From Casetext: Smarter Legal Research

Dudum v. City County of San Francisco

United States District Court, N.D. California, San Francisco Division
Sep 9, 2010
No. C 10-00504 RS (N.D. Cal. Sep. 9, 2010)


No. C 10-00504 RS.

September 9, 2010



Ron Dudum and five other registered voters in San Francisco seek injunctive relief against the City and County of San Francisco, its Department of Elections (the "Department"), and Director of Elections John Arntz. Plaintiffs argue San Francisco's "restricted" instant runoff system ("restricted IRV") employed in certain municipal elections arbitrarily and illegally prevents a significant number of voters from having their votes counted. The plaintiffs contend the selection system, facially and as applied, runs afoul of their First Amendment rights to free association, the equal protection and due process clauses of the Fourteenth Amendment, and the Civil Rights Act, 42 U.S.C. § 1983. This Court denied plaintiffs' motion for a preliminary injunction on April 16, 2010, finding that they had not demonstrated a likelihood of success on the merits. See Dudum v. San Francisco, No. 10-0504, 2010 WL 1532365 (N.D. Cal. Apr. 16, 2010).

The parties now file cross motions for summary judgment. Both sides agree that the material facts are not in dispute and insist judgment as a matter of law is appropriate. What each party seeks is a determination of the constitutionality of restricted IRV in San Francisco generally, or at least as applied in races with more than four candidates. Although the underlying factual and legal arguments were "rehearsed," as it were, at the preliminary injunction phase earlier this year, the parties have now honed their constitutional analyses and further developed the factual record.

Defendants seek leave to file a "rebuttal" expert study. Plaintiffs object to the supplemental filing as untimely and further argue that if considered, issues of fact may arise which would warrant further evidentiary proceedings. As the supplemental report was not timely presented, it will not be considered in ruling on the parties' cross motions.

This Court previously found that, in the form adopted by San Francisco, the burden restricted IRV imposes on a voter's associational right is not severe and its effects are not fundamentally unfair. A review of the record as it now stands mandates the reaffirmation of that conclusion. The City has pointed (as it must) to several important interests that justify the burden imposed by its version of restricted IRV. Accordingly, defendants' motion for summary judgment must be granted and plaintiffs' cross-motion for summary judgment correspondingly denied.


The City's Department of Elections oversees all federal, state and local elections in San Francisco. The Department serves more than 450,000 registered voters and controls roughly 560 polling sites each election. Sixteen permanent employees maintain the work of the Department throughout most of the year. These officers rely on more than 3,000 temporary workers, come election time. The Department performs myriad tasks in each electoral cycle: it orders ballot supplies; prepares and prints ballots; mails overseas and other absentee ballots; prints and distributes information to voters; trains the election cycle's batch of temporary employees and poll workers; tests the accuracy of every voting machine; counts and tabulates votes cast in the election; and conducts a mandatory, post-election hand count to verify results.

Eight years ago, the citizens of San Francisco County adopted Proposition A, which introduced a new electoral structure for certain municipal elections. That proposition replaced the two-step election system then in place with instant runoff voting (variously dubbed instant runoff or ranked choice voting and abbreviated in the parties' papers as "IRV" or "RCV"). Before the 2002 change, the City's Board of Supervisors, Sheriff, Mayor, District Attorney, Public Defender and several others were selected via a general election followed by a runoff conducted on a later date between the two candidates who garnered the most votes (assuming no candidate received majority support in the first instance).

Proposition A introduced a ranked choice system whereby all voters may rank candidates by order of preference. Once the polls close, the Department (or, rather, its voting machines) counts all first choice rankings. If a candidate receives a majority of votes, he or she wins the election. If not, the candidate who received the fewest first choice rankings is eliminated. The second choice rankings of voters who supported the eliminated candidate are then redistributed and the Department "retabulates" all votes. If no candidate with majority support emerges, the process repeats.

A handful of other cities in the United States employ some variant of instant runoff voting. Cambridge, Massachusetts and Minneapolis, Minnesota are two examples. Both systems have withstood constitutional challenge before state supreme courts. See, e.g., Minn. Voters Alliance v. Minneapolis, 766 N.W.2d 683 (Minn. 2009); McSweeney v. City of Cambridge, 665 N.E.2d 11 (Mass. 1996). These courts considered the constitutionality of IRV in the context of unlimited ranking. Accordingly, the constitutionality of restricted IRV, where the number of candidates exceeds the voter's ranking options, is apparently a matter of first impression.

Proposition A by its terms contemplated a system under which a voter could rank in order of preference all candidates who run in a field, but also contained the caveat that the Director could limit voters to no more than three choices if the "voting system, vote tabulation system or similar related equipment . . . cannot feasibly accommodate choices equal to the total number of candidates for each office." S.F. Charter § 13.102(b). Defendants assert that this is precisely the situation here, and IRV elections in San Francisco have always limited voters to three choices. Defendants rely on several independent reasons for implementation of the three choice limitation: the voting machines currently in use are not equipped to tabulate unlimited rankings, cost and logistical realities make accommodating the unlimited option untenable, and encouraging voters to rank every candidate in a large field might foment confusion or error. Certainly, requiring voters to rank all candidates might also raise constitutional questions.

Plaintiffs' constitutional arguments target the following prospect: where more than four candidates run in an election, it is possible that a voter will rank the three least popular candidates. If no candidate receives majority support after three rounds, these voter's candidates are eliminated, and his or her ballot is thereby "exhausted." As the City Charter describes the phenomenon, an exhausted ballot is "not counted in further stages of the tabulation, if all of the choices have been eliminated. . . ." S.F. Charter § 13.102. Vote exhaustion can take several forms: a voter may choose to rank only a first or second choice candidate, may mismark his or her ballot or commit some other nullifying error, or may exercise all three available choices but fail to rank a single candidate who survives into the final rounds.

Plaintiffs emphasize that their constitutional claim focuses on an individual citizen's right of equal and meaningful access to the franchise. Their attention is not directed to whether or not restricted IRV alters elections. That said, they suggest evidence of the latter phenomenon may reflect an inherent access problem. Both sides present election results and experts who make various extrapolations from them. Plaintiffs point out, for example, that in the 2004 supervisorial elections, 13,144 ballots (or 37.44 percent of those cast) in Supervisorial District Five were "exhausted" prior to the final round of voting. Of these, voters exercised all three available choices in 5,693 ballots (or 16.2 percent of all ballots). Plaintiffs theorize that voters who used all three choices would have continued to rank candidates. Assuming this is true, plaintiffs characterize these ballots as being exhausted "involuntarily." The victor in that 2004 election won by a margin of 311 votes. The suggestion is that the 5,693 individuals behind the involuntarily exhausted ballots might have had an impact on who the ultimate winner would be if unlimited ranking were possible. More broadly, plaintiffs note that, of the ten supervisorial elections conducted since 2004, the margin of victory in the final round was less than the number of involuntarily exhausted ballots half the time.

