*316DISSENT
STRAS, Justice(dissenting).
In an alleged effort to improve policing and police accountability within the City of Minneapolis, a group of citizens sought to include a proposed amendment to Minneapolis’s City Charter on the general-election ballot. The amendment would have required police officers to, among other things, procure professional-liability insurance. Even though the group satisfied all of the procedural requirements for placement of the question on the general-election ballot, the Minneapolis City Council struck the question based on an opinion by the City Attorney that the proposed amendment would, if passed, be “preempted by state law and conflict[] with state public policy,” The citizen group challenged the City Council’s decision in Hen-nepin County District Court, seeking a ruling requiring the question to be placed on the ballot. The district court dismissed the petition, concluding that the proposed amendment was preempted by state statute. In my view, the district court properly dismissed the citizen group’s petition, but it did so for the wrong reason. Instead of addressing the petition on the merits, as the district court did, I would conclude that this dispute is nonjusticiable.1
The citizen group’s petition called for nothing more than a classic advisory opinion because, prior to the election, the city-charter amendment was not the law. It was the citizen-initiative equivalent of a bill winding its way through the legislative process. It had the potential to become law, to be sure, but until it garnered the necessary votes, it was nothing more than an idea. We do not give advice to the Legislature when a bill is being considered for passage—an uncontroversial proposition that traces back to the nation’s founding—nor can we give advice to a citizen group or a city council on whether a proposed law will violate the Minnesota Constitution or a state statute. See Hayburn’s Case, 2 U.S. (2 Dall.) 409, 410, 1 L.Ed. 436 (1792). This is a task that we may undertake only after a proposal actually becomes law and once we are presented with a case or controversy involving adverse parties. Just as we are not a junior-varsity legislature, neither are we the legal counsel or the research division of the Minneapolis City Council. In re Guardianship of Tschumy, 853 N.W.2d 728, 756 (Minn. 2014) (Stras, J., dissenting).
The prohibition on the issuance of advisory opinions traces back to 1792, when the Supreme Court of the United States decided Haybum’s Case. In Haybum’s Case, a circuit court, and later the Supreme Court, were asked to decide whether William Hayburn, a war veteran, was entitled to disabled-veteran benefits. 2 U.S. (2 Dall.) at 409. The federal statute under which the parties proceeded contained what we would now view as an odd provision: it required a court to decide whether, in its view, a particular claimant’s family should receive benefits, but the decision was only advisory to the Secretary of War, who would make the final determination of eligibility. Id. at 410. One circuit-court decision, which was discussed extensively by the Supreme Court, said that “the business assigned to this court, by the act, is not judicial, nor directed to be performed judicially,” Id. Citing another circuit-court decision, Hayburn’s Case fur*317ther categorically stated that “the business directed by this act [was] not of a judicial nature.” Id. at 411. Although the Court never actually decided that the statute was unconstitutional, five of the six justices, each of whom considered the question while riding circuit, had concluded that the non-binding, advisory nature of the decision was inconsistent with the exercise of judicial power. See id. at 410-13; see also North Carolina v. Rice, 404 U.S. 244, 246, 92 S.Ct. 402, 30 L.Ed.2d 413 (1971) (“Early in its history, this Court held that it had no power to issue advisory opinions.” (citing Hayburn’s Case, 2 U.S. (2 Dall.) 409, 1 L.Ed. 436 (1792))).
Our advisory-opinion doctrine, built from cases like Hayburn’s Case, requires us to decline to answer legal questions until we are presented with a concrete case or controversy. See In re Application of the Senate, 10 Minn. 78, 81, 10 Gil. 56, 57-58 (1865); see also Lee v. Delmont, 228 Minn. 101, 110, 36 N.W.2d 530, 537 (1949) (“Issues which have no existence other than in the realm of future possibility are purely hypothetical and are not justicia-ble.”). To do otherwise would exceed the scope of “judicial power” in Article VI, Section 1, of the Minnesota Constitution by making us “advisers of the legislature, nothing more.” In re Application of the Senate, 10 Minn. at 81, 10 Gil. at 58. We have consistently adhered to the lesson of Haybum’s Case, including discussing it in our very first decision on the advisory-opinion doctrine, In re Application of the Senate, which we decided in 1865, less than 7 years after Minnesota became a state. 10 Minn. at 81, 10 Gil. at 57-58; see also State ex rel Young v. Brill, 100 Minn. 499, 521, 111 N.W. 639, 648 (1907) (“[A] statute which authorized either branch of the legislature to call for the opinion of the supreme court, or any one of the judges thereof,, was held unconstitutional.”).
