 149.) 17 DISCUSSION 18 I. Legal Standard 19 Rule 12(b)(1) of the Federal Rules of Civil Procedure provides that a defendant may 20 move to dismiss an action for “lack of subject-matter jurisdiction.” Courts “have an 21 independent obligation to determine whether subject-matter jurisdiction exists.” Arbaugh 22 v. Y&H Corp., 546 U.S. 500, 514 (2006). See also Fed. R. Civ. P. 12(h)(3) (“If the court 23 determines at any time that it lacks subject-matter jurisdiction, the court must dismiss the 24 action.”). “Under Rule 12(b)(1), a defendant may challenge the plaintiff’s jurisdictional 25 allegations in one of two ways. A ‘facial’ attack accepts the truth of the plaintiff’s 26 allegations but asserts that they are insufficient on their face to invoke federal 27 jurisdiction.” Leite v. Crane Co., 749 F.3d 1117, 1121 (9th Cir. 2014) (citation omitted). 28 “A ‘factual’ attack, by contrast, contests the truth of the plaintiff’s factual allegations,
1 usually by introducing evidence outside the pleadings.” Id. The plaintiff bears the burden 2 of establishing that subject-matter jurisdiction exists. Kokkonen v. Guardian Life Ins. Co. 3 of Am., 511 U.S. 375, 377 (1994). 4 To survive a motion to dismiss under Rule 12(b)(6), “a party must allege ‘sufficient 5 factual matter, accepted as true, to state a claim to relief that is plausible on its face.’” In 6 re Fitness Holdings Int’l, Inc., 714 F.3d 1141, 1144 (9th Cir. 2013) (quoting Ashcroft v. 7 Iqbal, 556 U.S. 662, 678 (2009)). “A claim has facial plausibility when the plaintiff pleads 8 factual content that allows the court to draw the reasonable inference that the defendant is 9 liable for the misconduct alleged.” Id. (quoting Iqbal, 556 U.S. at 678). “[A]ll well- 10 pleaded allegations of material fact in the complaint are accepted as true and are construed 11 in the light most favorable to the non-moving party.” Id. at 1144-45 (internal quotation 12 marks omitted). However, the court need not accept legal conclusions couched as factual 13 allegations. Iqbal, 556 U.S. at 679-80. Moreover, “[t]hreadbare recitals of the elements of 14 a cause of action, supported by mere conclusory statements, do not suffice.” Id. at 679. 15 The Court also may dismiss due to “a lack of a cognizable theory.” Mollett v. Netflix, Inc., 16 795 F.3d 1062, 1065 (9th Cir. 2015) (citation omitted). 17 II. Analysis 18 A. Facial Versus As-Applied Challenge 19 “A facial challenge is a claim that the legislature has violated the Constitution, while 20 an as-applied challenge is a claim directed at the execution of the law.” Young v. Hawaii, 21 992 F.3d 765, 779 (9th Cir. 2021). 22 Throughout its briefing, the State argues that Plaintiffs “have not pled a viable facial 23 claim.” (Doc. 118 at 17. See also Doc. 76 at 10, 21-22.) In a nutshell, the State contends 24 that (1) Plaintiffs cannot argue the challenged legislation is “particularly burdensome as 25 applied to particular groups” because their claim is facial, not as-applied, and “they have 26 not joined any such affected voters” as plaintiffs (Doc. 118 at 12); and (2) any facial claim 27 is necessarily invalid because there are obvious scenarios in which application of the 28 challenged legislation would not raise constitutional concerns (id. at 17). In response,
1 Plaintiffs accuse the State of “fundamentally misunderstand[ing]” how the Supreme Court 2 has considered “a facial versus as-applied challenge in this context” and argue that “[t]he 3 distinction between facial and as-applied challenges is irrelevant at the pleading stage.” 4 (Doc. 99 at 18-20 & n.12.) 5 The State is not entitled to dismissal based on its arguments related to facial versus 6 as-applied challenges. As an initial matter, the Supreme Court has emphasized that “the 7 distinction between facial and as-applied challenges is not so well defined that it has some 8 automatic effect or that it must always control the pleadings and disposition in every case 9 involving a constitutional challenge. The distinction . . . goes to the breadth of the remedy 10 employed by the Court, not what must be pleaded in a complaint.” Citizens United v. Fed. 11 Election Comm’n, 558 U.S. 310, 331 (2010). It is difficult to reconcile the State’s 12 position—that Plaintiffs’ claims must be dismissed at the pleading stage due to Plaintiffs’ 13 imprecision in defining the nature of their challenge—with this principle. 14 At any rate, Plaintiffs’ argument that the challenged legislation creates burdens 15 disproportionately borne by particular groups does not necessarily mean they are raising 16 an as-applied challenge. As discussed in more detail below, Plaintiffs contend those 17 uneven burdens are evidence of a constitutional violation (the legislature’s discriminatory 18 motive or an undue burden on the right to vote) that infects the legislation itself. The Ninth 19 Circuit has suggested that such claims may be raised as part of a facial challenge to a voting 20 law. Pub. Integrity All., Inc. v. City of Tucson, 836 F.3d 1019, 1024 n.2 (9th Cir. 2016) 21 (“Under Burdick, courts are to assess the ‘character and magnitude’ of the asserted burden, 22 the proven strength of the state’s interest, and whether the extent of the burden is 23 ‘necessary’ given the strength of that interest, so as to ferret out and reject unconstitutional 24 restrictions. . . . [I]n so doing, courts may consider not only a given law’s impact on the 25 electorate in general, but also its impact on subgroups, for whom the burden, when 26 considered in context, may be more severe.”) (citations omitted). Proceeding under this
1 theory may make it more difficult to prevail on a facial challenge, 5 but it does not 2 necessarily transform the challenge into an as-applied challenge. 3 The State also fails to cite any cases suggesting that Plaintiffs must join individual 4 voters as parties before raising the type of challenges being raised here. Nor would such a 5 rule make sense. As alleged in the complaint, Plaintiffs are organizations dedicated to 6 serving the groups that are affected in disproportionate ways by the challenged legislation 7 and Plaintiffs will suffer various injuries due to those disproportionate burdens. The State 8 does not challenge the sufficiency of these allegations for purposes of standing. It is 9 therefore unclear why Plaintiffs would need to formally join individual voters as parties as 10 a prerequisite to raising the type of challenges they seek to raise here. 11 United States v. Salerno, 481 U.S. 739 (1987), does not compel a different 12 conclusion. The State cites Salerno for the proposition that those who bring a facial claim 13 against legislation must allege that “no set of circumstances exists under which the Act[s] 14 would be valid.” (Doc. 76 at 10.) The State argues that, for example, a voter who has 15 never used his PEVL privileges since joining the list in 2007 could be removed from the 16 PEVL without violating the Constitution. (Id. at 22.) But even assuming the State is 17 correct, this would at most affect the scope of relief that Plaintiffs could obtain at the 18 conclusion of this action. Cf. Frank v. Walker, 819 F.3d 384, 386-87 (7th Cir. 2016) (“The 19 district court had held in 2014 that, because some voters face undue difficulties in obtaining 20 acceptable photo IDs, Wisconsin could not require any voter to present a photo ID. . . . We 21 reversed that injunction . . . [because] an across-the-board injunction would be improper 22 because the application of the statute to the vast majority of Indiana voters is amply 23 justified . . . . The argument plaintiffs now present is different. Instead of saying that 24 inconvenience for some voters means that no one needs photo ID, plaintiffs contend that 25 26 5 In Crawford v. Marion County Election Bd., 553 U.S. 181 (2008), the Supreme 27 Court rejected a challenge to a voter-identification statute in part because “petitioners have not demonstrated that the proper remedy—even assuming an unjustified burden on some 28 voters—would be to invalidate the entire statute.” Id. at 202-03 (citations and internal quotation marks omitted).
1 high hurdles for some persons eligible to vote entitle those particular persons to relief. 2 Plaintiffs’ approach is potentially sound . . . .”) (cleaned up). 3 B. Count One 4 In Count One of the complaint, Plaintiffs argue that S.B. 1485 and S.B. 1003, 5 individually and collectively, are unconstitutional because they create an undue burden on 6 the right to vote. (Doc. 1 ¶¶ 127-35.) The State argues (Doc. 76 at 15-16), and Plaintiffs 7 seem to agree (Doc. 99 at 13), that the Anderson-Burdick framework governs such a claim.6 8 The Anderson-Burdick framework7 “prescribes a sliding-scale level of scrutiny for 9 evaluating governmental actions that burden the right to vote. At one end of the spectrum, 10 severe restrictions must be narrowly drawn to advance a state interest of compelling 11 importance. At the other end of the spectrum, important state regulatory interests are 12 generally sufficient to justify reasonable, nondiscriminatory restrictions. Thus, the 13 burdening of the right to vote always triggers a higher level of scrutiny than rational basis 14 review, but does not always trigger strict scrutiny.” Tedards v. Ducey, 951 F.3d 1041, 15 1066 (9th Cir. 2020) (cleaned up). 16 The Ninth Circuit has explained that the Anderson-Burdick inquiry should be 17 conducted in two sequential steps. The first step is to assess the magnitude of the burden 18 created by the challenged regulation. See, e.g., Mecinas v. Hobbs, 30 F.4th 890, 904 (9th 19 Cir. 2022) (“[T]he first step . . . is to consider the character and magnitude of the asserted 20 injury . . . .”) (citations omitted); Soltysik v. Padilla, 910 F.3d 438, 446 (9th Cir. 2018) 21 (“Having established the extent of the burden Soltysik has pleaded, we turn now to the 22 other side of the scale.”); Pub. Integrity All., 836 F.3d at 1027 (“We now apply the Burdick 23 balancing approach, assessing first the burden imposed on Tucson voters by its hybrid 24 system”). During this step, courts evaluate where the burden falls on a minimal-to-severe 25 6 This framework draws its name from Anderson v. Celebrezze, 460 U.S. 780 (1983), and Burdick v. Takushi, 504 U.S. 428 (1992).
26 7 The tentative ruling issued before oral argument contained a more abbreviated 27 summary of the Anderson-Burdick framework. Because the Court has now altered aspects of the tentative ruling’s analysis of Count One to address the points raised during oral 28 argument, it is helpful to begin with a more comprehensive summary of how the Anderson-
Burdick framework operates.
