Harvard Law Review Harvard Law Review Harvard Law Review

Constitutional Law

Semple v. Griswold

Tenth Circuit Holds that Stricter Ballot Initiative Signature Requirements Do Not Violate the Equal Protection Clause.

Since the Supreme Court’s landmark decision in Baker v. Carr,1×1. 369 U.S. 186 (1962). the “one person, one vote” principle has been a contentious aspect of Fourteenth Amendment jurisprudence.2×2. See, e.g., Grant M. Hayden, The False Promise of One Person, One Vote, 102 Mich. L. Rev. 213, 214 (2003) (arguing that confidence in the one person, one vote principle as “a judicially manageable way of parsing out political power . . . is wholly misplaced”). At its core, the doctrine serves to protect citizens’ suffrage rights against the “dilution of the weight of [their] vote[s].”3×3. Reynolds v. Sims, 377 U.S. 533, 555 (1964). Most suffrage cases have arisen in the familiar context of representative democracy — citizens claim that their votes are being unevenly weighted in choosing elected officials.4×4. See, e.g., Baker, 369 U.S. at 192 (addressing a Tennessee map of state legislative districts that had not been updated in more than sixty years and thus resulted in severe malapportionment). Mechanisms of direct democracy, however, can also trigger such concerns.5×5. One of the primary mechanisms of direct democracy is the ballot initiative, which allows citizens to propose either legislation or constitutional amendments and then vote on them directly. See Thomas E. Cronin, Direct Democracy: The Politics of Initiative, Referendum, and Recall 2 (1989). Recently, in Semple v. Griswold,6×6. 934 F.3d 1134 (10th Cir. 2019). the Tenth Circuit upheld an amendment to the Colorado Constitution that instituted stricter requirements for petitions to make it onto the ballot by mandating signatures from two percent of voters in every state senate district.7×7. Id. at 1136–37. Relying heavily on the Supreme Court decision Evenwel v. Abbott,8×8. 136 S. Ct. 1120 (2016). the court held that variances in the number of registered voters in each district did not violate the Equal Protection Clause.9×9. Semple, 934 F.3d at 1141–42. By extending Evenwel to the direct democracy context, the Tenth Circuit overlooked key differences between representational and direct democratic models. As a result, Semple will curtail methods of direct legislation — such as ballot initiatives — that give voters a stronger voice in state governance.

Under Article V of the Colorado Constitution, citizens may directly adopt constitutional amendments through the ballot initiative process.10×10. Colo. Const. art. V, § 1(2). In 2016, Colorado voters used that process to approve Amendment 71.11×11. Semple, 934 F.3d at 1137. Previously, initiative proponents had to collect signatures from registered voters totaling at least five percent of the number of votes cast in the previous election for secretary of state.12×12. Id. But Amendment 71 added a requirement that proponents gather signatures from at least two percent of registered voters in each state senate district.13×13. Id. Amendment 71 also added a supermajority requirement that most citizen-initiated amendments be approved by at least fifty-five percent of voters. Id. at 1137 n.1. That provision was not at issue in Semple. Id. The express purpose of the amendment was “to make it more difficult to amend [the] constitution.”14×14. Colo. Const. art. V, § 1(2.5). After its approval, the new signature requirement became part of the Colorado Constitution as section 1(2.5), referred to as “Section 2.5.”15×15. Semple, 934 F.3d at 1137.

Following the adoption of Amendment 71, a group of plaintiffs involved in the ballot initiative process filed a complaint challenging the constitutionality of Section 2.5 on two grounds.16×16. Id. First, they claimed that the provision violated the one person, one vote requirement of the Equal Protection Clause because the equally apportioned districts had different populations of registered voters — with differences as great as fifty thousand17×17. See Semple v. Williams, 290 F. Supp. 3d 1187, 1189–90 (D. Colo. 2018). In 2017, District 23 had 132,222 voters, while District 21 had only 80,499 voters, a variance of over sixty percent. Id. — thereby giving some signatures more weight than others.18×18. Id. at 1194. Second, the plaintiffs claimed that Section 2.5 violated their First Amendment rights of political association.19×19. Id. at 1190.