The court relies on a single statute, Minn. Stat. § 204B.44(a)(1) (2016), to conclude that the district court had jurisdiction over the citizen group’s petition. The court’s analysis is, at best, partially complete. It is true that section 204B.44(a) provides statutory authority for the district court’s actions, but subject-matter jurisdiction requires more. See Zweber v. Credit River Twp., 882 N.W.2d 605, 608 (Minn. 2016). What is missing here is the constitutional authority to decide the case, and in particular, the presence of a concrete case or controversy. See Schowalter v. State, 822 N.W.2d 292, 298 (Minn. 2012). When the citizen group filed its petition in August with the district court, its request presented nothing more than a hypothetical question about the possible legal difficulties facing a city-charter amendment that had not yet passed. The petition therefore failed the first requirement for a justiciable controversy: it did not present “ ‘definite and concrete assertions of right that emanate[d] from a legal source.’” Schowalter, 822 N.W.2d at 298-99 (emphasis added) (quoting Onvoy, Inc. v. ALLETE, Inc., 736 N.W.2d 611, 617 (Minn. 2007)). As we noted in Schowalter, it makes no difference whether a statute gives us the authority to decide a dispute if constitutional authority is absent. Id. at 298 (“By virtue of this statute, we have original jurisdiction in this matter. But we do not issue advisory opinions and we do not ‘decide cases merely to establish precedent.’” (emphasis added) (citations omitted)).
It is true that some of our case law beginning in the 1930s purports to allow us to determine the legality of a ballot initiative in advance, but these cases cannot be squared with our longstanding jurisprudence prohibiting courts from issuing advisory opinions. This leaves us with two possible choices: either adhere to the prohibition on advisory opinions that traces *318back to our nation’s (and state’s) founding or follow a line of more contemporary cases that appears to give courts the authority to decide hypothetical ballot questions. I choose the former path because it is the only one that is consistent with the Minnesota Constitution.
Even so, it is not clear, despite the court’s assertion to the contrary, that our cases really require us to choose between these two seemingly distinct paths. Our preenactment-review cases begin with Winget v. Holm, 187 Minn. 78, 244 N.W. 831 (1982). In Winget, a citizen sought to restrain the Secretary of State from placing a proposed constitutional amendment on the general-election ballot. Id. at 81, 244 N.W. at 332. We held that a court should have the authority to “interpose to save the trouble and expense of submitting a proposed constitutional amendment to a vote, if it be not proposed in the form demanded by the constitution, so that, though approved by the electors, the courts would be compelled to declare it no part of the constitution.” Id. at 81, 244 N.W. at 332 (emphasis added). Importantly, Winget dealt only with a procedural defect in the proposed constitutional amendment, not its underlying validity. The objection to the proposed constitutional amendment was its lack of compliance with then-Article XIV, Section 1, of the Minnesota Constitution, which required that, when two or more amendments were submitted to the voters at the same time, the voters be allowed to vote on the amendments separately. This procedural defect, which was distinct from the amendment’s underlying validity, was the only challenge that we considered in Winget.2 See also State ex rel. Andrews v. Beach, 155 Minn. 33, 35, 191 N.W. 1012, 1013 (1923), (“We do not hold that an amendment to a charter must be submitted even though it is manifestly unconstitutional. That question is not now before us and consideration is not necessary to the determination of this appeal.”); McConaughy v. Sec’y of State, 106 Minn. 392, 401, 119 N.W. 408, 411 (1909) (listing examples of courts exercising authority in such cases, all of which involved procedural defects).
We have therefore held, quite correctly in my view, that courts have the constitutional authority to strike questions from the ballot when the party sponsoring the initiative has failed to follow the procedural requirements for placing it on the ballot. E.g., Winget, 187 Minn. at 81, 244 N.W. at 332. This line of cases makes sense because evaluating whether a party has followed the procedures for placing a question on the ballot does not require a court to prejudge the question’s legality. There is less authority, however, for the remarkable proposition adopted by the court in this case, which is that a court can prejudge the legality of a proposed initiative that may never become law.
