concurring.
I.
¶ 31 I have previously suggested that our separate amendment rule case law needed clarification. Clean Elections Inst., Inc. v. Brewer, 209 Ariz. 241, 248 ¶¶ 28-30, 99 P.3d 570, 577 (2004) (Hurwitz, J., concurring). Today’s opinion undertakes that task and admirably clears out a considerable amount of our jurisprudential underbrush.
¶32 Today the Court appropriately re-tens to first principles — the test articulated more than seventy years ago in Kerby v. Luhrs, 44 Ariz. 208, 36 P.2d 549 (1934). That test, as the Court explains, has two components. First, all provisions of a proposed amendment must relate to the same “general topic.” Id. at 221, 36 P.2d at 554. This is simply a test of germaneness. Although most proposed constitutional amendments will easily pass through this initial screen, a proposed amendment with separate provisions relating to, for instance, education and worker’s compensation would plainly fail this test.
¶33 But Kerby requires more. In that case, the various provisions of the proposed constitutional amendment all related to the same general topic — taxation. This Court, however, found that the proposal did not pass muster under the separate amendment rule because the various provisions were not sufficiently interrelated. This second part of the Kerby test requires that “logically speaking,,” the various provisions “should stand or fall as a whole.” Id. (emphasis added).
¶ 34 Our cases have generally applied this second prong of the Kerby test in a straightforward fashion. Tilson v. Mofford, 153 Ariz. 468, 737 P.2d 1367 (1987), is a paradigm. Tilson involved a proposed constitutional amendment with four provisions authorizing the Legislature to regulate tort damages. The Court first noted that the various propositions “all related to the same topic of tort damages.” Id. at 472, 737 P.2d at 1371. The first part of the Kerby test— topicality — was thus satisfied. The Court then went on to note that the various provisions in the amendment “all logically related to each other.” Id. (emphasis added).
¶ 35 Slayton v. Shumway, 166 Ariz. 87, 800 P.2d 590 (1990), is to the same effect. In upholding an initiative involving victims’ *128rights against a separate amendment rule challenge, the Court again applied the two-step Kerby analysis. It concluded that all provisions of the initiative dealt with the same topic — “victims’ proposals.” Id. at 92, 800 P.2d at 595. This satisfied the requirement of topicality. The Court also dealt with the argument that one provision of the proposed amendment, which gave the Legislature the power to promulgate rules relating to victims’ rights, was in reality a separate amendment and thus violated Article 21, Section 1. The Court rejected that argument because it agreed with the proponents of the initiative that this provision was “more than ‘reasonably related’ to the rest of the proposition.” Id. This satisfied the second prong of the Kerby test.
¶ 36 In this case, there is no doubt that the two parts of the Kerby test are satisfied. As the Court notes, both provisions of Proposition 107 relate to the same general topic— marriage. Op. ¶ 8. And, it is also clear that the two provisions have a “logical relationship.” Id. ¶ 17. One provision defines marriage as only being between a man and a woman; the second enforces the first by preventing governmental agencies from enacting marriage substitutes — relationships substantively identical to marriage but called by a different name.
¶ 37 To be sure, the second provision in the initiative before us today is not necessarily required by the first. It is quite possible to limit the institution of marriage to persons of different gender while allowing same-sex couples to enter into marriage-like relationships. But Kerby does not require that various provisions of a proposal all be required by the others. If that were the case, the initiative in Slayton, to use but one example, would not have passed muster; it was not necessary to extend rule-making power to the Legislature to protect victims’ rights. But it was logical and reasonable to do so, and the rule-making provision therefore met the interrelatedness test. The provisions here also do so.
¶ 38 I therefore concur in the Court’s conclusion that Proposition 107 does not violate the separate amendment rule. I write separately to address two points briefly.
II.
A.
¶ 39 The Court’s conclusion that the two provisions of Proposition 107 “share a logical relationship,” Op. ¶ 17, more than suffices to establish the interrelationship required by the second prong of the Kerby test. That conclusion should end the analysis. I would leave for another day the question of whether in some future ease the second prong could alternatively be satisfied by establishing one of the four “objective factors” listed in Korte v. Bayless, 199 Ariz. 173, 177 ¶ 11, 16 P.3d 200, 204 (2001), or by some other showing. See Op. ¶¶ 10-17.