1 spectrum. See, e.g., Soltysik, 910 F.3d at 444-45 (concluding first-step analysis by stating 2 that “the burden the California statutes impose on Soltysik’s First and Fourteenth 3 Amendment rights. . . is not severe”); Short v. Brown, 893 F.3d 671, 677 (9th Cir. 2018) 4 (noting during first-step analysis that “[t]o the extent that having to register to receive a 5 mailed ballot could be viewed as a burden, it is an extremely small one”); Pub. Integrity 6 All., 836 F.3d at 1027 (concluding first-step analysis by stating that “[i]f a burden exists at 7 all, which we doubt, it is at best very minimal”); Dudum v. Arntz, 640 F.3d 1098, 1113-14 8 (9th Cir. 2011) (concluding first-step analysis by stating that “if the [challenged 9 regulations] impose any burdens on the right to vote, they are minimal at best”). 10 The second step of the Anderson-Burdick inquiry is to “identify and evaluate the 11 precise interests put forward by the State as justifications for the burden imposed by its 12 rule.” Short, 893 F.3d at 676 (citation omitted). The degree of scrutiny applied during this 13 step turns on the magnitude of the burden established during the first step. Nader v. 14 Brewer, 531 F.3d 1028, 1034 (9th Cir. 2008) (“[T]he severity of the burden the election 15 law imposes on the plaintiff’s rights dictates the level of scrutiny applied by the court.”). 16 Severe burdens “must meet strict scrutiny” while non-severe burdens “trigger less exacting 17 review.” Mecinas, 30 F.4th at 904 (citations omitted). 18 The “less exacting review” applied in cases involving non-severe burdens may, 19 itself, take different forms. In cases involving “minimal” or “slight” burdens, the 20 challenged regulation will be upheld so long as it advances “an important regulatory 21 interest.” Arizona Democratic Party v. Hobbs, 18 F.4th 1179, 1192 (9th Cir. 2021). See 22 also Short, 893 F.3d at 679 (“Given that the burden is so slight for voters outside the all- 23 mailed ballot election system counties, California’s general interest in increasing voter 24 turnout and specific interest in incremental election-system experimentation adequately 25 justify the VCA’s geographic distinction.”); Dudum, 640 F.3d at 1117 (rejecting challenge 26 because the burdens imposed by the challenged voting system were “minimal at best” and 27 “the City has advanced valid, sufficiently-important interests to justify using its system”). 28 In contrast, in cases involving burdens that are “not severe . . . [but] more than slight,” the
1 resulting degree of scrutiny “is neither strict nor wholly deferential.” Soltysik, 910 F.3d at 2 445. Under this form of review, the government may need to establish that “less 3 burdensome . . . alternatives would not accomplish the goal” of the challenged regulation. 4 Id. at 447. 5 1. S.B. 1003 6 a. Hobbs 7 Before turning to Plaintiffs’ Count One challenge to S.B. 1003, it is necessary to 8 address the elephant in the room—the Ninth Circuit’s decision in Arizona Democratic 9 Party v. Hobbs, 18 F.4th 1179 (9th Cir. 2021), which was issued as the motion-to-dismiss 10 briefing process was unfolding. 11 Hobbs involved a challenge to a provision of Arizona’s then-applicable EPM that 12 ordered county recorders not to count ballots with unsigned affidavits, but rather to “make 13 a reasonable and meaningful attempt” to contact the voter, and to allow such voters to cure 14 the missing signature until 7:00 PM on election day. Arizona Democratic Party v. Hobbs, 15 485 F. Supp. 3d 1073, 1082 (D. Ariz. 2020). The provision at issue in Hobbs was identical, 16 in substance, to S.B. 1003—the only difference is that the former appeared in the EPM 17 while the latter codifies the same rule in Arizona’s statutory law. The plaintiffs in Hobbs 18 argued that the prohibition on curing missing signatures after election night unjustifiably 19 burdened the right to vote. Id. The district court granted the plaintiffs’ request for a 20 preliminary injunction but the Ninth Circuit reversed. 21 As for the burden created by the challenged rule, the Ninth Circuit held during the 22 first step of the Anderson-Burdick inquiry that “the election-day deadline for submitting a 23 completed ballot imposes, at most, a minimal burden.” Hobbs, 18 F.4th at 1187. This 24 determination meant that Arizona only needed to establish that the challenged rule was 25 supported by “important regulatory interests.” Id. at 1186-87, 1190. The court held this 26 standard was satisfied in light of “the State’s important regulatory interest in reducing 27 administrative burdens on poll workers.” Id. at 1194. Thus, the court concluded that 28 “Arizona’s law is constitutional.” Id. at 1196. However, in reaching these conclusions,
1 the court emphasized that “Plaintiffs have not alleged that the burden of signing the 2 affidavit falls disproportionately on a discrete group of voters, thereby implicating 3 heightened constitutional concerns” or that “the burden of fixing a missing signature—that 4 is, casting a replacement or provisional ballot—falls disproportionately on a discrete group, 5 thereby implicating heightened constitutional concerns.” Id. at 1190. The court 6 characterized these omissions as “[i]mportant to our analysis.” Id. 7 b. The Parties’ Arguments 8 The State argues that “this Court should dismiss the virtually identical Anderson- 9 Burdick claim here for the reasons explained by Hobbs.” (Doc. 83 at 2.) The State asserts 10 that Hobbs conclusively establishes that the rule codified by S.B. 1003 imposes a minimal 11 burden on voting rights that is outweighed by the State’s important interest in reducing 12 administrative burdens on poll workers. (Id. at 2-3.)8 13 Plaintiffs make three points in response: (1) Hobbs is distinguishable because the 14 plaintiffs there did not argue that the burdens of the challenged law fell disproportionately 15 on discrete groups of voters; (2) Hobbs did not (and could not) review evidence from the 16 November 2020 election, which strengthens their burden argument; and (3) Hobbs did not 17 rule on the cumulative burden created by S.B. 1003 and S.B. 1485, as is asserted here. 18 (Doc. 99 at 19-20.) 19 The State makes five points in reply: (1) because Plaintiffs lack standing to bring 20 as-applied challenges, they cannot argue that S.B. 1003 is particularly burdensome when 21 applied to particular groups; (2) Plaintiffs do not allege that the burden of signing the 22 affidavit falls disproportionately on discrete groups; (3) even if there are racial disparities 23 regarding the burden of curing an unsigned ballot affidavit, the burdens are still minimal; 24 (4) Plaintiffs may not rely on “cumulative” burdens because the two laws are not 25 meaningfully cumulative; and (5) Plaintiffs may not rely on 2020 voting evidence because 26 27 8 28 The State also asserted collateral estoppel and res judicata claims that have since been withdrawn. (Doc. 83 at 2; Doc. 118 at 29.)
1 they have not offered specific factual allegations concerning that evidence. (Doc. 118 at 2 12-16.) 3 c. Analysis 4 The Court agrees with the State that Plaintiffs’ challenge to S.B. 1003 in Count One 5 is subject to dismissal. In Hobbs, the Ninth Circuit rejected a challenge to what is 6 essentially the same law, and although Plaintiffs purport to identify various differences 7 between this case and Hobbs, those differences do not compel a different outcome. 8 As for Plaintiffs’ contention that they “intend to introduce evidence based on how 9 the cure period was actually implemented in the 2020 election” (Doc. 99 at 19-20), the 10 problem is that none of the allegations in the complaint pertaining to the 2020 election 11 touch upon (let alone call into question) the two core factual determinations underlying 12 Hobbs: first, that the election-night curing deadline creates only a minimal burden on the 13 right to vote; and second, that the deadline furthers Arizona’s important regulatory interest 14 of reducing administrative burdens on poll workers. Although Plaintiffs point to 15 paragraphs 64-68 of their complaint as “a crucial backdrop” for the challenged legislation, 16 those paragraphs only allege that the State’s putative interests in preventing election fraud 17 are unfounded or misleading. Hobbs did not mention the word “fraud”—instead, the court 18 upheld the law based on the State’s interest in reducing administrative burdens. Thus, 19 Plaintiffs have not alleged facts that would distinguish this case from Hobbs. 20 As for Plaintiffs’ contention that Hobbs is distinguishable because it only involved 21 a challenge to one election law, whereas here S.B. 1003 and S.B. 1485 must be considered 22 “collectively” and are “cumulative” (Doc. 99 at 20), the problem once again is that the 23 complaint is devoid of well-pleaded facts that might support this conclusion. Nor do 24 Plaintiffs elsewhere explain how the combination of the two laws burdens Arizona’s voters 25 more than the laws do individually. The Court can imagine how a pair of election laws 26 could be more onerous when combined: for instance, one law that required voters to use 27 their own pen to mark ballots, and another that prevented voters from returning to the 28 polling place if they forgot to bring a pen. Reviewing those laws separately might not
1 recognize the burden they create together. But the laws challenged here are not obviously 2 cumulative. Indeed, S.B. 1485 is being challenged under the theory that it will prevent 3 voters from receiving a mailed ballot, whereas S.B. 1003 is being challenged because it 4 will make it difficult for voters who receive a mailed ballot (notwithstanding the obstacles 5 created by S.B. 1485) to go further and cast that ballot in a meaningful way. Tellingly, 6 Plaintiffs’ only response is that the State’s objection “rings hollow” because the State’s 7 own briefing mentions other election laws that are not challenged here. At minimum, this 8 response does not explain why the two challenged laws must be considered cumulatively. 9 More fundamentally, even accepting that the cumulative burden arising from two 10 challenged voting laws may be greater than the individual burden arising from either law, 11 it doesn’t follow that the cumulative burden is necessarily severe or even moderate. In 12 Hobbs, the Ninth Circuit held that “[t]he election-day deadline for submitting a completed 13 ballot imposes, at most, a minimal burden.” 18 F.4th at 1187. As discussed elsewhere in 14 this order, the burden arising from S.B. 1485 is also, at most, minimal. The sum of these 15 two burdens remains minimal. Cf. Henderson v. Kennedy, 253 F.3d 12, 19 (D.C. Cir. 2001) 16 (“For this argument to prevail, one would have to conclude that . . . the combination of two 17 untenable claims equals a tenable one. But in law as in mathematics zero plus zero equals 18 zero.”). 19 Plaintiffs’ final basis for distinguishing Hobbs is that they are making the sort of 20 “disparate impact on discrete groups of voters” claim that was absent in Hobbs. (Doc. 99 21 at 19.) Although this argument presents a closer call, it ultimately does not provide a basis 22 for avoiding dismissal. 23 As noted, the Ninth Circuit in Hobbs took pains to note that the plaintiffs were not 24 arguing that the burdens arising from the signature-curing deadline fell disproportionately 25 on discrete groups of voters. 18 F.4th at 1190. The court held that such disproportionate 26 burdens “are more likely to raise constitutional concerns” and characterized the absence of 27 allegations on this point as “[i]mportant to our analysis.” Id.
1 Unfortunately, Hobbs sheds little light on how, exactly, a district court should go 2 about evaluating a claim that a voting law does create a disproportionate burden on discrete 3 groups of voters. Nor do earlier Ninth Circuit cases touching on this topic.9 Thus, although 4 Plaintiffs’ disproportionate-burden allegations in this case may be “important” and “more 5 likely” to raise constitutional concerns than the allegations in Hobbs, these observations do 6 not resolve how to evaluate Plaintiffs’ claim during the first step of the Anderson-Burdick 7 analysis. As discussed above, this step requires courts to determine where the alleged 8 burden falls on a minimal-to-severe spectrum. 9 In the tentative ruling issued before oral argument, the Court stated that, because 10 Hobbs cited Anderson v. Celebrezze, 460 U.S. 780 (1983), and Bullock v. Carter, 405 U.S. 11 134 (1972), as two examples of the type of evidence was not present in Hobbs, those 12 decisions could provide a useful benchmark when deciphering the type of evidence and 13 allegations that would be sufficient to prevail on a disproportionate-burden-on-discrete- 14 subgroups claim in the Ninth Circuit. The tentative ruling went on to explain that, because 15 Anderson and Bullock both involved voting regulations that necessarily affected all 16 members of a discrete group (independent voters and candidates in Anderson, “less 17 affluent” candidates in Bullock) but did not affect members of other groups, whereas 18 Plaintiffs acknowledge that S.B. 1003 would not affect all members of the discrete groups 19 at issue here (Native American, black, Latino, and disabled voters) and would affect some 20 members of other groups, Anderson and Bullock suggest that Plaintiffs have not pleaded 21 anything more than a minimal burden under Hobbs. 22 9 23 In Short, similar to Hobbs, the Ninth Circuit emphasized the absence of any allegation that a challenged voting regulation had a disproportionate impact on discrete 24 groups of voters and suggested this omission was “important.” 893 F.3d at 679 (“Importantly, the appellants do not argue that the VCA’s distinction along county lines is 25 a proxy for some other form of discrimination—that it is a racial or political gerrymander disguised as a geographic distinction.”). However, the court did not go further and explain 26 how a district court should go about evaluating such a challenge. In Public Integrity Alliance, although the Ninth Circuit observed in a footnote that “courts may consider not 27 only a given law’s impact on the electorate in general, but also its impact on subgroups, for whom the burden, when considered in context, may be more severe,” the court did not go 28 further and provide concrete guidance on how courts should evaluate such a claim. 836 F.3d at 1024 n.2.
1 During oral argument, Plaintiffs’ counsel explained why Hobbs’s references to 2 Anderson and Bullock should not be viewed as a signal that the Ninth Circuit has placed 3 limits the types of voting regulations that may be challenged under a disproportionate- 4 burden-on-discrete-subgroups theory. Although the Court is not necessarily persuaded that 5 the analytical approach in the tentative ruling was incorrect, the Court also acknowledges 6 that Plaintiffs’ arguments on this point have some force. Thus, in an abundance of caution, 7 the Court will set forth other ways to evaluate Plaintiffs’ burden allegations for purposes 8 of Anderson-Burdick’s first step. 9 Each side provided suggestions on this topic during oral argument. Plaintiffs argued 10 that any voting regulation that has a “material disproportionate impact” on discrete 11 subgroups of voters should automatically be viewed as creating a non-minimal burden that 12 triggers heightened scrutiny under the second step of the Anderson-Burdick inquiry and 13 cannot be dismissed at the pleading stage.10 Meanwhile, the State suggested at times that 14 because § 2 of the Voting Rights Act (“VRA”) already allows for disparate-impact claims, 15 such claims should not be separately cognizable under the Anderson-Burdick framework.11 16 10 During oral argument, Plaintiffs’ counsel was asked: “It seems to me your position 17 is in any case in which a voting regulation is challenged under Anderson-Burdick, your position is that if the plaintiffs can show [that] a facially neutral law has a disproportionate 18 impact on protected groups, no matter how small that disparate impact is in the overall scheme of things, automatically you’re out of this really low-level review and heightened 19 scrutiny applies. Is that correct?” In response, counsel stated: “I would say material disproportionate impact. . . . [I]f it is a material disproportionate impact . . . I think that 20 claim would survive a motion to dismiss in all or nearly all circumstances.” 21 11 The State initially argued: “[W]e don’t think disparate impacts are the ‘get out of 12(b) free’ card that Plaintiffs think they are. It simply isn’t the case that other . . . decisions 22 are . . . essentially chucking all of Anderson-Burdick doctrine out of the window the second that any plaintiff alleges disparate impact. And that’s really what we’re starting to get to 23 here. And . . . part of the reason for that is . . . there is a whole separate kind of claim designed to address disparate impact, and that’s the results-based test under the Voting 24 Rights Act, under Section 2 of it, and that is the claim actually designed to deal with racial disparate impacts. . . . Congress specifically enacted it [not only] because it thought both 25 that the intentional discrimination standard of the Fourteenth Amendment and Voting Rights Act was insufficient, but also because Anderson-Burdick doctrine was not doing 26 what Plaintiffs now say it does. And that is the reason Congress in 1982 enacted the Voting Rights Act amendment that created the results-based test. And you can’t take that test and 27 try to squeeze it into Anderson-Burdick standard. That isn’t how these cases works.”
However, the State later clarified: “It’s not our position that disparate impacts are 28 categorically irrelevant to addressing the burden. . . . But just because it’s a relevant factor doesn’t mean it’s a trump card.”
