742 So.2d 1276 (1999)
Donald G. RAY, Louis P. Kalivoda, Sybil C. Mobley, David W. Bowers, and Clarence Fort, Appellants,
v.
Sandra MORTHAM,[1] Florida Secretary of State in her capacity as Florida's Chief Elections Officer, Appellee.
No. 94,653.
Supreme Court of Florida.
September 2, 1999.
Robert J. Boyd and Laura Boyd Pearce of MacFarlane, Ferguson & McMullen, Tallahassee, Florida, for Appellants.
Robert A. Butterworth, Attorney General, and George Waas, Assistant Attorney General, Tallahassee, Florida, for Appellee.
Talbot D'Alemberte, Tallahassee, Florida, for Alan C. Sundberg, Amicus Curiae.
Frank A. Shepherd, Miami, Florida, and John H. Findley, Sacramento, California, for Pacific Legal Foundation, Amicus Curiae.
PER CURIAM.
We have on appeal the trial court's order granting final summary judgment in favor of appellee, denying appellants' complaint for injunctive and declaratory relief that sought to prevent enforcement of article VI, section 4(b) of the Florida Constitution, which limits the number of consecutive terms for which certain candidates for state offices can appear on the ballot. Appellants, who are registered voters residing in the state senate districts of Senators Thomas, Kirkpatrick and Hargrett, appealed to the First District Court of Appeal, which invoked this Court's pass-through jurisdiction for issues of great public importance requiring immediate resolution by this Court. We have jurisdiction. See art. V, § 3(b)(5), Fla. Const.
FACTS
In 1992, the citizens of this State amended the Florida Constitution to include article VI, section 4(b). This amendment limits the number of consecutive terms of office for state legislators, federal legislators, the Lieutenant Governor and members of the Florida cabinet. It provides, in pertinent part, that no person may appear on the ballot for reelection to any of these offices "if, by the end of the current term of office, the person will have served (or, but for resignation, would have served) in that office for eight consecutive years."[2]
This amendment was placed on the ballot via a citizens' initiative petition, pursuant to article XI, section 3, Florida Constitution, which reserves to the people of this State "[t]he power to propose the revision or amendment of any portion or portions of th[e] constitution by initiative ... provided that, any such revision or amendment... shall embrace but one subject and matter directly connected therewith." The citizens' initiative petition explained the purpose and goals of the initiative petition:
The people of Florida believe that politicians who remain in office too long may become preoccupied with re-election and become beholden to special interests and bureaucrats, and that present limitations on the President of the United States and Governor of Florida show that term limitations can increase voter participation, citizen involvement in government, and the number of persons who will run for elective office.
Advisory Opinion to the Attorney General Limited Political Terms in Certain Elective Offices, 592 So.2d 225, 226 (Fla. 1991) (hereinafter Limited Political Terms) (quoting citizens' petition). To meet these goals, the petition requested that the Florida Constitution be amended "to the extent permitted by the Constitution of the United States." Id. The initiative petition contained a severability clause providing that:
If any portion of this measure is held invalid for any reason, the remaining portion of this measure, to the fullest extent possible, shall be severed from the void portion and given the fullest possible force and application. The people of Florida declare their intention that persons elected to offices of public trust will continue voluntarily to observe the wishes of the people as stated in this initiative in the event any provision of this initiative is held invalid.
Id.
The constitution and laws of this state require that every citizens' initiative petition be submitted to this Court for an opinion on its compliance with article XI, section 3 and section 101.161, Florida Statutes (1997). See art. IV, § 10, Fla. Const.;[3] § 16.061(1), Fla. Stat. (1997).[4] Article XI, section 3, dictates that any amendment placed on the ballot via citizen's initiative petition "shall embrace but one subject and matter directly connected therewith." Section 101.161(1) requires that when an amendment is submitted to the voters, the substance of the amendment must appear on the ballot in "clear and unambiguous language," not exceeding 75 words, explaining the "chief purpose" of the measure. The measure must also include a ballot title not exceeding 15 words. See id.
Regarding the single-subject requirement, we explained in Limited Political Terms that a "proposed amendment meets the single-subject requirement if it has `a logical and natural oneness of purpose.'" 592 So.2d at 227 (quoting Fine v. Firestone, 448 So.2d 984, 990 (Fla.1984)). Stated differently, we explained that a proposed amendment is valid if it may be "logically viewed as having a natural relation and connection as component parts or aspects of [this] single dominant plan or scheme." Id. As we explained in Fine, "the single-subject restraint on constitutional change by initiative proposals is intended to direct the electorate's attention to one change which may affect only one subject and matters directly connected therewith." 448 So.2d at 989.