In support of its license to strike the proposed city-charter amendment in this *319ease, the court relies on several post-Wm-get cases, but each of these cases requires the satisfaction of a condition that is conspicuously absent from this case: manifest and obvious unconstitutionality. See Minneapolis Term Limits Coal. v. Keefe, 535 N.W.2d 306, 308 (Minn. 1995); Hous. & Redevelopment Auth. v. City of Minneapolis, 293 Minn. 227, 234, 198 N.W.2d 531, 536 (1972). The court nowhere holds— nor can it given the nature of the challenge in this ease—that the proposed amendment is manifestly unconstitutional. In fact, adjudicating the challenge in this case is particularly troubling precisely because it takes us a step beyond deferentially reviewing proposed legislation for manifest and obvious constitutional invalidity. It requires us to interpret a state statute—one that we have never considered before— and give it a definitive interpretation in light of a proposed city-charter amendment that does not, and may never, carry the force of law. Such review manages to at the same time undermine the limits on judicial power in the Minnesota Constitution and place the judiciary in the unfamiliar role of gatekeeper of local citizen initiatives and referenda. See Minn. Const. art. XII, § 5.3
The court nevertheless insists that its conclusion is the only one supported by stare decisis, and that I advocate overruling “our longstanding precedent” that has “stood the test of time over almost 100 years.” Nothing could be further from the truth. First, the court makes only a halfhearted attempt to explain why the action it takes today does not require an advisory opinion, which is a departure from even older and more longstanding precedent. Second, even if I were to accept that Homing & Redevelopment Authority and Keefe permit substantive preenactment review despite the advisory-opinion doctrine (which neither opinion addresses), those two cases still do not provide authority for the court’s opinion today. Rather, these cases only permit courts to preemptively strike questions from the ballot after finding manifest unconstitutionality, a conclusion that is nowhere to be found in today’s majority opinion.4 Given that the court to*320day ignores the limits on substantive pre-enactment review from Housing & Redevelopment Authority and Keefe, the court’s argument is really a form of faux stare decisis: it looks real at first, but when you peel back the layers, the only reasonable conclusion is that it is not.
The fact that the court’s rule in this case is inconsistent with the Minnesota Constitution and a close reading of our cases is alone sufficient to cast doubt on the court’s decision. But in addition, most jurisdictions that allow preenactment review limit it to procedural defects and have rejected invitations to expand it to encompass other types of challenges. See, e.g., Winkle v. City of Tucson, 190 Ariz. 413, 949 P.2d 502, 505 (1997) (“‘[I]t is only in cases where the initiative petition is defective in form ... where the procedure prescribed has not been followed that the court has authority to intervene and enjoin its enactment.’ ” (quoting Williams v. Parrack, 83 Ariz. 227, 319 P.2d 989, 991 (1957))); Stewart v. Advanced Gaming Techs., Inc., 272 Neb. 471, 723 N.W.2d 65, 76 (2006) (“We have thus made a distinction between substantive constitutional challenges to an initiative which do not become justiciable until the proposal is approved by voters and procedural challenges to the legal sufficiency of an initiative petition which may be determined prior to an election.” (emphasis added)); Herbst Gaming, Inc. v. Heller, 122 Nev. 877, 141 P.3d 1224, 1228-31 (2006) (permitting preelection challenges “based on asserted procedural defects,” but refusing to consider challenges “that the measure, if enacted, would violate substantive federal or state constitutional provisions”); Foster v. Clark, 309 Or. 464, 790 P.2d 1, 5 (1990) (stating that courts have jurisdiction to consider whether a proposed initiative is procedurally authorized to be placed on the ballot, but not to consider “general questions of constitutionality, such as whether the proposed measure, if enacted, would violate some completely different portion of the constitution”); Coppernoll v. Reed, 155 Wash.2d 290, 119 P.3d 318, 321-22 (2005) (permitting challenges to “a ballot measure’s noncompliance with procedural requirements,” but not those objecting to its substantive validity). And even those jurisdictions that allow substantive preenactment review limit the inquiry to whether a proposed law is manifestly or clearly unconstitutional, a far different inquiry than the court engages in here. See, e.g., Hessey v. Burden, 615 A.2d 562, 574 (D.C. 1992) (“We agree with the majority of courts which hold that [preenaetment] review is imprudent. But there may be extreme cases in which it would be both appropriate and efficient to decide the constitutionality of a proposed initiative.”); In re Initiative Petition No. 360, 879 P.2d 810, 814 (Okla. 1994) (“[W]e have limited such pre-election review to clear or manifest facial constitutional infirmities.... ”).
Our premature consideration of this case also artificially circumscribes our own review. Ordinarily, courts have the authority to sever unconstitutional portions of a statute or ordinance and allow the remaining, constitutional provisions to continue *321in effect. Minn. Stat. § 645.20 (2016). This is particularly true in cases involving preemption. See, e.g., Schwann v. FedEx Ground Package Sys., Inc., 813 F.3d 429, 432 (1st Cir. 2016); Cellco P’ship v. Hatch, 431 F.3d 1077, 1083 (8th Cir. 2005). Yet by striking the city-charter amendment before it is enacted, we have foreclosed the possibility of severance. It would be improper and impracticable to sever portions of a proposed city-charter amendment before it even becomes law. The unavailability of severance, a traditional remedy, provides additional evidence that preenactment review is inconsistent with the notion of judicial power.