Defendants characterize plaintiffs' figures as distorting and misleading. They suggest the 2004 district supervisorial race for District Five is an outlier, as it contained the highest proportion of exhausted ballots of any restricted IRV race in which voters ranked three candidates. The defendants match plaintiffs' figures with two other races — the 2008 supervisorial race for District Nine and the 2006 supervisorial race for District Six — where only 993 (3.4 percent) and 34 (0.2 percent) ballots were exhausted involuntarily. At a more basic level, defendants criticize plaintiffs' foundational assumption that voters really would rank more than three candidates, given the option.

Defendants insist that the multi-ranking system actually enhances the expressive capacity of voters because it supplies all citizens with more choices. They point out that, indisputably, every voter is afforded the same opportunity to rank three candidates out of a diverse field. Also, each voter can weigh in on the candidates who ultimately survive to the dispositive rounds; the voter just cannot know who that candidate will be. An exhausted vote, defendants therefore contend, only fails to "count" in the same way a vote for a losing candidate in a winner-take-all election does not "count," or is "wasted." That is, even if it does not help elect the winner, election officials nevertheless count it. In sum, defendants characterize the three choice limitation as a neutral, non-discriminatory regulation designed to deliver the benefits of IRV in an efficient, non-confusing way. They answer plaintiff's suggestion that the three choice restriction disadvantages those who celebrate unpopular candidates with a practical point: to the extent "unpopular" is not tethered to a particular viewpoint, sect, or candidate but refers to a failure to garner support, all elections work this way.

More immediately, defendants contend that offering an unlimited number of choices in each election is not technologically feasible. Worse, they suggest it would be undesirable from an accuracy or voter accessibility standpoint. The Department currently relies on two types of voting equipment: the first is an optical scan machine that tabulates paper ballots and the second is a touch-screen machine designed for voters with disabilities or for those who prefer not to use a paper ballot. The Department estimates that 99 percent of voters record their votes on paper ballots, which are then fed into, and tabulated by, the optical scan machines (this figure includes all absentee ballots). Department employees also transfer votes cast via the touch-screen machine to paper ballots. The optical scan machines tabulate these votes as well. The Department may not implement a voting system until it has been vetted and approved by the California Secretary of State. Cal. Elec. Code § 19201(a). Moreover, the Secretary of State will not certify any voting system unless it has received prior federal qualification from the United States Election Assistance Commission. This means that voting machines, software and ballot design, must all survive a rigorous and lengthy public examination process before they may be used in an election.

The voting machines currently in use in San Francisco underwent this review process and were approved and implemented in time for the April elections of 2004. From the outset, the review process appears to have assumed that voters would be limited to a first, second and third choice, even if more than three candidates run in any given field. Accordingly, the optical-scan machines selected are designed to read ballots with three columns and dimensions of a certain length and width. The machines are also equipped with memory packs to record votes cast on each ballot. Defendants express concern in general that an unlimited number of rankings might exceed the amount of data these packs can hold but do not offer a more precise description of the memory packs' capacities.

The Department posits that unlimited voting would be logistically impossible to implement using San Francisco's voting machines. In order to use the current machines, defendants envision a system whereby voters must fill out multiple ballot cards. They then list a parade of horribles: the physical separation of these cards may easily occur (in the machines' storage bins, in transportation, or at the polling location if and when a machine rejects a ballot card due to error) and dangerous (it might confuse any record of the voter's choice and jeopardize the accuracy of the post-election hand count or any recount); printing and storing sufficient paper ballots would be wasteful, voluminous and expensive; and the Department would need to hire more staff and rent more storage space. On the other hand, defendants lament that a single ballot card vast enough to handle the largest San Francisco election held since 2004 (that is, large enough to include space for 22 candidates as well as write-ins), would require either an eight-foot-long ballot with 23 columns or a massive 23-by-23 matrix with 529 selection bubbles. Apparently, state, federal and local regulations impose detailed font and typeface requirements, insist that all text appear in three languages, and mandate sufficient space for write-in candidates. Defendants predict an unrestricted IRV ballot would be seven times the length of its restricted counterpart and, at least for San Francisco's current technology, unreadable. They contend, moreover, that the Secretary of State would be unlikely to approve either envisioned ballot modifications catered to unlimited ranking.

In response to defendants' contention that no redesigned voting system is practicable, plaintiffs note that Proposition A provided a backup if IRV were deemed unworkable: the City could then revert to the two-step, general/runoff system used in the past. See S.F. Charter § 13.102(i).

In 2005, the Department invited one of its equipment vendors to participate in a pilot program. The vendor tested a four rank limitation (although the ballots still contained only three columns). Participants were reportedly dissatisfied with the ballot and described it as confusing, difficult to mark, and likely to cause error. Apparently in response, the Department continued to employ the three rank limitation. Consequently, every IRV election in San Francisco has limited voters to three choices, no matter the number of candidates seeking election to a particular office.


A. The Legal Standard

Summary judgment is proper "if the pleadings and admissions on file, together with the affidavits, if any, show that there is no genuine issue as to any material fact and that the moving party is entitled to judgment as a matter of law." Fed.R.Civ.P. 56(c). The purpose of summary judgment "is to isolate and dispose of factually unsupported claims or defenses." Celotex v. Catrett, 477 U.S. 317, 323-24 (1986). The moving party "always bears the initial responsibility of informing the district court of the basis for its motion, and identifying those portions of `the pleadings and admissions on file, together with the affidavits, if any' which it believes demonstrate the absence of a genuine issue of material fact." Id. at 323. If it meets this burden, the moving party is then entitled to judgment as a matter of law when the non-moving party fails to make a sufficient showing on an essential element of the case with respect to which that party bears the burden of proof at trial. Id. at 322-23.

The non-moving party "must set forth specific facts showing that there is a genuine issue for trial." Fed.R.Civ.P. 56(e). The non-moving party cannot defeat the moving party's properly supported motion for summary judgment simply by alleging some factual dispute between the parties. To preclude the entry of summary judgment, the non-moving party must bring forth material facts, i.e., "facts that might affect the outcome of the suit under the governing law. . . . Factual disputes that are irrelevant or unnecessary will not be counted." Anderson v. Liberty Lobby, Inc., 477 U.S. 242, 247-48 (1986). The opposing party "must do more than simply show that there is some metaphysical doubt as to the material facts." Matsushita Elec. Indus. Co. v. Zenith Radio, 475 U.S. 574, 588 (1986).