¶ 40 In my view, Korte did not correctly state or apply the Kerby interrelatedness test. As Chief Justice Zlaket cogently noted in dissent, there was simply no logical or reasonable relationship among many of the provisions of the proposed amendment at issue in Korte. 199 Ariz. at 179-80 ¶¶ 19-26, 16 P.3d at 206-07 (Zlaket, C.J., dissenting). For example, one provision of the proposed amendment allowed long-term leases of grazing land without public auction while another permitted school districts to obtain trust land at no cost. Id. at 179 ¶ 20, 16 P.3d at 206. The Korte majority did not find these provisions logically related to each other or to any other provision of the subject initiative; it instead concluded that the various provisions constituted a “multifaceted approach” to the complex issue of managing state lands wisely and thus had a single purpose. Id. at 178 ¶ 15, 16 P.3d at 205. But such can be said of virtually any proposed constitutional provision that meets the topicality prong of the Kerby test — each provision can be viewed as one facet of improving the constitution’s treatment of that topic or solving a complex problem.
¶ 41 I think that the second prong of the Kerby test requires more — a reasonable or logical relationship of the various provisions with each other, and not simply with the broader topic that they cover. It is this interrelatedness which, in the words of Ker-by, ensures that “logically speaking, they should stand or fall as a whole,” 44 Ariz. at *129221, 36 P.2d at 554, so that the provisions form one amendment, not several.
¶ 42 In finding the two provisions of Proposition 107 sufficiently interrelated to pass separate amendment rule scrutiny, the Court today cites language in Korte suggesting that the separate amendment rule is satisfied when “the various provisions are qualitatively similar in their effect on either procedural or substantive law.” Op. ¶¶ 10, 16-17. Korte in turn cited Slayton in support of this formulation, Korte, 199 Ariz. at 177 ¶ 11, 16 P.3d at 204, but this language does not appear in Slayton, nor do I believe that it accurately characterizes the analysis in Slayton. Rather, as noted above, I think that Slayton straightforwardly applied the topicality/logical relationship test set forth in Kerby.
¶ 43 A great virtue of the Court’s decision today is the return to the historic Kerby test. Because a logical relationship between two provisions plainly satisfies the second prong of that test, I would not today attempt to tease out of our post-Kerby cases other “objective factors” establishing interrelationship, and I am particularly reluctant to use Korte as an avatar.
B.
¶ 44 Perhaps the most useful aspect of the Court’s opinion is its interment of the “reasonable voter” test. Op. ¶¶ 18-20. I have previously noted my discomfort with that test, which requires “a judicial determination of whether a voter supporting one part of a proposed amendment would ‘be expected to support the principle of the others’” and thus “involves the Court in a prediction of voter preferences and behavior that is often somewhat subjective.” Clean Elections, 209 Ariz. at 248 ¶ 29, 99 P.3d at 577 (Hurwitz, J., concurring) (quoting Kerby, 44 Ariz. at 221, 36 P.2d at 554). Moreover, the test is in some ways unrelated to the true purpose of the separate amendment rule. It may be empirically true, for example, that all voters who oppose gun control also oppose trade with China. Yet these common preferences would not suffice to protect a constitutional amendment addressing both issues from an attack under Article 21, Section 1.
¶45 As the Court notes, our past cases applying the reasonable voter test have also relied on other parts of the Kerby rubric in determining whether a particular proposal satisfied the separate amendment rule. Op. ¶ 20. Whether each of those cases was correctly decided is not before us today, and there is no reason to revisit each of our prior opinions to speculate whether they would have come out the same way in the absence of the reasonable voter test. But, at the very least, our past focus on the mythical reasonable voter has required us to engage in a predictive exercise for which judges are ill-suited and which had the potential of producing inconsistent results. By limiting our analysis to two far more objective factors— topicality and interrelatedness — the Court’s opinion should add greater predictability to our future separate amendment rule jurisprudence.