The majority of the Court held in Limited Political Terms that the proposed amendment complied with article XI, section 3, concluding that the amendment addressed the "sole subject" of "limiting the number of consecutive terms that certain elected public officers may serve." 592 So.2d at 227. In so holding, we adhered to our requirement in Fine of "strict compliance with the single-subject rule in the initiative process for constitutional change." 448 So.2d at 989.
In Limited Political Terms, we also reviewed the ballot title and summary for the amendment to assure compliance with the statutory ballot summary requirements. The ballot summary stated as follows:
LIMITED POLITICAL TERMS IN CERTAIN ELECTIVE OFFICES
Limits terms by prohibiting incumbents who have held the same elective office for the preceding eight years from appearing on the ballot for re-election to that office. Offices covered are: Florida Representative and Senator, Lieutenant Governor, Florida Cabinet, and U.S. Senator and Representative. Terms of office beginning before approval are not counted.
Limited Political Terms, 592 So.2d at 228 (quoting ballot title and summary). The Court found that this ballot title and summary satisfied the requirement that it be "fair and advise[d] the voter sufficiently to enable him [or her] intelligently to cast his [or her] ballot," as we have interpreted section 101.161 to require. Id. at 228 (quoting Askew v. Firestone, 421 So.2d 151, 155 (Fla.1982)). Accordingly, the amendment was placed on the ballot.
The amendment was approved by the electorate in the 1992 elections.[5] Three years later, in 1995, the United States Supreme Court issued its decision in U.S. Term Limits, Inc. v. Thornton, 514 U.S. 779, 115 S.Ct. 1842, 131 L.Ed.2d 881 (1995), holding that state attempts to restrict candidacy for membership in either house of the United States Congress violate the Qualifications Clause of the Tenth Amendment to the United States Constitution and are therefore unconstitutional.
Thereafter, appellants brought this complaint for declaratory and injunctive relief requesting that the trial court strike the entire amendment.[6] Both parties filed motions for summary judgment asserting that there was no material issue of fact. The trial court rejected the attack on the amendment and granted appellee's motion for summary judgment. On appeal, appellants argue that the trial court was in error and that the entire amendment must be stricken because: (1) the unconstitutional portions relating to federal legislators cannot be severed from the remainder of the amendment; and (2) the amendment violates the First and Fourteenth Amendments to the United States Constitution. The amici curiae[7] arguing in support of appellants also assert that we should revisit our earlier decision in Limited Political Terms and strike the amendment because it actually encompasses more than one subject.
SEVERABILITY
We begin our analysis with two uncontroverted facts. First, there is no question but that, based on the United States Supreme Court's opinion in Thornton, section 4(b)(5) and (6) of article VI, placing limits on the terms of the U.S. Representatives and U.S. Senators, are unenforceable as violative of the United States Constitution's Qualifications Clause. See Advisory Opinion to the Attorney General re Term Limits Pledge, 718 So.2d 798, 801 n. 1. (Fla.1998) (hereinafter Term Limits Pledge); Thornton, 514 U.S. at 837-38, 115 S.Ct. 1842. Second, there is no question but that nearly 77% of those voting on the amendment approved it. See supra note 5.
Whether the doctrine of severability applies to constitutional provisions is a question of first impression in this state, although the doctrine has been applied by this Court to legislative enactments. See, e.g., Smith v. Department of Ins., 507 So.2d 1080, 1089 (Fla.1987). Severability is a judicial doctrine recognizing the obligation of the judiciary to uphold the constitutionality of legislative enactments where it is possible to strike only the unconstitutional portions. See State v. Calhoun County, 126 Fla. 376, 383, 170 So. 883, 886 (1936). This doctrine is derived from the respect of the judiciary for the separation of powers, and is "designed to show great deference to the legislative prerogative to enact laws." Schmitt v. State, 590 So.2d 404, 415 (Fla.1991).
The severability analysis answers the question of whether "the taint of an illegal provision has infected the entire enactment, requiring the whole unit to fail." Schmitt, 590 So.2d at 414. Stated simply: "The severability of a statutory provision is determined by its relation to the overall legislative intent of the statute of which it is a part, and whether the statute, less the invalid provisions, can still accomplish this intent." Martinez v. Scanlan, 582 So.2d 1167, 1173 (Fla.1991) (quoting Eastern Air Lines, Inc. v. Department of Revenue, 455 So.2d 311, 317 (Fla.1984)).
Appellants urge that the severability of invalid portions of constitutional amendments should be analyzed under the test applied in Smith v. Department of Insurance to a legislative enactment. That test provides that:
When a part of a statute is declared unconstitutional the remainder of the act will be permitted to stand provided: (1) the unconstitutional provisions can be separated from the remaining valid provisions, (2) the legislative purpose expressed in the valid provisions can be accomplished independently of those which are void, (3) the good and the bad features are not so inseparable in substance that it can be said that the Legislature would have passed the one without the other and, (4) an act complete in itself remains after the invalid provisions are stricken.
