WINFREE, Justice.

Margaret Paton Walsh and Laura Fox, Assistant Attorneys General, Anchorage, and Kevin G. Clarkson, Attorney General, Juneau, for Appellants.

Scott M. Kendall, Jahna M. Lindemuth, and Samuel G. Gottstein, Holmes Weddle & Barcott, P.C., Anchorage, for Appellee.

Before: Bolger, Chief Justice, Winfree, Stowers, Maassen, and Carney, Justices.

OPINION

WINFREE, Justice.

I. INTRODUCTION

The Alaska Constitution provides that all political power is inherent in Alaska's people and "founded upon their will only." The people may exercise this political power in a number of ways. The people have the constitutional right to vote in any state or local election, and "it is basic to our democratic society that the people be afforded the opportunity of expressing their will on the multitudinous issues which confront them." As a corresponding check to the constitutional right to elect officials, and a check on those elected officials' conduct, the people have the constitutional right to petition to recall elected officials. And, as a check on legislative action or inaction, the people have the constitutional rights to reject legislative acts by referendum and to legislate directly by initiative. This appeal arises from the State's action limiting the people's constitutional right to legislate directly by initiative.

A proposed initiative instituting three substantive changes to Alaska's election laws was submitted to the lieutenant governor for review, certification, and printing signature booklets. Determining the initiative violated a constitutional requirement that proposed initiative bills be confined to one subject, the lieutenant governor denied certification. The initiative's sponsors filed a superior court action challenging that decision. The superior court concluded, contrary to the lieutenant governor, that the initiative's various provisions were confined to the single subject of "election reform" and it accordingly should be certified; the court directed that the State distribute petition booklets for the sponsors to collect signatures for placing the initiative on a future election ballot.

The lieutenant governor and the State's elections office appeal the superior court's decision. But because the court correctly adhered to our prior interpretation of the relevant constitutional provisions — and because we reject the request to reverse precedent that the people's power to initiate laws generally is equivalent to that of the legislature — we affirm the court's decision.

II. CONSTITUTIONAL BACKDROP

A. Law-Making By Initiative

Understanding the one-subject rule's application to an initiative first requires understanding the people's law-making power under the Alaska Constitution. Article XI, section 1 provides in simple fashion: "The people may propose and enact laws by the initiative, and approve or reject acts of the legislature by the referendum."

The constitutional convention delegates debated extensively whether to include an initiative provision in the Alaska Constitution. Delegates recognized that an initiative provision would be a check on the legislature. But they also recognized that an initiative provision should not apply to certain areas of law-making, and article XI specifies limitations in section 7: "The initiative shall not be used to dedicate revenues, make or repeal appropriations, create courts, define the jurisdiction of courts or prescribe their rules, or enact local or special legislation."

The delegates voted late in the constitutional drafting process to include the initiative provision. But the style and drafting committee had been using "by law" and "by the legislature" interchangeably in the constitutional text, raising concern there could be confusion whether the phrase "by law" applied to both the legislature's power and the people's initiative power. To avoid confusion, the delegates included article XII, section 11, a general provision regarding law-making power: "Unless clearly inapplicable , the law-making powers assigned to the legislature may be exercised by the people through the initiative, subject to the limitations of [a]rticle XI." (Emphasis added.)

We previously have explained article XII, section 11's language: "The phrase ‘unless clearly inapplicable’ was included in the Alaska Constitution ‘so that the initiative would not replace the legislature where the legislature's power serves as a check on other branches of government, such as legislative power to define courts' jurisdiction or override judicial rules.’ " Common sense about law-making determines when, under article XII, section 11, the people's law-making power is not co-equal with the legislature's: "To test whether the initiative is ‘clearly inapplicable,’ one must ask whether ‘even 55 idiots would agree that it was inapplicable.’ " The constitutional provisions and the delegates' debates on the initiative thus make clear that the delegates intended the people's initiative law-making power be equivalent to the legislature's lawmaking power, except in specifically enumerated and other "clearly inapplicable" circumstances.

It is worth noting that the initiative is only one of numerous check and balance mechanisms found in our separation of powers form of government. This particular people's check on government has much in common with constitutional referendum and recall rights:

The initiative and referendum are devices that permit the electorate to participate directly in the law-making process. Through the initiative the voters may enact legislation, and through the referendum

they may veto laws passed by a recent legislature. Through the recall, the voters may remove an elected official from office. The initiative and referendum are known as "direct democracy" provisions. They first appeared in this country during the populist reform movement of the early twentieth century, and they are found in one form or another in about half of the state constitutions.^{[ ]}

There are corresponding checks on the people's right to initiate laws. One, involving restrictions on subject matter, is described above. Another is the legislature's power to effectively terminate an initiative by passing "substantially the same" legislation prior to an election. We will later return to this checks and balances theme.