Ultimately, plaintiffs ask this Court to enjoin the use of restricted IRV in San Francisco's municipal elections. According to well-established principles of equity, a plaintiff seeking a permanent injunction must satisfy a four-factor test before a court may grant such relief. The plaintiffs must demonstrate: (1) that they have suffered an irreparable injury; (2) that remedies available at law, such as monetary damages, are inadequate to compensate for that injury; (3) that, considering the balance of hardships between the plaintiffs and defendants, a remedy in equity is warranted; and (4) that the public interest would not be disserved by a permanent injunction. eBay Inc. v. MercExchange, LLC, 547 U.S. 388, 391 (2006). Injunctive relief is an "extraordinary remedy" and may only be awarded upon a clear showing that the plaintiff is entitled to such relief. Winter v. NRDC, Inc., 129 S. Ct. 365, 376 (2008) (citation omitted).

B. Irreparable Injury

Plaintiffs are registered voters of either the City or County of San Francisco. Each plaintiff represents the intention to participate in the municipal elections scheduled for 2010 and 2011. Moreover, each has expressed the intent to vote in an election where involuntary ballot exhaustion is mathematically possible. Plaintiffs also insist they would choose to rank more than three candidates in the absence of the limitation. Five districts are electing members of the Board of Supervisors in 2010; there are already five or more candidates in four of these districts. District Ten has as many as 22 declared candidates and District Six has 24. The mayoral race of 2011 already has five declared candidates. Accordingly, plaintiffs have presented a "realistic danger" of ballot exhaustion and have satisfied the requirements for individual standing. See Babbitt v. United Farm Workers Nat. Union, 442 U.S. 289, 298 (1979); Lujan v. Defenders of Wildlife, 504 U.S. 555, 560 (1992).

Restricted IRV, according to plaintiffs, is unconstitutional facially and as applied to all races with more than four candidates. As to their First Amendment and equal protection claims, they contend that San Francisco's implementation of IRV severely burdens the right to vote because it denies some percentage of voters meaningful access to final election rounds. They also claim restricted IRV is fundamentally unfair and runs afoul of the Fourteenth Amendment's substantive due process guarantee. Under either rubric, plaintiffs suggest that restricted IRV merits the most exacting scrutiny. They argue defendants must, but cannot, show that the system is narrowly tailored to serve compelling governmental interests. Even assuming strict scrutiny is not warranted, plaintiffs insist defendants have not presented governmental interests of sufficient import to outweigh even a burden that is less than severe. The validity of each claim and the appropriate standard of review are discussed in turn below.

Plaintiffs also ground their claims on 42 U.S.C. § 1983. As it is premised on underlying constitutional violations, the section 1983 claim can be analyzed under the same analytical umbrella. See, e.g., LaRouche v. Fowler, 152 F.3d 974, 987 (D.C. Cir. 1998) ("[W]e have previously recognized that the case law relating to section 1983 claims, and that relating to claims brought directly under the Constitution, have been assimilated in most . . . respects.") (internal quotation marks omitted). Moreover, the Supreme Court has instructed that, at least in the context of election regulation, courts also may view associational rights claims with equal protection implications through the same analytical lens. See, e.g., Anderson v. Celebrezze, 460 U.S. 780, 787 n. 7 (1983) (eschewing a separate equal protection analysis and opting instead to analyze election regulation under the auspices of the Court's First and Fourteenth Amendment associational rights jurisprudence).

1. First and Fourteenth Amendment Associational Right

The Constitution does not compel a "fixed method of choosing state or local officers or representatives." Rodriguez v. Popular Democratic Party, 457 U.S. 1, 9 (1982). After all, "[t]he science of government is the most abstruse of all sciences. . . . It is a science of experiment." Anderson v. Dunn, 19 U.S. 204, 226 (1821). States and municipalities are free to experiment with reform in an endeavor to craft more perfect or workable systems. Save and unless a state, county, or municipal government runs afoul of a federally protected right, it has vast leeway in the management of its internal affairs. Sailors v. Board of Ed. of Kent City, 387 U.S. 105, 111 (1967). Local elections are a plumb example of such a core affair. As emphasized in this Court's prior Order, it is not for a court to comment on the wisdom of electoral reform generally or San Francisco's particular approach toward that end. It is instead the system's constitutional effects that are subject to federal judicial scrutiny.

When a state or municipality provides for the election of public officials, "a citizen has a constitutionally protected right to participate . . . on an equal basis with other citizens in the jurisdiction." Dunn v. Blumstein, 405 U.S. 330, 336 (1972). In an early statement on election regulations, the Supreme Court explained that because the "right to exercise the franchise in a free and unimpaired manner is preservative of other basic civil and political rights, any alleged infringement of the right of citizens to vote must be carefully and meticulously scrutinized." Kramer v. Union Free School Dist. No. 15, 395 U.S. 621, 626 (1969). See also Wesberry v. Sanders, 376 U.S. 1, 17 (1964) ("No right is more precious in a free country than that of having a voice in the election of those who make the laws under which, as good citizens, we must live.").

"It does not follow, however, that the right to vote in any manner and the right to associate for political purposes through the ballot are absolute." Burdick v. Takushi, 504 U.S. 428, 433 (1992). "Common sense, as well as constitutional law," compels "the conclusion that government must play an active role in structuring elections; as a practical matter, there must be a substantial regulation of elections if they are to be fair and honest and if some sort of order, rather than chaos, is to accompany the democratic processes." Id. ( quoting Storer v. Brown, 415 U.S. 724, 730 (1974)).

Election regulations of all stripes inevitably burden the individual voter. Id. "[T]o subject every voting regulation to strict scrutiny and to require that the regulation be narrowly tailored to advance a compelling state interest . . . would tie the hands of States seeking to assure that elections are operated equitably and efficiently." Id. With this observation in mind, the Supreme Court in Burdick advocated a "flexible" standard for reviewing election regulations. Id. at 434. A court must weigh "the character and magnitude of the asserted injury to the rights protected by the First and Fourteenth Amendments that the plaintiff seeks to vindicate" against "the precise interests put forward by the [government] as justifications for the burden imposed by its rule, taking into consideration the extent to which those interests make it necessary to burden the plaintiff's rights." Id. The strength of the burden determines the rigorousness of the inquiry. Id. "[W]hen those rights are subjected to `severe' restrictions, the regulation must be `narrowly drawn to advance a state interest of compelling importance.'" Id. ( quoting Norman v. Reed, 502 U.S. 279, 289 (1992)). When an election regulation imposes only reasonable, nondiscriminatory restrictions, however, the municipality's important regulatory interests are generally sufficient to justify the restriction. Id.