507 So.2d at 1089 (quoting Cramp v. Board of Pub. Instruction, 137 So.2d 828, 830 (Fla.1962)). Appellee counters that it is unnecessary to perform a severability analysis and that the unconstitutional provisions of the amendment simply should not be enforced.
Because certain portions of the term limits amendment are undoubtedly void under the United States Constitution, see Thornton, 514 U.S. at 837-38, 115 S.Ct. 1842, we find it necessary to address the effect of the void provisions on the remainder of the amendment. However, we are also mindful that the initiative power of fully informed citizens to amend the Constitution must be respected as an important aspect of the democratic process. Therefore, just as we view the severability of laws with deference to the legislative prerogative to enact the law, we conclude that we must afford no less deference to constitutional amendments initiated by our citizens and uphold the amendment if, after striking the invalid provisions, the purpose of the amendment can still be accomplished.
Thus, we agree with appellants that a severability analysis is required, but disagree that the amendment at issue cannot survive such an analysis. Appellants maintain that they need only cast doubt on whether the amendment would have passed without the provisions relating to federal legislators and that unless appellee can "prove" that the voters would have adopted the amendment had it not contained provisions limiting the terms of U.S. Representatives and U.S. Senators the amendment must be stricken. We conclude that this would be an inappropriate burden to place on appellee, and one that we have not imposed, even when analyzing a legislative enactment. See, e.g., Waldrup v. Dugger, 562 So.2d 687, 694 (Fla. 1990); Smith, 507 So.2d at 1090.
The issue of severability arises only after an amendment already approved by voters has been challenged. Rather than ignoring the results of the election and requiring the Secretary of State to show that voters would have approved an amendment without the unconstitutional provisions, the burden is properly placed on the challenging party. The analysis urged by appellants would be the antithesis of the purpose underlying severability to preserve the constitutionality of enactments where it is possible to do so. Accordingly, we conclude that we should adopt the severability analysis that we have applied to legislative enactments.
Appellants do not dispute that prongs (1) and (4) of the severability analysis are satisfied, but assert that the amendment must fail because prongs (2) and (3) cannot be established. As to prong (2), which requires an analysis of whether the purpose expressed in the valid provisions can be accomplished independently of the void provisions, appellants assert that the purpose of the amendment cannot be satisfied without the provisions regarding term limits for federal legislators.
Appellants base this argument on our decision in Limited Political Terms, where we held that the amendment satisfied the single-subject requirement. Appellants maintain that because we held in Limited Political Terms that the amendment represents "one purpose," the purpose of the amendment cannot be accomplished without all the provisions. Thus, they argue that the void provisions cannot be severed and the entire amendment must fail.
We do not agree with appellants that the amendment's compliance with single-subject requirement of article XI, section 3 mandates that the entire amendment be stricken. The single-subject requirement was designed to "protect[] against multiple `precipitous' and `cataclysmic' changes in the constitution by limiting to a single subject what may be included in any one amendment proposal." Term Limits Pledge, 718 So.2d at 801 (quoting Advisory Opinion to the Attorney General, 705 So.2d 1351, 1353 (Fla. 1998)). By contrast, the severability analysis concerns whether, absent the invalid portions, the statute (or in this case, amendment) can nonetheless accomplish its purpose. See Martinez, 582 So.2d at 1173.
Therefore, simply because an amendment satisfies the single-subject requirement does not mean that the provisions of the amendment are so mutually dependent on one another that the overall purpose of the amendment cannot be accomplished absent the invalid provisions. Because each citizens' initiative amendment must satisfy the single-subject requirement before being placed on the ballot, appellants' argument would require us to invalidate the entirety of every citizens' initiative amendment whenever a portion was declared invalid after the passage of the amendment.
In this case, as set forth in the initiative petition, the overall purpose to be accomplished by this amendment was the limitation of terms for elected officials, be they state legislators, federal legislators, or cabinet officials. Simply because the purpose cannot be accomplished as to the class of federal legislators does not mean that the overriding purpose of the amendment cannot be accomplished as to the remaining offices. Therefore, we conclude that the severance of the unconstitutional provisions relating to federal legislators is not fatal to the amendment.
In reaching this conclusion we reject appellants' reliance on the Nebraska Supreme Court's decision in Duggan v. Beermann, 249 Neb. 411, 544 N.W.2d 68 (1996), where the Nebraska Supreme Court concluded that the valid portions of Nebraska's term limits amendments relating to state legislators could not be severed from the invalid portions relating to federal legislators, so that all the provisions must fail. We find that case distinguishable both factually and because, unlike Florida, Nebraska does not have a mandated pre-ballot judicial review procedure.