B. One-Subject Rule

1. Constitutional history

Alaska's constitutional one-subject rule is contained in article II, section 13, entitled "Form of Bills":

Every bill shall be confined to one subject unless it is an appropriation bill or one codifying, revising, or rearranging existing laws. Bills for appropriations shall be confined to appropriations. The subject of each bill shall be expressed in the title. The enacting clause shall be: "Be it enacted by the Legislature of the State of Alaska."

The one-subject rule's origins lie with the Roman Empire. The Lex Caecilia Didia , enacted in 98 B.C., was a legislative procedural law prohibiting unrelated provisions being combined within one bill (lex satura ). The one-subject rule has been used over the centuries to maintain clarity and transparency in law-making and to eliminate logrolling and riders. "Log[ ]rolling consists of deliberately inserting in one bill several dissimilar or incongruous subjects in order to secure the necessary support for passage of the measure." "Riders" are provisions often "unrelated to the main purpose of the bill" that are "attached to bills that are popular and so certain of adoption that the rider will secure adoption not on its own merits, but on the merits of the measure to which it is attached."

The first state to adopt a constitutional one-subject rule was Illinois, in 1818. By 1958, the year before Alaska officially reached statehood, nearly 40 states had adopted varying constitutional one-subject provisions. And the one-subject rule existed in Alaska well before statehood. Perhaps because the one-subject rule was such a common fixture, the delegates agreed with little fanfare to its inclusion in the Alaska Constitution.

During the Constitution's creation, standing committees drafted proposed constitutional articles that were submitted to the delegates for consideration. The Committee on Legislative Branch's proposal included a one-subject rule. The committee explained that, despite deference to the legislature on form and procedure for enactment, the Constitution would require a bill to be "confined to one subject." The committee also stated: "The use of riders on appropriation bills is prohibited, because of the abuses that have arisen in the Congress and in some states which do not have such a restriction."

The delegates only briefly explained the one-subject rule's inclusion in the Alaska Constitution; they expressly recognized that most states had adopted similar rules and that a one-subject rule would establish "minimum safeguards" in the bill enactment process. Delegate Steve McCutcheon stated:

The theory of requiring that all bills be confined to one subject with certain exceptions ... is that nothing can be gotten through the legislature under the guise of some other things. Often times a bill that is very popular and has a great deal of public support and sentiment will have a rider attached to it which may defeat the very purpose of the bill or may pertain to some other idea entirely, and the theory behind the requirement that each bill be confined to one subject indicates th[at] thinking.^{[ ]}

Returning to the checks and balances theme discussed earlier, it also is worth noting that in conjunction with other constitutional provisions — like the "direct democracy" provisions for the people to "veto" legislatively enacted laws through the referendum and directly initiate laws through the initiative — the one-subject rule works to implement checks and balances in our form of government. For example, under the Alaska Constitution a governor has the power to veto a legislative bill enacting a law, but only in its entirety, and the power to veto or reduce individual items in legislative appropriations bills, which are immune from the one-subject rule. Like the one-subject rule, the constitutional "item veto" for appropriations bills "originated as a reform measure to prevent legislators from ‘logrolling’ when they enact appropriation bills which necessarily address many subjects and need not be confined to a single subject, and to give governors some ability to limit state expenditures." If the legislature could pass bills enacting laws about multiple subjects in the same way it can pass appropriation bills with multiple subjects, a "governor's veto power would be compromised because the legislature could pair a subject that the governor opposed with one that [the governor] favored."

2. Our case law applying the one-subject rule

The delegates described the one-subject rule as a minimum safeguard to law-making, but they did not define "one subject." We therefore have identified its parameters when asked to apply the one-subject rule.

a. Gellert v. State

Our seminal one-subject-rule decision is Gellert v. State . That case involved the legislature's enactment of a bill to issue general obligation bonds funding (1) small boat harbor projects for coastal towns and villages and (2) a Fairbanks flood control project. The bond proposition was ratified at a general election. A lawsuit sought to block the bonds' issuance, in part on the ground that flood control and boat harbor projects were two distinct subjects in violation of the constitutional one-subject rule. After a trial the superior court rejected this constitutional challenge. On appeal we generally agreed that constitutional one-subject rules are primarily meant to restrain legislative logrolling. But we explained that such provisions should "be construed with considerable breadth" to avoid undue restrictions on the "scope and permissible subject matter" of legislation, which otherwise would result in "multiplying and complicating the number of necessary enactment[s] and their interrelationships." We stated that one-subject-rule decisions "must be made on a basis of practicality and reasonableness" and generally must be analyzed "only in terms of the particular facts of each case." We ultimately adopted the Minnesota Supreme Court's "germaneness" test:

