Multnomah County v. Mehrwein

[295] Argued and submitted November 14, 2019; judgment of circuit court affirmed in part and reversed in part, and case remanded to circuit court for further proceedings April 23, 2020 In the Matter of Validation Proceeding to Determine the Regularity and Legality of Multnomah County Home Rule Charter Section 11.60 and Implementing Ordinance No. 1243 Regulating Campaign Finance and Disclosure. MULTNOMAH COUNTY, Petitioner-Appellant, and Elizabeth TROJAN, Moses Ross, Juan Carlos Ordonez, David Delk, James Ofsink, Ron Buel, Seth Alan Woolley, and Jim Robison, Intervenors-Appellants, and Jason KAFOURY, Intervenor, v. Alan MEHRWEIN, Portland Business Alliance, Portland Metropolitan Association of Realtors, and Associated Oregon Industries, Intervenors-Respondents. (CC 17CV18006) (CA A168205) (SC S066445) 462 P3d 706 The county initiated a validation proceeding over its campaign finance ordinances. The trial court held that the county’s campaign contribution limits, independent expenditure limits, and disclosure rules were unconstitutional. The Court of Appeals certified the appeal to the Supreme Court. Held: (1) Vannatta v. Keisling, 324 Or 514 , 931 P2d 770 (1997), is overruled; (2) the county’s contribution limits are not facially invalid under Article I, section 8, of the Oregon Constitution; (3) the case is remanded for the trial court to decide whether the contribution limits violate the First Amendment to the United States Constitution; (4) the county’s independent expenditure limits violate both constitutions; and (5) a subsequent amendment to the county’s disclosure rules makes the question of their validity in their prior form moot. The judgment of the circuit court is affirmed in part and reversed in part, and the case is remanded to the circuit court for further proceedings. 296 Multnomah County v. Mehrwein En Banc On certification from the Court of Appeals under ORS 19.405.* Katherine Thomas, Multnomah County Attorney’s Office, Portland, argued the cause for appellant Multnomah County. Jenny M. Madkour, Multnomah County Attorney, filed the briefs. Also on the briefs was Katherine Thomas. Daniel W. Meek, Portland, argued the cause and filed the briefs for intervenors-appellants Moses Ross, Juan Carlos Ordonez, James Ofsink, Seth Alan Woolley, and Jim Robison. Linda K. Williams, Portland, argued the cause and filed the briefs for intervenors-appellants Elizabeth Trojan, David Delk, and Ron Buel. Gregory A. Chaimov, Davis Wright Tremaine LLP, Portland, argued the cause and filed the brief for intervenorsrespondents. Adam Kiel, Kafoury McDougal Law Firm, Portland, filed the briefs on behalf of amici curiae Derek Cressman, Sightline Institute, Asian Pacific American Network of Oregon, Bernie PDX, League of Women Voters of Oregon, League of Women Voters of Portland, Portland Forward, Portland Jobs with Justice, Alliance for Democracy, and Unite Oregon. Carson L. Whitehead, Assistant Attorney General, Salem, filed the brief on behalf of amicus curiae Kate Brown, Governor. Also on the brief were Ellen F. Rosenblum, Attorney General, and Benjamin Gutman, Solicitor General. Cody Hoesly, Larkins Vacura Kayser LLP, Portland, filed the brief for amici curiae Independent Party of Oregon, Oregon Progressive Party, Pacific Green Party, and Honest Elections Oregon. Kelly K. Simon, ACLU Foundation of Oregon, Inc., Portland, filed the brief on behalf of amicus curiae American ______________ * On certified appeal from a judgment of the Multnomah County Circuit Court, Eric J. Bloch, Judge. Cite as 366 Or 295 (2020) 297 Civil Liberties Union Foundation of Oregon, Inc. Also on the brief were Katherine McDowell, McDowell Rackner Gibson PC, Portland, and Daniel Belknap Bartz, Eugene. Denis M. Vannier, Senior Deputy City Attorney, Portland, filed the brief on behalf of amicus curiae City of Portland. Also on the brief was Naomi Sheffield, Deputy City Attorney. Steven C. Berman, Stoll Stoll Berne Lokting & Shlachter P.C., Portland, filed the brief on behalf of amicus curiae Planned Parenthood of Oregon. Also on the brief were Nadia H. Dahab and Lydia Anderson-Dana. Kyle Markley, Hillsboro, filed the brief on behalf of amicus curiae Kyle Markley. Owen Yeates, Institute for Free Speech, Alexandria, Virginia, filed the brief on behalf of amici curiae Taxpayers Association of Oregon and Taxpayers Association of Oregon Political Action Committee. Also on the brief was Allen Dickerson. WALTERS, C. J. The judgment of the circuit court is affirmed in part and reversed in part, and the case is remanded to the circuit court for further proceedings. 298 Multnomah County v. Mehrwein WALTERS, C. J. In the November 2016 election, Multnomah County voters approved Measure 26-184, an amendment to the Multnomah County Home Rule Charter containing campaign finance provisions. Multnomah County then adopted new ordinances, Multnomah County Code (MCC) §§ 5.200 - 203, mirroring and implementing those charter provisions. The first substantive section, MCC § 5.201, restricts campaign contributions. It limits the amount of money that donors may contribute and the amount that a candidate or campaign organization may receive from a particular donor. The second section, MCC § 5.202, limits what are known as independent expenditures. It sets a cap on the amount that individuals, acting independently of a campaign, can spend on communications supporting a candidate and forbids entities from spending any amount on communications supporting a candidate. The third section, MCC § 5.203, contains disclosure rules, which require that disclaimers about the sources of funding be attached to communications in support of a candidate. We consider the validity of those ordinances under the free speech provisions of both the Oregon and United States Constitutions—Article I, section 8, and the First Amendment. As we explain, we reach four conclusions: (1) the county’s contribution limits do not, on their face, violate Article I, section 8, of the Oregon Constitution; (2) we must remand this case to the trial court for factual findings and to consider, in the first instance, whether the contribution limits violate the First Amendment; (3) the county’s expenditure limits are invalid under both constitutional provisions; and (4) the parties’ dispute with respect to the disclosure provisions is moot. I. PROCEDURAL BACKGROUND In May 2017, Multnomah County initiated this validation action in the circuit court. Under ORS 33.710, the county sought judicial examination of MCC §§ 5.200 - 203, its new campaign finance ordinances, and a judgment upholding their legality. Two groups of intervenors joined in that court proceeding. See ORS 33.720(3) (permitting interested Cite as 366 Or 295 (2020) 299 parties to appear in the validation proceeding). Respondents1 appeared in the action to contest the validity of the county’s ordinances, arguing that they violate Article I, section 8, of the Oregon Constitution and the First Amendment of the United States Constitution. Trojan2 intervened in the action to support the county’s position that the county’s ordinances are valid. Article I, section 8, provides that “[n]o law shall be passed restraining the free expression of opinion, or restricting the right to speak, write, or print freely on any subject whatever; but every person shall be responsible for the abuse of this right.” With respect to that provision, respondents’ arguments centered on Vannatta v. Keisling, 324 Or 514 , 931 P2d 770 (1997) (Vannatta I), a decision in which this court struck down limits on campaign contributions and expenditures. The proponents acknowledged that Vannatta I was unfavorable precedent but urged the trial court to reject its reasoning in light of this court’s decision in Vannatta v. Oregon Government Ethics Comm., 347 Or 449 , 222 P3d 1077 (2009) (Vannatta II), which had distinguished Vannatta I and upheld restrictions on the receipt of gifts by public officials. The trial court ruled that, under Article I, section 8, all three sections of the county ordinances were facially invalid. The trial court considered Vannatta I controlling on the contribution and expenditure limit issues. Because the court resolved the case on state constitutional grounds, it did not address the ordinances’ validity under the First Amendment. State v. Copeland, 353 Or 816, 821 , 306 P3d 610 (2013) (“[W]e consider state constitutional issues before we consider federal claims.”). The county and Trojan appealed, and the Court of Appeals certified the appeal to this court. See ORS 19.405. We begin with the question of whether the county’s contribution limits violate the Oregon Constitution. [1] Alan Mehrwein, Portland Business Alliance, Portland Metropolitan Association of Realtors, and Associated Oregon Industries (respondents). [2] Elizabeth Trojan, Moses Ross, Juan Carlos Ordonez, David Delk, James Ofsink, Ron Buel, Seth Woolley, and Jim Robison (Trojan). On appeal, Trojan, Delk, and Buel raise assignments of error relating to the contribution and expenditure limits while Ross, Ordonez, Ofsink, Woolley, and Robison principally raise assignments of error relating to the disclosure rules. 300 Multnomah County v. Mehrwein II. CONTRIBUTION LIMITS A. Article I, Section 8 MCC § 5.201 limits the amount of money that donors may contribute in county elections and the amount that a candidate or campaign organization may receive from a particular donor. Much of the briefing in this case, from the parties and the 11 amici curiae, focuses on the role of campaign contributions in our political system and the asserted harms that are remediated, or not, by the county’s ordinance. Under the First Amendment, it is not unusual for courts to approach campaign finance cases in part by weighing those harms, and the government’s interest in abating them, against the importance of campaign contributions and expenditures to political expression. See Buckley v. Valeo, 424 US 1, 19-22, 26-27 , 96 S Ct 612 , 46 L Ed 2d 659 (1976) (adopting that approach); Citizens United v. Federal Election Comm’n, 558 US 310, 364 , 130 S Ct 876 , 175 L Ed 2d 753 (2010) (discussing the effects of corporate expenditure limits in light of circumvention and changing technology). We need not wade into that thicket, however, to determine the validity of the contribution limits under the Oregon Constitution. In this validation proceeding, the only question that is before us is whether MCC § 5.201 is unconstitutional on its face. As we will explain, not all laws are subject to a facial challenge under Article I, section 8, and we have an established framework for determining which laws may be so challenged. Using that framework, the question presented is whether the contribution limits are “written in terms directed to the substance of any ‘opinion’ or any ‘subject’ of communication.” State v. Robertson, 293 Or 402, 412 , 649 P2d 569 (1982). In considering that question, we do not write on a clean slate. Respondents rely, as they did in the trial court, on Vannatta I and on our determination that the limits that we considered in that case were subject to a facial challenge and our holding that those limits were unconstitutional. We agree that, if the analysis and the holding in Vannatta I are controlling, then the contribution limits at issue here also are subject to facial challenge and unconstitutional. To decide whether to adhere to that aspect of Vannatta I we Cite as 366 Or 295 (2020) 301 must examine not only Vannatta I and Vannatta II, but also the Robertson framework and the limited category of laws that we have held to be subject to facial challenge under Article I, section 8. That, therefore, is where we begin. 