1 Neither suggestion is persuasive. The problem with Plaintiffs’ approach is that the 2 first step of the Anderson-Burdick inquiry is fundamentally about measuring the “character 3 and magnitude” of the burden created by a challenged voting regulation. Burdick, 504 4 U.S. at 434. After all, “‘every electoral law and regulation necessarily has some impact on 5 the right to vote.’” Dudum, 640 F.3d at 1113 (quoting Weber v. Shelley, 347 F.3d 1101, 6 1106 (9th Cir. 2003)). In the Court’s view, these principles mean it is not enough for a 7 plaintiff bringing an Anderson-Burdick challenge to show that a discrete subgroup of voters 8 would be more burdened by the challenged voting regulation than other groups of voters— 9 instead, the plaintiff must go further and establish, on an absolute level, the magnitude of 10 the burden on the discrete subgroup. Courts outside the Ninth Circuit appear to follow this 11 approach. See, e.g., Mays v. LaRose, 951 F.3d 775, 784-85 (6th Cir. 2020) (“Precedent 12 from this court and the Supreme Court suggests that we must evaluate this burden from the 13 perspective of only affected electors and within the landscape of all opportunities that Ohio 14 provides to vote.”). The Supreme Court, too, has suggested that relative comparisons 15 between subgroups can be “highly misleading” in the voting context and that the better 16 approach is to evaluate the absolute effect on the burdened subgroup. Brnovich v. 17 Democratic Nat’l Comm., 141 S. Ct. 2321, 2344-45 (2021) (“[T]he racial disparity in 18 burdens allegedly caused by the out-of-precinct policy is small in absolute terms. The 19 District Court accepted the plaintiffs’ evidence that, of the Arizona counties that reported 20 out-of-precinct ballots in the 2016 general election, a little over 1% of [minority] voters 21 who voted on election day cast an out-of-precinct ballot. For non-minority voters, the rate 22 was around 0.5%. A policy that appears to work for 98% or more of voters to whom it 23 applies—minority and non-minority alike—is unlikely to render a system unequally open. 24 The Court of Appeals attempted to paint a different picture, but its use of statistics was 25 highly misleading . . . . [It] concluded that ‘minority voters in Arizona cast [out-of- 26 precinct] ballots at twice the rate of white voters’ . . . . [But] the statistics show only a 27 small disparity that provides little support for concluding that Arizona’s political processes 28 are not equally open.”).
1 Meanwhile, the most extreme version of the State’s proposed approach is foreclosed 2 by Ninth Circuit law. In Crawford v. Marion County Election Bd., 553 U.S. 181 (2008), 3 which involved a challenge to an Indiana voter-identification statute under the theory that 4 “a small number of voters who may experience a special burden under the statute,” id. at 5 200, several Justices characterized as “irrelevant” the “petitioners’ premise that the voter- 6 identification law ‘may have imposed a special burden on’ some voters” and stated that 7 “weighing the burden of a nondiscriminatory voting law upon each voter and 8 concomitantly requiring exceptions for vulnerable voters would effectively turn back 9 decades of equal-protection jurisprudence.” Id. at 204-07 (Scalia, J., concurring). 10 However, that position did not attract five votes, and as noted above, the Ninth Circuit has 11 subsequently recognized (in Public Integrity Alliance, Short, and Hobbs) that a plaintiff 12 challenging a voting regulation under the Anderson-Burdick framework may rely on the 13 theory that the regulation places a disproportionate burden on discrete subgroups. 14 Although, as further noted above, the Ninth Circuit’s decisions in this area provide very 15 little in the way of concrete guidance concerning how to evaluate such a claim, they 16 undermine the State’s suggestion that Plaintiffs’ theory is categorically impermissible. 17 With this backdrop in mind, the question is how to measure the magnitude of the 18 burden placed by S.B. 1003 on the discrete subgroups of voters identified in Plaintiffs’ 19 complaint—Native American, black, Latino, and disabled voters. The well-pleaded facts 20 in Plaintiffs’ complaint that are relevant to the disproportionate burden arising from S.B. 21 1003 are as follows: 22 ▪ Arizona election officials already deprive many Native American voters who live on reservations of reasonable access to polling places and election 23 offices. Lack of access to post offices and street addresses slows the process 24 by which they obtain and submit ballots and would similarly slow the process by which they receive notice of deficiencies. (Doc. 1 ¶ 91.)
25 ▪ Black and Latino voters are also burdened by lack of reasonable access to 26 polling places and election offices. Many such voters live in neighborhoods 27 with unequal and insufficient infrastructure. People of color are nearly twice as likely to lack access to a car than white Americans. In Maricopa County, 28 for example, voters in neighborhoods with high concentrations of black and
1 Latino voters would have to travel up to two hours one way by public 2 transportation in order to provide a missing signature. It would cost such voters between $7 and $19.50 each way to take a taxi to the election office.
3 In Yuma County, neighborhoods with high concentrations of black and Latino voters are between 25 and 90 minutes away from election offices by 4 car, and a cab could cost nearly $75. (Id. ¶ 92.)
5 ▪ Lack of language access substantially increases the likelihood that voters 6 will miss the signature requirement and be unaware of the need for curing.
Language barriers also make it more challenging for these voters to 7 understand any notice of the ballot’s deficiency, communicate with elections 8 officials, or learn what they need to do to cure their ballots. No provision in S.B. 1003 provides for notice to be given in a language spoken by the voter.
9 (Id. ¶ 93.) 10 ▪ Disabled voters often struggle to access public transportation. Voters who 11 are unable to provide a physical signature, or whose marks are not recognized as a signature, risk having their ballots discarded without sufficient time to 12 cure their ballots. (Id. ¶ 94.) 13 The putative disparate burdens can thus be split into five categories: (1) limited access to 14 traditional polling places (disproportionately burdening Native American, black, and 15 Latino voters); (2) inconsistent mail service (disproportionately burdening Native 16 American voters); (3) limited access to transportation and election offices 17 (disproportionately burdening black, Latino, and disabled voters); (4) inability to read and 18 understand a ballot’s signature requirement and/or instructions for curing a missing 19 signature (disproportionately burdening voters with language barriers); and (5) inability to 20 validly sign a ballot (disproportionately burdening disabled voters).12 21 Even accepting these facts as true and viewing them in the light most favorable to 22 Plaintiffs, they do not establish that S.B. 1003 will place anything more than a minimal 23 burden on the ability of Native American, black, Latino, and disabled Arizonans to vote. 24 As an initial matter, voting by mail is simply one method by which Arizonans may choose 25 to vote. Although it has become a popular method in recent years, Arizonans may also 26 choose to vote in the traditional manner—by voting in person at a polling place. And 27 12 The tentative ruling did not separately identify the complaint’s allegations regarding 28 lack of access to traditional polling places, but during oral argument, Plaintiffs’ counsel explained why those allegations provide context for the other allegations in the complaint.
1 although the complaint alleges that Arizona has greatly reduced the number of traditional 2 polling places in recent years, Plaintiffs do not independently challenge Arizona’s practices 3 in this area or otherwise allege that there is a constitutional right to vote by mail. Courts 4 have concluded that where a challenged voting regulation simply affects one method of 5 voting but leaves other alternative voting methods unaffected, this reduces (if not 6 eliminates) any claim of undue burden. See, e.g., Crawford, 553 U.S. at 199 (even though 7 Indiana’s voter-identification law placed “a somewhat heavier burden . . . on a limited 8 number of persons,” including the elderly and homeless persons, “[t]he severity of that 9 burden is, of course, mitigated by the fact that, if eligible, voters without photo 10 identification may cast provisional ballots that will ultimately be counted”); Ohio 11 Democratic Party v. Husted, 834 F.3d 620, 628 (6th Cir. 2016) (“The undisputed factual 12 record shows that it’s easy to vote in Ohio. Very easy, actually. Viewing S.B. 238 as one 13 component of Ohio’s progressive voting system, and considering the many options that 14 remain available to Ohio voters, even accepting the district court’s focus on the changes 15 wrought by S.B. 238, the removal of Golden Week [a six-day opportunity for same-day 16 registration and voting] can hardly be deemed to impose a true ‘burden’ on any person’s 17 right to vote. At worst, it represents a withdrawal or contraction of just one of many 18 conveniences that have generously facilitated voting participation in Ohio.”). See also 19 Mays, 951 F.3d at 785 (“[W]e must evaluate the burden on disparately treated voters 20 considering all available opportunities to vote.”). This is particularly true where, as here, 21 there is no constitutional right to use the alternative voting method that is affected by the 22 challenged law. McDonald v. Bd. of Election Comm’rs of Chicago, 394 U.S. 802, 807-08 23 (1969) (“It is thus not the right to vote that is at stake here but a claimed right to receive 24 absentee ballots. Despite appellants’ claim to the contrary, the absentee statutes, which are 25 designed to make voting more available to some groups who cannot easily get to the polls, 26 do not themselves deny appellants the exercise of the franchise . . . .”). See also Griffin v. 27 Roupas, 385 F.3d 1128, 1129-30 (7th Cir. 2004) (affirming dismissal of lawsuit brought 28 by “working mothers who contend that because it is a hardship for them to vote in person
1 on election day, the U.S. Constitution requires Illinois to allow them to vote by absentee 2 ballot” in part because the plaintiffs’ argument “that the Constitution requires all states to 3 allow unlimited absentee voting . . . ignores a host of serious objections to judicially 4 legislating so radical a reform in the name of the Constitution”). 5 Nevertheless, even if the ability to vote by mail were considered in isolation, without 6 regard to the other voting methods that Native American, black, Latino, and disabled 7 residents of Arizona may continue to use, the burdens imposed on those subgroups by S.B. 8 1003 would remain minimal. An Arizona voter wishing to vote by mail must sign an 9 affidavit that is printed on the outside of the envelope. Hobbs, 18 F.4th at 1181-82 (“To 10 vote by mail, a voter must return a completed ballot in a specially provided, postage-paid 11 envelope, and the voter must sign an affidavit that is printed on the envelope. The affiant 12 declares, under penalty of perjury, that he or she ‘voted the enclosed ballot.’”) (citations 13 omitted). The signature requirement is “prominent” and “the instructions include the 14 signature requirement in English, Spanish, and pictograph form.” Id. at 1182. In Hobbs, 15 the Ninth Circuit noted that “Plaintiffs have not alleged that the burden of signing the 16 affidavit falls disproportionately on a discrete group of voters, thereby implicating 17 heightened constitutional concerns. To the contrary, ‘[f]orgetfulness is an involuntary state 18 that any voter might reasonably experience.’” Id. at 1190 (citation omitted). 19 Plaintiffs here, like the plaintiffs in Hobbs, do not challenge the signature 20 requirement or contend that it disproportionately affects Native American, black, Latino, 21 and disabled voters. Instead, Plaintiffs’ theory is that the undue burden arises from S.B. 22 1003’s alleged curtailment of the curing opportunities that are available to voters who 23 violate the signature requirement.13 This theory is unpersuasive for three reasons. First, 24 13 25 In Hobbs, the Ninth Circuit emphasized that it was not addressing the merits of the theory that Plaintiffs are pursuing here: “Nor have Plaintiffs argued that the burden of 26 fixing a missing signature—that is, casting a replacement or provisional ballot—falls disproportionately on a discrete group, thereby implicating heightened constitutional 27 concerns.” Id. at 1190. The Court does not construe this passage as an instruction that courts should uncritically grant relief in all future cases raising such a claim. Instead, the 28 passage simply reflects that the Hobbs court was limiting its analysis to the issues properly before it.
