MINNESOTA VOTERS ALLIANCE, et al., Appellants, v. The CITY OF MINNEAPOLIS, et al., Respondents, Mark Ritchie, in his official capacity as the Secretary of State for the State of Minnesota or his successor, et al., Defendants, FairVote Minnesota, Inc., intervenor-defendant, Respondent.
This is a declaratory judgment action challenging the constitutionality of the instant runoff voting (IRV) election methodology adopted by the City of Minneapolis for its municipal elections. Appellants Minnesota Voters Alliance, a nonprofit organization, and six Minneapolis voters contend that the IRV method violates their rights to vote, to associate for political purposes, and to equal protection under both the United States and the Minnesota Constitutions. On cross-motions for summary judgment, the district court ruled that IRV does not infringe on appellants' constitutional rights and rejected the challenge to IRV. We affirm.
The City of Minneapolis conducts municipal elections in odd-numbered years, electing officers for the positions of Mayor, City Council, the Minneapolis Park and Recreation Board (Park Board), and the Minneapolis Board of Estimate and Taxation (Board of Estimate). The elections for Mayor and City Council are single-seat elections. The Park Board has six commissioners representing individual park districts. These commissioners are elected in single-seat elections. The Park Board also has three at-large commissioners. Two of the at-large commissioners are elected in a two-seat election every four years, and the other at-large commissioner is elected in a single-seat election two years later. The Board of Estimate includes two members elected city-wide every four years in a multiple-seat election.1
Prior to the change to IRV, Minneapolis city elections used a nonpartisan primary and general election format. For a single-seat election, all qualifying candidates ran in the same nonpartisan primary, and the top two vote-getters in the primary election qualified to be on the general election ballot. In a single-seat election, whether primary or general, each voter was allowed to vote for one candidate for that office. In multiple-seat elections, each voter was allowed to vote for as many candidates as there were seats to fill. For example, in a two-seat election, the top four vote-getters in the primary would qualify for the general election ballot, and the top two vote-getters in the general election would be elected.
On November 6, 2006, the voters in the City of Minneapolis voted on a referendum to approve of a new methodology for municipal elections. The question on the ballot read:
Should the City of Minneapolis adopt Single Transferable Vote, sometimes known as Ranked Choice Voting or Instant Runoff Voting, as the method for electing the Mayor, City Council, and members of the Park and Recreation Board, Library Board, and Board of Estimate and Taxation without a separate primary election and with ballot format and rules for counting votes adopted by ordinance?
Voters approved the referendum by a 65-35 percent margin. Based on the referendum, the Minneapolis City Charter was amended to read as follows:
Section 5B. Voting Method. The elected officers shall be elected by the method of Single Transferable Vote, sometimes know as Ranked Choice Voting or Instant Runoff Voting. The City Council shall, by ordinance, establish the ballot format and rules for counting the votes. The method shall be used for the first municipal election after adoption and all subsequent elections unless the City Council certifies, by ordinance, no later than four months prior to the election, that the City will not be ready to implement the method in that election. Such certification must include the reasons why the City is not ready to implement the method.
Minneapolis City Charter, ch. 2, § 5B.
On April 18, 2008, the City Council passed a comprehensive ordinance detailing the procedures for conducting municipal elections under the IRV method. The ordinance prescribes the method of counting votes for single- and multiple-seat elections. Minneapolis, Minn., Code of Ordinances (MCO) ch. 167 (2008).
In general terms, the IRV methodology eliminates the process of separate primary and general elections in favor of a single election in which voters may rank all candidates for a particular office in order of the voters' preference. Counting of the ballots then simulates a series of runoff elections, each narrowing the field of candidates until a candidate achieves the designated threshold number of votes to be elected.
In both single-seat and multiple-seat elections, all candidates are listed on the ballot, and each voter can rank the candidates in order of preference. MCO § 167.20 (definition of “Ranked-choice voting”). A voter may rank as many or as few candidates as she chooses.
In both single-seat and multiple-seat elections, a threshold is calculated to determine the number of votes needed for a candidate to win the election. Threshold is defined by the ordinance.
Threshold means the number of votes sufficient for a candidate to be elected. In any given election, the threshold equals the total votes counted in the first round after removing partially defective ballots, divided by the sum of one (1) plus the number of offices to be filled and adding one (1) to the quotient, disregarding any fractions. Threshold = (Total votes cast)/ (Seats to be elected +1) +1.
MCO § 167.20. Thus, in a single-seat election, the threshold is a majority of the votes cast (total votes cast, divided by 1 seat plus 1, or 2, plus 1 more vote). In multiple-seat elections, however, the threshold is less than a majority of votes. The threshold required for election is not adjusted in subsequent rounds when the number of votes cast may decrease because not all voters continued to rank candidates, but continues to be based on the total number of votes counted in the first round. The vote-counting methodology operates somewhat differently for single-seat and multiple-seat elections.