1. The Robertson framework Under Robertson, a law restricting speech falls into one of three categories. The first Robertson category encompasses any law “that is ‘written in terms directed to the substance of any “opinion” or any “subject” of communication.’ ” State v. Babson, 355 Or 383, 393-94 , 326 P3d 559 (2014) (quoting Robertson, 293 Or at 412 ). Laws in that category are unconstitutional on their face, “unless the restriction is wholly confined within an historical exception.” Id. at 394. In State v. Moyer, 348 Or 220 , 230 P3d 7 (2010), we examined the validity of such a law—a statute that makes it an offense to make a campaign contribution in a name “other than that of the person who in truth provides the contribution.” ORS 260.402 (2003). We reasoned that “the falsity that the statute prohibits can only be achieved through expression—through one person’s communication of a falsehood to another person,” and, for that reason, we concluded that “the statute must be classified as a Robertson category one law.” Moyer, 348 Or at 232 (emphasis added). Although we ultimately concluded that the law fell within a recognized historical exception, and upheld it on that basis, id. at 237-38 , that case nonetheless provides a good example of a law that is directed at expression and only expression and therefore falls within the first Robertson category. The first category is, by design, a limited one, as it includes only laws that expressly prohibit speech. The second category shares that limitation. To fall into the second category, the law also must expressly regulate speech but do so only insofar as that speech is linked to a particular harm—that is, where “the actual focus of the enactment is on an effect or harm that may be proscribed, rather than on the substance of the communication itself.” State v. Stoneman, 323 Or 536, 543 , 920 P2d 535 (1996) (emphases in original). For example, in State v. Moyle, 299 Or 691 , 705 P2d 740 (1985), we considered the constitutionality of a law that forbids making a threat that is expected 302 Multnomah County v. Mehrwein to and does cause alarm. We concluded that “[s]peech and writing are merely the means, albeit the only prohibited means, of achieving the forbidden effect—actual and reasonable alarm.” Id. at 699 . Laws that fall within the second category are analyzed for overbreadth and are held facially invalid if they are overbroad. Compare Robertson, 293 Or at 435-37 (striking down a second-category law as overbroad), with Moyle, 299 Or at 704-05 (concluding that the statute under consideration was not overbroad). Most laws fall into the third Robertson category— which includes laws that do not expressly restrict speech but that may have the effect of prohibiting or limiting it. Robertson third-category laws are not facially invalid, but they are subject to as-applied challenges. Babson, 355 Or at 404 . We have considered the difference between laws that expressly restrict speech—those in the first two Robertson categories—and those that do not—placing them in the third Robertson category—on many occasions. This case, too, requires particular attention to that distinction, and it behooves us to examine it in greater detail, beginning with our 1992 decision in State v. Plowman, 314 Or 157 , 838 P2d 558 (1992). In Plowman, the defendant was convicted of firstdegree intimidation, a crime then defined as “two or more persons, acting together, [who] ‘[i]ntentionally, knowingly, or recklessly cause physical injury to another because of their perception of that person’s race, color, religion, national origin or sexual orientation.’ ” Plowman, 314 Or at 159 (quoting ORS 166.165(1)(a)(A) (1991)). The defendant had, during an assault, exclaimed racial slurs at the victims and loudly yelled “white power” or “white pride.” Id. at 160 . He argued that the law was facially invalid under Article I, section 8, because it provided for an enhanced punishment—relative to the lesser offense of fourth-degree assault—on the basis of his beliefs. Id. at 163 . This court recognized that the relevant question was whether the law was “ ‘written in terms directed to the substance of any “opinion” or any “subject” of communication’ ” and concluded that it was not. Plowman, 314 Or at 165 (quoting Robertson, 293 Or at 412 ). We explained that Cite as 366 Or 295 (2020) 303 the law did not proscribe opinion or speech, that “[p]ersons can commit that crime without speaking a word, and holding no opinion other than their perception of the victim’s characteristics.” Id. at 165. Instead, we explained, the law proscribed a forbidden effect: “the effect of acting together to cause physical injury to a victim whom the assailants have targeted because of their perception that that victim belongs to a particular group. The assailants’ opinions, if any, are not punishable as such. ORS 166.165(1)(a)(A) proscribes and punishes committing an act, not holding a belief.” Id. Thus, in Plowman, any proscription of speech was not express. Although the law certainly punished conduct that could be expressive in nature—as the defendant’s conduct in that case appears to have been—the possibility, or even certainty, that the law would punish some expressive conduct did not bring the law within the first Robertson category. We reached an analogous conclusion in 2014, in Babson. That case concerned a challenge to a rule prohibiting overnight use of the steps in front of the state capitol. 355 Or at 386 . The defendants were protestors who had conducted a vigil on the steps after 11:00 p.m., in violation of the rule. Id. They argued that the rule was an express restriction on speech and either facially invalid as a Robertson category one law or overbroad as a Robertson category two law. Babson, 355 Or at 394 . This court rejected both contentions. First, this court explained, the rule was not “ ‘written in terms’ directed at expression or the content of expression.” Id. at 395 . Although the rule had the effect of prohibiting the defendants’ vigil, it was not written in those terms; it was written to bar all use, including nonexpressive use, of the capitol steps during certain hours. Because a person could violate the rule without engaging in any expressive activities—by, for instance, using the steps as a shortcut while crossing the capitol grounds at a time when the legislature was not conducting business—the rule was not a Robertson category one law. Id. at 396-97 . Second, this court concluded that the rule was not subject to an overbreadth challenge as a Robertson category 304 Multnomah County v. Mehrwein two law. Id. at 398 . Again, the terms of the rule did not include “expression as an element or ‘proscribed means’ of causing targeted harm.”3 Id. We rejected the defendants’ argument that apparent applications to speech were sufficient to make the rule one that “directly refer[s] to speech” within the second Robertson category: “Similarly, here, although the guideline does not directly refer to speech, the guideline does have apparent applications to speech, as defendants contend. A restriction on use of the capitol steps will prevent people like defendants from protesting or otherwise engaging in expressive activities on the capitol steps overnight. That fact alone, however, does not subject the guideline to Article I, section 8, scrutiny under the second category of Robertson. The guideline is not simply a mirror of a prohibition on words. The guideline also bars skateboarding, sitting, sleeping, walking, storing equipment, and all other possible uses of the capitol steps during certain hours. Thus, because the guideline does not expressly refer to expression as a means of causing some harm, and it does not ‘obviously’ prohibit expression within the meaning of Moyle, it is not subject to an overbreadth challenge under the second category of Robertson.” Babson, 355 Or at 403-04 ; see also State v. Illig-Renn, 341 Or 228, 236-37 , 142 P3d 62 (2006) (“In summary, the state is correct that only statutes that by their terms proscribe the exercise of the constitutionally protected rights of assembly or expression are susceptible to a facial overbreadth challenge under Article I, sections 8 and 26.”). In so holding, we noted only one exception, that where a law used “creative wording that does not refer directly to expression, but which could only be applied to expression, would be scrutinized under the first two categories of Robertson.” Babson, 355 Or at 403 (emphasis in original). As a result, laws that restrict conduct that only sometimes has an expressive component— and that do not refer to the expressive component in defining the conduct that is restricted—are not laws directed at speech. [3] The first two Robertson categories require an express restriction on speech. We need not decide, in this case, whether those requirements are identical, because we think it clear that the requirement attendant to the first Robertson category is at least as demanding as that applicable to the second. Cite as 366 Or 295 (2020) 305 We now turn to an examination of the Vannatta cases, and this court’s application of the Robertson framework in those cases. We first analyze whether Vannatta I correctly applied the Robertson framework, considering both the reasoning of that court and the reasoning of the court in Vannatta II. As we explain, we conclude that Vannatta I erred in treating the campaign contribution limits at issue there as Robertson category one laws. We then apply stare decisis principles and conclude that that reasoning must be abandoned and that Vannatta I’s holding that those laws were unconstitutional on that basis must be overruled. Finally, we apply Robertson to the county’s contribution limits at issue here and conclude that the text of the ordinances does not expressly refer to speech and that they are not facially invalid under Article I, section 8. 