All that is necessary is that [the] act should embrace some one general subject; and by this is meant, merely, that all matters ... should fall under some one general idea, be so connected with or related to each other, either logically or in popular understanding, as to be parts of, or germane to, one general subject.^{[ ]}

Turning to the case's facts, we declined to adopt the State's argument, accepted by the superior court, that with bond propositions the general subject is "the issuance of bonds for capital improvements"; we discerned a narrower single subject. We noted trial evidence showing that both projects would receive federal funding and be administered by the United States Army Corps of Engineers, that both projects were "part of a continuing program of federal-state cooperation in water resources development," and that preexisting law authorized the State to enter into both projects. We concluded that the two "topics pertain[ed] to one ongoing plan for the development of water resources and to the method of funding that plan" and that the plan logically and conveniently could encompass the flood control and boat harbor projects without violating the one-subject rule.

b. North Slope Borough v. Sohio Petroleum Corp.

Our next one-subject-rule case, North Slope Borough v. Sohio Petroleum Corp. , involved a legislative act that included provisions regarding income tax credits for fuel expenses, amendments to a state excise tax on cigarettes, and restrictions on municipalities' ability to tax. The superior court concluded that the act violated the one-subject rule because it included two separate subjects: municipal taxation and state taxation. The State appealed, arguing that Gellert 's one-subject rule was "very liberal," that the challenged subject of "taxation" was more limited than Gellert 's subject of water resources development, and that the legislative record showed no indication of logrolling, inadvertence, stealth, or fraud.

We reiterated that what constitutes one subject is to be liberally construed and that no act will be set aside for failure to comply with the one-subject rule "except where the violation is both substantial and plain." Because the state and municipal tax provisions were intertwined and had a direct impact on state tax revenue, we held that "[s]tate taxation is not an unduly broad category under the one[-]subject rule" and concluded that there was no constitutional violation. c. Short v. State

In a subsequent one-subject-rule case, Short v. State , we stated: "[I]t is apparent that the Gellert test requires no more than that the various provisions of [a] single legislative enactment fairly relate to the same subject, or have a natural connection therewith." That case involved another voter-approved legislative bond package, unsuccessfully challenged in superior court under the one-subject rule for combining public safety and correctional facilities capital projects. The public safety buildings were for "state troopers, fish and wildlife protection, a motor vehicles division, a fire prevention division" — essentially all public safety services "the Fairbanks and Soldotna regions" needed. The correctional facilities were "pre-trial detention facilities, juvenile offender institutions and new jail facilities in various regions of the state."

On appeal we concluded that the bond proposition projects fell within the single subject of "general public safety function of protecting life and property." We expressly rejected the challenger's argument that the one-subject rule requires a stricter standard when measures are submitted to voters, reasoning that Gellert also involved a bill ratified by voters and that the policies behind the one-subject rule are the same whether the law is before voters or the legislature:

The argument that measures submitted to the voters are deserving of enhanced protection under the one-subject rule of art[icle] II, [section] 13 of the state constitution has little merit. ... The one-subject rule is not restricted to those legislative acts which also must be approved by the voters; rather, it applies equally to all enactments of the legislature. Further, we have applied the provision to bond propositions in the past in the same way as it has been applied to other types of legislation. Thus, there does not appear to be any valid purpose to be served by adopting a more restrictive interpretation of article II, [section] 13 of the state constitution in cases where voters either initiate or ratify proposed legislation in their capacity as the larger legislative body of the state.^{[ ]}

d. State v. First National Bank of Anchorage

Questions about the Gellert test's feasibility were raised in State v. First National Bank of Anchorage . In that case the State sued certain defendants under provisions of the Uniform Land Sales Practices Act for misleading statements and omissions related to the sale of real property. One defendant claimed the Act violated the one-subject rule because of amendments enacted in 1977. The amendments (1) "br[ought] in-state sales of subdivided land within the Act's scope and ... add[ed] a general antifraud section" and (2) changed separate Alaska Land Act provisions "pertain[ing] to the leasing of state-owned lands and the Division of Land's zoning power." The superior court declined to decide the single-subject challenge, ruling on other grounds. We considered the challenge on appeal.