Initially, all first-choice votes are counted. See MCO § 167.60(a)(1)(a). If any candidate receives the threshold number of votes (a majority), that candidate is elected and no additional rounds are counted. Id. If no candidate receives the threshold number of votes in the first round, the candidate who received the lowest number of first-choice votes is eliminated, and a second round of counting proceeds. MCO § 167.60(a)(1)(a)-(d). In the second round, the first-choice votes of all the continuing candidates are counted and the second-choice votes of the voters for whom the eliminated candidate was the first choice are counted and allocated to the continuing candidates. Id. If allocation of those second-choice votes does not give any of the continuing candidates the threshold number of votes, a third round is triggered. Id. Once again, the candidate who received the fewest votes in the previous round is eliminated and her votes are allocated to the continuing candidates based on the next ranked choice on those ballots. Id. Additional “instant runoff” rounds continue in this same way until a candidate achieves the threshold number of votes. Id. If only two candidates remain and neither achieves the threshold amount, the candidate with more votes wins. MCO § 167.60(a)(1)(e).
Only one vote, or candidate ranking, is counted for each ballot in each round of counting votes. MCO § 167.60(a)(1)(a). A voter may choose to rank only his first-choice candidate and indicate no subsequent rankings. In that event, his vote will count in the first round and any subsequent rounds in which the first-choice candidate has not been eliminated. See id. But if the voter's single ranked candidate has been eliminated, the voter's ballot does not count in any subsequent rounds. MCO § 167.60(a)(2). A voter may also skip a single ranking, indicating, for example, first- and third-choice candidates, but no second choice. In that event, if the voter's first-choice candidate is eliminated, in the next round, when the voter's second choice would be counted, the third choice is counted. Id. If a voter skips two rankings, the ballot is treated as exhausted. Id.
In multiple-seat elections, such as those for the Park Board and Board of Estimate in which two seats are filled, the basic counting methodology is the same.2 A threshold number of votes needed to win a seat is calculated using the same formula used for a single-seat election (total votes divided by total seats plus 1, plus 1 more vote). See MCO § 167.70. In a two-seat election, for example, the threshold is one-third of the total votes cast plus one vote (total votes cast, divided by 2 seats plus 1, or 3, plus 1 vote). See MCO § 167.20. Voters may rank all candidates on the ballot in order of their preferences, and serial runoff rounds are counted until the seats have been filled. MCO § 167.70(a)(1)(a). In any round in which no candidate reaches the threshold, the candidate who received the fewest votes is eliminated for the next round, with that candidate's votes allocated to the voters' next ranked choices. MCO § 167.70(a)(1)(f). Thus, in the first round, all first-choice votes are counted, and if no candidate reaches the threshold amount, the candidate with the fewest votes is dropped for the next round. Id.
The difference in methodology for multiple-seat elections, as compared to single-seat elections, occurs when one candidate receives the threshold number of votes and one or more seats remain to be filled. When that occurs, the votes that the winning candidate received above the threshold are deemed to be “surplus” votes. MCO § 167.70(a)(1)(d)-(e); see also MCO § 167.20 (definition of “Surplus”). Those surplus votes are reallocated to the subsequent-choice candidates listed on the ballots of those who voted for the winning candidate. MCO § 167.70(a)(1)(e). The subsequent-choice votes are allocated among the other candidates in the proportion of the winning candidate's total first-choice votes that were surplus. MCO § 167.20 (definition of “Surplus fraction of a vote”). That is, if one-third of the winning candidate's votes were surplus, the equivalent of one-third of the total votes is allocated to the subsequent-choice candidates on those ballots.
Plaintiffs, appellants here, Minnesota Voters Alliance and six Minneapolis voters filed the complaint in this case on December 20, 2007, in Hennepin County District Court and filed an amended complaint on August 28, 2008. The defendants, respondents here, are the City of Minneapolis and its Mayor, R.T. Rybak (collectively “the City”). FairVote Minnesota, Inc., an organization that supports IRV, intervened as a defendant and is also a respondent here.3 Both sides moved for summary judgment. In an order and memorandum filed January 13, 2009, the district court, the Honorable George McGunnigle, denied plaintiffs' motion for summary judgment and granted the motions of the City and FairVote.
Plaintiffs appealed to the court of appeals, and the City then petitioned for accelerated review in this court, which plaintiffs did not oppose. We granted accelerated review, with an expedited briefing and argument schedule.
This case was decided in the district court on cross-motions for summary judgment, based on the parties' separate statements of undisputed facts. None of the parties contend that there is a genuine issue of material fact that would preclude summary judgment. Accordingly, the question presented is which parties are entitled to judgment as a matter of law. Minn. R. Civ. P. 56. We review issues of law de novo. E.g., Miller v. One 2001 Pontiac Aztek, 669 N.W.2d 893, 895 (Minn.2003).