Plaintiffs caution against the adoption of rational basis review of those burdens the Court does not find "severe." They argue Burdick is correctly read literally: it contemplates a balancing test where the rigor of the Court's review increases with the degree of burden imposed, not a two-tier system where severe burdens are met with strict scrutiny and all others — be they moderate burdens or de minimis ones — must survive only rational basis review. For support, plaintiffs cite to the Supreme Court's divided opinions in Crawford v. Marion County Election Board, 553 U.S. 181 (2008). Justice Stevens, writing the lead opinion joined by two other Justices, recited the Burdick standard but then emphasized that, "[h]owever slight [the] burden may appear, . . . it must be justified by relevant and legitimate state interests sufficiently weighty to justify the limitation." Crawford, 553 U.S. at 191 (internal quotation marks omitted). Three other Justices, writing in dissent, also seemed to reject a two-tiered approach. The Ninth Circuit has applied the Burdick test on various occasions after Crawford but has never specified whether it has adopted something more stringent than rational basis review. Other circuits, however, have read Burdick in combination with Crawford to require a balancing test that, depending on the burden at issue, may be less exacting than strict scrutiny but also more rigorous than rational basis review. See, e.g., Common Cause/Georgia v. Billups, 554 F.3d 1340, 1352-53 (11th Cir. 2009). This Order aims to follow the plain text in Burdick: the Court explained that a district court should weigh "the character and magnitude of the asserted injury to the rights protected by the First and Fourteenth Amendments that the plaintiff seeks to vindicate" against "the precise interests put forward by the [government] as justifications for the burden imposed by its rule," taking into consideration "the extent to which the burden is necessary to serve important governmental interests." Burdick, 504 U.S. at 434.

It is crucial to keep in mind that "[n]o bright line separates permissible election regulation[s] from unconstitutional infringements on First [and Fourteenth] Amendment freedoms." Timmons v. Twin Cities Area New Party, 520 U.S. 351, 359 (1997). See also Storer, 415 U.S. at 730 ("[T]he rule fashioned by the Court to pass on constitutional challenges . . . provides no litmus-paper test separating those restrictions that are valid from those that are invidious. . . ."). "[A] court must identify and evaluate the interests put forward by the State as justifications for the burden imposed by its rule, and then make the `hard judgment' that our adversary system demands." Crawford v. Marion County Elec. Bd., 553 U.S. 181, 190 (2008).

Accordingly, the analysis begins with an understanding of the nature of the burden imposed by the three choice limitation. All eligible voters may express three preferences in any manner they wish: they may opt to use all three available choices or none, they may vote purely according to preference or try to support or undermine strategically a perceived frontrunner. Moreover, each voter has an identical opportunity at the outset to express a preference between those few candidates who survive into the final rounds of tabulation. The burden, though, seems related to the fact that no voter can know which candidates this will include at the time he or she casts a ballot. So, some voters will select the "right" candidate at the outset and can say with certainty that they expressed a choice amongst the ballot leaders. The three choice limitation, however, means that some voters simply might never express a preference between these candidates. The burden is not that they cannot express such a preference, but rather that the system operates in a manner that makes it possible that they might not. As detailed in the fact segment, "involuntary" ballot exhaustion has occurred in IRV elections in non-trivial numbers. Because all IRV elections in San Francisco have employed the limitation, it is impossible to know whether voters would use more than three choices. The six named plaintiffs, at least, assert that they would.

Although neither party can point to any conclusive facts (and they agree that no conclusive facts exist), defendants suggest that the named plaintiffs' view is not representative of the City's general population. For support, defendants point to the pilot program conducted in 2005 to test a four choice limitation. Participants in that study did characterize the four choice option as confusing. Defendants also point out that San Francisco voters often forgo the use of all three options. They note that in the seven races for seats on the Board of Supervisors in 2004, 18.4 percent of voters selected only one candidate, and another 8.3 percent did not vote for any candidate. They add that, in the four races for seats on the Board of Supervisors in 2006, 34.6 percent of voters selected only a single candidate. Finally, defendants cite to the results gleaned in a series of exit polls conducted by defense rebuttal expert Cook. In the particular survey on which defendants rely, Cook found that 96 percent of those citizens polled expressed no desire to select more than three candidates for a single office. Plaintiffs counter that the particular race involved was not one where involuntary vote exhaustion as contemplated by their constitutional argument was even possible (more simply, the results happened to reflect fields of four or fewer candidates). To be clear, the exit poll data discussed here refers only to information included in Cook's first declaration. His supplemental declaration, as noted above, was filed impermissibly late and will not be considered here.

While no "litmus-paper test" separates a severe burden from all others, both parties direct this Court to the Ninth Circuit's guidance in Lemons v. Bradbury, 538 F.3d 1098 (9th Cir. 2008). There, the Ninth Circuit observed that the Supreme Court has characterized as severe two types of voting regulations: (1) regulations that "unreasonably deprive some residents in a geographically defined governmental unit from voting in a unit wide election," Id. at 1104 ( citing Green v. City of Tucson, 340 F.3d 891, 899, 900 (9th Cir. 2003)); and (2) regulations that "contravene the principle of `one person, one vote' by diluting the voting power of some qualified voters within the electoral unit," id. Plaintiffs argue that the three choice limitation employed in San Francisco is functionally equivalent to one or the other classically problematic types. They argue the limitation unreasonably deprives meaningful access to the franchise to some, but not all, voters or, in the alternative, ascribes different weights to different ballots. This Court previously disagreed, finding that nothing in the three-candidate limitation "unreasonably differentiates between or automatically cabins-off voters," "controls the way a voter exercises his or her preferences," or ascribes unequal weight to different ballots. Dudum, 2010 WL 1532365, at *10. Plaintiffs ask the Court to revisit these conclusions here.

i. Unreasonable Deprivation

At least as historically understood, an election regulation that effects an "unreasonable deprivation" of the franchise conditions the right to participate in an election on membership in some definable or discreet group. See, e.g., Dunn v. Blumstein, 405 U.S. at 335-37 (upending statute that conditioned voter registration on durational residency requirement); City of Phoenix v. Kolodziejski, 399 U.S. 204, 209 (1970) (finding unconstitutional state statute conditioning participation in municipal bond vote on property ownership). In short, this breed of regulation denies access to the franchise to otherwise qualified voters. Plaintiffs suggest San Francisco's system conditions participation in the final, dispositive round of voting on a voter's ability to anticipate accurately which candidates will still be in the running. They argue this means the limitation either systematically denies complete access to voters who tend to favor unpopular candidates or forces them to adopt positions on unpalatable candidates in order to participate in a meaningful fashion. While this Court in its Prior Order readily agreed with the plaintiffs that election officials may not exclude otherwise qualified voters from participating in municipal elections, it found that San Francisco's IRV system does not operate this way. Plaintiffs have still not demonstrated otherwise. Even if a voter does not rank a single candidate who survives into the final rounds, he or she still participates in and affects the election. So too does this voter participate on equal footing with the rest of the electorate: from the outset, every voter may consider and rank in order of preference any three out of a complete field of candidates. Even where a voter fails to help elect the winner, he or she nonetheless shares an identical opportunity for meaningful participation. No matter how or for whom a voter casts his or her preferences, every ballot is counted and every ballot affects the election.