In Duggan, there were numerous proposed amendments, poorly and confusingly drafted, purporting to amend four existing provisions and add four new sections to Nebraska's constitution. The amendments were written in such a way that the provisions relating to state offices were "intertwined" with, and incorporated by reference in the provisions relating to congressional offices. See id. at 79. The amendment, which had not been subject to a pre-ballot review by the Nebraska Supreme Court, was found to have other constitutional flaws as well, not the least of which was that the petition's "object clause" failed to inform voters that the measure applied to state legislators. Id. at 80.
In contrast, in this State there is an extensive statutorily and constitutionally mandated pre-ballot review of citizen initiatives to determine if the proposed amendment satisfies the single-subject requirement and if the ballot title and summary are clear and unambiguous. Unlike the Nebraska amendment, this ballot summary clearly and unambiguously advised the voters that it applied to state and federal legislators.
We find this case to be more akin to the term limits amendment analyzed in the Arkansas Supreme Court's decision in U.S. Term Limits, Inc. v. Hill, 316 Ark. 251, 872 S.W.2d 349 (1994), aff'd sub nom. U.S. Term Limits, Inc. v. Thornton, 514 U.S. 779, 837-38, 115 S.Ct. 1842, 131 L.Ed.2d 881 (1995). In Hill, the Arkansas Supreme Court held that the unconstitutional portions of the amendment to the Arkansas Constitution relating to federal legislators could be severed from the portions relating to state legislators. 872 S.W.2d at 358-59. The amendment in that case was similar in structure to ours, being divided into sections for state legislators, congressional legislators, and executive branch officials. See id. at 351-52. The court concluded that the portions relating to state legislators were clearly "functionally independent" of the portions relating to federal legislators and that the amendment in that case served one overall purpose, even without the invalid portions relating to federal legislators, which was "the limitation of public service terms." Id. at 358.
Likewise, we find that the portions of this amendment are functionally independent. The unconstitutional provisions of this amendment can be stricken without disrupting the integrity of the remaining provisions. Further, the overall purpose of limiting political terms can still be accomplished after the unconstitutional portion is stricken.
We next address prong (3), that is, whether the valid and invalid features are so inseparable in substance that it can be said that the citizens would not have passed the one without the other. As to appellants' argument that the amendment must be stricken because it is impossible to be certain that the voters would have adopted the amendment had it not contained provisions limiting the terms of federal legislators, this argument misplaces the burden of proof, which is properly on the challenging party.
Appellants' argument that "it is more likely that the people would not have voted for term limits for state representatives and senators had they known of the unconstitutionality of term limits for U.S. Representatives and U.S. Senators," Appellants' Initial Brief at 15, is based on nothing more than conjecture and speculation. Although the amici arguing in support of appellants point to several media reports suggesting that the backers of the amendment often focused on the excesses of career federal legislators, see Brief of Amici Curiae in Support of Appellants, according to the statistical evidence in the record, voters nationwide supported measures for term limits in nearly equivalent proportions, whether directed at federal legislators, state legislators or both.
Moreover, it is clear from the initiative petition that severability was anticipated by the voters. The initiative petition in this case specifically contained a severability clause, which is persuasive of the fact that the framers intended severability to save the amendment in case portions of it were declared invalid. Cf. Moreau v. Lewis, 648 So.2d 124, 127 (Fla. 1995); Smith v. Department of Ins., 507 So.2d at 1080. The full text of the amendment, including the severability clause, was disseminated throughout the state prior to the election and was posted conspicuously in every precinct on the day of the election,[8] thus providing voters with clear notice of severability.
Finally, we find the reasoning of the Arkansas Supreme Court relevant to our analysis of this prong:
The fact that one category of persons is eliminated [from the amendment] does not mean that the voters did not intend it to apply to the remaining two categories. Nor do we consider term limits on federal legislators to be the bait which enticed voters to vote aye on the amendment as a whole. There is nothing to suggest that this was the case. In short, we are confident that [the amendment] would have passed even without the inclusion of the [section relating to federal legislators] in that the majority was voting for a conceptthe limitation of public service terms.
Hill, 872 S.W.2d at 358 (emphasis supplied).
In conclusion, we find that Florida's term limits amendment is viable and complete, even when the invalid portions are stricken. Even without the invalid provisions, the amendment accomplishes its purpose of limiting political terms of state officeholders. Finally, appellants have failed to establish that the voters' purpose in passing the amendment was only to limit terms of federal officeholders and the voters would not have approved the term limits amendment if only state officeholders had been included. We thus conclude that the amendment survives a severability analysis and should not be stricken. We next consider whether we should revisit our decision in Limited Political Terms.