Appellants assert constitutional challenges to the City Charter and ordinance provisions that provide for IRV. The Charter and ordinance are presumed constitutional, and the burden of proving that they are unconstitutional is on the appellants. See City of St. Paul v. Dalsin, 245 Minn. 325, 329, 71 N.W.2d 855, 858 (1955).
Because IRV has not yet been implemented, appellants challenge the municipal law on its face, rather than as applied. “[A] plaintiff can only succeed in a facial challenge by ‘establish[ing] that no set of circumstances exists under which the Act would be valid,’ i.e., that the law is unconstitutional in all of its applications.” Wash. State Grange v. Wash. State Republican Party, --- U.S. ----, ----, 128 S.Ct. 1184, 1190, 170 L.Ed.2d 151 (2008) (quoting United States v. Salerno, 481 U.S. 739, 745, 107 S.Ct. 2095, 95 L.Ed.2d 697 (1987)); see also Soohoo v. Johnson, 731 N.W.2d 815, 821 (Minn.2007) (stating that a facial challenge to the constitutionality of a statute requires a showing that no set of circumstances exists under which the statute would be valid).
Appellants argue that the IRV methodology violates their right to vote, right to political association, and right to equal protection under one-person, one-vote principles. They argue that because these are fundamental rights, the ordinance is subject to strict scrutiny, and can only survive if it is narrowly tailored to serve compelling governmental interests.
The United States Supreme Court has recognized that states have authority to establish their own election processes, see, e.g., Wash. State Grange, 128 S.Ct. at 1191, and that election regulations typically impose some level of restrictions on the right to vote and the concomitant right to political association, see Anderson v. Celebrezze, 460 U.S. 780, 788, 103 S.Ct. 1564, 75 L.Ed.2d 547 (1983).4 See also Clayton v. Kiffmeyer, 688 N.W.2d 117, 128 (Minn.2004). Courts do not apply strict scrutiny simply because legislation imposes some burden on the right to vote. Burdick v. Takushi, 504 U.S. 428, 433-34, 112 S.Ct. 2059, 119 L.Ed.2d 245 (1992); Clayton, 688 N.W.2d at 129. Rather, strict scrutiny is reserved for circumstances where the state imposes a “severe” restriction on the right to vote. Wash. State Grange, 128 S.Ct. at 1191; Clayton, 688 N.W.2d at 129. If a statute imposes only modest burdens, then “the State's important regulatory interests are generally sufficient to justify reasonable, nondiscriminatory restrictions” on election procedures. Anderson, 460 U.S. at 788, 103 S.Ct. 1564. Courts must evaluate the burdens, if any, imposed by the state's regulation and then “weigh the asserted injury to the right to vote against the ‘precise interests put forward by the State as justifications for the burden imposed by its rule.’ ” Crawford v. Marion County Election Bd., --- U.S. ----, ----, 128 S.Ct. 1610, 1616, 170 L.Ed.2d 574 (2008) (quoting Burdick, 504 U.S. at 434, 112 S.Ct. 2059); see also Clayton, 688 N.W.2d at 129.
We first examine the burdens that appellants contend IRV imposes on the right to vote. Appellants argue that IRV burdens the right to vote in several ways:
• by giving some votes more weight than others,
• by diluting some votes for the benefit of another,
• by allowing the second choice of one voter to harm the first-choice vote of another voter,
• by reallocating proportional “surplus” second-choice votes of voters who voted for a winner while second-choice votes of voters for continuing candidates are not counted,
• by allowing fractions of a vote to go to different candidates, and
• by creating the possibility that casting a vote for a preferred candidate may harm the chances for that candidate to win office.
The first four assertions comprise appellants' argument that IRV weights some votes more heavily than others. The fifth assertion (fractional votes) relates to the proportional reallocation of surplus votes in multiple-seat elections; and the final assertion relates to appellants' argument that IRV burdens the right to vote because it is non-monotonic.
We address these contentions in turn, in light of the legal standards that govern facial constitutional challenges.
A. Unequal weighting of votes.
1. Counting second- and subsequent-choice votes after candidates are eliminated.
The first three ways in which appellants claim IRV burdens the right to vote are variations of appellants' contention that under IRV some votes count more in determining the outcome of an election than others. This contention focuses primarily on the method for counting votes in second and subsequent rounds in which the candidate who garnered the fewest votes in the previous round has been eliminated and subsequent choices of those who had voted for the eliminated candidate are counted.
The central premise of appellants' unequal weighting argument is that in the second round, first-choice votes cast for continuing candidates were exhausted in the first round and have no further opportunity to affect the election. Appellants claim that, in contrast, voters who cast their first-choice vote for the eliminated candidate get a second chance to influence the election by having their second-choice votes, for a different candidate, counted in the second round. Appellants assert that the same is true in subsequent rounds-voters for continuing candidates have exhausted their ability to affect the election, while voters who had selected the next eliminated candidate get yet another opportunity, as their next choice is counted.