1 just as there is no constitutional right to vote by mail, there is no constitutional right to cure 2 a missing signature on a mailed ballot. In Hobbs, the Ninth Circuit noted that “[m]ost 3 forms of voter negligence have no remedy” and that “[a] voter who neglects to sign the 4 affidavit bears all responsibility for the error.” Id. at 1188, 1193-94. Although “Arizona 5 law offers a measure of grace” for “voters who forget to sign the affidavit,” id. at 1188, it 6 is difficult to see how S.B. 1003’s imposition of temporal limitations on this “measure of 7 grace” could somehow be characterized as placing anything more than a minimal burden 8 on the right to vote (or, more precisely, on an alternative form of voting that itself is not 9 constitutionally required). 10 The Ninth Circuit’s decision in Lemons v. Bradbury, 538 F.3d 1098 (9th Cir. 2008), 11 underscores this conclusion. Lemons involved a challenge to Oregon’s two-tiered system 12 of signature-verification requirements, under which “[c]ounty elections officials do not 13 notify voters after rejecting referendum signatures as non-matching. In contrast, when 14 county elections officials reject a signature on a vote-by-mail election ballot, they give the 15 voter ten-days notice and an opportunity to submit an updated voter registration card, or 16 otherwise provide proof that the signature is valid.” Id. at 1103. The plaintiffs in Lemons 17 argued this approach was unconstitutional because, inter alia, Oregon’s failure to provide 18 an opportunity to cure mismatched signatures on referendum petitions created an undue 19 burden on the ability to participate in the referendum process. Id. at 1102 (“[T]hey argue 20 that the method of verifying referendum signatures burdens their fundamental right to 21 vote.”). The Ninth Circuit disagreed, explaining that “[a]lthough county elections officials 22 do not notify voters after rejecting non-matching signatures, referendum petition cover 23 sheets instruct voters to ‘[s]ign your full name, as you did when you registered to vote.’” 24 Id. at 1104. Given the presence of this and other procedural safeguards, the court 25 concluded that, “[w]hen viewed in context, the absence of notice and an opportunity to 26 rehabilitate rejected signatures imposes only a minimal burden on plaintiffs’ rights.” Id. 27 If Oregon’s refusal to allow any form of signature curing during the referendum process— 28 where signing the referendum is the only way to participate—only amounted to a minimal
1 burden on the right to participate in that process in light of the provision of clear 2 instructions regarding the signature requirement, it follows that Arizona’s practice of 3 allowing signature curing (albeit with time limitations) during the vote-by-mail process— 4 which is simply an alternative voting method that is not constitutionally required, and 5 which includes the same clear instructions regarding the signature requirement that were 6 present in Lemons—qualifies as no more than a minimal burden. Although this minimal 7 burden may affect Native American, black, Latino, and disabled residents of Arizona more 8 often than it affects members of other groups, it remains minimal in the context of those 9 subgroups’ members’ overall ability to vote. 10 A second, related reason why any burden posed by S.B. 1003 must be considered 11 minimal arises from a comparison of Arizona’s voting practices to the voting practices of 12 other states. In Hobbs, the Ninth Circuit recognized that “the challenged cure provision is 13 more lenient than that of many other States.” 18 F.4th at 1185. It would be anomalous to 14 conclude that Arizona’s approach of allowing voters to cure missing signatures on mailed 15 ballots, so long as the curing occurs by election night, qualifies as a more-than-minimal 16 burden on the right to vote where many of its sister states do not allow for curing at all. Cf. 17 Brnovich, 141 S. Ct. at 2338-39 (“Because every voting rule imposes a burden of some 18 sort, it is useful to have benchmarks with which the burdens imposed by a challenged rule 19 can be compared. . . . We doubt that Congress intended to uproot facially neutral time, 20 place, and manner regulations that . . . are in widespread use in the United States.”). 21 Plaintiffs’ counsel acknowledged during oral argument that, under the logic underlying 22 Plaintiffs’ challenge to S.B. 1003 in Count One, every state that doesn’t allow in-person 23 voting—and there are several in the Ninth Circuit 14 —would likely be required under 24 Anderson-Burdick to allow post-election night curing of missing signatures. This 25 underscores some of the Court’s concerns with Plaintiffs’ position. See generally Short, 26 893 F.3d at 676 (“No one disputes that the right to vote is fundamental. But not all election 27 14 Hoffard v. Cnty. of Cochise, 2020 WL 6555235, *4 n.4 (D. Ariz. 2020)
28 (“Washington, Oregon, Hawaii, Colorado, and Utah do not offer in-person voting on Election Day and instead conduct their elections by mail.”).
1 laws impose constitutionally suspect burdens on that right. And states retain broad 2 authority to structure and regulate elections.”) (citations omitted). True, Arizona could 3 have made it even easier to cure missing signatures, such as by enacting a later curing 4 deadline, but the existence of a less-restrictive option does not mean that Arizona’s chosen 5 option is constitutionally suspect. After all, “the Constitution merely sets a floor. Nothing 6 in our opinion should be construed as dissuading Arizona, or other States, from providing 7 a more generous deadline than the Constitution requires. . . . We are not called upon to 8 express our political views; instead, we merely decide the narrow question before us: 9 whether this one voting regulation violates the Constitution.” Hobbs, 18 F.4th at 1195 10 (citations omitted). 11 The third reason why any burden posed by S.B. 1003 must be considered minimal, 12 even from the perspective of the discrete subgroups identified by Plaintiffs, stems from 13 how infrequently missing-signature issues arise in Arizona. In Hobbs, the Ninth Circuit 14 noted that “Arizona election officials reject only a small fraction—approximately one-tenth 15 of one percent—of the total number of ballots due to a missing signature. Missing 16 signatures led officials to reject 3,079 ballots in 2016 and 2,435 ballots in 2018.” 18 F.4th 17 at 1190.15 Given that a vanishingly small percentage of voters experience any difficulty 18 with the signature requirement in the first place, Plaintiffs’ allegation that only some 19 members of some demographics of a minuscule fraction of voters will be burdened 20 disproportionately cannot plausibly convert “at most, a minimal burden” into a severe one. 21 To be clear, Plaintiffs are correct that they “do not need to offer ‘evidence’ at the 22 motion-to-dismiss stage.” (Doc. 99 at 20.) However, Plaintiffs do need to offer well- 23 pleaded factual allegations supporting their claim. Iqbal, 556 U.S. at 679. Here, Plaintiffs 24 have failed to plead a disproportionate burden that would distinguish their case from 25 Hobbs. Thus, there would be no point in allowing discovery with respect to Plaintiffs’ 26 15 Plaintiffs do not allege otherwise in this action—although the complaint alleges that 27 “[t]he burdens of [S.B. 1003] are especially severe for people of color, non-English speakers, disabled voters, and voters who live on reservations” (Doc. 1 ¶ 90), the complaint 28 does not purport to quantify the number of voters from these discrete subgroups who will miss out on curing opportunities as a result of S.B. 1003.
1 challenge to S.B. 1003 in Count One—even if the discovery process ended up generating 2 evidence that fully supported all of the factual allegations in the complaint, that evidence 3 still would not establish anything more than a minimal burden on the right to vote. 4 Given this conclusion, it follows that S.B. 1003 is not unconstitutional under the 5 Anderson-Burdick framework. In Hobbs, the Ninth Circuit held that “[t]he State has an 6 important regulatory interest in reducing the administrative burden on poll workers, and 7 Plaintiffs’ proposed relief would increase that burden in a meaningful manner.” 18 F.4th 8 at 1192. In reaching this conclusion, the court acknowledged that Arizona could allow for 9 post-election night curing of missing signatures and that Arizona already allows for such 10 curing of mismatched signatures but held that these features of Arizona law did not 11 undermine the bottom-line conclusion that “the State’s important regulatory interest in 12 reducing administrative burdens on poll workers sufficiently justifies the minimal burden 13 on a voter to sign the affidavit or to correct a missing signature by election day.” Id. at 14 1192-94. Here, similarly, none of the well-pleaded factual allegations in the complaint 15 undermine the State’s contention that requiring Arizona to extend the current deadline for 16 curing missing signatures would result in increased administrative burdens. Indeed, 17 Plaintiffs don’t appear to challenge this aspect of Hobbs in their response to the motion to 18 dismiss—all of their efforts to distinguish Hobbs turn on the Ninth Circuit’s analysis during 19 the first step of the Anderson-Burdick inquiry, which evaluates the severity of the burden, 20 and not the second step, which evaluates the State’s regulatory interests (including 21 administrative burdens). 22 “[V]oting regulations are rarely subjected to strict scrutiny.” Dudum, 640 F.3d at 23 1106. Rarer still is the situation in which a district court is presented with the Ninth 24 Circuit’s definitive answer to a nearly identical question but finds reason to chart its own 25 course. The Court concludes that Plaintiffs’ Anderson-Burdick challenge to S.B. 1003 in 26 Count One must be dismissed for failure to state a claim. 27 … 28 …
1 2. S.B. 1485 2 a. The Parties’ Arguments 3 The State argues that “Plaintiffs have no right to vote by mail at all—let alone 4 remain on an EVL despite chronic non-voting.” (Doc. 76 at 16.) “To the extent Plaintiffs’ 5 claim is cognizable at all,” the State argues that “the actual burden imposed is truly 6 minimal,” especially when compared to early voting programs in other states, and that the 7 State’s important interests in reducing administrative burdens and costs and securing 8 elections are sufficient to sustain S.B. 1485. (Id. at 17-21.) Finally, the State argues that 9 Plaintiffs have not satisfied the requirements to bring a facial claim. (Id. at 21-22.) 10 Plaintiffs respond that the Court cannot resolve this claim on the pleadings without 11 a fully developed evidentiary record, particularly because the law is alleged to place a 12 disproportionate burden on an identifiable segment of voters and is thus more likely to raise 13 constitutional concerns. (Doc. 99 at 14.) Plaintiffs argue that “[n]one of the cases the 14 [State] cites show that SB 1485’s burden can be deemed ‘minimal’ on a motion to dismiss.” 15 (Id. at 15.) Plaintiffs also assert that, even if the actual burden is minimal, the State’s 16 putative interests are insufficient: the cost of printing, processing, and sending all mail-in 17 ballots for Maricopa County is not enough to “deprive citizens of their constitutional right 18 to vote” and the interest in secure elections “is a pretext and lacks a rational basis.” (Id. at 19 16-17.) 20 In reply, the State contends that, under Short v. Brown, 893 F.3d 671 (9th Cir. 2018), 21 and given that Arizona’s voting regime is much more generous than the norm, the burden 22 imposed by S.B. 1485 is extremely minimal. (Doc. 118 at 17-20.) The State also argues 23 that its interest in reducing administrative burdens is “concrete and substantial” and that its 24 interest in securing elections is a compelling interest that cannot be defeated by allegations 25 about the subjective motivations of legislators. (Id. at 22-23.) 26 b. Analysis 27 As with S.B. 1003, the Court does not write on a clean slate when analyzing 28 Plaintiffs’ challenge to S.B. 1485. Just as Hobbs serves as the starting point (and, in many
1 respects, the ending point) for the analysis concerning S.B. 1003, the Ninth Circuit’s recent 2 decision in Short v. Brown serves as the starting point for the analysis concerning S.B. 3 1485. 4 Short involved a challenge to California’s Voter’s Choice Act (“VCA”). 893 F.3d 5 at 674. The VCA created an “all-mailed ballot system,” under which “a ballot is 6 automatically mailed to every registered voter twenty-nine days before the election date,” 7 but it did not call for statewide implementation of the new system in one fell swoop. Id. at 8 674-75. Instead, the VCA authorized 14 of California’s 58 counties to opt into the new 9 system in 2018 and authorized the remaining counties to opt into the new system in 2020. 10 Id. In Short, the plaintiffs argued that because it was undisputed that “election participation 11 will be higher under the all-mailed ballot election system than it would be under the 12 traditional polling-place system,” it followed that “the VCA’s county-by-county 13 structure—permitting voters in some counties to receive a ballot by mail automatically, 14 while requiring voters in other counties to register to receive a ballot by mail—inequitably 15 ‘dilutes’ votes in ‘disfavored’ counties.” Id. at 675, 677. The district court denied the 16 plaintiffs’ request for a preliminary injunction and the Ninth Circuit affirmed. First, when 17 assessing the severity of the burden created by the VCA, the Court held that “[a]s for voters 18 outside the counties that have opted in to the all-mailed system, . . . [t]o the extent that 19 having to register to receive a mailed ballot could be viewed as a burden, it is an extremely 20 small one, and certainly not one that demands serious constitutional scrutiny.” Id. at 677. 21 In reaching this conclusion, the court emphasized (similar to the court in Hobbs) that the 22 plaintiffs did “not argue that the VCA’s distinction along county lines is a proxy for some 23 other form of discrimination—that it is a racial or political gerrymander disguised as a 24 geographic distinction.” Id. at 679. Finally, given that the burden created by the challenged 25 law was “so slight,” the court concluded that California’s important regulatory interests 26 (which included “incremental election-system experimentation”) were sufficient to 27 immunize the law from constitutional challenge. Id.