REVISITING OUR DECISION IN LIMITED POLITICAL TERMS
The amici arguing in support of appellants in this case have asked us to revisit our decision in Limited Political Terms, where we held that this amendment complied with single-subject and ballot title and summary requirements. See 592 So.2d at 227-29. The amici argue that our earlier decision was in error, that Justices Kogan and Overton were correct in their dissenting opinions, and that we should now conclude that the petition did not in fact embrace one single subject and was therefore improper.
In arguing that we should revisit our decision in Limited Political Terms, the amici have suggested that advisory opinions from this Court have little precedential value and are only persuasive, citing to Ervin v. City of North Miami Beach, 66 So.2d 235, 236-37 (Fla.1953), Lee v. Dowda, 155 Fla. 68, 19 So.2d 570 (1944), and Collins v. Horten, 111 So.2d 746, 751 (Fla. 1st DCA 1959). However, none of these cases concern advisory opinions required by section 16.061, Florida Statutes (1997). At the outset, we point out that when our "advisory" opinions conclude that there is a defect in the ballot title and summary or a violation of the single-subject requirement, the effect of our "advice" is the removal of the amendment from the ballot. See, e.g., Term Limits Pledge, 718 So.2d at 804; Advisory Opinion to the Attorney General re People's Property Rights Amendments, 699 So.2d 1304, 1312 (Fla.1997).
We addressed the precedential effect of our advisory opinions on single-subject and ballot title and summary issues in Florida League of Cities v. Smith, 607 So.2d 397 (Fla.1992). In Smith, the Attorney General had previously requested this Court's opinion regarding the proposed initiative petition's compliance with article XI, section 3 and section 101.161. 607 So.2d at 398. We held that the petition satisfied both these requirements. See In re Advisory Opinion to the Attorney General Homestead Valuation Limitation, 581 So.2d 586, 588 (Fla.1991).
Subsequent to our advisory opinion on the proposed amendment, interested parties brought a petition for writ of mandamus requesting that the proposed amendment be removed from the ballot, bringing to the Court's "attention an issue not addressed in our prior opinion regarding the ballot title and summary." Smith, 607 So.2d at 398.[9] Before considering the merits of the petition, we first addressed whether the previous advisory opinion precluded us from considering the petition for mandamus. We acknowledged the principle that advisory opinions "are not binding judicial precedent." Id. at 399 n. 3. However, we stressed that
relitigation of issues expressly addressed in an advisory opinion on a proposed amendment is strongly disfavored and almost always will result in this Court refusing to exercise its discretionary jurisdiction. Renewed litigation will be entertained only in truly extraordinary cases, such as in the present case where a vital issue was not addressed in the earlier opinion.
Id. at 399 (emphasis supplied). Thus, we have previously made clear that although our advisory opinions are not strictly binding precedent in the most technical sense, only under extraordinary circumstances will we revisit an issue decided in our earlier advisory opinions.
This case does not present such a circumstance. The argument on whether the amendment complies with the single-subject requirement is not new argument on an important issue not addressed in our earlier opinion. Along with the ballot title and summary discussion, the single-subject question was the precise issue covered in our prior opinion. The subsequent invalidity of the provisions relating to federal legislators due to the Supreme Court's decision in Thornton does not present a basis for us to recede from our earlier decision. Therefore, we conclude that our earlier decision in Limited Political Terms is binding on the single-subject issue and we find no reason to reconsider the issue and recede from that decision.
FEDERAL CONSTITUTIONAL ANALYSIS
Finally, in the Limited Political Terms opinion, this Court identified the amendment as one imposing a qualification on holding office. 592 So.2d at 228. Therefore, the validity of this amendment under the United States Constitution should be analyzed like other candidate qualification rules such as age, residency, mental competency or restoration of civil rights for persons previously convicted of a felony. Candidate qualification rules, such as term limits, implicate the right to associate for the advancement of political beliefs and the right of qualified voters to cast their votes effectively, which are rights protected by the First and Fourteenth Amendments. See Anderson v. Celebrezze, 460 U.S. 780, 786-87, 103 S.Ct. 1564, 75 L.Ed.2d 547 (1983).
In Burdick v. Takushi, 504 U.S. 428, 434, 112 S.Ct. 2059, 119 L.Ed.2d 245 (1992), the United States Supreme Court explained the analysis to be applied in challenges to restrictions on candidacy such as candidate qualification rules. The Court held that in analyzing whether the restrictions are unconstitutional, the reviewing court must weigh the "character and magnitude of the asserted injury to the rights protected by the First and Fourteenth Amendments" against the interest put forward by the state in justifying the burden on these rights. Id. (quoting Anderson v. Celebrezze, 460 U.S. at 789, 103 S.Ct. 1564). Both the Arkansas and California Supreme Courts have concluded that the electorate's interest in limiting terms outweighs the interests of candidates for reelection and the voters who wished to vote for those particular candidates. See Hill, 872 S.W.2d at 360; Legislature of Cal. v. Eu, 54 Cal.3d 492, 286 Cal.Rptr. 283, 816 P.2d 1309 (1991).