2. The Robertson framework and the Vannatta cases In Vannatta I, this court considered challenges to several campaign finance laws, including laws limiting both contributions and expenditures. The state conceded that the expenditure limits were express restrictions on expression, but it made no equivalent concession as to contributions. Rather, the state “argued that campaign contributions merely are gifts which in themselves are devoid of political expression and, as such, constitute conduct that permissibly may be regulated.” Vannatta I, 324 Or at 521 . Vannatta I rejected that argument, concluding that the campaign contribution limits fell into the first Robertson category.4 Vannatta I began its analysis without reference to Robertson, instead considering, in the abstract, whether contributions to political campaigns and candidates “are a form of expression under Article I, section 8.” 324 Or at 522 . The court concluded “that many—probably most—are.” Id. In laying out the principal reasoning supporting that conclusion, the court again emphasized that “[w]e think that it [4] Vannatta I also considered and rejected several other arguments about the validity of the contribution limits, including arguments resting on constitutional provisions other than Article I, section 8. In this opinion, however, we reconsider Vannatta I only to the extent that it placed the contribution limits in the first Robertson category, and we therefore do not summarize the portions of Vannatta I that do not bear on that issue. 306 Multnomah County v. Mehrwein takes little imagination to see how many political contributions constitute expression.” Id. at 523 . Vannatta I reached that conclusion in two ways. First, the court explained that a campaign contribution is expression by the contributor, the equivalent of a citizen standing on a street corner and announcing “ ‘I support candidate X.’ ” Id. at 524 . Second, it reasoned that, just as an individual’s purchase of a newspaper ad in favor of a candidate was speech, so was an individual’s contribution to a “collective ‘pot’ ” that could be used for such expression. Id. at 523-24: “We assume, for example, that no one would deny the right of a citizen to purchase individually a newspaper ad that urges others to support a particular candidate or cause. And, if the individual can persuade enough neighbors and friends to join in the effort, the resulting spending power may produce much larger ads or television or radio commercials. No one, we take it, would gainsay the right of the individual to amplify his or her voice through collective buying power—gaining adherents for one’s views is the essential purpose of political advocacy. It then follows ineluctably that the contribution of the collective ‘pot’ thus collected is expression, just as the individual’s ad was. Indeed, it does not even matter if the money goes directly into an ad created by the contributors themselves or, instead, the money goes to professionals who create the ad for a fee. The outcome is the same—‘expression,’ for the purposes of Article I, section 8. “Viewed in the foregoing way, expenditures and contributions can be better seen for what they are—not opposite poles, but closely related activities.” Vannatta I, 324 Or at 523-24 . It is significant that, in setting out the two prongs of its reasoning, this court did not assert that campaign contributions are always expressive. With respect to the first prong, the court stated somewhat categorically that a “contribution, in and of itself, is the contributor’s expression of support for the candidate or cause—an act of expression that is completed by the act of giving and that depends in no way on the ultimate use to which the contribution is put.” Id. at 522 (emphasis in original). But the court also qualified its Cite as 366 Or 295 (2020) 307 conclusion, stating only that “many—probably most” contributions are expressive. Vannatta I, 324 Or at 522 . And the court also acknowledged that a contribution may not be intended as speech, as when money is given to a politician without anticipating that it will be put toward a political campaign. Id. at 522 n 10. In the second prong of its reasoning, the court made it even more clear that “many political contributions constitute expression.” Id. at 523 (emphasis added). The court did not say that campaign contributions can only be used to express the views of the contributor. After all, campaign contributions may be used to amplify one’s voice, but they also may be used for other purposes, such as currying influence with a candidate.5 Moreover, the court acknowledged that a contribution “may never be used to promote a form of expression by the candidate; instead, it may (for example) be used to pay campaign staff or to meet other needs not tied to a particular message.” Id. at 522 . Nevertheless, the court concluded, from the fact that “many” campaign contributions are expressive, that “campaign contributions” are expression. Id. at 523-24 . Only after that abstract consideration of the nature of campaign contributions did the court turn to Robertson. Although the court could have concluded, as it had in Plowman, that a person can make a contribution to a candidate without saying a word and without expressing any opinion, it did not. Instead, without discussing Plowman, the court stated its conclusion in two short sentences: “All the listed provisions of Measure 9 either expressly limit, or ban outright, campaign contributions that may be given to or that may be accepted by a candidate. By their terms, those provisions are targeted at protected speech.” Id. at 537-38 . That is, relying on its prior conclusion that campaign contributions are speech—a conclusion itself [5] The United States Supreme Court has observed that, “in 1996 and 2000, more than half of the top 50 soft-money donors gave substantial sums to both major national parties, leaving room for no other conclusion but that these donors were seeking influence, or avoiding retaliation, rather than promoting any particular ideology.” McConnell v. Federal Election Comm’n, 540 US 93, 148 , 124 S Ct 619 , 157 L Ed 2d 491 (2003), overruled by Citizens United, 558 US 310 (emphasis in original). 308 Multnomah County v. Mehrwein premised on an observation that many or most contributions are expressive—the court concluded that limits on contributions fall into the first Robertson category. That reasoning was erroneous. As was established before Vannatta I in Plowman and reaffirmed afterward in Babson, a law that is directed at conduct that is only sometimes, rather than necessarily, expressive is not subject to a facial challenge as a law “written in terms directed to the substance of any ‘opinion’ or any ‘subject’ of communication.” Robertson, 293 Or at 412 . Treating a law as an express restriction of speech because many or even most of its applications restrict expression not only calls into question the specific results in Plowman and Babson, it also substantially expands the first Robertson category. We have previously observed that “most purposive human activity communicates something about the frame of mind of the actor[.]” Huffman and Wright Logging Co. v. Wade, 317 Or 445, 449-50 , 857 P2d 101 (1993). As the governor’s amicus brief argues, a rule that laws directed at conduct that is typically expressive fall into the first Robertson category— and thus are valid only under narrow circumstances—would reach far beyond campaign contributions: “For example, a parent can express affection for a child by giving that child a large inheritance. And yet, under Vannatta I’s rationale, an inheritance tax presumably would be a restriction on speech itself because it affects the expression embodied by that transfer of property—a proposition that cannot be true.” The court’s observation that campaign contributions often may be used by a candidate to communicate a message also fails to convert campaign contributions into conduct that is necessarily expressive. Vannatta I noted that contributed “money may never be used to promote a form of expression by the candidate; instead, it may (for example) be used to pay campaign staff or to meet other needs not tied to a particular message.” Vannatta I, 324 Or at 522 . That statement was correct: Although the personal use restrictions applicable to funds received as contributions eliminate some nonexpressive uses of contributions, they do not winnow down the possible uses such that only expressive uses Cite as 366 Or 295 (2020) 309 remain. See ORS 260.407(1) (limiting permissible uses of campaign contributions by candidates and principal campaign committees). Under Oregon law, campaign contributions need not be used for campaign expenses at all; they may be used for expenses incurred as a holder of public office. ORS 260.407(1)(a)(A), (1)(b)(A). Money contributed to a campaign may ultimately be used to finance expression, but that does not distinguish money given to a political campaign from money given to a politician as a gift—or from money in general. Thus, when we now look at Vannatta I, it is apparent that the court’s reasoning and resulting determination that the campaign contribution limits at issue there were facially invalid as Robertson category one laws was erroneous. Our analysis of Vannatta I is complicated, however, by that opinion’s partial reconsideration in Vannatta II. That case, decided more than 10 years after Vannatta I, involved what could be seen as a reductio ad absurdum of the reasoning employed in Vannatta I: an argument that Article I, section 8, protected the right of a lobbyist to give money to politicians. More concretely, the case involved facial challenges to several provisions of Oregon’s ethics laws which, among other things, prohibited public officials from receiving gifts above particular amounts and prohibited individuals, including lobbyists, from offering gifts to politicians. Vannatta II, 347 Or at 453-54 . For example, ORS 244.025(1) prohibited (and prohibits) public officials, candidates, and their relatives from receiving gifts in excess of $50 from a source with a legislative or administrative interest. See also ORS 244.025(4)(a); ORS 244.042(1) - (2). This court’s treatment of the gift receipt limits in Vannatta II highlighted the tension between Plowman and Vannatta I.6 First, Vannatta II laid out the elements of the statute limiting gifts and relied on Plowman to explain why that statute did not fall into the first Robertson category. The court emphasized that, just as in Plowman, “[a] public official who is subject to restrictions on the receipt of gifts [6] Vannatta II separately analyzed the restrictions on offering gifts, concluding that those restrictions expressly regulated speech and violated Article I, section 8. 