1 S.B. 1485 is similar to the VCA in that both regulate how voters may choose to vote 2 by mail. Even though it was undisputed in Short that the VCA would, as a statistical matter, 3 make it more difficult to vote in certain California counties that still allowed in-person 4 voting, the Ninth Circuit held that the burden faced by voters in those counties was “so 5 slight,” “an extremely small one,” and “certainly not one that demands serious 6 constitutional scrutiny” because the only thing such voters needed to do to receive a mailed 7 ballot was “to register.” Id. at 677, 679. Similarly, although S.B. 1485 will lead to some 8 voters being removed from the PEVL, those voters can be reinstated and continue receiving 9 mailed ballots simply by re-registering—the very step that was found to create an 10 “extremely small” and “slight” burden in Short. 11 Plaintiffs argue that Short is distinguishable because it arose in the context of a 12 request for a preliminary injunction, whereas this case involves a challenge to the 13 pleadings. (Doc. 99 at 15.) But Short’s determination that a registration requirement to 14 receive a mailed ballot creates a “slight” and “minimal” burden on the right to vote was 15 not, at least in the Court’s estimation, some sort of fact-bound determination that was based 16 on the specific evidentiary record in Short—rather, it was a common-sense observation 17 that has the force of law and must be applied by district courts in the Ninth Circuit in future 18 cases. Moreover, although the Ninth Circuit has suggested that a “fully developed 19 evidentiary record” is sometimes necessary to evaluate an Anderson-Burdick challenge to 20 a voting regulation, Soltysik, 910 F.3d at 447-50, the Ninth Circuit has also held that the 21 dismissal of a challenge to a voting regulation may be permissible at the pleading stage 22 where—as here—the well-pleaded facts only establish the existence of a minimal burden 23 on the right to vote. Tedards, 951 F.3d at 1067 (“Relying on Soltysik more generally, 24 Plaintiffs argue that all of Defendants’ arguments fail at the motion to dismiss stage 25 because an evidentiary hearing is necessary to apply something more than rational basis 26 review. We disagree. This case is distinguishable from Soltysik because, compared to the 27 [minimal] burden at issue here, the burden in Soltysik fell higher on the Burdick sliding 28 scale between ‘reasonable, nondiscriminatory’ and ‘severe. . . . Under these circumstances,
1 we held that further development of the evidentiary record was necessary to determine 2 whether there were ‘more precise ways’ to address the State’s alleged interest in preventing 3 voter confusion”). 4 Another potential difference between this case and Short is that the plaintiffs in 5 Short did not argue the VCA would create a disparate impact on discrete groups of voters, 6 whereas Plaintiffs raise such a claim here. But the analysis as to this issue mirrors the 7 analysis concerning S.B. 1003 and Hobbs. The Court accepts that the presence of these 8 allegations means Short is not fully on point, but their presence does not mean that 9 Plaintiffs have necessarily alleged the existence of a moderate or severe burden on the right 10 to vote. Instead, the Court again notes that the disparate burdens alleged in the complaint 11 are dissimilar to the type of disparate burdens at issue in Anderson and Bullock 16 and 12 concludes that those allegations do not, at any rate, establish anything more than a minimal 13 burden on the ability of Latino, black, Native American, and disabled Arizonans to vote. 14 Like S.B. 1003, S.B 1485 at most affects (but does not eliminate) the ability of 15 Arizona residents to use an alternative voting method—voting by mail—that is not 16 constitutionally required. Thus, Plaintiffs’ allegation that S.B. 1485 will cause Latino, 17 black, and Native American voters to be removed from the PEVL at a higher rate than 18 white voters (Doc. 1 ¶ 77) is not dispositive. Such voters may still re-register for the PEVL 19 and/or vote in person. Crawford, 553 U.S. at 199; McDonald, 394 U.S. at 807-08; Husted, 20 834 F.3d at 628; Mays, 951 F.3d at 785. True, the PEVL-removal rate may be higher 21 16 22 In Short, the Ninth Circuit seemed to identify Obama for America v. Husted, 697 F.3d 423 (6th Cir. 2012), as another example of a voting regulation that would not pass 23 muster under the Anderson-Burdick framework. 893 F.3d at 678. But Husted does not support Plaintiffs’ position for the same reasons that Anderson and Bullock don’t support 24 Plaintiffs’ position—there, the challenged regulation (an Ohio statute that shortened the early-voting period for the general population but not for military personnel) directly 25 affected all members of one discrete group but no members of other discrete groups.
Although it is true, as Plaintiffs noted during oral argument, that Husted contains a few 26 references to how the challenged law “disproportionately” affected voters who were “women, older, and of lower income and education attainment,” 697 F.3d at 431, the 27 opinion ultimately turned on the fact that “Ohio’s statutory scheme is not generally applicable to all voters.” Id. at 434. See also id. at 435 (“Equally worrisome would be the 28 result if states were permitted to pick and choose among groups of similarly situated voters to dole out special voting privileges.”).
1 within certain groups, but it is “highly misleading” to rely on such relative comparisons 2 when the relevant inquiry is the absolute effect of the challenged regulation on members 3 of the subgroups at issue. Brnovich, 141 S. Ct. at 2344-45. 4 Like S.B. 1003, S.B 1485 also implicates an alternative voting method that is not 5 offered in many other states. Arizona has made an undisputed and judicially noticeable 6 showing that “[a] full thirty states have no EVL for anyone.” (Doc. 76 at 18.) In Brnovich, 7 the Supreme Court emphasized that Arizona’s practice of allowing “[a]ll Arizonans [to] 8 vote by mail,” without the need for a “special excuse,” simply by “ask[ing] to be sent an 9 early ballot automatically in future elections” meant that “Arizona voting law . . . generally 10 makes it quite easy for residents to vote.” 141 S. Ct. at 2333-34. Those features of Arizona 11 voting law remain in place after the enactment of S.B. 1485. As discussed above with 12 respect to S.B. 1003, it would be anomalous if Arizona’s modification of an optional, 13 alternative voting practice that, even after modification, remains substantially more 14 “generous” than the voting practices of more than half of its sister states could be said to 15 qualify as a severe or even moderate burden on the right to vote. 16 Plaintiffs make much of the fact that S.B. 1485 represents a retraction from 17 Arizona’s previous approach to the PEVL, but as other courts have recognized, the 18 Anderson-Burdick framework does not stand for the proposition “that any expansion of 19 voting rights must remain on the books forever.” Husted, 834 F.3d at 635. “Such a rule 20 would have a chilling effect on the democratic process: states would have little incentive 21 to pass bills expanding voting access if, once in place, they could never be modified in a 22 way that might arguably burden some segment of the voting population’s right to vote. 23 Accepting the ‘long recognized . . . role of the States as laboratories for devising solutions 24 to difficult legal problems,’ . . . imposing such a one-way ratchet is incompatible with the 25 ‘flexible’ Anderson-Burdick framework.” Id. (citations omitted). 26 Because S.B. 1485 creates only a minimal burden on the right to vote, it need only 27 be supported by important regulatory interests. Here, the State identifies two such interests:
1 (1) reducing administrative costs and burdens and (2) promoting election integrity. (Doc. 2 76 at 19-21.)17 3 As for the former, the State has submitted judicially noticeable materials 4 establishing that the cost of mailing each early ballot is $2 to $3. (Doc. 76 at 19, citing 5 Doc. 68 at 430.)18 According to the allegations in the complaint, S.B. 1485 will result in 6 the elimination of 125,000 to 150,000 voters from the PEVL. (Doc. 1 ¶ 76.) Thus, 7 according to Plaintiffs’ own allegations, S.B. 1485 may result in a savings of up to 8 $450,000 per election cycle in the cost of printing and mailing ballots. 9 In their response to the motion to dismiss, Plaintiffs tellingly do not dispute that S.B. 10 1485 will result in a net cost-savings benefit to the State—they simply dismiss that benefit 11 as “remote” and “not in any sense necessary to the proper administration of [Arizona’s] 12 election laws.” (Doc. 99 at 16-17, citations and internal quotation marks omitted).19 These 13 arguments are unavailing because, as the Ninth Circuit and other courts have recognized, 14 cost savings qualify as an important regulatory interest that may justify a regulation that 15 creates only a minimal burden on the right to vote. See, e.g., Dudum, 640 F.3d at 1116 16 (“The City points to evidence that restricted IRV will save money . . . . The interest in 17 17 In Short, the regulatory interest that was deemed sufficient was California’s interest 18 in “incremental election-system experimentation.” 893 F.3d at 679. Because Arizona does not seek to justify S.B. 1485 based on this interest, Short’s analysis of the second prong of 19 the Anderson-Burdick test is less useful here.
18 Because Plaintiffs did not oppose the State’s request for judicial notice, the Court 20 summarily granted it under LRCiv 7.2. (Doc. 89.) Additionally, although Plaintiffs assert in their response to the motion to dismiss that the judicially noticeable materials merely 21 establish “the costs of printing, processing and sending all mail-in ballots for Maricopa County for the November 2020 election” (Doc. 99 at 17), Plaintiffs specifically allege in 22 the complaint that “[o]ver two million of the 2.6 million registered voters in Maricopa County—Arizona’s largest county—are on the permanent early voting list.” (Doc. 1 ¶ 69.)
23 Thus, the State’s calculation of a per-ballot cost of $2-3 is a simple function of dividing the judicially noticeable overall cost of $4,956,384 for “Early Ballot Printing and 24 processing and Mailing” (Doc. 68 at 430) by the number of printed ballots (2 million)
alleged in the complaint.
25 19 During oral argument, Plaintiffs’ counsel suggested that S.B. 1485 may not result 26 in a net financial benefit because the $450,000 in ballot-printing savings may be offset by the additional costs of compliance created by S.B. 1485. However, counsel acknowledged 27 that the complaint does not attempt to quantify those additional costs or otherwise allege that they exceed the ballot-printing savings. If Plaintiffs include such allegations in a future 28 iteration of the complaint, the analysis regarding S.B. 1485 under the second step of the Anderson-Burdick inquiry may change.
1 alleviating the costs and administrative burdens of conducting additional elections can be 2 ‘a legitimate state objective’ that also justifies the use of IRV, given the minimal at best 3 burdens the system imposes on voters’ constitutional rights to vote.”); Weber v. Shelley, 4 347 F.3d 1101, 1106 (9th Cir. 2003) (identifying “saving money” as one of the “positive 5 changes” that may justify a voting regulation that does not “severely restrict[] the right to 6 vote”); Husted, 834 F.3d at 634 n.8 (“Though saving tens of thousands of dollars may be 7 a ‘minimal’ benefit when compared to the overall election budgets, we reject the district 8 court’s dubious and blanket proposition that ‘where more than minimal burdens on voters 9 are established, the State must demonstrate that such costs would actually be burdensome.’ 10 Fiscal responsibility, even if only incrementally served, is undeniably a legitimate and 11 reasonable legislative purpose.”). Cf. Clements v. Fashing, 457 U.S. 957, 965 (1982) 12 (“States have important interests in . . . avoiding the expense and burden of run-off 13 elections.”). 14 Given this determination, the Court need not decide whether the State’s other 15 proffered regulatory interest (promoting election integrity) would also be sufficient to 16 insulate S.B. 1485 from challenge under Anderson-Burdick. Plaintiffs’ challenge to S.B. 17 1485 in Count One must be dismissed for failure to state a claim. 18 C. Counts Two And Three—S.B. 1003 19 In Counts Two and Three of the complaint, Plaintiffs challenge S.B. 1485 and S.B. 20 1003 under the theory that each law was enacted with a discriminatory purpose. In 21 response, the State not only argues that these counts fail to state a claim but also argues 22 that Plaintiffs (and Plaintiffs-Intervenors) lack standing to the extent they wish to challenge 23 S.B. 1003 under a discriminatory-purpose theory.20 Because “a federal court generally 24 25 20 The State did not, in contrast, move to dismiss Plaintiffs’ challenge to S.B. 1003 in Count One based on a lack of standing. At oral argument, the State asserted that challenges 26 to voting regulations under the Anderson-Burdick framework implicate different standing considerations than challenges brought under Arlington Heights or the VRA. Given that 27 explanation, and in light of the fact that the Ninth Circuit in Hobbs reached the merits of the Anderson-Burdick challenge to the functional equivalent of S.B. 1003 (which implies, 28 if weakly, that the court found no issue with standing), the Court will confine its standing analysis to the challenge to S.B. 1003 in Counts Two and Three.