In this case, according to the citizens' petition, the interest of the voters in limiting terms was three-fold: (1) to increase voter participation, (2) to increase citizen involvement in government, and (3) to increase the number of persons who will run for elective office. See Limited Political Terms, 592 So.2d at 226 (quoting the initiative petition). The voters have also expressed a belief that "politicians who remain in office too long may become preoccupied with re-election and become beholden to special interests and bureaucrats." Id. As the Arkansas Supreme Court recognized, "[i]t is not the function of this court to agree or disagree with the purpose and rationale" behind the term limits amendment. Hill, 872 S.W.2d at 360. Applying the Burdick analysis and weighing the degree of infringement against the state interest, we agree with the conclusion of the Arkansas and California Supreme Courts that this type of amendment passes muster under the First and Fourteenth Amendments. See Hill, 872 S.W.2d at 360; Eu, 286 Cal.Rptr. 283, 816 P.2d at 1314-29.
We reject appellants' argument that the amendment violates the equal protection clause by diluting the votes of voters who reside in rural counties that do not have as many senators representing the counties' interests as voters in urban counties. We conclude that the constitutionality of this amendment is better analyzed under the Supreme Court's decision in Burdick. See Anderson, 460 U.S. at 787 n. 7, 103 S.Ct. 1564 (declining to apply an equal protection analysis on candidate qualification restrictions).[10]
CONCLUSION
Based on the foregoing, we hold that a severability analysis applies to constitutional amendments, that the amendment satisfies that analysis, and that the valid portions can be severed from the invalid portions. We also conclude that the amendment withstands scrutiny under the First and Fourteenth amendments to the United States Constitution and find it unnecessary to revisit our decision in the earlier advisory opinion regarding the amendment's compliance with the single-subject requirement.
Accordingly, the decision of the trial court is affirmed.
It is so ordered.
HARDING, C.J., and SHAW, WELLS, ANSTEAD, PARIENTE and QUINCE, JJ., concur.
LEWIS, J., concurs in result only with an opinion.
LEWIS, J., concurring in result only.
I concur in the majority's result only because it is now too late to revisit our almost decade-old decision in Advisory Opinion to the Attorney General-Limited Political Terms in Certain Elective Offices, 592 So.2d 225 (Fla.1991) (finding that the underlying initiative petition met the single-subject requirement of article XI, section 3 of the Florida Constitution). A change in midstream would be an affront to our governing charter and the democratic process. Nevertheless, I write to express my complete agreement with Justice Kogan's dissent to that finding and Justice Overton's conclusion (with which Justice Kogan concurred) that, in appropriate cases, this Court's review of ballot initiatives should include a determination of "whether or not the proposed amendment to [the Florida] constitution meets constitutional requirements of validity under the Constitution of the United States." Id. at 229 (Overton, J., concurring in part and dissenting in part). In my view, the circumstances in this case point to a serious constitutional flaw, if we adhere to the principle announced in Fine v. Firestone, 448 So.2d 984 (Fla.1984), that this Court should not be placed in the position of redrafting a proposed constitutional amendment by judicial construction.
In Fine, this Court recognized that the one-subject analysis applicable to laws enacted by the legislature is not the same as the one-subject analysis applicable to initiatives seeking to amend the Florida Constitution. See 448 So.2d at 988-89. This Court indicated that it should not be placed in the position of being required to redraft substantial portions of the constitution by judicial construction, see id. at 989, which has now become necessary if any portion of the amendment at issue here is to be upheld. However, based upon this Court's prior approval of the form presented to the voters, we must engage in some interpretation or analysis in order to afford respect to the initiative power in the democratic process.
As a threshold issue, Justice Overton pointed out, in Limited Political Terms, that article IV, section 10 of the Florida Constitution requires this Court to consider the validity of an initiative petition presented to the electorate under article XI, section 3. 592 So.2d at 229. While acknowledging this Court's tasks regarding single subject and ballot title and summary, Justice Overton nonetheless maintained "that those provisions do not limit our responsibility in considering whether or not the proposed amendment to [the Florida Constitution] meets constitutional requirements of validity under the Constitution of the United States." Id. Justice Overton reasoned:
A review at this time, should this legal issue be resolved adverse to the proponents of the amendment, would save both proponents and opponents of the amendment considerable expense and the considerable expense to the state of a futile election. To allow the people to vote and then, if adopted, hold the provision unconstitutional on its face perpetuates a fraud on the voting public. I find that both our constitution and case law recogniz[e] our authority to resolve this strictly legal issue now, without further court proceedings.