Like the district court, we reject the central premise of appellants' unequal weighting argument: that the vote for a continuing candidate is exhausted in the first round in which it is exercised and then is not counted and is of no effect in subsequent rounds. On the contrary, the vote for a continuing candidate is carried forward and counted again in the next round. Just because the vote is not counted for a different candidate in the new round (as is the vote originally cast for an eliminated candidate), does not mean that the ballot was exhausted, that the vote for the continuing candidate is not counted in the subsequent rounds, or that the voter has lost the ability to affect the outcome of the election. See Stephenson v. Ann Arbor Bd. of Canvassers, No. 75-10166 AW (Mich.Cir.Ct. Nov. 1975) (rejecting a claim that an IRV system for election of mayor gave more weight to votes of some voters than others because those who voted for an eliminated candidate had their second choice counted while the second choice of voters whose candidate remained in the race were not counted).5 Indeed, it is only because votes for continuing candidates are carried forward and combined with subsequent-choice votes of voters for eliminated candidates that any candidate can eventually win.
Moreover, this aspect of the IRV methodology is directly analogous to the pattern of voting in a primary/general election system. In a nonpartisan primary election, each voter's vote counts in determining which two candidates survive to reach the general election. In essence, those primary votes are the voters' first-choice ranking of the candidates. As a result of the primary, all but the top two candidates are eliminated. Then, in the general election, voters who voted for candidates eliminated in the primary are allowed to cast another ballot, which necessarily will be for a different candidate-presumably, their second choice. This is no different than the counting of the second-choice votes of voters for eliminated candidates in instant runoff voting. At the same time, in the general election, voters who voted in the primary for either of the two surviving candidates are allowed to vote again, and they are most likely to vote again for their choice in the primary (unless, perhaps, they were voting strategically in the primary and did not vote for their actual first choice in an effort to advance a weaker opponent for their first choice to the general election). This is the equivalent of the continuing effect of the first-choice votes for continuing candidates in instant runoff. A vote in the general election still counts and affects the election, even though it is for the same candidate selected in the primary. Appellants attempt to distinguish the primary/general election system on the basis that those elections are separate, independent events, but the effect in terms of the counting of votes is the same.
Appellants argue that our decision in Brown v. Smallwood, 130 Minn. 492, 153 N.W. 953 (1915), is binding precedent that compels the conclusion that IRV is unconstitutional. In Brown, we held unconstitutional a preferential cumulative voting system that the City of Duluth adopted for municipal elections. Id. at 502, 153 N.W. at 957. Appellants contend that the IRV system violates the principles articulated in Brown in two ways. First, they assert that in ranking multiple candidates, the voter has impermissibly cast more than one vote. Second, appellants argue that a voter's vote for a continuing candidate is impermissibly exhausted in the first round and then unfairly opposed by subsequent choices of voters for eliminated candidates. Neither contention has merit.
The first assertion, that the ranking of multiple candidates improperly allows a voter more than one vote, is based on a broad interpretation of Brown that would preclude any form of ranked-choice voting under the Minnesota Constitution. We do not understand Brown to establish such a sweeping proscription. Indeed, we expressly disclaimed such intent in Brown:
Men of serious purpose have given thought to the preferential and other systems of voting and are of the opinion that the prevailing system of voting by ballot is not effective․ Our concern is with the constitutionality of the act before us and not with the goodness of other systems or with defects in our own.
Id. at 501-02, 153 N.W. at 957 (emphasis added). Consistent with that stated focus on the “act before us,” we identified specific characteristics of the Duluth system that offended constitutional requirements. As we explain below, those characteristics resulted from the cumulative vote-counting method of the Duluth system. They are not universal to preferential voting systems, and are not present in the IRV system challenged here.
Appellants' second point is that Brown prohibits the unequal weighting of votes that appellants perceive to occur in the Minneapolis IRV system. We did state in Brown that the word “vote” as used in the Minnesota Constitution has “never meant that the ballot of one elector, cast for one candidate, could be of greater or less effect then [sic] the ballot of another elector cast for another candidate. It was to be of the same effect.” Id. at 498, 153 N.W. at 956. We went on, however, to explain more fully the nature of our concern:
It was never thought that with four candidates one elector could vote for the candidate of his choice, and another elector could vote for three candidates against him. The preferential system directly diminishes the right of an elector to give an effective vote for the candidate of his choice. If he votes for him once, his power to help him is exhausted. If he votes for other candidates he may harm his choice, but cannot help him. Another candidate may vote for three candidates opposed to him.
Id. at 498, 153 N.W. at 956 (emphasis added).
An examination of the preferential voting system at issue in Brown reveals the basis for these specific concerns. The Duluth system permitted ranking of multiple candidates. 130 Minn. at 496, 153 N.W. at 955. The voter could indicate for each candidate that he was the voter's first choice, second choice, or an “additional” choice. Id. at 496, 153 N.W. at 955. In the first round, all first-choice votes were counted. If no candidate garnered a majority of the first-choice votes, all second-choice votes were added to the tally for each candidate. Id. at 496, 153 N.W. at 955. No candidate or votes were eliminated. Rather, each candidate would be credited with the cumulative total of his or her first and second-choice votes. If no candidate garnered a majority of those cumulative votes, all the “additional” choice votes for each candidate were then added to their totals, and the candidate with the most cumulative votes won the election. Id. at 496, 153 N.W. at 955. Under this system, the total votes counted could exceed the total number of ballots cast.