1 may not rule on the merits of a case without first determining that it has jurisdiction over 2 the category of claim in suit (subject-matter jurisdiction),” Sinochem Int’l Co. v. Malaysia 3 Int’l Shipping Corp., 549 U.S. 422, 430-31 (2007), the Court begins by addressing the 4 State’s standing challenge. 5 1. Legal Standard 6 “[S]tanding is an essential and unchanging part of the case-or-controversy 7 requirement of Article III.” Lujan v. Defenders of Wildlife, 504 U.S. 555, 560 (1992). 8 “[T]he irreducible constitutional minimum of standing contains three elements. First, the 9 plaintiff must have suffered an ‘injury in fact’—an invasion of a legally protected interest 10 which is (a) concrete and particularized and (b) actual or imminent, not conjectural or 11 hypothetical. Second, there must be a causal connection between the injury and the 12 conduct complained of—the injury has to be fairly traceable to the challenged action of the 13 defendant, and not the result of the independent action of some third party not before the 14 court. Third, it must be likely, as opposed to merely speculative, that the injury will be 15 redressed by a favorable decision.” Id. at 560-61 (cleaned up). 16 “[T]he ‘fairly traceable’ and ‘redressability’ components for standing overlap and 17 are ‘two facets of a single causation requirement.’” Washington Env’t Council v. Bellon, 18 732 F.3d 1131, 1146 (9th Cir. 2013) (citation omitted). However, they are distinct in that 19 traceability “examines the connection between the alleged misconduct and injury, whereas 20 redressability analyzes the connection between the alleged injury and requested relief.” Id. 21 Redressability is satisfied so long as the requested remedy “would amount to a significant 22 increase in the likelihood that the plaintiff would obtain relief that directly redresses the 23 injury suffered.” Renee v. Duncan, 686 F.3d 1002, 1013 (9th Cir. 2012). 24 2. The Parties’ Arguments 25 The State argues that “prior to S.B. 1003 passing, preexisting Arizona statutory law 26 affirmatively precluded counting mail-in ballots not signed or cured by poll-close time, 27 thereby barring a post-election cure period.” (Doc. 76 at 7.) Specifically, the State points 28 to A.R.S. §16-548(A), which provides that a ballot affidavit “must be received . . . [by]
1 7:00 pm on election day,” and A.R.S. § 16-552(B), which provides that “[i]f the affidavit 2 is insufficient, the vote shall not be allowed.” The State contends that the combination of 3 these two provisions means “that ballots must have arrived with their respective ballot 4 affidavits by poll-close time, and that if they are not sufficient then, the vote accordingly 5 ‘will not be allowed.’” (Id.) The State argues that “because Plaintiffs’ alleged injury is the 6 inability to cure non-signatures after election day, and even granting their requested relief 7 as to S.B. 1003 in its entirety would not allow them do so, Plaintiffs cannot satisfy their 8 burden to establish redressability.” (Id. at 8.) The State adds that Plaintiffs fundamentally 9 ask the Court to “conjure a law providing for . . . curing” because there “is no pre-existing 10 law to revert to that would permit such curing,” and because neither the Reconstruction 11 Amendments nor the VRA permit that result, the Court cannot redress Plaintiffs’ putative 12 injury. (Id. at 9.) Finally, the State contends that it “raised this precise argument in” Hobbs. 13 (Id. at 8.) 14 Plaintiffs respond that, regardless of the Court’s interpretation of preexisting 15 Arizona law, “a favorable ruling will redress the harm of racial discrimination,” which 16 itself constitutes an Article III injury. (Doc. 99 at 21.) Plaintiffs also argue that the State’s 17 interpretation of state law presents, “at best, a disputed issue.” (Id. at 22.) Plaintiffs assert 18 that the State’s proffered interpretation “conflates two separate processes” and “violates 19 bedrock rules of statutory construction.” (Id. at 22-23.) Plaintiffs also point to Defendant 20 Hobbs’s answer, which purportedly states that, “prior to SB 1003, nothing in Arizona law 21 prohibited election officials from allowing voters to cure unsigned early ballots after 22 election day.” (Id. at 22 [citing Doc. 63 ¶ 53].) Plaintiffs support Defendant Hobbs’s 23 position by reference to the 2021 EPM, which “implemented S.B. 1003 by requiring curing 24 of unsigned early ballots by election day. It does not cite any other, pre-existing Arizona 25 law as the basis of its authority to impose that requirement.” (Id. at 30.) Finally, Plaintiffs 26 argue that “the Court need not resolve these questions of Arizona law” because “[e]njoining 27 S.B. 1003 would, at the very least, leave Arizona officials and judges free to resolve
1 whether post-election curing is permitted under state law without being dictated to choose 2 one result by an unconstitutionally discriminatory law.” (Id. at 23.) 3 The State replies that Hobbs recognized that “Arizona always has imposed the 4 election-day deadline on voters to submit a signed ballot” and that Plaintiffs’ contrary 5 contentions fail. (Doc. 118 at 7-8.) The State also asserts that Plaintiffs’ arguments about 6 the 2021 EPM are categorically false.21 (Id. at 8-9.) The State also interprets Plaintiffs’ 7 arguments about “the harm of racial discrimination” as a “stigmatic injury” and argues that 8 because Plaintiffs do not allege they were personally subjected to discriminatory treatment, 9 and because bare discriminatory intent alone is “abstract stigmatic injury” that is not 10 cognizable, Plaintiffs lack standing. (Id. at 17-18.) 11 3. Analysis 12 Plaintiffs lack standing to pursue their challenges to S.B. 1003 in Counts Two and 13 Three. As discussed in Part I.F of the Background section of this order, Plaintiffs are 14 nonprofit organizations who allege that, because S.B. 1003 will make voting more difficult 15 for members of the communities they serve, they will need to divert their resources in 16 various ways to support educational and ballot-curing efforts. Similarly, Plaintiffs- 17 Intervenors are political organizations who allege that, because “[i]t is inevitable that 18 Democrats or those who would support Democratic candidates will not have their vote 19 counted as a result of” S.B. 1003, they will suffer “harms to their competitive interests” as 20 a result of S.B. 1003 and will also need to divert resources. (Doc. 55 ¶¶ 24-26.) At bottom, 21 all of these theories of harm are predicted on the notion that an order enjoining or otherwise 22 nullifying S.B. 1003 would, alone, make it easier for certain voters to have their votes 23 counted (because such voters could, but-for S.B. 1003, take advantage of post-election day 24 opportunities to cure missing signatures).22 25 21 Following the submission of the State’s reply, Plaintiffs filed a notice in which they 26 acknowledged that their response “mistakenly characterized the legal status of the draft 2021 EPM.” (Doc. 122 at 2.)
27 22 Plaintiffs and Plaintiffs-Intervenors do not seek an order affirmatively requiring Arizona election officials to permit the curing of missing signatures after election day—
28 the only forms of relief sought in their complaints are injunctive and declaratory relief directed at S.B. 1003. (Doc. 1 at 31; Doc. 55 at 31.) In contrast, the plaintiffs in Hobbs
1 The difficulty with this approach is that, even if S.B. 1003 were enjoined and 2 declared invalid, it is entirely speculative that an Arizona voter who submitted a mailed 3 ballot with a missing signature would be allowed to cure the missing signature after 4 election day. As noted in Hobbs, Arizona has never allowed unsigned affidavits to be cured 5 after election day. 18 F.4th at 1183 (“So far as the record in this case reveals, in the nearly 6 century of early voting in Arizona, no county recorder ever has allowed a voter to correct 7 a ballot with a missing signature after election day. Arizona always has imposed the 8 election-day deadline on voters to submit a signed ballot.”). Although the parties dispute 9 whether pre-existing statutes prevent county recorders from extending the cure period for 10 unsigned affidavits to five days after the election, it is apparently undisputed that no 11 preexisting statute requires county recorders to do so. 12 Plaintiffs argue that “[e]njoining SB 1003 would, at the very least, leave Arizona 13 officials and judges free to resolve whether post-election curing is permitted under state 14 law without being dictated to choose one result by an unconstitutionally discriminatory 15 law.” (Doc. 99 at 23.) But “[a] litigant must demonstrate . . . a substantial likelihood that 16 the judicial relief requested will prevent or redress the claimed injury to satisfy the ‘case 17 or controversy’ requirement.” Duke Power Co. v. Carolina Env’t Study Grp., Inc., 438 18 U.S. 59, 79 (1978). Thus, “[t]o establish redressability, a plaintiff must show that it is 19 ‘likely, as opposed to merely speculative, that the injury will be redressed by a favorable 20 decision.’” M.S. v. Brown, 902 F.3d 1076, 1083 (9th Cir. 2018) (quoting Lujan, 504 U.S. 21 at 561). Here, given that Arizona has never—throughout 110 years of statehood—allowed 22 post-election day curing of missing signatures, the Court cannot find that the simple 23 elimination of S.B. 1003 would, without more, “amount to a significant increase in the 24 likelihood that the plaintiff would obtain relief that directly redresses the injury suffered.” 25 Mecinas v. Hobbs, 30 F.4th 890, 900 (9th Cir. 2022). 26 27 sought a declaration that “all voters who submit a ballot without a signature must be allowed the same opportunity to cure that defect as is allowed to voters who submit a mail 28 ballot with a signature mismatch.” Complaint for Injunctive and Declaratory Relief, Hobbs, No. 2:20-CV-01143, 2020 WL 5535933 (D. Ariz. 2020).
1 In many respects, the situation here mirrors the situation in Arizonans for Fair 2 Elections v. Hobbs, 454 F. Supp. 3d 910 (D. Ariz. 2020). There, although the plaintiffs 3 sought to challenge certain Arizona statutes that require in-person signature gathering for 4 initiative petitions, they did not challenge the provisions of the Arizona constitution that 5 impose the same requirements. Id. at 917. As a result, the State argued that the plaintiffs 6 had failed to establish redressability. Id. (“In the State’s view, this creates a standing 7 problem—even if Plaintiffs succeed in arguing that Title 19 is unconstitutional, the Arizona 8 constitution would stand and Plaintiffs’ injury would not be redressed.”). In response, the 9 plaintiffs argued that “the requested relief would still redress their injury because, once [the 10 statutory] requirements are stripped away, the Arizona courts would be free to” rule in their 11 favor. Id. at 917-18. This Court disagreed, holding that because it was “entirely 12 speculative that the Arizona courts” would eliminate the requirement of in-person signature 13 gathering based solely on the invalidation of the challenged statutes, if followed that 14 plaintiffs had failed to establish redressability. Id. at 918-20. Afterward, the Ninth Circuit 15 denied the plaintiffs’ request for emergency relief for the same reason, holding that 16 “Appellants, having failed to challenge the Arizona constitutional requirement of in-person 17 signatures, cannot get the redress from the court they now seek by only challenging the 18 statute at issue.” Arizonans for Fair Elections v. Hobbs, 9th Cir. No. 20-15719, Dkt. 37 19 (May 5, 2020). 20 Here, although there is no provision of the Arizona constitution that specifically 21 prohibits the curing of missing signatures after election day, other statutes (which existed 22 before S.B. 1003 was enacted) arguably prohibit such curing and Hobbs makes clear that 23 such curing has never been allowed in Arizona. Under these circumstances, it is “merely 24 speculative,” as opposed to “likely,” that an order nullifying S.B. 1003 would result in a 25 change to Arizona’s historical practice of prohibiting the curing of missing signatures after 26 election day. M.S., 902 F.3d 1076 at 1083.23 27 23 There is no merit to Plaintiffs’ contention that the fact of S.B. 1003’s enactment shows the Arizona legislature intended to change existing law. (Doc. 99 at 22-23.) The 28 legislature expressly noted in the text of S.B. 1003 that it was clarifying existing law, not changing it. See 2021 Ariz. Legis. Serv. Ch. 343 (S.B. 1003), available at
1 For these reasons, this case is distinguishable from Mecinas. There, the plaintiffs 2 challenged Arizona’s “Ballot Order Statute” under the theory that “it gives candidates the 3 benefit of appearing first on the ballot, not on the basis of some politically neutral ordering 4 (such as alphabetically or by lot), but on the basis of political affiliation.” 30 F.4th at 894. 5 In response, the defendants argued that redressability was lacking because the Secretary of 6 State’s “ability to adhere to a court’s injunction may be stymied by the governor or the 7 attorney general, both of whom must approve the [EPM] before it can go into effect.” Id. 8 at 900. The Ninth Circuit disagreed, holding that the defendants’ redressability concerns 9 were “of no moment” because “an injunction against the Secretary would ‘significant[ly] 10 increase’ the likelihood of relief.” Id. In reaching this conclusion, the court emphasized 11 that, in previous decisions, it had recognized “that a challenged Arizona election law [is] 12 traceable to the Secretary.” Id. But here, the issue isn’t whether some unspecified Arizona 13 official might, following a ruling in Plaintiffs’ favor regarding the validity of S.B. 1003, 14 attempt to obstruct Defendants’ efforts to authorize post-election day curing of unsigned 15 affidavits. Rather, the issue is that post-election day curing of unsigned affidavits would 16 remain impermissible following such a ruling due to Arizona’s settled, century-long 17 practice of disallowance (as recognized in Hobbs). 18 Alternatively, Plaintiffs contend that “being subject to a racially discriminatory law 19 inherently constitutes an Article III injury, and an injunction against such a law remedies 20 that injury.” (Doc. 99 at 21.) But even if Plaintiffs have associational standing to assert 21 such an injury on behalf of the members of the communities they serve,24 the stigmatizing 22 https://www.azleg.gov/legtext/55leg/1R/bills/SB1003S.pdf (“The legislature intends that 23 the amendments made by this act to 44 sections 16-547 and 16-550, Arizona Revised Statutes, are clarifying changes only and do not provide for any substantive change in the 24 law.”). The Arizona courts have held that the legislature may engage in such acts of clarification. Enter. Leasing Co. of Phoenix v. Ariz. Dep’t. of Revenue, 211 P.3d 1, 4 (Ariz.
25 Ct. App. 2008) (“In this case, the Legislature indicated its curative intent as directly as possible. It stated that the revisions were intended to be ‘clarifying changes and are 26 consistent with the legislature’s intent when those sections were enacted’ . . . .”).
24 27 The tentative ruling stated that Plaintiffs could not assert such an injury because their complaints did not raise claims of associational standing. Plaintiffs’ (and Plaintiffs-
28 Intervenors’) presentation during oral argument persuades the Court that this component of the tentative ruling may have been incorrect.