Id. at 229-30. Accordingly, Justice Overton concluded that the portions of the amendment relating to United States Senators and Representatives were violative of the Qualifications Clause of the United States Constitution, Article I, Sections 2 and 3. Id. at 231.
Subsequent events vindicated Justice Overton's prescience regarding the constitutionality of congressional term limits. Thus, it was not a bolt from the blue when the United States Supreme Court decided the issue as it did in U.S. Term Limits, Inc. v. Thornton, 514 U.S. 779, 115 S.Ct. 1842, 131 L.Ed.2d 881 (1995). If Justice Overton's advice had been heeded, we would not be writing this opinion, and our finite judicial resources would be focused on other cases. Limited Political Terms, 592 So.2d at 230 ("Deciding [the constitutional validity issue] now would further judicial economy."). Additionally, we would not need to establish precedent for judicial construction of voters' intentions, a precedent described by this Court in Fine as "dangerous." 448 So.2d at 989.
As to our single-subject analysis, which bequeathed the unusual situation now confronting us, I believe Justice Kogan pinpointed the fatal flaw in approving catchall initiatives such as this. As I cannot improve on Justice Kogan's cogent analysis, I quote him at length:
The single-subject requirement contained in article XI, section 3 of the Florida Constitution states that
any ... revision or amendment shall embrace but one subject and matter directly connected therewith.
Art. XI, Sec. 3, Fla. Const. As the majority correctly notes, we traditionally have stated that this constitutional provision requires an initiative to contain a logical and natural "oneness of purpose." E.g., Fine v. Firestone, 448 So.2d 984 (Fla.1984). However, the erratic nature of our own case law construing article XI, section 3 shows just how vague and malleable this "oneness" standard is. What may be "oneness" to one person might seem a crazy quilt of disparate topics to another. "Oneness," like beauty, is in the eye of the beholder; and our conception of "oneness" thus has changed every time new members have come onto this Court.
I think the only proper way to resolve this issue is by looking to the fundamental policies underlying article XI, section 3. Why was the single-subject clause put into this provision?[[11]]
The obvious and unmistakable purpose underlying article XI, section 3 is to reserve to the voters the prerogative to separately decide discrete issues. Therefore, one way of deciding the question before us today is to determine whether the proposed initiative contains more than one separate issue about which voters might differ.5 In other words, is there at least one discrete, severable portion of the ballot language that reasonable voters6 might reject if given the choice, even while accepting the remainder of the ballot language? If the answer is yes, then this Court must find that the initiative contains more than one subject and lacks "oneness." Accord Evans v. Firestone, 457 So.2d 1351, 1354 (Fla.1984).
The policy underlying this requirement is self-evident. Where reasonable voters may differ, then the voters should not be placed in the position of accepting an all-or-nothing grab-bag initiative. Each discrete issue should be placed separately on the ballot so that voters can exercise their franchise in a meaningful way. No person should be required to vote for something repugnant simply because it is attached to something desirable. Nor should any interest group be given the power to "sweeten the pot" by obscuring a divisive issue behind separate matters about which there is widespread agreement. Accord Evans v. Firestone.
I believe the present initiative clearly and unmistakably violates these principles, rendering it conclusively defective. Here, the voters of Florida are being asked to approve or disapprove an initiative designed to limit the terms of persons who hold public office at many different levels of government. Under the proposed ballot language, the voter can only decide to limit all, or limit none. Those voters who might desire, for example, to limit the terms of state legislators but not members of Congress have no meaningful way to make this choice, even though there are many valid reasons for taking such a position.
Limited Political Terms, 592 So.2d at 231-32 (Kogan, J., concurring in part and dissenting in part) (emphasis added)(footnote added).[12]
Note 5: I do not suggest that an initiative contains multiple subjects if reasonable voters might disagree with some integral component by which the initiative achieves its purposes. Rather, such disagreement must be about matters that, if severed, would leave at least two complete and workable proposals. If so, the component is discrete and not integral. If the disagreement is about a matter that cannot be severed without rendering the remainder absurd, then the initiative must stand or fall as a unit when put to the voters.
Note 6: Obviously, the role of this Court is not to determine how the vote will go, but merely whether at least one reasonable and legitimate controversy might exist that voters should decide separately from the rest.
Taking that analysis a step further, it should be beyond debate that any initiative petition containing a severability clause telegraphs the message that even its proponents realize it does not contain a "single," discrete subject. In other words, if a proposal truly had a "logical and natural oneness of purpose," how could severability be relevant? Anything that is divisible is not, by definition, "single." It is a sum of the parts that can be sorted into discrete and separate units. If a significant or substantial portion of a proposed amendment is declared invalid, and "an act complete in itself remains after the invalid provisions are stricken," that necessarily means that a single, discrete proposal was not presented. Needless to say, if a "single" subject is declared unconstitutional, nothing remains to be severed. Accordingly, a severability clause should act as a red flag in this Court's review of the constitutionality of initiative petitions.