Citing North Slope Borough and Gellert , the State argued that the challenger had not met his substantial burden to prove a one-subject rule violation. The State emphasized the "liberal construction in favor of the validity" we previously applied to legislative acts challenged under the one-subject rule, arguing that even if the challenged amendments to the Act contained numerous provisions under the broad title of "land" they all were "obviously and logically connected." Countering the challenger's argument on appeal that if "land" can be a proper subject the one-subject rule has no meaning, the State observed that an act relating to "land" or "transportation" would not be logically connected if it were to "include provisions dealing with unemployment compensation, child labor laws, state funding of abortions, or state usury laws," thus making the one-subject rule meaningful.

We determined that the Act's challenged amendments satisfied the one-subject rule under the subject of "land," but we acknowledged:

Were we writing on a clean slate, we would be inclined to find this subject impermissibly broad. Permitting such breadth under the one-subject rule could conceivably be misconstrued as a sanction for legislation embracing "the whole body of the law." Nevertheless, while the issue is indeed close, we are unable to say that the legislature has transgressed the limits of article II, section 13 established by prior decisions of this court.^{[ ]}

e. Yute Air Alaska, Inc. v. McAlpine

Yute Air Alaska, Inc. v. McAlpine is our first case directly addressing the one-subject rule in the initiative context. The proposed initiative would "repeal statutes regulating motor and air carriers ..., open the carrier business to ... all financially responsible persons, prohibit municipalities from regulating these activities, and require the governor to seek repeal of the federal statute ... which requires the use of United States vessels for shipping goods between United States ports." The lieutenant governor's initiative certification was unsuccessfully challenged in superior court, in part on the one-subject rule, and allowed on the ballot. The challengers appealed, arguing that combining provisions about state and local transportation regulations with provisions about a federal maritime statute violated the one-subject rule. The State countered by agreeing with our Short dictum that the constitutional one-subject rule applies equally to legislation and initiatives, arguing that "the Alaska Constitution makes the law[-]making power equal between the legislature and the electorate." The State again emphasized our case law setting out a liberal construction of the one-subject rule and the requirement that alleged violations of the rule "must be substantial and plain before the court will declare a bill invalid." The State then argued that the proposed initiative focused on the single subject of alleviating the negative economic effects of unnecessary transportation regulation on commercial enterprise in Alaska, and particularly on abolishing anti-competitive practices. Arguing that the case's facts were more like Gellert than First National Bank of Anchorage , the State countered the challengers' argument by stating that "the logrolling of similar subjects is not a proscribed activity."

We relied on a common sense reading of the initiative, noting: "To the miner at Minto who wants to bring his supplies from Seattle, the interaction and interrelation [of Alaskan intrastate and federal interstate provisions] is more than just self-evident — it is glaringly so." We concluded, in light of previously approved legislation covering broad subject matter, that it was rational and convenient to combine the initiative's various provisions into one bill under the subject of "transportation." We further explained that the initiative's provisions all contained "a common thread narrower than ‘transportation’ ": the provisions related to eliminating "regulations and statutes thought to create needless transportation costs."

f. Evans ex rel. Kutch v. State

In Evans ex rel. Kutch v. State we briefly discussed the one-subject rule when upholding 1997 tort reform legislation. The legislation was codified into various sections of the Alaska Statutes and

included many new tort law provisions, including caps on noneconomic and punitive damages, a requirement that half of all punitive damages awards be paid into the state treasury, a ten-year "statute of repose," a modified tolling procedure for the statute of limitations as applied to minors, comparative allocation of fault between parties and non-parties, a revised offer of judgment procedure, and partial immunity for hospitals from vicarious liability for some physicians' actions.^{[ ]}

After noting that we, and the court of appeals, had upheld "legislation that was in some cases very broad," we concluded that the various tort reform provisions fell "within the single subject of ‘civil actions.’ "

g. Croft v. Parnell

We struck a proposed initiative for violating the one-subject rule in Croft v. Parnell . In that case the lieutenant governor denied certification of a "clean elections" initiative that would publicly fund state elections through a new oil production tax, with a non-binding directive to transfer excess tax funds to the Permanent Fund Dividend. The superior court upheld the lieutenant governor's certification denial, explaining that there was no connection between the type of revenue created and the type of program the initiative established.

On appeal the State argued that the one-subject rule extends to initiatives and that the purpose of the rule is to prevent logrolling. The State acknowledged our deferential standard for determining whether the rule has been violated, but it did not ask that we change that standard or that we adopt a stricter standard for initiatives. The State instead argued that "[e]ven under the most deferential of standards ... oil taxes and campaign finance are different subjects," that there was "no logical connection between oil taxation and campaign finance," and that logrolling was a large concern in light of the focus on oil taxes.