1 effect of a racially discriminatory law is not, alone, an injury that is sufficient to confer 2 standing on any plaintiff who wishes to challenge the law. Instead, a plaintiff seeking to 3 challenge a law under the theory that it was enacted with discriminatory intent must show 4 that he has suffered (or will suffer) an injury as a consequence of the law that will be 5 redressed by a favorable decision. As for the reasons stated above, Plaintiffs cannot make 6 such a showing here. 7 The decision in National Association for Advancement of Colored People v. Horne, 8 2013 WL 5519514 (D. Ariz. 2013), supports this conclusion. There, various civil rights 9 organizations asserted a constitutional challenge to an Arizona statute that prohibits race- 10 and gender-based abortions. Id. at *1-2. Although the organizations did “not claim that 11 their members have been denied abortions because of the Act or face prosecution or 12 liability under the Act,” they argued they had associational standing due to “the fact that 13 the Act stigmatizes and denigrates their members on the basis of race and gender.” Id. at 14 *4. When evaluating this claim, the district court looked to the Supreme Court’s decisions 15 in Allen v. Wright, 468 U.S. 737 (1984), and Heckler v. Mathews, 465 U.S. 728 (1984). 16 From those decisions, the court distilled the principle that “stigmatizing injury alone is not 17 sufficient for standing in equal protection cases. Plaintiffs must also show that their 18 members personally have been denied equal treatment by the Act.” Id. at *4-5. Put another 19 way, only “a person actually denied benefits under a law has standing to challenge that law 20 on equal protection grounds.” Id. at *5. Because “[s]tigmatic injury does not suffice under 21 Allen, and Plaintiffs fail[ed] to identify any personal injury suffered by them as a 22 consequence of the alleged constitutional error, other than the psychological consequence 23 presumably produced by observation of conduct with which one disagrees,” the court 24 concluded that the organizations lacked standing to pursue their challenge. Id. at *8 25 (citation and internal quotation marks omitted). The Ninth Circuit affirmed, agreeing with 26 the district court that a “stigmatic injury caused by being a target of official discrimination 27 is not itself a personal denial of equal treatment” and that although “Plaintiffs purport to 28 present an alternate basis for standing resulting from being ‘the targets of . . . discriminatory
1 intent’ . . . [t]hat theory is a mere repetition of Plaintiffs’ stigmatic injury, which does not 2 support standing.” Nat’l Ass’n for Advancement of Colored People v. Horne, 626 F. App’x 3 200, 201 (9th Cir. 2015). 4 Here, Plaintiffs’ assertion that “being subject to a racially discriminatory law 5 inherently constitutes an Article III injury” is simply another variant of the stigma argument 6 rejected in Horne. Under Allen and its progeny, standing requires a redressable personal 7 injury arising from the law apart from the stigmatizing effect that inherently flows from 8 the enactment of a discriminatory law. See also Cato v. United States, 70 F.3d 1103, 1109- 9 10 (9th Cir. 2005) (“No plaintiff has standing ‘to complain simply that their Government 10 is violating the law.’ Neither does Cato have standing to litigate claims based on the 11 stigmatizing injury to all African Americans caused by racial discrimination.”) (quoting 12 Allen, 468 U.S. at 755).25 Because Plaintiffs have not established such a redressable injury 13 here, they lack standing to pursue their challenges to S.B. 1003 in Counts Two and Three. 14 DaimlerChrysler Corp. v. Cuno, 547 U.S. 332, 352 (2006) (“A plaintiff must demonstrate 15 standing for each claim he seeks to press.”).26 16 D. Counts Two And Three—S.B. 1485 17 a. Legal Standard 18 A legislature acts in violation of the Fourteenth and Fifteenth Amendments when “a 19 discriminatory purpose [is] a motivating factor” in the legislature’s action. Vill. of 20 25 Ellis v. Costco Wholesale Corp., 657 F.3d 970 (9th Cir. 2012), which Plaintiffs cited during oral argument, is not to the contrary. There, plaintiff Sasaki had standing to pursue 21 a claim that “Costco’s promotional practices discriminate based on gender” in part because Sasaki had “been denied promotion to GM” as a result of the challenged practices and 22 “Costco’s failure to promote Sasaki (a current AGM) to GM is a ‘concrete and particularized’ harm, because the injury affects her ‘in a personal and individual way.’” Id.
23 at 974, 978-79. Additionally, “[b]ecause Costco’s challenged promotion practices for GM and AGM positions have not changed, there is a likelihood that Costco will continue to fail 24 to promote Sasaki . . . [and] if Costco’s allegedly discriminatory policies are enjoined, Sasaki’s claimed threat of being passed over for promotion due to her gender is likely to 25 be redressed.” Id. at 979. These are precisely the sort of concrete, non-speculative showings of redressability that are lacking here.
26 26 Intervenor-Plaintiffs also assert an injury to their “competitive standing.” (Doc. 55 27 ¶ 21.) In Mecinas, the Ninth Circuit held that an injury to competitive standing is not redressable if the court’s action will not significantly increase the likelihood of relief. 30 28 F.4th at 898-99. Thus, Intervenor-Plaintiffs’ competitive standing injury is not redressable for the reasons set forth above.
1 Arlington Heights v. Metro. Hous. Dev. Corp., 429 U.S. 252, 265-66 (1977). A plaintiff 2 asserting a discriminatory purpose claim need not prove that “the challenged action rested 3 solely on racially discriminatory purposes . . . ; racial discrimination is not just another 4 competing consideration. When there is proof that a discriminatory purpose has been a 5 motivating factor in the decision, . . . judicial deference is no longer justified.” Id. 6 (emphasis added). In Arlington Heights, “[t]he Supreme Court articulated the following, 7 non-exhaustive factors that a court should consider in assessing whether a defendant acted 8 with discriminatory purpose: (1) the impact of the official action and whether it bears more 9 heavily on one race than another; (2) the historical background of the decision; (3) the 10 specific sequence of events leading to the challenged action; (4) the defendant’s departures 11 from normal procedures or substantive conclusions; and (5) the relevant legislative or 12 administrative history.” Arce v. Douglas, 793 F.3d 968, 977 (9th Cir. 2015). A plaintiff 13 need not establish any particular element in order to prevail. Pac. Shores Properties, LLC 14 v. City of Newport Beach, 730 F.3d 1142, 1156 (9th Cir. 2013) (“[A]ny indication of 15 discriminatory motive may suffice to raise a question that can only be resolved by a 16 factfinder.”). 17 Like the Fourteenth and Fifteenth Amendments, Section 2 of the Voting Rights Act 18 (“VRA § 2”) prohibits voting laws and practices adopted with a discriminatory purpose. 19 Chisom v. Roemer, 501 U.S. 380, 394 n.21 (1991). Thus, a showing of intent “sufficient 20 to constitute a violation of the [F]ourteenth [A]mendment” also suffices “to constitute a 21 violation of [S]ection 2.” McMillan v. Escambia Cnty., 748 F.2d 1037, 1046 (5th Cir. 22 1984). See also Democratic Nat’l Committee v. Hobbs, 948 F.3d 989, 1038 (9th Cir. 2020), 23 rev’d on other grounds by Brnovich, 141 S. Ct. at 2321 (“[Arlington Heights] provides the 24 framework for analyzing a claim of intentional discrimination under Section 2.”). 25 b. The Parties’ Arguments 26 The State argues that “Plaintiffs’ allegations fall short of plausibly establishing 27 discriminatory intent, especially when considered against the background of the strong 28 presumption of good faith for state legislatures.” (Doc. 76 at 11.) The State asserts that
1 Plaintiffs’ claims “rely almost exclusively on mere ‘awareness’ of disparate impacts, [and 2 thus] they have failed to plead a cognizable intentional discrimination claim.” (Id. at 12.) 3 The State also argues that Plaintiffs’ allegations, which are limited to (1) historical 4 discrimination in Arizona, (2) ambiguous legislator statements, and (3) discriminatory 5 effect, are insufficient to defeat the presumption of legislative good faith. (Id. at 12-14.) 6 In particular, the State urges the Court to reject application of the “cat’s paw” theory27 to 7 legislator statements. (Id. at 14-15.) 8 Plaintiffs respond that they have pleaded a cognizable intentional discrimination 9 claim under Arlington Heights by alleging facts that touch upon the discriminatory impact 10 of the challenged legislation, legislative history and contemporaneous statements, 11 precipitating events and departures from practice, and Arizona’s history of discrimination 12 preceding the official action. (Doc. 99 at 5-9.) Plaintiffs also argue that this Court need 13 not give excessive deference to legislative good faith at the motion-to-dismiss stage, 14 otherwise “no plaintiff could state an intentional discrimination claim in which legislators 15 were smart enough not to announce their discriminatory animus.” (Id. at 11.) On that note, 16 Plaintiffs assert that Representative Kavanagh’s statement “suggests a discriminatory 17 purpose and gives the lie to the notion that the legislature was really concerned about 18 election integrity,” and even if the Court were otherwise inclined to credit Representative 19 Kavanagh’s attempt to “pass it off as something else,” the Court could not do so at the 20 pleading stage where all inferences must be drawn in their favor. (Id.) Plaintiffs also assert 21 that the Supreme Court’s dismissal of the “cat’s paw” theory for legislators is irrelevant 22 when considering the plausibility of the complaint’s allegations. (Id. at 12.) 23 27 “The term ‘cat’s paw’ derives from a fable conceived by Aesop . . . and injected into 24 United States employment discrimination law by Judge Posner in 1990. . . . In the fable, a monkey induces a cat by flattery to extract roasting chestnuts from the fire. After 25 the cat has done so, burning its paws in the process, the monkey makes off with the chestnuts and leaves the cat with nothing. [The fable] observes that the cat is similar to 26 princes who, flattered by the king, perform services on the king’s behalf and receive no reward.” Staub v. Proctor Hosp., 562 U.S. 411, 416 n.1 (2011). See also Brnovich, 141 27 S. Ct. at 2325 (“The Court of Appeals concluded that the District Court committed clear error by failing to apply a ‘cat’s paw’ theory—which analyzes whether an actor was a 28 ‘dupe’ who was ‘used by another to accomplish his purposes.’ That theory has its origin in employment discrimination cases and has no application to legislative bodies.”).
1 In its statement of interest, the United States argues that “Plaintiffs’ allegations 2 regarding the foreseeable discriminatory impact SB 1003 and SB 1485 will have on Native 3 American, Latino, and Black voters . . . when viewed as a whole, raise a strong inference 4 that the adverse effects were desired, weighing in favor of a finding of discriminatory 5 purpose.” (Doc. 78 at 12.) The United States further asserts that “whether [the presumption 6 of legislative good faith] stands or yields is an inherently fact-based question best suited 7 for the merits stage of litigation.” (Id. at 14.) 8 In reply, the State argues that “most of Plaintiffs’ key allegations are conclusory” 9 and that the remaining “nonconclusory allegations fail to plausibly allege intentional 10 discrimination.” (Doc. 118 at 24-25.) According to the State, statistical patterns supporting 11 Plaintiffs’ disparate impact theory are explainable on grounds other than race; the 12 legislative statements at issue here cannot impugn the entire legislature and are less 13 probative than other statements held to be insufficient; the alleged “departures from 14 practice” do not qualify as the sort of procedural irregularities described by Arlington 15 Heights; and the history cited by Plaintiffs “does nothing to advance their claim of 16 plausibility.” (Id. at 26-28.) Finally, the State emphasizes the “extraordinary caution” the 17 Court should apply when adjudicating claims that a state has acted on the basis of race and 18 argues that Plaintiffs have failed to bring forth the “specific, non-conclusory allegations” 19 to defeat the presumption of legislative good faith, even at the pleading stage. (Id. at 29.) 20 c. Analysis 21 To survive the State’s motion to dismiss under Rule 12(b)(6), Plaintiffs must allege 22 sufficient factual matter, accepted as true, to state a claim of intentional racial 23 discrimination that is plausible on its face. Iqbal, 556 U.S. at 678. Although it presents a 24 somewhat close call, the Court concludes that Plaintiffs have met this threshold. 25 First, “an invidious discriminatory purpose may often be inferred from the totality 26 of the relevant facts, including the fact, if it is true, that the law bears more heavily on one 27 race than another.” Washington v. Davis, 426 U.S. 229, 242 (1976). As discussed in Part 28 II.B.2 above, Plaintiffs plausibly allege that voters of color will be disproportionately
1 affected by S.B. 1485. Although the alleged burden is not significant or even moderate for 2 purposes of the first step of the Anderson-Burdick inquiry, it will still (according to 3 Plaintiffs’ well-pleaded allegations) be felt disproportionately by protected groups. 4 None of the cases cited by the State—Brnovich, Regents of the Univ. of California 5 v. DHS, 140 S. Ct. 1891 (2020), and Ramos v. Wolf, 975 F.3d 872 (9th Cir. 2020)—compels 6 a contrary conclusion. In Brnovich, the Supreme Court held that a facially neutral voting 7 rule with a long pedigree that reasonably advanced important state interests should not be 8 “taken down” by the mere fact of disparate outcomes that could be explainable by 9 differences in employment, wealth, and education. 141 S. Ct. at 2343. But this analysis 10 appeared in the portion of Brnovich addressing the plaintiffs’ disparate impact claim under 11 VRA § 2. In the separate portion of the decision addressing the plaintiffs’ discriminatory 12 purpose claim, the Court simply noted that “the law’s impact on different racial groups” is 13 one of multiple factors courts must consider when evaluating such a claim before affirming 14 the district court’s overall finding that the challenged law “was not enacted with a racially 15 discriminatory purpose.” Id. at 2348-50. Here, because Plaintiffs have plausibly alleged 16 that S.B. 1485 does, in fact, have a disparate impact on certain racial groups, they have 17 plausibly alleged one of the considerations that might, in combination with other 18 considerations, support a discriminatory purpose claim under Arlington Heights and VRA 19 § 2. 20 In Regents, the Supreme Court addressed whether the rescission of the Deferred 21 Action for Childhood Arrivals (“DACA”) program violated the Fifth Amendment because 22 it was “motivated by animus.” 140 S. Ct. at 1915-16. The Court concluded that the 23 plaintiffs had not sufficiently pleaded such a claim because their three proffered indicia of 24 animus—“(1) the disparate impact of the rescission on Latinos from Mexico, who represent 25 78% of DACA recipients; (2) the unusual history behind the rescission; and (3) pre- and 26 post-election statements by President Trump”—did not “either singly or in concert[] 27 establish[] a plausible equal protection claim.” Id. With respect to the first factor, the 28 Court stated that “because Latinos make up a large share of the unauthorized alien
1 population, one would expect them to make up an outsized share of recipients of any cross- 2 cutting immigration relief program. Were this fact sufficient to state a claim, virtually any 3 generally applicable immigration policy could be challenged on equal protection grounds.” 4 Id. (citations omitted). But here, the State has not identified a similar “unsurprising” reason 5 why S.B. 1485 would have a disparate impact on minorities. This case involves a voting 6 regulation, not an immigration policy. Additionally, the Court does not construe the cited 7 passage from Regents as a holding that a showing of disparate impact can never form one 8 component of a discriminatory purpose claim under Arlington Heights. Rather, Regents 9 holds that a showing of disparate impact alone is insufficient to state a plausible 10 discriminatory purpose claim. Here, Plaintiffs do not rely solely on their allegations of 11 disparate treatment but identify those allegations as one component of a multi-factor 12 showing. 13 Finally, in Ramos, the plaintiffs raised a discriminatory purpose challenge to the 14 government’s decision to terminate temporary protected status (“TPS”) for refugees from 15 certain “non-white, non-European” countries. 975 F.3d at 898. The Ninth Circuit held that 16 this claim “fail[ed] predominantly due to the glaring lack of evidence tying the President’s 17 alleged discriminatory intent to the specific TPS terminations.” Id. at 897. Additionally, 18 the court held that the plaintiffs’ purported evidence of disparate impact did “not help 19 [them] much” because “[w]hile the four countries at issue in this case are ‘non-European’ 20 with predominantly ‘non-white’ populations, the same is true for the four other countries 21 whose TPS designations were extended by the Trump Administration during the same 22 period. In fact, virtually every country that has been designated for TPS since its inception 23 has been ‘non-European’ . . . and most have majority ‘non-white’ populations. Under the 24 district court’s logic, almost any TPS termination in the history of the program would bear 25 ‘more heavily’ on ‘non-white, non-European’ populations and thereby give rise to a 26 potential equal protection claim. This cannot be the case, as the Supreme Court recently 27 pointed out in rejecting the disparate impact argument in Regents.” Id. at 898. But once 28 again, the situation is very different here.