347 Or at 468 . That portion of the analysis in Vannatta II did not implicate Vannatta I. 310 Multnomah County v. Mehrwein can violate the restrictions without saying a word, without engaging in expressive conduct, and regardless of any opinion that he or she might hold.” Vannatta II, 347 Or at 459 . The court further emphasized that the receipt restrictions “do not focus on the content of speech or writing, or on the expression of any opinion.” Id. That is, because the restriction could be violated without engaging in expressive conduct, they were not express restrictions on expressive conduct. That reasoning involved a straightforward, and correct, application of Plowman. Were it not for Vannatta I, the court could have left the matter there. However, the Vannatta II plaintiffs resisted that conclusion in several ways, one of which is particularly pertinent here: an argument that “any constitutional protection for political contributions should apply equally to gifts to legislative officials because they are indistinguishable from political contributions.” Vannatta II, 347 Or at 459 . That is, the plaintiffs asked the court to adopt reasoning analogous to that followed in Vannatta I.7 This court addressed those arguments and, in the process, attempted to limit and to distinguish Vannatta I. Unfortunately, it did so in ways that were less than clear. In brief, it concluded that Vannatta I had relied on two premises: first, that contributions are the expression of the contributor, and second, that contributions are inextricably intertwined with the speech of the campaign or candidate. Vannatta II, 347 Or at 464-65 . Vannatta II characterized the first as being primarily a response to Buckley, the United States Supreme Court decision that had approved of limits on campaign contributions but disapproved of limits on expenditures and independent expenditures, relying in part on a conclusion that the expression involved in campaign contributions was less significant, 424 US at 21-22 . Vannatta II [7] In Vannatta II, the state principally argued that the gift receipt statutes were Robertson category two laws, that they fell within an exception to the Robertson framework, or that they were time, place, and manner restrictions. The argument that the court adopted—that the gift provisions were not an express restriction on speech at all—was advanced by the state primarily as a fallback argument. See Brief on the Merits of Respondents on Review at 31, Vannatta II. The state embraced the validity of the reasoning in Vannatta I and attempted to distinguish gifts from contributions. Id. at 36 . Cite as 366 Or 295 (2020) 311 concluded that the first premise on which Vannatta I was based was therefore not essential to the holding and withdrew it. 347 Or at 465 . Vannatta II then explained that the second premise—that contributions are inextricably intertwined with speech—was not applicable to gifts: “Giving a gift to a public official is not inextricably linked with a public official’s ability to carry out official functions. Public officials can speak whether or not lobbyists have given them gifts, which distinguishes this case from Vannatta I and its focus on the connection between the restriction on campaign contributions and the candidate’s or campaign’s ability to communicate a political message.” Vannatta II, 347 Or at 465 . Vannatta II did not discuss whether Vannatta I’s reasoning comported with Robertson, nor did it discuss how the distinction that it drew between contributions and gifts mapped onto the Robertson framework. In hindsight, that treatment of Vannatta I was unsatisfactory. First, Vannatta II’s characterization of Vannatta I’s holding was not entirely accurate. It is true that, at one point, Vannatta I discussed its view of contributions as an expression of support by the contributor in the context of its disagreement with Buckley. But Vannatta I returned to that theme without mention of Buckley later in its analysis of whether campaign contributions constitute expression. 324 Or at 524 . Vannatta II’s explanation for its withdrawal of the first premise—that it was not part of Vannatta I’s holding was therefore strained. And Vannatta II’s discussion of the second prong of the reasoning in Vannatta I also is unclear. Vannatta II stated that Vannatta I’s holding “assumed that restricting campaign contributions restricts a candidate’s or a campaign’s ability to communicate a political message.” Vannatta II, 347 Or at 465 . But Vannatta I emphasized that “a contribution is protected as an expression by the contributor, not because the contribution eventually may be used by a candidate to express a particular message.” Vannatta I, 324 Or at 522 (emphasis in original). Both of the premises for Vannatta I’s reasoning rested on the contribution being the expression of the contributor. Vannatta I did not discuss the practical impact of campaign contribution restrictions on political expression by a candidate or campaign. 312 Multnomah County v. Mehrwein Second, and more importantly, Vannatta II distinguished Vannatta I in a way that deepened, rather than resolved, the tension between Vannatta I and Plowman. Vannatta II distinguished the contribution limits at issue in Vannatta I from the gift limits at issue in Vannatta II based on a claimed difference in their effects on speech. Vannatta II stated that Vannatta I’s holding rested on an assumed “symbiotic relationship between the making of contributions and the candidate’s or campaign’s ability to communicate a political message.” Vannatta II, 347 Or at 465 . Vannatta II explained that the giving of gifts did not create that assumed relationship: “Public officials can speak whether or not lobbyists have given them gifts, which distinguishes this case from Vannatta I and its focus on the connection between the restriction on campaign contributions and the candidate’s or campaign’s ability to communicate a political message.” 347 Or at 465 . Thus, Vannatta II drew a distinction between campaign contributions and gifts in terms of their effects on the ultimate ability of politicians to speak. But Vannatta II did not explain why that distinction was meaningful in the Robertson analysis. In determining that the limits on gifts to public officials did not fall into the first Robertson category and holding that those limits are constitutional, Vannatta II adhered to the Robertson framework. However, in its discussion of Vannatta I and the distinction that it drew between gifts to public officials and campaign contributions, Vannatta II deviated from that framework by implying that a law that is not an express limit on speech can fall into the first Robertson category. That deviation is confusing, and much of the briefing filed in this case reflects that confusion. Trojan and the county ask us to overrule Vannatta I in light of Vannatta II, by asking that we place further emphasis, as they do, on the distinction that Vannatta II drew between contributions and gifts. For example, the county reads Vannatta II as holding “that any protection for campaign contributions must stem from an inextricable link between the contribution and a candidate’s ability to engage in political speech.” The county therefore argues that the key question in this case is whether the county’s contribution limits are so low Cite as 366 Or 295 (2020) 313 as to prevent candidates from being able to effectively engage in expression. That is not an unreasonable reading of Vannatta II, but it is a reading that takes us far from ordinary applications of Robertson. If a law falls into the first Robertson category as an express restriction on speech, it cannot be defended based on the availability of alternative modes of expression. Conversely, a law that is not an express restriction on speech is not subject to a facial challenge at all. We conclude that, just as Vannatta I’s reasoning is inconsistent with Robertson, so too is Vannatta II’s effort to shore up Vannatta I in the process of distinguishing it. To the extent that Vannatta II can be understood as interpreting Vannatta I to place laws within the first Robertson category when they are not written in terms directed to the substance of any opinion or subject of communication, but instead may have an “effect” on such expression, that interpretation too would be at odds with Robertson and its progeny and would be erroneous. Under the Robertson framework, a law that restricts conduct without expressly regulating speech is not a Robertson category one law directed toward expression, even if the law may affect a person’s ability to speak. And the fact that contributions may enable speech also does not turn the conduct of making a campaign contribution into conduct that is necessarily expressive. Therefore, limitations on campaign contributions that regulate conduct and, in doing so, make it either easier or more difficult for a person to speak also are not properly analyzed as Robertson category one limitations. We conclude that both Vannatta I and Vannatta II were erroneous in reasoning that the contribution limits at issue in Vannatta I are Robertson category one laws. As a result, we also conclude that Vannatta I erred in holding that those laws are facially invalid on that basis. 