Just as in the motion for preliminary injunctive relief, plaintiffs' deprivation theory draws upon a specific characterization of instant runoff voting. They argue the best way to think about IRV is as a condensed version of the two-step system restricted IRV replaced. In such a system, the general and runoff elections clearly are independent events. Assuming a voter's preferred candidate does not garner sufficient votes to appear in the runoff, he or she may weigh in anew on those who remain (of course, even assuming the preferred candidate survives the general election, a voter is also free to make a different choice in the runoff). Plaintiffs contend that each tabulation round in an IRV system operates just like a distinct election, as redistributed votes are weighed "anew" against surviving votes. Accordingly, they insist involuntary ballot exhaustion is akin to an outright ban in a traditional runoff election of all those citizens who in the general election did not vote for the leaders.

This "series of separate elections" concept is not plaintiffs' invention but represents something of a favorite analogy for observers faced with explaining IRV systems in the abstract. Notably, the Minnesota Supreme Court highlighted the similarity of unlimited IRV to a primary and general election scheme when it upheld the ranked-choice voting system employed in Minneapolis. See, e.g., Minn. Voters Alliance v. Minneapolis, 766 N.W.2d 683, 690 (Minn. 2009) (finding methodology of IRV "directly analogous to the pattern of voting in a primary/general election system"). The plaintiffs in Voters Alliance advanced a "mirror-image" of the argument plaintiffs present here. The Voters Alliance plaintiffs suggested that IRV affords voters who prefer losing candidates an unequal and indeed enhanced opportunity to affect elections. The court rejected the argument and reasoned that first-choice preferences cast for surviving candidates are "carried forward" into further rounds and therefore continue to "affect" the election. The court then drew the separate election analogy on which plaintiffs rely: votes that are "carried forward" in IRV elections are analogous to votes re-cast in a runoff for the same candidates favored in the general election. Id. at 691. That is, for every voter who is able to "weigh in" with a second-choice, all others get to "reassert" their first choices.

In the Voter Information Pamphlet that accompanied Proposition A, San Francisco also touted the system as a "simulation" of the multi-step runoff method it replaced. The Court took judicial notice of the pamphlet in the Prior Order, pursuant to Federal Rule of Evidence 201.

Defendants suggest that Minneapolis also employs a three choice limitation to implement its IRV system. There is no clear indication in the Voters Alliance opinion, however, that the court considered the limitation or its constitutional effects.

If IRV is best understood as a series of separate but distinct elections, plaintiffs insist the three choice limitation blocks some voters from access to final rounds. This, they have insisted from the outset and reiterate here, effects a severe burden. For support, plaintiffs rely on the analyses developed by two courts asked to review the effect of a regulation that conditioned a voter's ability to vote in one election on participation in a separate one. Both identified a severe burden in such a system. See Ayers-Schaffner v. Distefano, 37 F.3d 726 (1st Cir. 1994); Partnoy v. Shelley, 277 F. Supp. 2d 1064 (S.D. Cal. 2003).

In Partnoy v. Shelley, two distinct questions appeared on a ballot. Voters were required first to vote — one way or the other — on whether to recall the sitting governor before election authorities would consider a vote cast for a potential successor. 277 F. Supp. 2d at 1069. The court found the condition occasioned a severe burden, particularly on voters who for various reasons wished not to adopt a position on the recall but still wanted to participate in the selection of a successor. Id. at 1075. In Ayers-Schaffner, election officials mandated a curative re-vote to remedy defective election results. When these officials attempted to limit participation to only those voters who cast ballots in the original race, the First Circuit intervened. 37 F.3d at 727. Non-participants in the first election, the court reasoned, did not lose or waive an interest in the outcome of the curative election any more than the original participants waived the right to alter their choices if their preferences changed. "This case," the court explained, "asks us to decide whether a state may condition the right to vote in one election on whether that right was exercised in a preceding election. So stated, it is hardly worthy of discussion." Id.

Plaintiffs suggest that, as in Ayers-Schaffner, San Francisco's IRV system effectively "conditions" participation in final rounds on the manner in which voters divvy up their three choices. As in Partnoy, plaintiffs argue a three choice limitation "forces" voters to estimate accurately (to the extent possible), and then rank, likely frontrunners if they expect their vote to "count." Both iterations of the their argument ignore, however, precisely what makes IRV distinct from the system it replaced and, for that matter, the elections in Ayers-Schaffner and Partnoy. IRV is unlike a series of separate elections in the important respect that it is condensed and confined to a single electoral question. Even if votes are "tabulated" and "redistributed" in each round, voting takes place only once. From the outset, access is identical. Contrary to a series of runoffs separated in time, voters cannot rethink their decisions or reflect upon a shifting political landscape. Each runoff is really not independent insofar as voters may not change course in response to the character of a winnowing field (in this vein, defendants comfortably distinguish Ayers-Schaffner). Regardless of the number of rounds that ensue, moreover, the end purpose is to reach a single winner, who fills a single seat. Unlike in Partnoy, voters are responding to a single inquiry: which candidate is best-suited for a particular office? As defendants accurately recount, all that happens between the completion of ballots and an announcement of final results is the process of sorting the winning candidate from the losing candidates, based on the amount of support he or she garners.

What all of this means is that, at the moment voting occurs, every voter has an opportunity to consider and then express a preference for or against all candidates, including the ones who survive into the final rounds. The fact that voters can choose between whatever candidates ultimately survive or even win, though, certainly does not mean that they must. Nothing in a system restricted to selecting no more than three candidates controls how a voter employs that choice. As the Prior Order recognized, it may be true that a voter will carefully try to cast his or her preferences with the final rounds in mind. Where there are enough candidates to make ballot exhaustion possible, voters might even aim to avoid the result by ranking at least one candidate who is likely to survive until the final rounds. The same voter is also free to rank three "unpopular" candidates who are better aligned with genuine preference and hope that at least one of them amasses enough support to win.

The fact that an election scheme may incentivize strategic voting surely does not operate as a severe burden on the franchise. If it did, voters in a plurality scheme who prefer third party candidates but are wary of "wasting" their votes and so vote strategically might all have winning constitutional claims, which they do not. See, e.g., Timmons v. Twin Cities Area New Party, 520 U.S. 351, 362 (1997) (acknowledging that while "many features of our political system — e.g., single-member districts [and] `first past the post' elections . . . make it difficult for third parties to succeed in American politics . . . the Constitution does not require States . . . to move to proportional-representation elections"). On a related note, the difficult fact that a vote for an "unpopular" candidate may ultimately be outnumbered by the more numerous votes cast for mainstream competitors cannot by itself constitute a severe burden. The phenomenon is to some degree characteristic of any election.