We noted the need to balance the one-subject rule's purpose against the need for legislative efficiency and indicated that "[o]ur solution" has been to apply the one-subject rule broadly, but we rejected the initiative sponsors' argument that the initiative's provisions could be related under the subject of "clean elections." We explained that in the initiative context the one-subject rule "protects the voters' ability to effectively exercise their right to vote by requiring that different proposals be voted on separately," "allows voters to express their will through their votes more precisely," and prevents logrolling, stealth, and fraud. We noted that the initiative "proposed the creation and ‘soft dedication’ of a new revenue source, and proposed

the creation of an entirely new government program." We concluded that, because the Alaska Constitution prohibits earmarking tax revenues, "a ‘soft dedication’ cannot be considered for purposes of a single-subject analysis and therefore cannot be used to make two independent provisions of an initiative address one subject." We ultimately held that combining the distinct proposals for an oil production tax and election campaign finance reform "r[an] afoul of the single-subject rule" by depriving voters of the opportunity to vote on each separate proposal.

III. THIS INITIATIVE

A. Facts And Proceedings

1. Initiative

Alaskans for Better Elections is a ballot initiative committee (Committee) seeking to place on a future ballot an initiative entitled:

An Act prohibiting the use of dark money by independent expenditure groups working to influence candidate elections in Alaska and requiring additional disclosures by these groups; establishing a nonpartisan and open top four primary election system for election to state executive and state and national legislative offices; changing appointment procedures relating to precinct watchers and members of precinct election boards, election district absentee and questioned ballot counting boards, and the Alaska Public Offices Commission; establishing a ranked-choice general election system; supporting an amendment to the United States Constitution to allow citizens to regulate money in Alaska elections; repealing the special runoff election for the office of United States Senator and United States Representative; requiring certain written notices to appear in election pamphlets and polling places; and amending the definition of "political party."

The initiative contains 74 sections, all but two of them amending current provisions of Title 15, Alaska's Election Code. The initiative most significantly changes Alaska's election laws by: (1) replacing Alaska's current party-based primary system with an open, nonpartisan primary; (2) establishing ranked-choice voting in general elections; and (3) adopting new disclosure and disclaimer requirements for independent expenditure groups and their donors.

2. Certification denial

In July 2019 the Committee submitted an initiative application to Lieutenant Governor Kevin Meyer. The Division of Elections designated the ballot initiative as "19AKBE." The lieutenant governor requested legal review by Attorney General Kevin Clarkson, who recommended certification denial. The attorney general concluded that the initiative was not in the proper form because it contained more than one subject in violation of the Alaska Constitution. Briefly, the attorney general's constitutional analysis focused on concerns we have expressed about the Gellert test's possible over-broadness and our rejection of the proposed initiative in Croft . The attorney general concluded that the initiative raised the concerns identified in Croft and therefore violated the constitutional one-subject rule. On August 30 the lieutenant governor denied certification based on the attorney general's opinion.

3. Superior court proceedings

Days later the Committee filed suit in superior court against the lieutenant governor and the Division of Elections (collectively, the State), seeking a declaration that the initiative is in proper form and does not violate the Alaska Constitution and requesting injunctive relief directing certification and distribution of initiative petition booklets. The parties agreed to an expedited briefing schedule and filed cross-motions for summary judgment. The superior court held oral arguments in mid-October.

The superior court ruled in the Committee's favor in late October, rejecting the State's contention that Croft provided "new guidance" on the one-subject test. The court instead quoted Gellert 's holding that Alaska's decades-old one-subject test is construed "with considerable breadth" and concluded that the test applies equally to legislation and initiatives. Explaining that the initiative's substantive provisions all related to election reform, the court quoted Yute Air and found "no indication that the [initiative's] provisions are targeted to different constituencies or that any of the provisions were calculated to obtain sufficient votes to pass the proposed initiative by attaching something of popularity ‘likely to carry along the enactment of whatever state law is attached for the ride.’ " The court ordered that the State immediately distribute petition booklets for the sponsors to collect signatures for placing the initiative on a future election ballot. The State moved for a stay pending appeal; the court denied the request. Final judgment was entered in early November.

4. This appeal

The State appealed, asking for "extremely expedited consideration of this appeal," arguing that because the Committee was gathering signatures for the initiative, any delay might cause us to feel "constrained by the sponsors' mounting reliance interests, which will increase every day." The State pointed to Yute Air 's language regarding our reluctance to invalidate an initiative because "the sponsors of the initiative ... relied on our precedents in preparing the present proposition and undertaking the considerable expense and time and effort needed to place it on the ballot."