The district court found that “[Section] 2.5 creates a classic vote-dilution problem, demanding strict scrutiny under the Equal Protection Clause.”20×20. Id. at 1202. It then concluded that Colorado had no interest “compelling enough to outweigh registered voters’ right not to have the value of their petition signatures diluted,” and thus Section 2.5 violated the EqualProtection Clause.21×21. Id. at 1202–03. Finding the Fourteenth Amendment claim dispositive, the court did not reach the First Amendment arguments.22×22. Id. at 1190.

The Tenth Circuit reversed.23×23. Semple, 934 F.3d at 1137. Writing for the court, Judge Murphy24×24. Judge Murphy was joined by Judge McHugh. found that Evenwel precluded the plaintiffs’ equal protection claim.25×25. Semple, 934 F.3d at 1140–42. In that case, the Supreme Court held that states are permitted to draw legislative districts on the basis of total population rather than eligible-voter population.26×26. Id. at 1140 (citing Evenwel v. Abbott, 136 S. Ct. 1120, 1123 (2016)). Judge Murphy observed that the plaintiffs’ challenge to Section 2.5 was analogous because they claimed that the provision resulted in vote dilution.27×27. Id. Just as the claim failed in Evenwel, it failed here too because one person, one vote does not require states to include an equal number of voters in each legislative district.28×28. Id. Rejecting the plaintiffs’ contention that Evenwel was not controlling because it involved a “representational equality” analysis, the court found that ballot initiatives also “implicate the principle of representational equality.”29×29. Id. at 1140–41. “In the direct democracy context,” the court reasoned, “voters are able to directly advance the interests of [nonvoters] . . . when they decide whether to support a citizen initiative.”30×30. Id. at 1141. Judge Murphy concluded, therefore, that “it is not unconstitutional to base direct democracy signature requirements on total population.”31×31. Id. In his view, requiring signatures from two percent of registered voters in each district was permissible because the districts had approximately the same number of residents.32×32. Id. at 1142.

The court then turned to the plaintiffs’ First Amendment challenges. Judge Murphy found that their ballot access claim was not cognizable because the Tenth Circuit had “rejected the proposition that the First Amendment is implicated by a state law that makes it more difficult to pass a ballot initiative.”33×33. Id. (citing Initiative & Referendum Inst. v. Walker, 450 F.3d 1082, 1099–1103 (10th Cir. 2006) (en banc)). Since Section 2.5 regulated the initiative process — not the content of plaintiffs’ speech — the court determined that it did not create a cognizable First Amendment claim.34×34. Id. Judge Murphy also rejected the plaintiffs’ contention that Section 2.5 impermissibly compelled political speech by forcing them to campaign in districts they would prefer to avoid.35×35. Id. at 1143. Since Section 2.5 did not bar the plaintiffs from expressing their ideas outside the initiative process, it did not create the kind of state-mandated penalty required for a compelled speech claim.36×36. Id.

Judge Briscoe dissented.37×37. Id. at 1144 (Briscoe, J., dissenting). In her view, the majority opinion “completely misse[d] the point of the plaintiffs’ Equal Protection challenge.”38×38. Id. She argued that the case was “categorically different” from Evenwel because it did not involve the principle of representational equality.39×39. Id. at 1146–47. When voters sign a petition, Judge Briscoe reasoned, they are representing only themselves, even if they might have the interests of others in mind.40×40. Id. at 1148. She also suggested that Evenwel did not completely abandon the concept of voter equality because its holding was limited to the issue of how jurisdictions may draw their legislative districts.41×41. Id. at 1148–49.

Having dismissed Evenwel as inapposite, Judge Briscoe observed that the Supreme Court has “long recognized” individual voters’ right to have their votes weighted equally.42×42. Id. at 1149. That right, she asserted, extends not only to cases involving malapportioned districts, but also to other contexts such as petition requirements that discriminate against residents of populous counties.43×43. Id. at 1149–50 (citing Moore v. Ogilvie, 394 U.S. 814, 819 (1969)). Judge Briscoe then concluded that Section 2.5 plainly violated the plaintiffs’ right to have their votes weighted the same as those of voters in other districts.44×44. Id. at 1155. Although she acknowledged that there were valid state interests behind the provision, she argued that those interests did not outweigh the burden imposed on the plaintiffs’ voting rights.45×45. Id. at 1155–56. In her view, therefore, Section 2.5 violated the Equal Protection Clause.46×46. Id. at 1156. Judge Briscoe also disagreed with the majority’s decision to reach the First Amendment challenges and would have remanded those claims to the district court for initial consideration. Id.