The Minneapolis IRV system differs from the Duluth system in ways that are significant. Because votes were cumulated in the Duluth system, after the first round a voter could have more than one vote counted at the same time. Under IRV, only one vote per voter can be counted in each round, just as in serial primary/general elections a voter may vote only once per election. Second, under the Duluth system, if a voter voted for second- or additional-choice candidates, those votes did in fact work against the voter's own first-choice candidate in subsequent rounds, because each voter is actually voting more than once as votes are accumulated. In IRV, a voter's subsequent choices are not counted unless the voter's higher-choice candidate has been eliminated (or elected, in a multiple-seat race), so a voter's subsequent choices cannot count against his first-choice candidate. Under the Duluth system, if a voter chose not to make more than a first-choice vote, so as not to hurt his first-choice candidate in subsequent rounds, the result was that the voter's one first-choice vote could be opposed not only by the first, but also by the second- and additional-choice votes of another voter, all at the same time, because of the cumulative counting system. In IRV, a first-choice vote for a continuing candidate may compete against a second or third choice of another voter, but only one at a time, and each time each voter's vote counts only as a single vote.
The cumulative vote-counting system in Brown made it possible that the ballot of one elector “could be of greater or less effect then [sic] the ballot of of another elector,” and that “one elector could vote for the candidate of his choice, and another elector could vote for three candidates against him.” Brown, 130 Minn. at 498, 153 N.W. at 956. Under the Duluth system, if a voter “votes for [the candidate of his choice] once, his power to help him is exhausted. If he votes for other candidates he may harm his choice, but cannot help him. Another elector may vote for three candidates opposed to him.” Id. at 498, 153 N.W. at 956.
The characteristics of the Duluth system that we found fatal in Brown do not exist in the IRV methodology, at least with respect to this facial constitutional challenge. Contrary to the suggestion of appellants at oral argument, those characteristics are not irrelevant factual differences, divorced from the “principles” we enunciated in Brown. Rather, the principles of Brown arise from and are defined by the cumulative voting system at issue in that case. This relationship is made abundantly clear by our repeated reference to the characteristics of cumulative voting each time we discussed our constitutional concerns in Brown, 130 Minn. at 498, 501, 153 N.W. at 956-57, and especially by our reiteration of the rationale for striking down the Duluth system in our ruling on petition for reargument, id. at 508, 153 N.W. at 959-60. Accordingly, we conclude that the Minneapolis IRV system does not contravene the principles we articulated in Brown. Nor does the system of counting subsequent choices of voters for eliminated candidates unequally weight votes. Every voter has the same opportunity to rank candidates when she casts her ballot, and in each round every voter's vote carries the same value.
2. Reallocation of “surplus” votes in multiple-seat elections.
Another aspect of appellants' claim that IRV unconstitutionally weights some votes more than others is based on the method for counting of votes in multiple-seat elections. Specifically, appellants contend that the reallocation of “surplus” votes gives the voters for a winning candidate a second opportunity to influence the outcome of the election by counting their second-choice votes for the next seat, while all voters for non-winning and non-eliminated candidates have only one opportunity to influence the election, because only their first-choice vote is counted.
As in a single-seat election, in a multiple-seat race, each voter may rank some or all of the candidates in order of preference, and a threshold number of votes needed to win is calculated based on the total first-choice votes cast and the number of seats to be filled. When a candidate in a multiple-seat election receives the threshold number of votes needed to win, all votes received by the winning candidate in that round of counting above the threshold number are considered “surplus” votes. MCO § 167.20. Rather than counting those surplus votes simply as additional votes for the winning candidate, the surplus votes are redistributed to the next-choice candidates. MCO § 167.70(a)(1)(e).
For example, if 100 votes is the threshold number needed to win, and Candidate A receives 200 first-choice votes in the first round, there are 100 surplus votes and therefore 100 second-choice votes to be reallocated. Instead of selecting 100 of the Candidate A first-choice ballots to treat as surplus for reallocation of their second-choice votes, the IRV system considers the second-choice votes on all of the Candidate A first-choice ballots and then reallocates those second-choice votes on a proportional basis. Id. In the example, the 100 surplus votes are proportionally 50 percent of the 200 total first-choice votes cast for the Candidate A.6 Accordingly, all of the second-choice votes on the Candidate A ballots are tallied and each second-choice candidate receives 50 percent of the second-choice votes cast for him or her. Mathematically, this is the equivalent of counting one-half of each of the 200 first-place votes for Candidate A for that candidate (giving her the 100 votes needed to win) and one-half of each of the second-choice votes on those ballots for the second-choice candidates (representing the 100 surplus votes).