We denied "extremely expedited" consideration but agreed to expedite the appeal, noting we would "give full and fair consideration to this appeal's legal merits, including [the State's] stated intent to ask us to reverse long-standing precedents." After expedited briefing, the parties argued the case to us on February 19, 2020.

IV. DISCUSSION

We turn now to the State's arguments, making clear their specific nature. At oral argument before us, the State conceded that, had the initiative bill been passed by the legislature, the bill would comply with the one-subject rule; that is a well-made concession, as we discuss further below. The State's arguments instead are that (1) in Croft we modified the Gellert test for an initiative's compliance with the one-subject rule — making the test more stringent for initiatives so voters have the ability to vote on separate topics — and, if not, (2) we should overrule our precedent holding that initiatives and legislation generally are on even footing for compliance with the one-subject rule and adopt a more stringent test for initiatives.

A. Croft Did Not Establish A Different Test For Initiatives.

The superior court ruled that Croft did not create a different one-subject-rule test for initiatives but rather was consistent with and followed our long-existing Gellert test precedent. That Croft did not overrule or modify previous one-subject-rule decisions to create a different test for initiatives is a simple and straightforward conclusion; in Croft we articulated tangible boundaries to help identify whether the initiative went too far beyond the parameters of "one subject." We first focused on the initiative's magnitude and means of achieving "clean elections." We reasoned that the proposed initiative created (1) a new oil-production tax and (2) a new and unrelated government program to fund state office election campaigns. We decided the two provisions were of such magnitude as to be distinct and unrelated subjects, and we additionally noted the strong appearance of logrolling different subjects to garner support for the initiative.

We did not completely rule out the possibility that the two distinct topics could be germane to each other, but we concluded that the connection was not obvious. We required a factual demonstration of a nexus, and we noted that the sponsors' submission of "two newspaper articles and a listing of the top groups lobbying the Alaska legislature" was "insufficient to demonstrate a clear or established connection between the oil industry and a need for public financing of state electoral campaigns." This inquiry was consistent with the Gellert test, requiring that "all matters treated" should be connected or related with each other "as to be parts of, or germane to, one general subject." Without evidence showing a link "between the oil industry and a need for public financing of state electoral campaigns," the initiative's distinct topics could not be joined under the subject of "clean elections."

We note that in Croft the State did not ask us to establish a different one-subject-rule test for initiatives, and we certainly did not say we were doing so. We did not differentiate between improper legislative and initiative logrolling. And we did not disturb our Yute Air ruling that "[a] one[-]subject rule for initiatives which is more restrictive than the rule for legislative action is not permitted." In short, we affirm the superior court's conclusion that Croft did not establish a new and different one-subject-rule test for initiatives.

B. We Decline To Overrule Yute Air .

The State alternatively argues that if the original Gellert test still applies to initiatives after Croft , we "should overrule [that precedent] and restore substance to the single-subject rule to protect voter choice in the initiative context." The State notes that some members of the court previously "expressed skepticism" because the Gellert test is "extremely lax," and the State specifically requests that we overrule our Yute Air holding that the Gellert test applies equally to legislation and initiatives. The State now argues — contrary to its argument to us in Yute Air — that the one-subject rule is only a procedural limitation to law-making, that the Alaska Constitution therefore allows a stricter standard for initiatives, and that a stricter one-subject standard for initiatives would allow voters to more precisely express their will. Finally, the State proposes a one-subject standard for initiatives that would "consider both how the parts of an initiative are inter-related and the overall significance of each reform."

The Committee counters that the State has not met its high burden of overcoming the presumption of stare decisis. The Committee argues: "For the people to have equal law[-]making power to the legislature, the single-subject rule must be equally applied." And the Committee contends that the State's proposed stricter one-subject standard for initiatives "is unworkable in practice, would hamstring the ability of Alaskans to enact laws by initiative, and would completely muddy the waters in an otherwise clear area of the law."

We agree with the Committee that imposing a stricter one-subject standard to initiatives than to legislation would run counter to the delegates' intent that the initiative serve as the people's check on the legislature. Under our system of checks and balances, when the legislature fails to pass laws the people believe are needed, the people have the initiative power to create those laws. And one driving force behind article XII, section 11 was identifying clearly the people's broad law-making power, preventing confusion about its extent.

The State nonetheless contends we should impose a stricter one-subject standard for initiatives that would (1) require interrelatedness between the various provisions and (2) "consider the overall significance of an initiative's provisions." The State notably made no suggestion in its briefing that the legislature should be subject to this stricter one-subject standard, and at oral argument before us the State conceded that it was not seeking a stricter one-subject standard that would affect the legislature. The State recognized at oral argument that applying a stricter one-subject standard would be impractical for the legislature's efficiency.