Semple unnecessarily restricts the protections available to citizens in the ballot initiative process. Contrary to the Tenth Circuit’s analysis, Evenwel’s concern with representational equality is not implicated by direct democracy and is therefore inapposite. While some scholars have suggested that voters have representational obligations in the context of direct democracy, that claim is not consistent with historical justifications for representative democracy and is arguably foreclosed by Supreme Court precedent. Thus, by relying on Evenwel, the Tenth Circuit failed to recognize key differences between representative and direct democracy. And the Semple majority overlooked another line of one person, one vote cases holding that vote dilution on the basis of geography is impermissible. In doing so, the court curtailed the Colorado electorate’s ability to affect policy through methods of direct legislation such as ballot initiatives.

Evenwel should not be read to apply to cases involving direct democracy for two reasons. First, the underlying facts of the case, along with the Court’s treatment of precedent, make clear that Evenwel is limited to the context of representation, not direct democracy. The plaintiffs, who lived in Texas state senate districts with relatively large eligible-voter populations, argued that drawing districts based on total population diluted their votes and thus violated the Equal Protection Clause.47×47. Evenwel v. Abbott, 136 S. Ct. 1120, 1125 (2016). The Court rejected that claim, noting that previous opinions consistently recognized states’ interest not only in preventing voter dilution, but also in promoting equality of representation,48×48. See id. at 1131 (noting that the Court had “described ‘the fundamental principle of representative government in this country’ as ‘one of equal representation for equal numbers of people’” (quoting Reynolds v. Sims, 377 U.S. 533, 560–61 (1964))). which is based on the notion that elected officials represent both voters and nonvoters in their districts.49×49. See id. at 1132. Nowhere in Evenwel, however, did the Court cite any precedents addressing voter equality challenges to mechanisms of direct democracy.

Second, the Evenwel Court relied on practical considerations that are relevant only in the context of representative democracy. The Court found that apportionment by total population “promotes equitable and effective representation” by “ensuring that each representative is subject to requests and suggestions from the same number of constituents.”50×50. Id. But no such considerations are present in the context of a ballot initiative because signatories are not subject to such “requests and suggestions.”51×51. As Judge Briscoe noted in her Semple dissent, a voter who signs a petition might have the interests of other people in mind, but that does not mean that the voter is “officially representing” anyone else. Semple, 934 F.3d at 1148 (Briscoe, J., dissenting). And the Evenwel Court’s contention that apportionment based on eligible voters “would upset a well-functioning approach to districting”52×52. Evenwel, 136 S. Ct. at 1132. is also not relevant because the Semple plaintiffs were not trying to uproot any existing districting models. They sought only to prevent the imposition of a new restriction on signature gathering.

The Tenth Circuit did not deny that Evenwel was about representation, but rather asserted that Semple also implicated the principle of representational equality.53×53. Semple, 934 F.3d at 1141 (“[V]oting-eligible citizens assume a role similar to that of elected representatives when those voters engage in the initiative process.”). Professors Michael Serota and Ethan Leib have taken a similar position. They argue that it is “because the direct democracy voter is authorized to make coercive law and bind her fellow citizens that she, like any other elected official in a democratic polity, must represent those citizens’ interests when she exercises that authority.”54×54. Michael Serota & Ethan J. Leib, Essay, The Political Morality of Voting in Direct Democracy, 97 Minn. L. Rev. 1596, 1605 (2013). In their account, it is this ability to coerce other citizens that “establishes a structural relationship of political representation.”55×55. Id. The holding in Semple aligns with this view: direct democracy voters are not purely self-interested and thus they may represent nonvoters — rather than only themselves — when casting a ballot.56×56. See Semple, 934 F.3d at 1141.