Appellants contend that this reallocation of surplus votes gives the voters for a winning candidate a second opportunity to influence the outcome of the election by counting their second-choice votes for the next seat, while all voters for non-winning and non-eliminated candidates have only one opportunity to influence the election, because only their first-choice vote is counted. This, appellants argue, is another example of IRV improperly weighting some votes more than others.
But reallocation of surplus votes will not inevitably occur in every multiple-seat election. For example, if in the first round a sufficient number of candidates receive at least the threshold number of votes to fill all the available seats, the election is over. MCO § 167.70(a)(1)(a) (“If the number of candidates whose vote totals equal or exceed the threshold is equal to the number of seats to be filled, the tabulation is complete.”). In that circumstance, there would be no additional rounds and no reallocation of surplus votes.
Addressing the circumstances in which surplus votes are reallocated, the City and FairVote respond that the reallocation of surplus votes for a winning candidate does not constitute giving extra weight, or an extra opportunity to influence the election, to the voters for the winning candidate. They assert that each voter's ballot counts only for one vote in each round, and this is true for reallocation of surplus votes as well.
We need not resolve whether respondents' claim-that surplus vote reallocation in multiple-seat races allows each voter but one full vote in each round-is accurate, and negates appellants' claim of weighting some votes more than others. This is a facial challenge, and appellants can succeed only if they have demonstrated that IRV violates constitutional principles in every application. See Wash. State Grange v. Wash. State Republican Party, ---U.S. ----, ----, 128 S.Ct. 1184, 1190, 170 L.Ed.2d 151 (2008). Because reallocation of surplus votes would not be needed if, in a multiple-seat election, all open seats were filled in the first round, appellants' claim of unequal weighting because of surplus vote reallocation does not even arise in every application of the IRV system. In a facial challenge, once a constitutional application is identified, it is inappropriate to speculate regarding other hypothetical circumstances that might arise, and we decline to do so. Accordingly, appellants have failed to carry their burden in this facial challenge IRV as applied in multiple-seat races.
B. Counting proportional surplus votes.
Appellants additionally contend that the reallocation of fractions of votes to different candidates violates constitutional principles. As explained, the reallocation of fractions of votes results from the proportional distribution of second-choice surplus votes in multiple-seat elections. Appellants argue that counting fractions of votes is improper for two reasons. First, they contend that Brown v. Smallwood requires that each vote, and every vote, must count as a numerical one, not a fraction of one. Second, appellants argue that distributing the surplus portion of a vote to the second-choice candidate interferes with the voter's right to political association.
Because these arguments challenging fractional votes are premised on the reallocation of surplus votes in multiple-seat elections, and because there is a circumstance in which there would be no need for reallocation of surplus votes in a multiple-seat election, we need not resolve the question of whether appellants' legal theories about fractional votes have any merit in order to uphold IRV on its face.
Finally, appellants argue that the fact that IRV is non-monotonic requires a conclusion that the system violates voters' right to vote and is therefore unconstitutional. A voting system is “monotonic” if voting for a preferred candidate or giving a preferred candidate a higher ranking, with the order of other candidates' ranking remaining the same, cannot hurt the preferred candidate's chances of winning. An election methodology is “non-monotonic” if the opposite is true. That is, a system is non-monotonic if voting for a preferred candidate or ranking the preferred candidate higher, without changing the order of ranking other candidates, can hurt the preferred candidate's chances of winning.
The parties agree, and the district court found, that the Minneapolis IRV system is non-monotonic. Notably, the district court also found that the City's nonpartisan primaries, which IRV is intended to replace, are also non-monotonic. Although the parties agree that the IRV system is non-monotonic, they disagree significantly whether that has any ramifications for the constitutionality of the system.
Appellants cite no case authority that applies monotonicity as a legal standard. Instead, they simply assert that when casting a vote for a preferred candidate may harm that candidate's chances to win, the system necessarily burdens the right to vote and is unconstitutional.
Respondents contend that monotonicity is merely a mathematical concept, and not a constitutional requirement. They explain that monotonicity is one of several characteristics identified by economist Kenneth Arrow as desirable in a democratic election system. See generally Kenneth Arrow, Social Choice and Individual Values (1951). Arrow proved mathematically, in what is known as Arrow's Theorem, that no voting system can satisfy all of the desired conditions that he identified. Respondents contend that because no election system can comply with all the characteristics, it is inappropriate to use any of them as a constitutional requirement. In particular, respondents point out that, as the district court found, even the Minneapolis primary/general election system was non-monotonic, and therefore the fact that IRV is non-monotonic cannot be fatal. Appellants respond that a plurality election system is monotonic, in that each additional vote for a candidate in such a system can only help that candidate.