The court repeatedly justifies its approach based on its concerns about futility, asserting that it is our job to prevent voters from undertaking a “futile” election on a ballot initiative that would be invalid if passed. But that, in fact, is not our job. It is equally futile to ask legislators to vote on proposed bills that are likely preempted by federal law or unconstitutional, and I do not understand the court to be suggesting that we have the power to give advice to the Legislature before it votes on proposed legislation. This case is no different.
No matter the policy considerations involved, we simply cannot exercise judicial authority that we do not have. Nowhere does the court cast doubt on the central claim that we are not empowered to prejudge the validity of proposals before they become law. Haybum’s Case and In re Application of the Senate remind us that constitutional limits are not a matter of convenience or preventing futility, but rather are fundamental constraints on our authority that preserve the separation of powers, a principle itself enshrined in the Minnesota Constitution. See Minn. Const. art. Ill, § 1. Because I would conclude that the question posed by this case is nonjusticiable, I would vacate the district court’s order and dismiss the petition.
I respectfully dissent.
. Although the City Council's decision remains in place under my approach, which appears at first glance to be the same as the result reached by the court, the court would affirm the district court’s decision, whereas I would vacate the district court’s decision. The differing dispositions mean that my separate writing is a dissent rather than a concurrence.
. The court says that nothing in Winget suggested that our statement about the "form” of a proposed amendment meant that only procedural, rather than substantive, review can occur prior to enactment. For support, the court seizes on the fact that we also mentioned the "object and purpose of [the amendment]” in Winget. However, the court omits perhaps the most important fact about Win-get: when we referred to the "object and purpose of [the amendment],” it was only to determine whether the amendment complied with the procedural requirement of Article XIV, Section 1, to have the voters consider separate amendments independently, not to determine the amendment’s underlying substantive validity. Winget, 187 Minn. at 85-86, 244 N.W. at 334. In fact, Winget involved only a procedural defect, bolstering the conclusion that Winget was all about ensuring that the amendment satisfied the procedural requirements for placement on the ballot, not about allowing courts to serve as the preenactment gatekeepers for constitutional amendments.
. In addition, the court's robust judicial review of citizen initiatives will undermine democracy by having a chilling effect on such measures, because its arguments apply to all initiatives, not just those proposed under Article XII, Section 5. The court, in effect, makes us a super city council, requiring us to decide which initiatives should be placed on the ballot. This is a determination that should be made by the legislative branches: either the Legislature itself, or a city council operating in a legislative capacity.
. The court suggests that my position "cannot be squared with” Schowalter and Minneapolis Federation of Men Teachers, Local 238 v. Board of Education, 238 Minn. 154, 56 N.W.2d 203 (1952). The court is wrong. One misconception permeating the court's analysis is its near-myopic focus on the presence of adversity in this case, which suggests that it believes that adversity—just one of the three requirements of the advisory-opinion doctrine—is sufficient for a justiciable controversy. See, e.g., Schowalter, 822 N.W.2d at 299 (focusing on the adversity requirement of the advisory-opinion doctrine). Even the one case that marginally supports the court's position by discussing one of the other advisory-opinion-doctrine requirements, Minneapolis Federation of Men Teachers, involved a situation in which the hypothetical nature of the controversy was a product of judicial intervention—• specifically, the grant of an injunction—not a case like this one involving a dispute that was inherently hypothetical from the very beginning. 238 Minn. at 157, 56 N.W.2d at 206. Our cases, again dating back to In re Application of the Senate, make clear that a case must satisfy all three requirements, not just one, to be justiciable.
An example illustrates the court’s error. Suppose that two parties took adverse positions under the Senate resolution at issue in In re Application of the Senate, arguing that a proposed bill that would impose land-use regulations violates the Minnesota Constitution. On one side of the dispute is a group of *320homeowners and on the other side is a municipality. The mere presence of adverse parties would no more permit us to decide such a hypothetical controversy than the case before us now. The reason is not the absence of adversity, which I concede is present here, but rather the fact that it would require us to adjudicate a controversy without a "definite and concrete assertion[ ] of [a] right that emanate[s] from a legal source” and one that is incapable of "resolution by judgment rather than presenting hypothetical facts that would form an advisory opinion.” Onvoy, Inc., 736 N.W.2d at 617-18 (emphasis added). The advisory-opinion doctrine is not, and never has been, a balancing test. Rather, it is a fundamental constraint on our exercise of judicial power under the Minnesota Constitution.