3. Stare decisis That does not mean, however, that we must or should overrule Vannatta I. In their defense of Vannatta I, respondents place their emphasis on stare decisis. They argue, correctly, that “a philosophical disagreement with a conclusion is not grounds for reconsideration” and that this court does, 314 Multnomah County v. Mehrwein and should, overrule its constitutional precedents in a very limited set of circumstances. Our most sustained consideration of when a constitutional precedent may be overturned was in Couey v. Atkins, 357 Or 460 , 355 P3d 866 (2015). In that case, emphasizing that “[s]tare decisis does not permit this court to revisit a prior decision merely because the court’s current members may hold a different view than its predecessors about a particular issue,” id. at 485 , we distilled from our prior decisions three categories of cases in which a constitutional precedent could warrant reconsideration: “First, there are cases in which a prior pronouncement amounted to dictum or was adopted without analysis or explanation. Second, there are cases in which the analysis that does exist was clearly incorrect—that is, it finds no support in the text or the history of the relevant constitutional provision. Third, there are cases that cannot be fairly reconciled with other decisions of this court on the same constitutional provision.” Id. at 485-86 (internal citations omitted). Respondents argue that those circumstances are not present here. We disagree. This case, like Couey itself, falls into the third category. As we noted above, and discuss in more detail below, Vannatta I adopted an approach to the Robertson analysis that conflicts with our other Article I, section 8, decisions. In Couey, we faced the question of whether to overrule our precedent on the subject of justiciability. We confronted a situation where two applicable cases pointed in opposite directions. Id. at 489 . Here we face an analogous situation. Were we to focus on the fact that the law before us is a contribution limit, we might reason that this case is controlled by Vannatta I and therefore conclude that the ordinance must be struck down. Conversely, if we focus on the text of the county’s ordinance and attempt to answer the question of whether it expressly proscribes speech, the holding of Vannatta II would be directly applicable, because the limits on financial transfers to candidates here are no more directed to the substance of any opinion or subject of communications than was the prohibition on financial transfers to candidates at issue in Vannatta II. Cite as 366 Or 295 (2020) 315 And we are not faced with inconsistency only between those two cases. We have decided many cases under Article I, section 8, and our application of Robertson to campaign contributions in Vannatta I conflicts not only with our application of Robertson to gifts in Vannatta II but with several other decisions of this court, before and since Vannatta I. Those cases include Plowman and Babson, but the tension is not confined to those decisions. State v. Ciancanelli, 339 Or 282 , 121 P3d 613 (2005), furnishes another example of that tension. In that case, we considered the validity of two statutes. The first was ORS 167.062 (2003), a prohibition on certain “live public sex show[s].” The law prohibited sadomasochistic abuse and sexual conduct, but only when that conduct took place in a live public show, defined as follows: “(a) ‘Live public show’ means a public show in which human beings, animals, or both appear bodily before spectators or customers. “(b) ‘Public show’ means any entertainment or exhibition advertised or in some other fashion held out to be accessible to the public or member of a club, whether or not an admission or other charge is levied or collected and whether or not minors are admitted or excluded.” ORS 167.062(5) (2003). The state argued that the law addressed “conduct,” rather than expression. We disagreed: “In arguing against the suggestion that ORS 167.062 is directed at expression, the state also relies on this court’s recognition, in Huffman and Wright Logging Co. v. Wade, 317 Or 445 , 857 P2d 101 (1993), that conduct is not protected expression under Article I, section 8, merely because the actor intends the conduct to convey a message. But, in so arguing, the state loses sight of the fact that the issue here is the overall constitutionality of a statute, not whether defendant can claim that his particular conduct is expressive and therefore immunized from any and all criminal liability. It may or may not be true that the sexual acts that defendant directed were conduct in the most basic sense and, as such, could be punished under some other statute. But the fact remains that the statute at issue here—ORS 167.062—prohibits and criminalizes those acts only when they occur in an expressive context, i.e., in a ‘live public 316 Multnomah County v. Mehrwein show.’ Under those circumstances, we cannot avoid the conclusion that the statute is directed primarily, if not solely, toward the expressive aspect of the conduct that it describes. That is, the statute is one restraining free expression.” Ciancanelli, 339 Or at 320-21 (emphasis in original). By contrast, in the same case, we rejected the defendant’s challenge to his conviction for promoting prostitution. We explained: “ORS 167.012 prohibits promoting prostitution—owning, controlling, managing, or supervising a prostitution enterprise—regardless of the presence or absence of any circumstances that might add an expressive element to the conduct. It is not targeted either at expression itself or at the expressive aspects of certain conduct. It therefore does not, in and of itself, raise an issue of facial unconstitutionality under Article I, section 8. Defendant’s contrary argument is not well taken.” Id. at 323 (emphasis added; footnote omitted). Those contrasting dispositions adhere to the rule discussed above. The first statute restricted nude dancing—conduct that may be but is not necessarily expressive—but did so only when it was expressive. That law was expressly directed at speech, fell into the first Robertson category, and was invalid. The second law restricted promoting prostitution— conduct that might be linked to or involve expression in some circumstances—regardless of whether any expressive component was present. That law was not expressly directed at speech, did not fall into the first Robertson category, and was sustained. Along similar lines, this court decided City of Portland v. Tidyman, 306 Or 174 , 759 P2d 242 (1988), a case involving a challenge to zoning regulations targeted at adult bookstores, which used the content of the publications sold by those establishments as its basis for zoning restrictions. We determined that the ordinance fell within the first Robertson category, and we therefore held that it violated Article I, section 8. But we emphasized that that decision was a consequence of the fact that the city’s ordinance had expressly based its restrictions on the content of the bookstores’ speech, and that zoning regulations that did not depend on content of communicative merchandise could be sustained against a facial challenge: Cite as 366 Or 295 (2020) 317 “Thus the city could regulate the location of a business that sells other merchandise, ‘adult’ or otherwise, even if it purveys communicative materials, as long as selling such other merchandise is not permitted at the location. A grocery store gains no privilege against a zoning regulation by selling The National Enquirer and Globe at its checkout counter. * * * Many regulations are not impermissible laws ‘restricting the right to speak, write, or print freely on any subject whatever,’ although they can be impermissibly applied in individual cases.” Tidyman, 306 Or at 182 . All of those cases point in the same direction: laws that proscribe conduct that is often, but not necessarily, expressive cannot be facially invalid under the Robertson framework. We affirmed that principle most clearly in Plowman, Vannatta II, and Babson, all of which involved challenges to laws that proscribed conduct that could, but need not, be expressive. In all three cases we held that those laws did not fall within the first Robertson category. By contrast, in Ciancanelli we were faced with a law that proscribed conduct only when it occurred in an expressive context. That, therefore, was the type of law that Robertson forbade, and we held it unconstitutional. The distinction between laws that expressly regulate speech and laws that restrict expressive activity in only some of their applications is a significant one. The former laws are those “restraining the free expression of opinion, or restricting the right to speak, write, or print freely on any subject whatever,” Article I, section 8—the category of laws that the constitutional provision most explicitly forbids and hence the laws which this court approves most infrequently. The latter category—those that do not expressly restrict speech but that are written to restrict a broader category of conduct that is sometimes but not always expressive—may encompass nearly every law to a greater or lesser degree. Laws falling into that latter category are subject to only as-applied challenges. “When a law does not expressly or obviously refer to expression, the legislature is not required to consider all apparent applications of that law to protected expression and narrow the law to eliminate them.” Babson, 318 Multnomah County v. Mehrwein 355 Or at 400 . The limits on the first Robertson category do not make it an empty set, but they do restrict its application, and the accompanying high standard for facial validity, to those laws that, in directly regulating speech, pose the most danger to the expression protected by Article I, section 8. The above limitation holds, and we have stuck to it, even when the law has readily apparent applications to speech. It may be, and may have been, that many of the assaults prohibited by the law at issue in Plowman had some expressive content—that they were intended to convey disapproval of individuals or to communicate hatred of certain groups. Similarly, in Babson, it may have been, and likely was, the case that very many foreseeable uses of the Capitol steps bore some connection to expressive activity and that, as a practical matter, certain forms of expression were limited. Babson, 355 Or at 403 (“although the guideline does not directly refer to speech, the guideline does have apparent applications to speech, as defendants contend”). Nevertheless, we did not place those laws into the first Robertson category. “We do not lightly decide to overrule an earlier constitutional decision.” State v. Savastano, 354 Or 64, 95 , 309 P3d 1083 (2013). But the inconsistency that Vannatta I has produced is comparable to that which previously has justified our abandonment of an aberrant constitutional decision. In Savastano, we overruled our prior decision in State v. Freeland, 295 Or 367 , 667 P2d 509 (1983), which had held that Article I, section 20, of the Oregon Constitution required prosecutors to develop coherent, systematic policies to govern certain charging decisions. In Savastano, reviewing our other Article I, section 20, cases, we concluded that Freeland could not be reconciled with decisions before and since that set a less stringent standard in similar contexts. 