Appellants' response fails to address the candidate-elimination function of the nonpartisan primary. It is at that stage that the primary/general election system is non-monotonic. This is illustrated by the fact that in some circumstances, a voter can increase her preferred candidate's chances to win office by voting in the primary for a non-preferred candidate who would be a weaker opponent for her preferred candidate. By helping the non-preferred, but weaker, candidate succeed in the primary, the voter can help her preferred candidate win the general election. Conversely, voting for the preferred candidate and denying the weaker, non-preferred candidate that primary vote, could allow a stronger opponent to advance-and the stronger opponent could defeat the preferred candidate in the general election. In that way, a vote in the primary for the preferred candidate could hurt her chances in the general election-a non-monotonic result.
Although it is disconcerting to acknowledge that a voter cannot be sure that his or her vote for a candidate will help, rather than hurt, that candidate, any system that involves a process for narrowing a field of three or more candidates has that potential. But this is not because a vote for the preferred candidate counts for less in some circumstances, but rather because of the consequent changes in the relative strength of the other candidates. Accordingly, the fact that IRV is non-monotonic does not establish that the system interferes with the right to vote.
Even if non-monotonicity were viewed as an indication that the right to vote may be burdened, a further problem with appellants' monotonicity argument is that they have provided no evidence, even on a hypothetical basis, of the frequency with which the non-monotonic effect is likely to occur in a real-world election-that is, what proportion of voters would be adversely affected. In the context of a facial challenge, this is significant, in two respects.
First, the risk of this adverse characteristic of IRV is at this stage purely hypothetical, not because IRV has not yet been implemented, but because the non-monotonic effect of IRV may or may not occur in the real world. As we have already noted, in a facial challenge to constitutionality, the challenger bears the heavy burden of proving that the legislation is unconstitutional in all applications. E.g., Wash. State Grange, 128 S.Ct. at 1190. Where the harm alleged is hypothetical and may or may not occur, the challenger has not met that burden. See our discussion supra Section II.A.2 regarding reallocation of surplus votes in a multiple-seat election.
Second, a key issue in a challenge to voting regulations is whether the regulations impose a severe burden on the right to vote. Wash. State Grange, 128 S.Ct. at 1191. The fact that there may be some burden is not enough to invoke strict scrutiny. Burdick, 504 U.S. at 433-34, 112 S.Ct. 2059. See our discussion infra Section III. The Supreme Court has recently reiterated that where the regulation and the burden imposed affect a limited number of voters, the burden cannot be characterized as severe. See Crawford, 128 S.Ct. at 1622-23. Although it is apparently undisputed that the IRV methodology has potential for a non-monotonic effect, there is no indication, much less proof, of the extent to which it might occur, and so there is no way to know whether the alleged burden will affect any significant number of voters. Accordingly, appellants have not established that non-monotonicity imposes a severe burden on the right to vote.
The second step in analyzing a challenge to government regulation of elections is to determine whether any burden imposed on the right to vote is justified by the interests served by the regulation. If government regulation severely burdens the right to vote, the regulation cannot survive unless it is narrowly tailored to serve compelling state interests. See, e.g., Wash. State Grange, 128 S.Ct. at 1191. But if regulation imposes only modest burdens, then “the State's important regulatory interests are generally sufficient to justify reasonable, nondiscriminatory restrictions” on election procedures. Anderson, 460 U.S. at 788, 103 S.Ct. 1564.7 For the reasons just discussed, we have determined that appellants have failed to establish that IRV on its face burdens the right to vote, and even if it could be construed as a burden, that burden is minimal. Accordingly, the question, if there is some burden imposed, is whether there are “important regulatory interests” that justify the burden.
Respondents identify several interests that they contend IRV serves and that are adequate to justify any burden imposed on the right to vote. The City asserts that IRV serves the following interests:
(1) Because the citizens of Minneapolis adopted IRV by referendum, IRV serves the purpose of respecting the democratic process;
(2) Because IRV requires only one election, rather than separate primary and general elections, IRV reduces the inconvenience and costs to voters, candidates, and taxpayers;
(3) IRV will increase voter turnout; and
(4) IRV encourages less divisive campaigns as candidates seek support for second- and subsequent-choice votes.
Respondent FairVote argues that IRV serves the following interests in addition to those identified by the City:
(1) IRV promotes the election of candidates with majority mandates, eliminating plurality winners in one-seat races;
(2) IRV eliminates the “spoiler” effect of third-party candidacies; and
(3) IRV helps insure more diverse representation by promoting minority representation in multiple-seat races.
Appellants do not directly address most of these proffered benefits of IRV. They argue that the only evidence of purpose in the record is in the report of the Instant Runoff Voting Task Force, which focused only on cost-saving. Appellants also argue that all the proffered benefits are hypothetical, and the City cannot be entitled to summary judgment based on hypothetical justifications for IRV.