1 Second, the legislative history of a statute, “especially where there are contemporary 2 statements by members of the decisionmaking body,” is sometimes probative when 3 evaluating a discriminatory purpose claim. Arlington Heights, 429 U.S. at 266. Here, 4 Plaintiffs point to statements by Representatives Grantham and Kavanagh. 5 On the one hand, even when viewing it in the light most favorable to Plaintiffs, the 6 Court does not see how Representative Grantham’s statement could be plausibly viewed 7 as betraying racial animus. The context is that Representative Bolding, a black man who 8 is the minority leader of the Arizona House, made a speech on the House floor. (Doc. 55 9 ¶ 114.) Representative Bolding stated that S.B. 1485 would make it harder for 10 “independent voters, seniors, Native Americans, Black, brown and low income people to 11 vote.” (Id.) In response, Representative Grantham stated, “I feel personally that motives 12 were [attributed to] members, including myself with regards to colored people, Black 13 people, whatever people this individual wants to single out and their ability to vote . . . I 14 think he should be sat down and he shouldn’t be allowed to speak.” (Id.) This statement 15 constitutes a denial by Representative Grantham that supporters of S.B. 1485 were 16 motivated by any sort of racial animus. It would be bizarre if such a denial could be twisted 17 into evidence of racial animus. Nor does a plausible inference of racial animus arise from 18 the fact that Representative Grantham is white and Representative Bolding is black. 19 Members of different races may disagree with each other about the merits of disputed 20 issues, including race-related issues, without raising an inference of racial animus. Finally, 21 without in any way defending Representative Grantham’s use of the phrase “colored 22 people,” that phrase’s inclusion in a broader statement denying the existence of racial 23 animus among supporters of S.B. 1485 does not—even when viewed in the light most 24 favorable to Plaintiffs—raise a plausible inference that Representative Grantham secretly 25 meant the exact opposite of what he was saying. 26 On the other hand, Representative Kavanagh’s statement, viewed in the light most 27 favorable to Plaintiffs, can be construed as supporting their claim. As noted, 28 Representative Kavanagh stated in an interview that:
1 Democrats value as many people as possible voting, and they’re willing to 2 risk fraud. Republicans are more concerned about fraud, so we don’t mind putting security measures in that won’t let everybody vote—but everybody 3 shouldn’t be voting. 4 *** 5 Not everybody wants to vote, and if somebody is uninterested in voting, that probably means that they’re totally uninformed on the issues. Quantity is 6 important, but we have to look at the quality of votes, as well. 7 Although Representative Kavanagh first addressed voter fraud, he then transitioned, for 8 unexplained reasons, into an argument for reducing the voting population based on the 9 “quality” of voters. There may be innocent reasons for this segue that will be revealed 10 during future stages of this case, but the latter part of Representative Kavanagh’s statement 11 could be viewed, when construed in the light most favorable to Plaintiffs (and in light of 12 Plaintiffs’ other allegations suggesting that it was well known that S.B. 1485 would 13 disproportionately affect minorities), as expressing the discriminatory trope that minorities 14 are uneducated voters.
15 Of course, “[c]ourts must use caution . . . when seeking to glean a legislature’s 16 motivations from the statements of a handful of lawmakers.” United States v. Machic-
17 Xiap, 552 F. Supp. 3d 1055, 1062 (D. Or. 2021) (citing Brnovich, 141 S. Ct. at 2349-50.)
18 The State argues that, even if the proffered statements are viewed unfavorably, Brnovich 19 compels this Court to reject a “cat’s paw” theory, which would attribute an individual 20 legislator’s statements to the legislature. “As a matter of law, then, one statement by one 21 legislator cannot be used to impute discriminatory purpose to the entire legislature.” (Doc.
22 118 at 5.)
23 This argument is unavailing. If contemporaneous statements made by legislators 24 can still be some evidence of discriminatory intent (and Brnovich does not suggest 25 otherwise), the State’s arguments are premature. In Brnovich, controversy surrounded a 26 “racially-tinged” video promoted by Representative Don Shooter, which allegedly 27 prompted the enactment of ballot collection legislation. 141 S. Ct. at 2349-50. The 28 Supreme Court upheld the district court’s conclusion, after a bench trial, that “the racially-
1 tinged video helped spur the debate about ballot collection, [but there was] no evidence 2 that the legislature as a whole was imbued with racial motives.” Id. at 2349-50. If the 3 Brnovich district court could only definitively interpret the import of the Shooter video 4 after a bench trial in which the State produced evidence that the legislature as a whole was 5 not imbued with racial motives, it would be inappropriate to short-circuit that inquiry here 6 by conclusively determining at the pleading stage that, notwithstanding Representative 7 Kavanagh’s statement, the legislature as a whole was not imbued with racial motives. As 8 the United States points out, “most of the relevant facts about . . . the purposes animating 9 . . . SB 1485 are possessed solely by the State, its counties, and other governmental actors. 10 Discovery may help illuminate those purposes.” (Doc. 78 at 15.) At this stage, a 11 contemporaneous statement that can be interpreted as “racially-tinged” in the light most 12 favorable to Plaintiffs, in concert with Plaintiffs’ other well-pleaded allegations, provides 13 plausible support for Plaintiffs’ overall claim. 14 Third, “[c]ourts may also consider the specific sequence of events leading to the 15 challenged action.” United States v. Carrillo-Lopez, 555 F. Supp. 3d 996, 1008 (D. Nev. 16 2021) (citing Arlington Heights, 429 U.S. at 265-68). Courts distinguish a “strange about- 17 face,” which might be evidence of discriminatory intent, from a “natural response to a 18 newly identified problem.” Regents, 140 S. Ct. at 1916. 19 Plaintiffs allege that “[i]n an effort to bolster false and discredited claims of fraud, 20 and provide a veneer of legitimacy for its discriminatory voter suppression efforts, the 21 Arizona Senate hired a Florida corporation called Cyber Ninjas, Inc. to conduct an ‘audit’ 22 of the 2020 election results in Maricopa County.” (Doc. 1 ¶ 58.) The State does not directly 23 address Plaintiffs’ mention of Cyber Ninjas but obliquely responds that “[t]hese are not the 24 sort of procedural irregularities that Arlington indicated would form meaningful 25 circumstantial evidence of intentional discrimination. Arlington instead referred to 26 departures from practice—i.e., changes in ordinary procedure that indicated some 27 invidious motive was at work.” (Doc. 118 at 28.)
1 The Court need not credit Plaintiffs’ conclusory assertion that the Cyber Ninjas 2 audit was “an effort to bolster false and discredited claims of fraud and provide a veneer of 3 legitimacy for its discriminatory voter suppression efforts” to conclude that the alleged 4 facts could give a reasonable factfinder some pause. The well-pleaded factual allegations, 5 which the Court must accept at this stage of the proceedings, reflect that there was no 6 legitimate evidence supporting the need for an audit, that several Arizona governmental 7 agencies vehemently confirmed the accuracy of the vote count shortly after the election, 8 and that Cyber Ninjas has no experience auditing elections. (Doc. 1 ¶¶ 57-59.) Plaintiffs 9 have thus plausibly alleged that the Senate’s decision to hire Cyber Ninjas with no evidence 10 of fraud and against the counsel of state experts was, to paraphrase the State, a “change[] 11 in ordinary procedure that indicated some invidious motive was at work.” (Doc. 118 at 12 28.) And as Plaintiffs pointed out during oral argument, this episode is plausibly (if 13 indirectly) related to the enactment of S.B. 1485 because one of the proffered justifications 14 for that law was the need to enhance voter integrity and combat electoral fraud. If that 15 justification was pretextual, as Plaintiffs allege, this can plausibly be viewed as 16 circumstantial evidence supporting Plaintiffs’ contention that one of the true, unexpressed 17 motivations for the law was discriminatory. Cf. Reeves v. Sanderson, 530 U.S. 133, 147 18 (2000) (“Proof that the defendant’s explanation is unworthy of credence is simply one form 19 of circumstantial evidence that is probative of intentional discrimination, and it may be 20 quite persuasive.”). 21 Finally, as for the State’s contention that any challenge to S.B. 1485 is barred by the 22 presumption of legislative good faith (Doc. 76 at 2, 11-14; Doc. 118 at 3-5, 29, 32), it is 23 not clear whether this presumption is a separate hurdle Plaintiffs must clear beyond 24 plausibly alleging intentional racial discrimination. Some passages in Arlington Heights 25 suggest the presumption is already incorporated into the framework for evaluating 26 intentional-discrimination claims: “When there is proof that a discriminatory purpose has 27 been a motivating factor in the decision, . . . judicial deference is no longer justified.” 429 28 U.S. at 265-66. Cf. Regents of Univ. of California v. Bakke, 438 U.S. 265, 319 (1978) (“In
1 short, good faith would be presumed in the absence of a showing to the contrary in the 2 manner permitted by our cases. See, e.g., Arlington Heights . . . .”). Nevertheless, even if, 3 as some courts have concluded, the presumption of legislative good faith stands as an 4 additional hurdle,28 that hurdle has been cleared here for Rule 12(b)(6) purposes. As noted, 5 Plaintiffs do not rely solely on the presence of disparate impacts or the statement of a single 6 legislator or the existence of a “strange about-face”—factors that, if considered 7 individually, would be insufficient to support a claim of intentional discrimination—but 8 instead rely on the collective weight of those factors. 9 For these reasons, Plaintiffs have pleaded a plausible purposeful discrimination 10 challenge to S.B. 1485 under the United States Constitution and under VRA § 2.29 11 III. Leave To Amend 12 Although Plaintiffs and Plaintiffs-Intervenors do not expressly request leave to 13 amend in their responses to the motion to dismiss (Docs. 99, 100), the State contends in its 14 reply that “[g]iven the Ninth Circuit’s mandate that plaintiffs should generally be given at 15 least one chance at amendment, the State does not oppose leave for Plaintiffs to attempt to 16 cure these deficiencies.” (Doc. 118 at 29.) Given this backdrop, and in light of the Ninth 17 Circuit’s recognition that, in certain circumstances, “a district court should grant leave to 18 amend even if no request to amend the pleading was made,” Ebner v. Fresh, Inc., 838 F.3d 19 958, 963 (9th Cir. 2016) (citation omitted), the Court will sua sponte grant leave to amend 20 to Plaintiffs and Plaintiffs-Intervenors. 21 … 22 … 23 28 See, e.g., League of Women Voters of Fla., Inc. v. Fla. Sec. of State, 32 F.4th 1363, 1373-74 (11th Cir. 2022) (“[T]he district court failed to properly account for what might 24 be called the presumption of legislative good faith. . . . [I]n its 288-page opinion, the district court never once mentioned the presumption. And while we do not require courts to incant 25 magic words, it does not appear to us that the district court here meaningfully accounted for the presumption at all.”).
26 29 The Court perceives no tension between this conclusion and the conclusion that 27 Plaintiffs’ challenge to S.B. 1485 in Count One fails to state a claim. Although the failure to plausibly allege anything more than a minimal burden on the right to vote may be fatal 28 to a claim under Anderson-Burdick, that factor forms only one piece of the puzzle for purposes of a claim of purposeful discrimination under Arlington Heights and VRA § 2.
1 Accordingly, 2 IT IS ORDERED that the State’s motion to dismiss (Doc. 76) is granted in part 3 and denied in part. 4 IT IS FURTHER ORDERED that Plaintiffs and Plaintiffs-Intervenors may file 5 amended complaints within 21 days of the issuance of this order. Any changes shall be 6 limited to attempting to cure the deficiencies raised in this order and Plaintiffs and 7 Plaintiffs-Intervenors shall, consistent with LRCiv 15.1(a), attach a redlined version of the 8 pleading as an exhibit. 9 Dated this 24th day of June, 2022.
Mi Familia Vota v. Fontes (Mi Familia Vota v. Fontes) — published by Counsel Stack Legal Research, free access to 12M+ legal documents.