But historical and legal evidence indicates that this conclusion is at odds with the nature of mechanisms of direct democracy such as ballot initiatives, which are by definition not representative. Critics of direct democracy have long noted that voters often fail to pursue the common good.57×57. See, e.g., The Federalist No. 10, at 73 (James Madison) (Clinton Rossiter ed., 2003) (“[In all pure democracies human passions have] divided mankind into parties, inflamed them with mutual animosity, and rendered them much more disposed to vex and oppress each other than to co-operate for their common good.”). In 1787, the Founders’ distrust of the people helped drive their decision to establish a representative democracy.58×58. See Michael J. Klarman, The Framers’ Coup 607 (2016) (observing that Federalists at the Constitutional Convention “had little confidence in the ability of ordinary people to participate in government decision-making”). Many prominent political leaders were strongly opposed to involving the masses in the business of government.59×59. See id. at 244. Roger Sherman, for instance, argued the people “should have as little to do as may be about the government,” id., and John Adams felt the “frailties and passions of ordinary men made them incapable of responsible participation in government,” Cronin, supra note 5, at 13. Since individual voters could not be relied on to consider the common good of the community, it was necessary to “refine and enlarge the public views by passing them through the medium of a chosen body of citizens, whose wisdom may best discern the true interest of their country.”60×60. The Federalist No. 10, supra note 57, at 76. In the Founders’ view, “the self-interest that prevails [among voters] . . . hinders achievement of the public interest under direct democracy.”61×61. Frank H. Easterbrook, The State of Madison’s Vision of the State, 107 Harv. L. Rev. 1328, 1331 (1994); see also Klarman, supra note 58, at 244 (noting that the Founders sought to create a government “as detached as possible from public opinion”). Representative government, conversely, had the potential to remedy some of those self-interest concerns.62×62. See Easterbrook, supra note 61, at 1331–32 (arguing that representative government may solve the self-interest problem because “a representative’s self-interest is not at stake in the vast majority of votes, and in any event, is not identical to the interests of the constituents”).

Furthermore, Serota and Leib’s view that voters in the direct democracy context represent the broader electorate is inconsistent withSupreme Court precedent. In Hollingsworth v. Perry,63×63. 570 U.S. 693 (2013). voters who had supported California’s Proposition 8 — which banned same-sex marriage — sought to defend its constitutionality after state officials chose not to do so.64×64. Id. at 705. But the Court found that those voters were “plainly not agents of the State” and did not owe the people of California the kind of fiduciary obligation necessary to create an agency relationship.65×65. Id. at 713–14. As Professor Theodore Rave argues, moreover, “it is difficult to see whom voters represent when they vote in a plebiscite.”66×66. D. Theodore Rave, Fiduciary Voters?, 66 Duke L.J. 331, 340 (2016). In other words, Hollingsworth stands for the proposition that “voters are not agents of the people.”67×67. Id. at 342. Thus, in adopting an overly broad reading of Evenwel and concluding that direct democracy implicates representational concerns, the Semple majority stepped out of line with the Court’s views on direct democracy.