If a stricter one-subject standard would be impractical for the legislature, and if the constitutional initiative provision were intended to provide co-equal law-making power to the people (subject to stated exceptions), we cannot employ a separate and stricter one-subject standard for initiatives. The purposes behind the initiative right — to serve as a check on the legislature and to pass needed laws — are not served if we impose a stricter one-subject standard on the people but allow the legislature to operate under a more liberal rule. In effect, the State asks us to put our judicial thumb on the scale to limit the people's constitutional check against legislative inaction, limiting the people's law-making power to only piecemeal legislation.

With all this in mind, we turn to our framework for deciding whether to overrule precedent. When we are asked to overrule our precedent, the importance of stare decisis cannot be overstated: "[S]tare decisis is a practical, flexible command that balances our community's competing interests in the stability of legal norms and the need to adapt those norms to society's changing demands." A party seeking reversal bears the "heavy threshold burden of showing compelling reasons for reconsidering the prior ruling." "We will overrule a prior decision only when clearly convinced that the rule was originally erroneous or is no longer sound because of changed conditions and that more good than harm would result from a departure from precedent." We need address only the first prong of the test to reject the State's argument.

The State dismisses Short and Yute Air as "originally erroneous," speculating that in both cases we felt compelled to uphold the measures because voters already had approved them. As we noted earlier, Short 's statement about the one-subject rule was dictum; Yute Air is the controlling authority, and we focus our attention on that decision.

Yute Air descended from Gellert , in which Justice Fitzgerald dissented. The dissent believed the combination of Fairbanks flood control and coastal small boat harbor projects was a "good example" of improper logrolling because the bill sought to "gather voter support for a project in the interior of Alaska by linking it with harbor projects dear to the coastal towns and villages." The dissent argued that the majority's test would render the one-subject rule "meaningless," although admitting that it is "difficult to determine whether a group of projects has one subject matter."

We expressed concern about the feasibility of the one-subject rule in subsequent cases, and in Yute Air we directly addressed whether there was a more workable alternative. In that case Justice Moore dissented, stating that we "mistakenly continued to give the rule such an extremely liberal interpretation that the rule has become a farce." The dissent noted that the one-subject rule's purpose was "to restrain ‘log[ ]rolling’ in the legislative process" and that the lax one-subject rule "actually allows disparate subjects to be enfolded within the cloak of a broad generality." The dissent also believed that in the initiative context "[t]here is a greater danger of logrolling, or the deliberate intermingling of issues to increase the likelihood of an initiative's passage, and there is a greater opportunity for ‘inadvertence, stealth and fraud’ in the enactment-by-initiative process."

The dissent presented two proposals for replacing our current one-subject standard. In one footnote the dissent suggested applying a "functional test," adopted by the Florida Supreme Court, when considering whether an initiative violates the one-subject rule. Courts applying the "functional test" consider whether the proposed initiative was merely an "aggregation of dissimilar provisions (designed) to attract support of diverse groups to assume its passage" and whether the "initiative performs the functions of different branches of government." The dissent also offered a "more useful" Gellert test: "A stronger and clearer version ... would read as follows: An act or initiative should embrace one subject. By this we mean that all matters treated should be logically connected." The dissent notably suggested that this new test apply prospectively to both the legislature and the initiative. But the dissent acknowledged that this new wording of the Gellert test would "not automatically turn this court away from the Anything Goes approach of the ‘merely ... germane’ standard embraced in Gellert ."

We responded to the dissent, acknowledging the argument that we had given the legislature too much deference in Gellert 's application of the one-subject rule. But we explained that we had not found a more workable stricter standard and that the dissent's proposed Florida rule would make enacting uniform codes prohibitively difficult. We emphasized the stability and predictability of our prior decisions, noting that both the initiative sponsors and the lieutenant governor had relied on Gellert . We further reiterated that "an initiative is an act of direct democracy guaranteed by our constitution" and "should be liberally construed." We believed stricter application of the one-subject rule to the initiative "ignore[d] the explicit constitutional directive" of article XII, section 11. We therefore concluded we would continue applying the Gellert test equally to the legislature and the initiative.