Whether a voter tries to exercise his or her preferences strategically or from the heart, he or she participates and affects the election even in the event of ballot exhaustion. The most meaningful misconception in plaintiffs' argument is that exhausted votes are not counted. In McSweeney v. City of Cambridge, the Massachusetts Supreme Judicial Court convincingly suggested a way to think about ballots that do not survive into IRV's final rounds: "exhausted ballots too are read and counted; they just do not count toward the election of any of the [ultimately] successful candidates. Therefore it is no more accurate to say that these ballots are not counted than to say that the ballot designating a losing candidate in a two-person, winner-take-all race are not counted." 665 N.E.2d 11, 14 (Mass. 1996). The reasoning applies here: even involuntarily exhausted ballots are counted just as if they were votes cast for losing candidates in a winner-take-all system. Where a ballot is exhausted, it means the voter's lowest-ranked, longest surviving candidate lost the election. It does not mean the voter's preferences were arbitrarily ignored or disregarded. Just like the voter whose first-choice candidate finished second, the voter behind an exhausted ballot still participates and "affects" the election as much as any voter who participates in, but in the end fails to select a winning candidate, in a plurality system. So, if there is strategic maneuvering, it is more accurately characterized as conduct designed to make winning more likely, not for the threshold ability to participate. As stated in the Prior Order, a vote for a losing candidate is not the same as disenfranchisement. Plaintiffs insist theirs is not a claim for a constitutional right to win elections but is instead about the right to participate equally. San Francisco's implementation of IRV does not infringe that right.

In McSweeney, the court analyzed an unlimited form of IRV. When the Massachusetts Supreme Judicial Court spoke of exhausted votes, it referred to those instances where a voter either chose not to rank all candidates or made an error by, for example, listing two candidates as his or her first choice. Accordingly, the exhausted vote problem in that instance did not pose the same possible dilemma as present here. In Cambridge, voters who chose not to rank all candidates, in a sense, opted out of the chance to participate in all possible rounds. Even so, the court's logic applies to restricted IRV: the exhausted votes were not arbitrarily thrown out or ignored but rather were cast for losing candidates.

Defendants suggest that, if anything, restricted IRV is a truer method for reflecting the expressive choice of each citizen than alternative models. As McSweeney characterized IRV, albeit in its unrestricted form: "[A] preferential scheme, far from seeking to infringe on each citizen's equal franchise, seeks more accurately to reflect voter sentiment and to provide for the representation of minority groups in the municipal council or to enlarge the possibility of a voter's being represented therein by giving [the voter] an opportunity to express more than one preference among candidates. This purpose is not a derogation from the principle of equality but an attempt to reflect it with more exquisite accuracy." 422 Mass. at 654 (internal citation and quotation marks omitted) (alteration in original).

ii. Vote Dilution

Plaintiffs argue in the alternative that even if every qualified voter may participate in San Francisco's elections, the three-choice cutoff "dilutes" the effectiveness of votes. See, e.g., Reynolds v. Sims, 377 U.S. 533, 555 (1964) ("[T]he right of suffrage can be denied by a debasement or dilution of the weight of a citizen's vote just as effectively as by wholly prohibiting the free exercise of the franchise."). In Reynolds v. Sims, the Supreme Court held that "the effect of state legislative districting schemes which give the same number of representatives to unequal numbers of constituents" was identical to "a law providing that certain of the State's voters could vote two, five or ten times for their legislative representatives, while voters living elsewhere could vote only once." 377 U.S. at 562-63. Where an urban individual's vote carried less weight than his rural counterpart, the Court found an obvious and "extraordinary" equal protection violation. Id. See also Moore v. Ogilvie, 394 U.S. 814, 818-19 (1969) (striking down statute that made it more difficult for residents of populous counties to nominate Electoral College candidates); Gary v. Sanders, 372 U.S. 368, 379-81 (1963) (condemning county system that improperly weighed rural votes more heavily than urban votes). The Court's reapportionment cases make clear that the "crucial consideration is the right of each qualified voter to participate on an equal footing in the election process." Wells v. Edwards, 409 U.S. 1095, 1098 (1973) ( quoting Hadley v. Junior Coll. Dist., 397 U.S. 50, 54-55 (1970)). Moreover, the Court has emphasized that, "[t]here is more to the right to vote than the right to mark a piece of paper and drop it in a box or the right to pull a lever in a voting booth. The right to vote includes the right to have the ballot counted . . . at full value without dilution or discount. . . ." Reynolds, 377 U.S. at 555 n. 29.

The Prior Order emphasized that, "San Francisco's system satisfies the one person, one vote principle [because] even if voters may rank up to three choices, only one of those choices ultimately `counts.' Even where a ballot is exhausted after three rounds, that vote plainly has been `counted.' In effect, it is cast for a losing candidate." Dudum, 2010 WL 1532365, at *6. A ballot is "exhausted" once the candidates ranked therein are eliminated. When that occurs, no preferences remain to "redistribute" and "retabulate." This does not mean, as explained above, that voters are deprived of their right to participate or that exhausted votes are uncounted:

Imagine an election with five candidates. Four rounds ensue before the winner is revealed. The voter whose first choice candidate ultimately wins uses — and election officials count — a single vote. The voter who ranks the three least popular candidates uses — and election officials count — a single vote (he or she spends it on the losing candidate eliminated last). . . . The concept of vote dilution, where the votes of some voters but not others, mathematically weigh more or less heavily is not an apt fit in the IRV context.
Id. Plaintiffs argue, though, that this reasoning does not square with the deprivation analysis viewing the IRV system as a single electoral event. Here, plaintiffs insist that the only sensible way to save IRV from a "one person, one vote" challenge is to draw upon their separate elections analogy. Plaintiffs argue that either voters whose top selections are eliminated early simply get "more votes" or that the system must really be a series of separate elections (where "recounting" is constitutionally unproblematic because every ballot is again in the mix). If the latter is the case, plaintiffs suggest the Court must rethink its deprivation analysis.

As detailed earlier, where at least one court has upheld an unlimited IRV system against a dilution challenge, it did so by drawing the separate election analogy. The Minnesota Supreme Court acknowledged that "the Court's one-person, one-vote cases do address the general issue of unequal weighting of votes" but reasoned that "they are inapposite" in the IRV context. 766 N.W.2d at 698. The fact that voters whose preferred candidates were eliminated early were able to "use" their second choices, the court reasoned, did not really afford them more than one vote. Instead, the IRV system redistributed the second choice rank and then retabulated the results. As the court explained, some first choice preferences were carried forward in each tabulation round, quite like what one might expect would happen in a primary and runoff election. While the separate election analogy does help to conceptualize how the redistribution and retabulation process works, plaintiffs push the analogy too far.