The Semple majority also overlooked another line of one person, one vote cases — including Gray v. Sanders68×68. 372 U.S. 368 (1963). and Moore v. Ogilvie69×69. 394 U.S. 814 (1969). — that suggest geographic discrimination in voting is inherently suspect.70×70. See id. at 818–19 (“It is no answer to the argument under the Equal Protection Clause that this law was designed to require statewide support for launching a new political party rather than support from a few localities.”); Gray, 372 U.S. at 379 (“How then can one person be given twice or ten times the voting power of another person in a statewide election merely because he lives in a rural area or because he lives in the smallest rural county?”). The majority cited Moore only in its background discussion of governing law, Semple, 934 F.3d at 1137, and did not mention Gray at all. As Judge Briscoe observed in her dissent, both of these cases illustrate that the Court has recognized an individual’s right to an equally weighted vote “in contexts other than those involving malapportioned maps.”71×71. Semple, 934 F.3d at 1149–50 (Briscoe, J., dissenting). In Gray, the Court struck down a Georgia voting scheme that awarded candidates votes based on the populations of the state counties they carried.72×72. See Gray, 372 U.S. at 370–71, 381 (“[O]nce the class of voters is chosen and their qualifications specified, we see no constitutional way by which equality of voting power may be evaded.” Id. at 381.). And in Moore, the Court concluded that an Illinois law requiring a set number of signatures — rather than a percentage — from counties throughout the state violated equal protection.73×73. See Moore, 394 U.S. at 815, 819 (finding that the law “discriminate[d] against the residents of the populous counties of the State in favor of rural sections,” id. at 819). Admittedly, these cases are not precisely the same as Semple because they both involved geographic weighting based on counties rather than state senate districts.74×74. See id. at 815; Gray, 372 U.S. at 370–71. Those counties were not equalized on the basis of either registered voters or total population,75×75. In Gray, Fulton County, which comprised 14.11% of Georgia’s population, received only 1.46% of the “unit vote,” while Echols County, comprising just 0.5% of the state’s population, received 0.48% of the unit vote. 372 U.S. at 371. This variance gave a single voter in Echols County the same influence as ninety-nine voters in Fulton County. Id. In Moore, 93.4% of the state’s registered voters lived in the forty-nine most populous counties, while just 6.6% of registered voters lived in the other fifty-three counties. 394 U.S. at 816. whereas Section 2.5 operates in the context of Colorado districts that have roughly the same population.76×76. See Semple, 934 F.3d at 1138. The broader principle of Gray and Moore, however, is still instructive: vote dilution based on geography is constitutionally suspect.77×77. See cases cited supra note 70.

The Tenth Circuit’s failure to address the vote dilution problem created by Section 2.5 will have harmful consequences for ballot access in the direct democracy context. If advocates for a petition are forced to collect signatures in every district across the state, the costs of getting an initiative on the ballot are likely to increase significantly, thereby disadvantaging causes without wealthy supporters behind them.78×78. See Richard J. Ellis, Signature Gathering in the Initiative Process: How Democratic Is It?, 64 Mont. L. Rev. 35, 69–71 (2003) (analyzing the influence wealthy donors can exert over whether an initiative makes it onto the ballot and suggesting that “concentrated spending by one or a few individuals at the qualification stage . . . sabotages the purpose of a signature requirement, which is to demonstrate intensity and breadth of popular support,” id. at 70). As opponents of Section 2.5 have noted, this change will require increased resources for ground operations, as well as either training more volunteers or hiring professionals to gather the required signatures.79×79. See Corey Hutchins & Kelsey Ray, Amendment 71, aka “Raise the Bar,” Explained, Colo. Indep. (Oct. 19, 2016), https://www.coloradoindependent.com/2016/10/19/colorado-raise-the-bar-amendment-71 [https://perma.cc/6F3B-K7W2]. These barriers to the initiative process are especially troubling due to the lack of responsiveness that is prevalent in representative institutions today.80×80. See, e.g., Martin Gilens, Affluence and Influence: Economic Inequality and Political Power in America 1 (2012) (“[U]nder most circumstances, the preferences of the vast majority of Americans appear to have essentially no impact on which policies the government does or doesn’t adopt.”). Direct democracy, in this context, can serve as a check on unresponsive legislatures.81×81. See Nathaniel Persily & Melissa Cully Anderson, Regulating Democracy Through Democracy: The Use of Direct Legislation in Election Law Reform, 78 S. Cal. L. Rev. 997, 1034 (2005) (“The initiative may provide an avenue of reform that under certain circumstances could allow reform-minded voters to get around the obstructionist, self-interested tactics of their legislators.”). In fact, the Supreme Court has stated that “[t]he very prospect of lawmaking by the people may influence the legislature when it considers (or fails to consider) election-related measures.”82×82. Ariz. State Legislature v. Ariz. Indep. Redistricting Comm’n, 135 S. Ct. 2652, 2677 (2015); cf. Rucho v. Common Cause, 139 S. Ct. 2484, 2507 (2019) (noting that voters in multiple states have approved constitutional amendments creating independent commissions to address the problem of partisan gerrymandering). But such prospects are weakened when citizens participating in government through direct democracy have their votes diluted based on their residence — precisely the scenario that the Tenth Circuit failed to redress in Semple.