354 Or at 91 . To be sure, in all reconsiderations of precedent, we must take into account the “undeniable importance of stability in legal rules and decisions.” Stranahan v. Fred Meyer, Inc., 331 Or 38, 53 , 11 P3d 228 (2000). In Savastano, we noted that Freeland had been a relative outlier and that “the cases that have followed Freeland have eroded its precedential Cite as 366 Or 295 (2020) 319 value and effectively returned to the more limited and historically grounded principle stated in [State v. Clark, 291 Or 231 , 630 P2d 810 , cert den, 454 US 1084 , 102 S Ct 640 , 70 L Ed 2d 619 (1981)].” Savastano, 354 Or at 96 . By contrast, in Ciancanelli, where we were asked to overrule Robertson and declined, we emphasized that “[m]any decisions of this court serve as precedent in later decisions. Thus, disavowing one case may undermine the precedential significance of several others,” 339 Or at 290—an observation that was especially true of Robertson, this court’s foundational decision on Article I, section 8. As we explained in Ciancanelli, “The contrast between Stranahan and this case illustrates the foregoing principle. In Stranahan, the allegedly erroneous decision had been rendered less than 10 years earlier, and few intervening precedents had relied on the earlier case, Lloyd Corporation v. Whiffen, 315 Or 500 , 849 P2d 446 (1993). The Stranahan majority simply acted at the earliest possible moment to correct what it perceived to be an analytical mistake made in the immediately preceding case, Lloyd Corporation. The present case, by contrast, involves a challenge not only to Robertson, but also to the many cases that this court has decided since 1983 that have utilized its methodology.” Id. at 290-91. In terms of the importance of stability in the law, Vannatta I resembles Lloyd Corporation. Although more time has passed between Vannatta I and this case than elapsed between Lloyd Corporation and Stranahan, the portion of Vannatta I’s holding in question here—its application of Robertson’s first category—has not been relied on in any other case. To the contrary, the only case to discuss it in detail, Vannatta II, distinguished Vannatta I and withdrew a material portion of its reasoning.8 And, in Vannatta II, no party asked this court to reconsider Vannatta I. Our other cases applying Robertson since Vannatta I, including [8] In Moyer we relied on different portions of Vannatta I, relating to the scope of the historical exception to Robertson. 348 Or at 236 . To be clear, Vannatta I was an opinion that discussed many issues, only some of which are contested in this case, and only one of which we reconsider in this opinion. We do not disavow all portions of Vannatta I, only those that, as we have explained, conflict with the Robertson framework. 320 Multnomah County v. Mehrwein Ciancanelli and Babson, have made no mention of Vannatta I, despite reasoning in ways that conflict with Vannatta I. And two cases on similar topics, Vannatta II, 347 Or at 464 , and Moyer, 348 Or at 230 , have attributed confusion by parties and lower courts to Vannatta I. Vannatta I is the “immediately preceding” case on campaign contributions. Although we have decided cases since Vannatta I touching on campaign finance, we have not reached the merits of an Article I, section 8, challenge in any of those cases.9 Respondents note that in those cases we referred to Vannatta I as governing precedent. But that is hardly surprising, as Vannatta I was governing precedent in each of those cases—a point that we made sure to distinguish, where appropriate, from the question of whether Vannatta I was correctly decided. Vannatta I derives little additional precedential force from the fact that, primarily in ballot title cases, we acknowledged its existence without endorsing its reasoning. In assessing the prudential factors that may counsel for or against overruling Vannatta I, we also consider the effect of a ballot measure that was submitted to the voters in 2006—Measure 46 (2006)—which “sought to amend the Oregon Constitution to permit the enactment of laws prohibiting or limiting electoral campaign ‘contributions and expenditures, of any type or description.’ ” Hazell v. Brown, 352 Or 455, 458 , 287 P3d 1079 (2012). Voters rejected Measure 46, although they approved Measure 47 (2006), a companion ballot initiative creating new campaign finance measures, conditional on a change in the constitutional limitation, such as the passage of Measure 46 or a judicial overruling of Vannatta I. Hazell, 352 Or at 462-63, 469 . Measure 46 plainly was directed at overruling the key holdings of Vannatta I, and it was rejected. Of course, a ballot measure that did not pass cannot change the meaning of the constitution or affect our duty to interpret it. Whether such a measure may affect our stare decisis analysis is a more nuanced [9] See Markley/Lutz v. Rosenblum, 362 Or 531, 533 , 413 P3d 966 (2018); Hazell v. Brown, 352 Or 455, 467-68 , 287 P3d 1079 (2012); Meyer v. Myers, 343 Or 399, 404-05 , 171 P3d 937 (2007); Meyer v. Bradbury, 341 Or 288 , 293 n 4, 142 P3d 1031 (2006). Cite as 366 Or 295 (2020) 321 issue. But even assuming that, in an appropriate case, an event like the failure of Measure 46 might weigh against overturning a precedent, it has little impact here. Just as the “legislature may decline to address a judicial decision for any number of reasons, none of which necessarily constitutes an endorsement of the decision’s reasoning or result,” Farmers Ins. Co. of Oregon v. Mowry, 350 Or 686, 696 , 261 P3d 1 (2011), so too may the people decline to adopt a proposed constitutional amendment for a myriad of reasons. And there are explanations for Measure 46’s failure other than the possibility that voters meant to express their approval of Vannatta I’s contribution limits holding. Most obviously, Vannatta I struck down both contribution limits and expenditure limits, and Measure 46 would have amended Article I, section 8, with respect to both. However, independent expenditures present a different, and potentially more difficult, constitutional problem than campaign contributions. For example, under the First Amendment, the United States Supreme Court has, since Buckley, held that expenditure limits place a greater burden on constitutionally protected expression than contribution limits do. Even some proponents of campaign finance reform view a constitutional amendment permitting expenditure limits as dangerous, in view of the effect that such an amendment might have on other areas of law. See Richard L. Hasen, Three Wrong Progressive Approaches (and One Right One) to Campaign Finance Reform, 8 Harv L & Pol’y Rev 21, 26-27 (2014). Thus, voters could have rejected Measure 46 because of its application to expenditure limits, even if they supported the measure to the extent that it applied to contribution limits. Moreover, parsing the amendment’s failure in this case is even more difficult, because of the simultaneous adoption of Measure 47 (2006), which would have created limits on both contributions and expenditures if Measure 46 was adopted or Vannatta I was overruled to a sufficient degree. See Hazell, 352 Or at 458 . One voter could have supported such limits but believed that the constitutional amendment went too far. Another could have preferred to have this court reconcile such limits with the constitution, rather than risking an amendment that could threaten speech. As a result, 322 Multnomah County v. Mehrwein the rejection of Measure 46 deserves no real weight in our stare decisis analysis. Ultimately, we do not believe that we can have one Article I, section 8, approach to laws restricting campaign contributions and another for all other laws. Vannatta I itself rejected any “distinctions based on the ‘centrality’ of particular forms of expression” when explaining why, contra Buckley, it would not apply a different standard to contribution limits than to expenditure limits. 324 Or at 521 . And we cannot honor stare decisis by expanding Vannatta I’s application of the first Robertson category to all laws— allowing facial challenges to laws restricting conduct that has an expressive component in many or most applications. That also would require overturning precedent, and more of it. Vannatta II, although correct in its own application of Robertson, did not successfully rehabilitate Vannatta I, nor did it supply a viable basis on which Vannatta I could be distinguished from this court’s other Article I, section 8, cases. Rather, it deepened the confusion surrounding Vannatta I’s basis and validity. Given the clear conflict between Vannatta I and our other cases, it is Vannatta I that must give way. We disavow the reasoning in Vannatta I that campaign contribution limits necessarily are Robertson category one laws. Vannatta I erred in holding contribution limits unconstitutional based on that reasoning. The correct inquiry, under Robertson, is whether such limits are “written in terms directed to the substance of any ‘opinion’ or any ‘subject’ of communication.” Robertson, 293 Or at 412 .10 The remaining question in this case is whether the county’s contribution limit ordinance is such an express restriction. 4. Application to this case We now undertake that inquiry and examine the text of the campaign contribution limits at issue here. MCC § 5.201 provides: “(A) An Individual or Entity may make Contributions only as specifically allowed to be received in this Section. “(B) A Candidate or Candidate Committee may receive only the following contributions during any Election Cycle: [10] Vannatta I did not conduct that inquiry, and we need not decide here what result would have obtained if it had. Cite as 366 Or 295 (2020) 323 “(1) Not more than five hundred dollars ($500) from an Individual or Political Committee other than a Small Donor Committee; “(2) Any amount from a Small Donor Committee; and “(3) No amount from any other Entity.” The county’s definition of “contribution,” contained in MCC § 5.200, cross-references the definitions in ORS 260.005(3) and ORS 260.007.11 ORS 260.005(3) provides: “Except as provided in ORS 260.007, ‘contribute’ or ‘contribution’ includes: “(a) The payment, loan, gift, forgiving of indebtedness, or furnishing without equivalent compensation or consideration, of money, services other than personal services for which no compensation is asked or given, supplies, equipment or any other thing of value: “(A) For the purpose of influencing an election for public office or an election on a measure, or of reducing the debt of a candidate for nomination or election to public office or the debt of a political committee; or “(B) To or on behalf of a candidate, political committee or measure; and “(b) The excess value of a contribution made for compensation or consideration of less than equivalent value.” However, the county argues that only a portion of that definition is operative in the context of MCC § 5.201. That contention requires a brief digression into statutory construction. There is no dispute that MCC § 5.201 sets limits on contributions to candidates and candidate committees. The plain text of MCC § 5.201(B) establishes that a candidate can accept a thing of value from an individual or entity only if (1) it falls within the contribution limits or (2) it is excluded from the definition of “contribution” by MCC § 5.200 or ORS 260.007. However, amicus Oregon Taxpayers Association (OTA) argues that MCC § 5.201(A) also prohibits contributions to ballot measure campaigns and contributions to be used in independent expenditures. [11] ORS 260.007 contains exceptions to the definitions of “contribution” and “expenditure.” 