Here, too, all voters may rank up to three candidates, but election officials ultimately "count" one. See also McSweeney, 422 Mass. at 652 (rejecting argument that, because voters who prefer unpopular candidates were able to exercise a second choice, IRV gave them extra opportunity with the observation that "no ballot can help elect more than one candidate"). No matter how many preferences a voter expresses, the Department only "counts" the preference associated with the candidate who survives the longest. Where a ballot is exhausted, it means the vote was effectively cast for the lowest ranked candidate. Even though that candidate ultimately loses as well, it is not true (as plaintiffs suggest) that the limitation operates such that a voter votes not at all or three times. The ability to rank multiple candidates means no more than that all voters get three chances to cast their vote in an effective manner. This is also plainly consistent with the idea that equal access at the outset undermines the disenfranchisement claim.

iii. Channeling Voters' Expressive Activity at the Polls

The discussion above has focused on the improper characterization of the burden imposed by the system at issue. So, how best to characterize it? The three choice limitation imposes a ceiling on the number of times, or the number of ways in which a voter may express his or her preferences. At least in this sense, the burden at issue in Burdick is similar. There, the Court recognized as less than severe Hawaii's ban on write-in voting. Specifically, the plaintiff wished to cast a protest vote favoring Donald Duck. The Court observed that it was not difficult for a qualified candidate to appear on Hawaiian ballots with a bit of preparedness and recourse to the proper channels. 504 U.S. at 436-37. The Court also emphasized that the "function of the election process is to winnow out and finally reject all but the chosen candidates," and cautioned that "[a]ttributing to elections a more generalized expressive function would undermine the ability of States to operate elections fairly and efficiently." Id. at 438. While "[r]easonable regulation of elections does not require voters to espouse positions that they do not support," the Court in Burdick found nothing inherently content-based about a flat ban on all write-in ballots. Id. The same is true here. While San Francisco's three choice limitation does have the "effect of channeling expressive activity at the polls," id., the mere fact that voters may not rank in order of preference every candidate who appears in a field does not force or block a particular point of view. This is instead a neutral, even-handed regulation. While it does impose a burden on the right to vote, this burden is not severe.

iv. Important Governmental Interests

As Burdick and its progeny instruct, important governmental interests must balance against even an election regulation's non-severe burden. Defendants highlight several legitimate and important interests served both by the IRV system and the three choice limitation in particular. At least as reflected in the balloting materials prepared by a Ballot Simplification Committee that accompanied Proposition A in 2002, IRV boasted four benefits: (1) elects municipal officers by a majority of voters; (2) avoids the low-voter turnout characteristic of runoff elections; (3) reduces negative campaigning; and (4) eliminates those costs associated with a second, runoff election. The City, while no longer relying on the negative campaigning justification, asserts that the three choice limitation is indeed necessary to implement IRV in an efficient, accurate and non-confusing manner. To this end, they highlight "difficult technical hurdles, the need for unprecedented regulatory approvals, concerns about widespread voter confusion, and serious logistical complications." (Defs.' Opp'n at 14:3-5.)

At the outset, the plaintiffs disagree that defendants may advance these later, particularized interests tied to the limitation. Because these defenses were not expressly included in any balloting material that accompanied Proposition A, plaintiffs warn that they are post hoc rationales designed to defend against this litigation. See, e.g., United States v. Virginia, 518 U.S. 515, 533 (1996) ("[A] justification must be genuine, not hypothesized or invented post hoc in response to litigation."). Defendants agree that they may not invent rationales, but insist the interests they advance to justify the three choice limitation were always implicit in the electoral materials. This is right. Proposition A expressly contemplated a limitation to no more than three choices if "the voting system, vote tabulation system or similar related equipment used by the City and County cannot feasibly accommodate choices equal to the total number of candidates running for each office. . . ." S.F. Charter § 13.102. Defendants insist this is the case and point out that the interests of efficiency, accuracy and clarity the limitation provides are all intimately bound up in the basic difficulty of tabulating more than three preferences per voter. Accordingly, plaintiffs' characterization of these interests as purely post hoc rationalizations is not supported by the record. All interests proffered by the defendants shall be considered.

As to the interest in candidates supported by a majority mandate, plaintiffs seem to agree that this is a legitimate and lofty goal but argue that restricted IRV undermines it. Ballot exhaustion, of course, can mean that the ultimate winner is elected with less than a true majority of votes. Plaintiffs point, for example, to the 2008 election where candidates were elected in all four supervisorial races in play with less than a true majority. The undisputed evidence also indicates that San Francisco's system often does produce winners with majority support. Whether San Francisco's IRV system really serves this rationale, then, seems at best a draw. Next, plaintiffs dispute whether voter turnout ever was problematically low for runoffs but also argue that the supposed benefit is unrelated to the three choice limitation. They lodge the same objection to the notion that a condensed election saves costs. As explained below, however, defendants insist that the three choice limitation is necessary to implement IRV, and accordingly that the limitation at least indirectly serves these two interests.

Regarding the limitation, the defendants assert that San Francisco's implementation of IRV is justified by a number of important interests related to the stability, integrity and orderly administration of elections. Certainly, these are important interests. See, e.g., Eu v. San Francisco County Cent. Comm., 489 U.S. 214, 231 (1989) ("A State indisputably has a compelling interest in preserving the integrity of its election processes."); Storer, 415 U.S. at 736 (characterizing the government's interest in "the stability of its political system" as "not only permissible, but compelling"); Lemons, 538 F.3d at 1104 (noting that a state's interest "in the orderly administration of elections [is] weighty and undeniable"). The City insists not only that the three choice limitation is for all practical purposes the only way to implement IRV, but also that the limitation is actually a sensible way to do so.

The City's first argument focuses on how its voting machines are currently not able to tabulate unlimited rankings with accuracy. Modifications to ballots, they suggest, might be possible but come with a host of negative side effects. The City's current voting machines are not designed to read ballots with more than three columns on each side of any single ballot card (the three columns, of course, allow for preferences in the order of a first, second, and third choice). To accommodate unlimited ranking in a large field (defendants use the 2004 contest for District Five Supervisor with 22 candidates as their example), each voter's ballot would need to include multiple (eight, in defendants' scenario) separate cards to capture not merely the first through third preferences, but also the fourth through sixth preferences, the seventh through ninth choices, and so on. The multiple ballot card prospect poses several problems. The City's machines are equipped with a finite memory capacity and defendants voice concern over whether the machines can capably record unlimited rankings. Defendants also reason that multiple cards might be separated or lost in the machine's storage bins, in transportation after the election, or at the polling place when a voting machine rejects one of a voter's ballot cards because of voter error. This, they suggest, would undermine the accuracy of a recount or the mandatory hand-count. Finally, the Department laments that the its already significant challenges in storing and moving over a million ballot cards to and from over 550 polling places on Election Day would be compounded significantly by multiple ballots.

As the City's machines are rigid at least with regard to ballot width, logic suggests they resort to longer ballots. Alas, state, federal and local regulations impose detailed font and typeface requirements, insist that all text appear in three languages and mandate sufficient space for write-in candidates. As a result, defendants predict an unrestricted IRV ballot would be seven times the length of its restricted counterpart and, at least for San Francisco's current technology, unreadable. New machines, moreover, would need to be vetted and approved by the California Secretary of State which, defendants insist, is a long and arduous endeavor.