We reaffirm our Yute Air conclusion. Contrary to the State's speculation that we were overly swayed by the initiative's passing at the previous general election held just before our decision was issued, our reasoning focused solely on whether the initiative's provisions satisfied our Gellert one-subject standard. Although the State also now argues that the one-subject rule is "procedural" rather than "substantive" and that the Alaska Constitution permits the State's alternative, stricter test, that argument — even assuming its validity — demonstrates nothing erroneous in our decision to apply the one-subject test equally to initiatives and legislation. When the State originally argued Yute Air before us, it argued for the same conclusion; we clearly explained in our decision why we rejected the dissent's approach. And our subsequent Croft decision, striking down a proposed initiative after application of the Gellert test, demonstrates that, although imperfect, our one-subject-rule test is workable in practice.

We see no basis to overrule Yute Air and create a different one-subject rule for the people's right to make laws through the initiative process. The approach we announced in Yute Air is consistent with the Alaska Constitution, and the State's proposed approach runs counter to the people's constitutional law-making right. The State has not met its burden to show that our Yute Air holding was originally erroneous.

C. The Initiative Passes The Gellert Test.

We turn now to the Committee's proposed initiative. The proposed initiative includes a "findings and intent" section setting out the goal of "increasing transparency, participation, access, and choice" in the electoral process. The initiative aspires to ensure that wealth does not unduly influence state elections and that Alaskans "know in a timely manner the source, quantity, timing, and nature of resources used to influence candidate elections in Alaska." The section further provides that "[i]t is in the public interest" to adopt an open, nonpartisan primary system and a general election ranked-choice voting system. The extremely detailed initiative title similarly explains that the act would, among other things, "prohibit[ ] the use of dark money by independent expenditure groups working to influence candidate elections in Alaska and requir[e] additional disclosures by these groups."

The parties agree that the initiative would make three substantive changes to Alaska election law: (1) replacing the party primary system with an open, nonpartisan primary; (2) establishing ranked-choice voting in the general election; and (3) mandating new disclosure and disclaimer requirements to existing campaign finance laws. A plain reading of the initiative shows that its provisions embrace the single subject of "election reform" and share the nexus of election administration. All substantive provisions fall under the same subject matter of elections, seek to institute an election reform process, and, as the superior court noted, change a single statutory title, Title 15, Alaska's Elections Code.

Although the initiative's inclusion of terms such as "dark money" and "true source" arguably could evoke "strong feelings" in voters, those feelings relate to the election reform topic; the initiative's text shows no transparent attempt to garner voter support through completely unrelated provisions. And nothing suggests the title is misleading or the initiative is so unclear that it gives rise to a concern about confusion, fraud, or inadvertence. The initiative's provisions substantively modify current election laws such that we can logically conclude they fall under the one subject of "election reform."

The State argues that the three substantive changes to the election laws "are not actually ‘connected’ through cross-references or other logical reliance" and that "[n]one depends on the others to function properly." The State's argument focuses primarily on whether voters could vote separately on each substantive legal reform. But neither Gellert nor Croft requires severing every provision that, in the abstract, could be voted on separately. And we have never required that a proposed law's "subject" be the most minute and discrete possible. We recognized instead in Gellert the convenience of classifying related matters under a single bill. In Short we further emphasized the need to give "great latitude in enacting comprehensive legislation" and cautioned that "[t]he one-subject provision should not be construed so as to unnecessarily restrict the scope and operation of laws, or to multiply their number excessively ." And in Croft we rejected the initiative not simply because its provisions were severable, but because they were distinct subjects and lacked a nexus. Thus, the question is not whether the initiative could be split into separate measures, but rather whether the various provisions "embrace some one general subject."

The initiative's provisions are logically related. The substantive changes relate to elections and are encompassed within Title 15. The open, nonpartisan primary system changes the status quo by forwarding four candidates for voters to rank in the general election by ranked-choice voting. These two substantive changes are interrelated because they together ensure that voting does not revert to a two-candidate system. The Committee also argues that "when moving away from party primary elections and allowing for more candidates on the general election ballot, it becomes more important than ever that voters have adequate and accurate information about who is paying for campaign communications to influence their vote." A provision increasing voter knowledge logically relates to election reform.

Unlike the Croft sponsors' juxtaposing oil industry taxation, campaign finance, and Permanent Fund Dividend payments into one "clean elections" initiative, this initiative's provisions are properly classified under "election reform" as a matter of both logic and common sense. They all relate to the elections process and share the common thread of reforming current election laws. We can logically conclude that the various initiative provisions substantively change (or reform) the state's elections.

The Gellert test, despite its imperfections, has guided our one-subject-rule precedent for over 40 years. The subject "election reform" is more constricted than subjects we previously have upheld. The initiative satisfies the Gellert test; it now is up to the people to decide whether the initiative's provisions should become law.

V. CONCLUSION

We AFFIRM the superior court's ruling and judgment reversing the lieutenant governor's certification denial of the Committee's proposed initiative.