324 Multnomah County v. Mehrwein The county disavows both applications and contends that MCC § 5.201(A) applies only to contributions to candidates and a principal candidate committee. We agree with the county that the contribution limits imposed by MCC § 5.201 apply only to contributions to candidates in Multnomah County candidate elections and their principal candidate committees. We approach the question using our usual methodology for statutory interpretation. State v. Gaines, 346 Or 160 , 206 P3d 1042 (2009). To begin with, it makes little sense to read the text of MCC § 5.201(A)—“[a]n Individual or Entity may make Contributions only as specifically allowed to be received in this Section”—as a complete prohibition on all contributions in all elections. The title of that provision—which was part of the text submitted to voters—is “CONTRIBUTIONS IN MULTNOMAH COUNTY CANDIDATE ELECTIONS” (uppercase in original). It therefore makes sense to read § 5.201(A), in context, as a prohibition on contributions in Multnomah County candidate elections, not a much more extensive proscription. That reading is confirmed by legislative history. As submitted to voters, the ballot measure explained that it would create a new charter provision limiting “[c]ontributions to political campaigns for candidates running for county elective offices.” Multnomah County Voters’ Pamphlet, General Election, November 8, 2016, M-28. The fact that the county’s ordinance cross-references a state statutory definition of “contribution” that refers to “measures” does not mean that the substantive limits imposed by the county provision apply beyond their intended scope. Similarly, OTA’s argument that MCC § 5.201(A) forbids all contributions to be used in independent expenditures proves unpersuasive in light of its context. Other provisions of the county’s ordinance clearly contemplate that such contributions will take place. For example, MCC § 5.202(C)(3) provides that “[a] Political Committee may make aggregate Independent Expenditures of not more than ten thousand dollars ($10,000), provided that the Independent Expenditures are funded by means of contributions to the Political Committee Cite as 366 Or 295 (2020) 325 by Individuals in amounts not exceeding five hundred dollars ($500) per Individual per year.” If OTA’s reading of MCC § 5.201(A) were correct, MCC § 5.202(C)(3) would be rendered a nullity, because no political committee other than a candidate committee could accept contributions of any size. The more plausible reading, and the only reading consistent with the legislative history quoted above, is that MCC § 5.201(A) restricts contributions only to candidates and candidate committees, the two entities to which the limits in MCC § 5.201(B) apply. Therefore, only a portion of the definition of “contribution” quoted above is operative in the context of MCC § 5.201. The other parts of that definition would have a role to play in a more expansive restriction of contributions— such as the contribution limits at issue in Vannatta I or OTA’s broad reading of MCC § 5.201(A). But, because the county’s contribution limits apply only to contributions to candidates or candidate committees, only a portion of the statutory definition is left with any role to play. With irrelevant or redundant portions omitted, the operative definition reads: “(a) The payment, loan, gift, forgiving of indebtedness, or furnishing without equivalent compensation or consideration, of money, services other than personal services for which no compensation is asked or given, supplies, equipment or any other thing of value: “* * * * * “(B) To or on behalf of a candidate [or] political committee * * *; and “(b) The excess value of a contribution made for compensation or consideration of less than equivalent value.” No portion of that definition contains an express reference to speech. Nor, as in Ciancanelli, does the definition target conduct only insofar as it is expressive. Instead, “contribution” is defined in terms of conduct that is not necessarily expressive. Vannatta II’s analysis of the gift limit, which was written in similar terms, controls here: 326 Multnomah County v. Mehrwein “A public official who is subject to restrictions on the receipt of gifts can violate the restrictions without saying a word, without engaging in expressive conduct, and regardless of any opinion that he or she might hold.” 347 Or at 459 .12 Accordingly, we conclude that MCC § 5.201(A) and (B) are not subject to facial challenge under Robertson. They are not, therefore, facially invalid under Article I, section 8. The county’s limits may be subject to as applied challenges. But, in this case, respondents have raised only a facial challenge to the county’s contribution limits, and we reject that challenge. B. Contribution Limits and the First Amendment Because the trial court concluded that the county’s contribution limits violated Article I, section 8, it did not reach the question of whether the limits violate the First Amendment. The First Amendment sets limits on the regulation of campaign contributions. In Buckley, the Supreme Court upheld individual contribution limits of $1,000, recognizing that, although contribution limits restricted speech to some extent, the government had a significant interest in preventing corruption or the appearance of corruption. 426 US at 26. Although Buckley upheld contribution limits of $1,000, the Supreme Court has since made clear that Buckley did not set a floor. In Nixon v. Shrink Missouri Government PAC, 528 US 377 , 120 S Ct 897 , 145 L Ed 2d 886 (2000), the Supreme Court rejected a First Amendment challenge to Missouri’s $1,075 individual contribution limit in statewide elections. That limit, when adjusted for inflation, was lower than that upheld in Buckley in 1968. Nevertheless, emphasizing that Buckley did not set a floor, the Court rejected the challenge. Shrink Missouri Government PAC, 528 US at 396-97 . Randall v. Sorrell, 548 US 230 , 126 S Ct 2479 , 165 L Ed 2d 482 (2006), was the first Supreme Court decision [12] We need not, and do not, consider whether a contribution limit in which ORS 260.007(a)(A)—and its reference to “the purpose of influencing an election for public office or an election on a measure”—were operative would be an express restriction of speech. For that reason, we do not consider whether Vannatta I’s result—which was to strike down a broader set of contribution limits where contributions were so defined—was incorrect. Cite as 366 Or 295 (2020) 327 to find a contribution limit facially invalid. In that case, six Justices, across three opinions, held that Vermont’s contribution limit scheme, which involved contribution limits for statewide races as low as $200, was unconstitutional. Although no rationale commanded a majority of the Court, Justice Breyer’s opinion provided the narrowest ground for the judgment and is therefore binding on this court. See Marks v. United States, 430 US 188, 193 , 97 S Ct 990 , 51 L Ed 2d 260 (1977) (“When a fragmented Court decides a case and no single rationale explaining the result enjoys the assent of five Justices, the holding of the Court may be viewed as that position taken by those Members who concurred in the judgments on the narrowest grounds.” (Internal quotations and citation omitted.)); Thompson v. Hebdon, ___ US ___, ___, 140 S Ct 348 , 350 & n (2019) (per curiam) (recognizing that Justice Breyer’s opinion in Randall is controlling). The opinion in Randall framed the question as being whether the contribution limits “prevent candidates from ‘amassing the resources necessary for effective [campaign] advocacy’; whether they magnify the advantages of incumbency to the point where they put challengers to a significant disadvantage; in a word, whether they are too low and too strict to survive First Amendment scrutiny.” Randall, 548 US at 248 (opinion of Breyer, J.) (quoting Buckley, 424 US at 21 ; alteration in Randall; internal citation omitted). The opinion in Randall answered that question using a two-staged approach. It explained that “where there is strong indication in a particular case, i.e., danger signs, that such risks exist (both present in kind and likely serious in degree), courts, including appellate courts, must review the record independently and carefully with an eye toward assessing the statute’s ‘tailoring,’ that is, toward assessing the proportionality of the restrictions.” Id. at 249. First, the opinion identified “danger signs,” principally that Vermont’s “contribution limits are substantially lower than both the limits we have previously upheld and 328 Multnomah County v. Mehrwein comparable limits in other States.” Id. at 253. Having reached that conclusion, the opinion turned to a closer examination of Vermont’s contribution limits, considering evidence concerning the likely effect of those limits. Ultimately, the opinion noted five characteristics that led it to conclude that Vermont had set its contribution limits too low, including an unusually expansive definition of “contribution” and the fact that the same limits that applied to individuals applied to political parties. Id. at 256-61. The opinion concluded that the scheme, as a whole, was facially invalid. Id. at 261. Since Randall, and after briefing in this case was completed, the Supreme Court weighed in once again. In Thompson, the court held that Alaska’s $500 contribution limit for all political candidates had “danger signs” similar to those found in Randall. ___ US at ___, 140 S Ct at 350-51 . Some of those “danger signs” are present here. Multnomah County’s contribution limits are substantially lower than those upheld by the Supreme Court in the