Appellants' arguments ignore the fact that legislation is presumed constitutional and the challenger has the burden of proof to rebut that presumption. See City of St. Paul v. Dalsin, 245 Minn. 325, 329, 71 N.W.2d 855, 858 (1955). Moreover, as the Supreme Court has explained, in constitutional litigation of this type, the Court does not “require elaborate, empirical verification of the weightiness of the State's asserted justifications.” Timmons v. Twin Cities Area New Party, 520 U.S. 351, 365, 117 S.Ct. 1364, 137 L.Ed.2d 589 (1997) (citing Munro v. Socialist Workers Party, 479 U.S. 189, 195-96, 107 S.Ct. 533, 93 L.Ed.2d 499 (1986)). Accordingly, the lack of evidence supporting the suggested beneficial effects of IRV is not fatal.
Reducing the costs and inconvenience to voters, candidates, and taxpayers by holding only one election, increasing voter turnout, encouraging less divisive campaigns, and fostering greater minority representation in multiple-seat elections are all legitimate interests for the City to foster.8 Whether and to what degree implementation of IRV will achieve those benefits remains to be seen. But it is plausible that IRV may advance one or more of these interests. In the context of this facial challenge, that possibility is sufficient to justify any minimal burden imposed by IRV.
In addition to arguing that IRV violates the rights to vote and to political association, appellants argue it violates their right to equal protection. This claim appears to be based primarily on the arguments about unequal weighting of votes, and as we have seen, there is no unequal weighting in the IRV system for single-seat races or in multiple-seat races where no allocation of surplus votes occurs. Appellants' equal protection claim fails as well because it is not supported by the legal authority on which it is premised, specifically, the Supreme Court's one-person, one-vote jurisprudence and Bush v. Gore, 531 U.S. 98, 121 S.Ct. 525, 148 L.Ed.2d 388 (2000) (per curiam).
Although the Court's one-person, one-vote cases do address the general issue of unequal weighting of votes, they are inapposite here. The one-person, one-vote cases had their origin in the malapportionment of legislatures. See Reynolds v. Sims, 377 U.S. 533, 84 S.Ct. 1362, 12 L.Ed.2d 506 (1964). That is, the number of voters in some districts electing one legislator was several multiples higher than in other districts, meaning that a vote in the smaller population district had more impact in terms of electing a legislator than a vote in the more populous district. See id. at 562-63, 84 S.Ct. 1362. No such vote inequality is created by IRV.
In addition, appellants contend that, under IRV, some votes are counted differently than others, and the system therefore violates the equal protection principles articulated in Bush. We agree with the district court that Bush is not controlling here. The essence of the equal protection problem addressed in Bush was that because there were no established standards under Florida law for discerning voter intent, in the recount process ballots were being judged differently from county to county, and even within individual counties. See Bush, 531 U.S. at 106, 121 S.Ct. 525. In contrast, in the IRV system, every ballot and every vote is counted by the same rules and standards.
Finally, it is worth reiterating the comment of Justice Hallam dissenting in Brown on the role of this court in addressing a constitutional challenge of this type: “Many reasons might be given why this legislation should not have been passed by the people of Duluth. With its wisdom we are not concerned. The only question is whether this community had the constitutional right to adopt this plan of election.” 130 Minn. at 504, 153 N.W. at 958 (Hallam, J., dissenting). The voters of Minneapolis chose to adopt the IRV method. We conclude that this facial challenge to the constitutionality of the IRV method fails.
1. The Board of Estimate also includes other city officials separately elected to their positions.
2. The referendum adopting IRV included elections for the Library Board, but there are no longer elections for that board because in 2008, the Minneapolis Public Library System merged with the Hennepin County Library System.
3. The initial complaint also named Secretary of State Mark Ritchie and other officials as defendants, but they have never appeared in the action and were not named in the amended complaint.
4. For simplicity, the rights to vote and to associate for political purposes will generally be referred to collectively as the right to vote in this opinion, unless separate reference to the right to association is warranted by the discussion.
5. Although there are a number of additional cases from other jurisdictions involving challenges to various forms of IRV, primarily dating from the early to mid-20th century, they addressed different issues and are not particularly helpful here. Most often the issue was whether any use of ranked voting for multiple candidates was permissible or whether a system that allowed only sequential ranking of candidates in a multiple-seat race violated the right to vote in “each election.” See, e.g., Maynard v. Bd. of Dist. Canvassers, 84 Mich. 228, 47 N.W. 756, 760 (1890) (holding constitution by implication forbids any voter to cast more than one vote for any candidate for any office); Reutener v. City of Cleveland, 107 Ohio St. 117, 141 N.E. 27, 33 (1923) (rejecting challenge that preferential voting system for multiple-seat election violates right to vote “at all elections”).
6. Total first choice votes received (200) less threshold needed to win (100) = surplus votes (100). Surplus votes (100) = 50% of total votes (200). See MCO § 167.20.
7. Appellants acknowledge that IRV is nondiscriminatory, in that it does not create invidious classifications.
8. Because Minneapolis did not employ a plurality election system before adoption of IRV, the proffered benefits of promoting majority mandates and eliminating the spoiler effect are not relevant here.
MAGNUSON, Chief Justice.