Assuming new machines were approved, though, defendants still maintain that including a mechanism for ranking all candidates on the same ballot is untenable. As noted above, even if the City were to reformat its ballot design to accommodate more columns, defendants suggest that a single ballot in a large field would resemble a massive matrix. In a field with 22 candidates, such a matrix would have 529 possible bubbles, plus space for write-ins. Both this imagined ballot or the multiple card version, defendants contend, threaten to confuse and frustrate voters, and increases the likelihood of voter error.

As discussed in the Prior Order, either alternative would seem to undermine the City's interest in an orderly, accurate election process. A single ballot card alleviates the risk that separated ballots might result in error but would also foist a potentially confusing and burdensome ballot onto voters. Cf. Rubin v. City of Santa Monica, 308 F.3d 1008, 1017 (9th Cir. 2002) (recognizing "legitimate goal of achieving a straightforward, neutral, non-confusing ballot"). Defendants also suggest that a confusing process or the possibility of voter error also might undermine voter confidence in the City's electoral system. Purcell v. Gonzalez, 549 U.S. 1, 4-5 (2006) ("Confidence in the integrity of our electoral processes is essential to the functioning of our participatory democracy."). While it would be clearly inappropriate to characterize San Francisco's electorate as easily confused or otherwise unable to rank more than three candidates in a large field, common sense (and a quick perusal of even the copy of a hypothetical four rank ballot included as an exhibit) does support defendants' suggestion that these ballots do seem overwrought, confusing and potentially frustrating to complete.

The defendants have amply demonstrated that the three choice limitation is necessary to serve what are unquestionably important governmental interests: the orderly, accurate and userfriendly implementation of IRV. Defendants have also pointed to several key interests that IRV serves more generally. They also remind the Court that a majority of voters in 2002 opted for IRV, in the place of the traditional, two-step runoff system. Whether these voters were wooed by IRV's cost and time saving function or perhaps by the prospect of expressing their preferences with greater nuance than alternative systems employed in the City, all are valid, legitimate interests.

2. Substantive Due Process

"[A]n election is a denial of substantive due process if it is conducted in a manner that is fundamentally unfair." Bennett v. Yoshina, 140 F.3d 1218, 1226 (9th Cir. 1998). That said, "[o]nly a pervasive error which undermines the `organic processes' of the ballot is sufficient to trigger constitutional scrutiny." Soules v. Kauaians for Nukolii Campaign Comm., 849 F.2d 1176, 1184 (9th Cir. 1988) ( citing Hennings v. Grafton, 523 F.2d 861, 864 (7th Cir. 1975)). Plaintiffs argue San Francisco's three choice limitation has substantive due process implications in two respects. The first involves a broad argument that assumes, at least in fields with more than four candidates, that involuntary ballot exhaustion blocks some voters from equal and meaningful participation in the franchise. Plaintiffs' second, narrower argument suggests that the informative materials distributed to voters in support of Proposition A were grossly misleading and, for that reason, implicate the substantive due process doctrine.

As to the first of plaintiffs' arguments, they insist a limitation to three choices disenfranchises a swath of eligible voters because its structure invites the possibility of involuntarily exhausted votes. Vote exhaustion, they argue, is the "pervasive error" that renders the system fundamentally unfair. As recognized in this Court's Prior Order, plaintiffs' argument relies on the same facts and overarching theory advanced to suggest the system improperly burdens a voter's ability to participate equally in the franchise. The "freedom to engage in association for the advancement of beliefs and ideas is an inseparable aspect of the `liberty' assured by the Due Process Clause of the Fourteenth Amendment, which embraces freedom of speech." NAACP v. Alabama, 357 U.S. 449, 460 (1958). Plaintiffs' constitutional claims, then, have both equality and liberty implications. As the Supreme Court has given more fulsome guidance in the associational right and equal protection context as to how courts should examine election regulations, it makes more sense to employ the analysis suggested there. See Anderson v. Celebrezze, 460 U.S, 780, 787 n. 7 (1983). As explained above, plaintiffs have still not persuasively demonstrated that restricted IRV arbitrarily blocks voter access and it would be "analytically inconsistent to rely on an almost identical argument to find restricted IRV fundamentally unfair." Dudum, 2010 WL 1532365, at *4. "Because the facts and legal arguments plaintiffs proffer under each constitutional theory substantially overlap, it also would be duplicative to reiterate the analysis here." Id.

Plaintiffs' second argument, in contrast, is new and cuts a narrower path. Plaintiffs suggest the ballot materials implied that voters could rank by order of preference all candidates in a field, even though the City should have known (and now insists) that logistical realities would necessitate a three choice limitation. The Ninth Circuit has recognized that misleading ballot materials may implicate due process rights, but has emphasized that the party asserting such a claim carries a high burden. The misrepresentation must "so upset the evenhandedness of the [election] that it work[s] a patent and fundamental unfairness. . . . Such an exceptional case can arise . . . only when the ballot language is so misleading that voters cannot recognize the subject of the amendment at issue." Nat'l Audubon Society, Inc. v. Davis, 307 F.3d 835, 858 (9th Cir. 2002) ( quoting Burton v. State of Georgia, 953 F.2d 1266 (11th Cir. 1992)). Defendants point out that voter information pamphlets included the text of Proposition A, which plainly discussed the three choice limitation and the possibility that the Director of Elections could adopt it. See, e.g., Caruso v. Yamhill County ex rel. County Com'r, 422 F.3d 848, 863 (9th Cir. 2005) (rejecting substantive due process claim where voters could locate accurate information in other available materials, including language of the measure at stake). Defendants also observe that the limitation was in the mix in public debate and conversation regarding Proposition A. It just does not follow, then, that voters would have been so uninformed or unaware of the three choice limitation that they would have materially misunderstood the electoral system they endorsed. Accordingly, language in the balloting materials that referred to unlimited forms of IRV as opposed to its restricted relative does not rise to the level of a due process violation.


While a limitation to no more than three preferences in a large field of candidates does exert some burden on voting rights, it is not severe. Defendants, for their part, have adequately identified important government interests that are well-served by the limitation. Accordingly, plaintiffs' request for either a permanent injunction or declaratory relief must be denied and defendants' motion for summary judgment must be granted.


Dated: 09/09/2010


Based on the order issued herewith, granting defendants' motion for summary judgment, the Court enters judgment in favor of defendants and against plaintiff in the above-entitled action.


Dated: 09/09/2010

Summaries of

Dudum v. City County of San Francisco

United States District Court, N.D. California, San Francisco Division
Sep 9, 2010
No. C 10-00504 RS (N.D. Cal. Sep. 9, 2010)
Case details for

Dudum v. City County of San Francisco

Case Details

Full title:RON DUDUM, et al., Plaintiffs, v. CITY AND COUNTY OF SAN FRANCISCO, et…

Court:United States District Court, N.D. California, San Francisco Division

Date published: Sep 9, 2010


No. C 10-00504 RS (N.D. Cal. Sep. 9, 2010)