past— adjusted for inflation, $500 is less than a third of the limit upheld in Shrink Missouri. See Thompson, ___ US at ___, 140 S Ct at 351 (doing similar math). Both Randall and Thompson treated similar comparisons as “danger signs,” although neither viewed that single factor as dispositive. See Thompson, ___ US at ___, 140 S Ct at 350-51 ; Randall, 548 US at 251 . And the $500 limit at issue here is effectively lower than the $500 limit found problematic in Thompson. Under the county’s ordinance, $500 is the maximum individual-to-candidate contribution over a two-year election cycle. MCC §§ 5.200, 5.201(B). By contrast, the $500 Alaska limit analyzed in Thompson limits individual-to-candidate contributions “to $500 per year,” Thompson, ___ US at ___, 140 S Ct at 348 (emphasis added), meaning that it “allows a maximum contribution of $1,000 over a comparable twoyear period,” id. at ___, 140 S Ct at 350-51 . Yet the county’s ordinance differs from the laws considered in Randall and Thompson in pertinent respects. The county’s $500 limits apply only to county elections. While in Randall and Thompson the Court emphasized that the laws it was considering were out of step with contribution limits set by other states, it is less clear that Multnomah County’s Cite as 366 Or 295 (2020) 329 limits are inconsistent with those set by comparable municipalities. See Zimmerman v. City of Austin, Texas, 881 F3d 378, 387 (5th Cir 2018), cert den, ___ US ___, 139 S Ct 639 , 202 L Ed 2d 492 (2018) (upholding Austin’s $350 city council contribution limits based, in part, on comparisons to contributions limits in other large municipalities). In addition, in both Randall and Thompson, the Court deemed it particularly problematic that the contribution limits at issue in those cases were not indexed for inflation. Multnomah County’s contribution limits are automatically adjusted for inflation in every odd-numbered year. MCC § 5.205. The controlling Supreme Court precedent makes it difficult to decide whether the county’s contribution limits violate the First Amendment without further proceedings in the trial court. In a First Amendment analysis, the constitutionality of a contribution limit depends not only on whether there are “danger signs,” but also on the government’s interest in imposing contribution limits and the effect the limits could have on candidates’ ability to conduct an effective campaign. See Randall, 548 US at 253-56 (opinion of Breyer, J.) (evaluating the likely effect of Vermont’s contribution limits in light of the evidence in the record); Thompson, ___ US at ___, 140 S Ct at 351 (remanding for consideration of whether Alaska had shown a special justification for its contribution limit). Here, the parties and amici submitted evidence on the problems that the contribution limits addressed and their likely effects, and the county relies in part on that evidence and other empirical support for its argument that its ordinances survive First Amendment scrutiny. Those arguments turn on facts that are not conceded and on particular inferences that may be—but need not be—drawn from that evidence. Because the trial court never reached the First Amendment issue, it did not make the factual findings that are necessary to that analysis. We therefore remand the case to the trial court to address the validity of the county’s contribution limits under the First Amendment in the first instance. III. EXPENDITURE LIMITS We next consider the validity of the county’s expenditure limits under Article I, section 8. The two provisions of 330 Multnomah County v. Mehrwein the county’s expenditure ordinance that the trial court held invalid state: “(A) No Individual or Entity shall expend funds to support or oppose a Candidate, except those collected from the sources and under the Contribution limits set forth in this Section. “* * * * * “(C) Only the following Independent Expenditures are allowed per Election Cycle to support or oppose one or more Candidates in any particular Multnomah County Candidate Election: “(1) An Individual may make aggregate Independent Expenditures of not more than five thousand dollars ($5,000). “(2) A Small Donor Committee may make Independent Expenditures in any amounts from funds contributed in compliance with Section 5.200. “(3) A Political Committee may make aggregate Independent Expenditures of not more than ten thousand dollars ($10,000), provided that the Independent Expenditures are funded by means of contributions to the Political Committee by Individuals in amounts not exceeding five hundred dollars ($500) per Individual per year.” The county’s expenditure limits cannot be distinguished from those held unconstitutional in Vannatta I. The county argues that Vannatta I did not fully consider the question of whether expenditures are protected expression and takes the position that its expenditure limits do not violate Article I, section 8. However, the county acknowledges that its expenditure limits violate the First Amendment under existing Supreme Court precedent13 and for that reason accepts that this is not an appropriate case in which to reconsider the validity of expenditure limits under Article I, section 8. Trojan also argues that the validity of expenditure limits was paid insufficient attention in Vannatta I and contends that we should uphold the county’s expenditure limits under Article I, section 8. [13] The county argues that the controlling federal cases were wrongly decided, but it acknowledges that this court is not the proper forum for that argument. Cite as 366 Or 295 (2020) 331 We decline to reconsider Vannatta I’s expenditures holding for three reasons. First, nothing that we have disavowed regarding Vannatta I’s reasoning concerning contribution limits calls into question Vannatta I’s conclusion that limits on independent expenditures are an express restriction on speech subject to a facial challenge under Robertson. The definition of “independent expenditure” refers expressly to the content of the communications whose funding it restricts, bringing the limits on independent expenditures within the first Robertson category. ORS 260.005(10) (2015), made applicable by MCC § 5.200 (“Unless otherwise indicated by the text or context of this Section, all terms shall have the definitions at Chapter 260 of Oregon Revised Statutes, as of November 8, 2016.”), defined an independent expenditure as “an expenditure by a person for a communication in support of or in opposition to a clearly identified candidate or measure” made independently of a candidate or campaign. That phrase is further defined to refer to a communication that “clearly and unambiguously urges the election or defeat of a clearly identified candidate for nomination or election to public office, or the passage or defeat of a clearly identified measure,” ORS 260.005(10)(c)(A)(i) (2015), or communications that “refer[ ] to a clearly identified candidate who will appear on the ballot or to a political party,” ORS 260.005(10)(c)(B)(ii) (2015). As in Ciancanelli, even if expenditures may be viewed as conduct, the county’s ordinance restricts them only insofar as they are expressive. Second, although the county, Trojan, and some amici have argued that Vannatta I should be overturned as to expenditure limits as well, the briefing that they have filed is more straightforwardly directed to the validity of the contribution limits, and the expenditure limit briefing is not as well developed. Third, as the county concedes, and we agree, the county’s expenditure limits unambiguously violate the First Amendment. Buckley held that the government cannot restrict independent expenditures by individuals, 424 US at 47-51 . Citizens United, 558 US 310 , held that independent restrictions by corporations and unions cannot be restricted either. The county’s ordinance restricts both. To be sure, we interpret the Oregon Constitution independently of the 332 Multnomah County v. Mehrwein First Amendment, and our free speech jurisprudence does not track the Supreme Court’s interpretation of the First Amendment. But, although Buckley and Citizens United are not relevant to the question of whether Vannatta I was correctly decided, the futility of reconsidering Vannatta I with respect to this plainly unconstitutional ordinance weighs against doing so in this case. See State v. Avila-Nava, 356 Or 600, 621 , 341 P3d 714 (2014) (Kistler, J., concurring in part and concurring in the judgment) (“There would be little point * * * in announcing a state constitutional rule that permits Oregon courts to consider evidence that the Fifth Amendment precludes them from considering.”). Accordingly, we affirm the trial court’s decision that MCC § 5.202(A) and (C) are invalid. IV. DISCLOSURE PROVISIONS Some of the intervenors assign error to the trial court’s decision on the disclosure provisions, but that issue is now moot: During the pendency of this proceeding, the county amended the disclosure provisions of its ordinance and a decision about the validity of the former provisions will have no practical effect. See Kerr v. Bradbury, 340 Or 241, 244 , 131 P3d 737 , adh’d to on recons, 341 Or 200 , 140 P3d 1131 (2006) (stating that case is moot when a decision will “no longer have some practical effect on the rights of the parties to the controversy” (internal citations and quotations omitted)). Although it is sometimes appropriate for an appellate court to vacate the affected portion of the trial court’s judgment, we have not been asked to employ the “ ‘equitable remedy of vacatur,’ ” id. at 249 (quoting U.S. Bancorp Mortgage Co. v. Bonner Mall Partnership, 513 US 18, 25 , 115 S Ct 386 , 130 L Ed 2d 233 (1994)), and we do not do so. V. CONCLUSION The trial court ruled that three provisions of the county’s ordinances violated Article I, section 8. We conclude that the contribution limits are not facially invalid under Article I, section 8, and therefore reverse that portion of the trial court’s decision and remand the case to the trial court so that it can consider whether the contribution limits are Cite as 366 Or 295 (2020) 333 valid under the First Amendment. We agree with the trial court that the expenditure limits violate Article I, section 8, and we affirm the trial court’s judgment as to those provisions. Although the trial court held that the disclosure rules violated Article I, section 8, that part of its decision became moot on appeal, and we decline to decide the now-theoretical question. The judgment of the circuit court is affirmed in part and reversed in part, and the case is remanded to the circuit court for further proceedings.