Finally, I am troubled by the majority's adoption of the legislative severability analysis, especially the fourth prong: "an act complete in itself remains after the invalid provisions are stricken." Majority op. at 1281(quoting Smith v. Department of Insurance, 507 So.2d 1080, 1089 (Fla. 1987)). This portion of the analysis returns us to Justice Kogan's identification of the underlying flaw in this Court's initial review of the amendment: "What may be `oneness' to one person might seem a crazy quilt of disparate topics to another. `Oneness,' like beauty, is in the eye of the beholder." Limited Political Terms, 592 So.2d at 231 (Kogan, J., concurring in part and dissenting in part). Under this prong of the analysis, we must again make a value judgment as to what constitutes "an act complete in itself," in other words, are the remaining provision(s) significant enough for inclusion in our basic governing charter, the Florida Constitution? This is an inherently subjective decision which in this case is, in essence, another single-subject review after the fact. However, because the amendment here met the single-subject requirement as broadly defined by a majority of this Court eight years ago, it is highly unlikely that the approved amendment would fail this prong of the analysis years later.
This circular process renders a severability analysis in this context a hollow exercise. The issue has been effectively decided by the initial single-subject review. However, if one accepts, under these circumstances, the notion that the constitutional concept of "one subject and matter" can be harmonized properly with the concept of "severability," which we are forced to address in this case, then, in order to afford respect to the initiative power in the democratic process, it becomes necessary to apply some form of severability analysis and to determine the proper severability analysis to be applied.
In my view, the elements of the legislative severability analysis adopted by the majority are inadequate to protect our basic governing charter from misadventures in the future. It is clear that, over the last fifty years, various subject matters determined to be constitutionally invalid by the United States Supreme Court have been the object of passionate debate, and it is inevitable that the same will arise in the future. Depending upon the power of any interest group at any given time, invalid but politically popular subjects can be coupled with connected, valid but divisive issues to effectuate a change in the very structure of government which would require redrafting of the constitutional amendment by judicial construction, unless invalidity of any part of a proposed amendment renders the entire proposal invalid.
In such a case, it would be impossible, pursuant to the third prong of the test adopted by the majority, to prove or determine after the fact, in accordance with any legally acceptable standard, how voters would have responded, had a substantial or significant part of a multifaceted but "connected" amendment been eliminated before vote. Therefore, an appropriate analysis should include elements 1, 2 and 4 of the majority's severability test, but it should also include consideration of whether a significant or substantial portion of the amendment has been invalidated. This consideration-which would provide the lens through which the other elements are viewed-would then require a balancing of that which has been invalidated with that which remains.
The test to be employed is essential to preservation of the integrity of the citizen initiative process. Its goal is to strike a balance between the importance of giving effect to the will of the people to make a change in their organic law, as expressed through their vote on the initiative ballot, and the importance of giving effect to the change in law which was actually voted on by the people, rather than some judicially crafted change.
In order to achieve this goal, one must look to the particular single-subject concept as defined by this Court in construing the proposed amendment at the time of the citizen initiative, to determine whether what has been invalidated is a significant or substantial part of that concept. If that which has been taken away is a significant or substantial part of the single-subject concept as thus defined, then what remains is not severable from it and cannot survive. If, on the other hand, that which was taken away is not a significant or substantial part of the single-subject concept at issue, then what remains is severable and will be upheld.
As applied to this case, a very close question is presented, but in looking to the broad definition of single subject which was adopted in this case (term limits), it cannot be said that what was taken away (one subset of elective officers) is so significant that it vitiates the single subject identified therein. However, if Justice Overton's and Justice Kogan's views had been followed, we would not have been required to accept this constitutional amendment by judicial construction.
For these reasons, I can only concur in the result reached by the majority.
NOTES
[1] As a result of the November 1998 elections, Katherine Harris replaced Sandra Mortham as Secretary of State, effective January 1999. The lawsuit was commenced prior to the time Katherine Harris became Secretary of State.
[2] The full text of the amendment reads as follows:
(b) No person may appear on the ballot for re-election to any of the following offices:
(1) Florida representative,
(2) Florida senator,
(3) Florida Lieutenant governor,
(4) any office of the Florida cabinet,
(5) U.S. Representative from Florida, or
(6) U.S. Senator from Florida if, by the end of the current term of office, the person will have served (or, but for resignation, would have served) in that office for eight consecutive years.
Art. VI, § 4(b), Fla. Const.
[3] Article IV, section 10 of the Florida Constitution provides:
The attorney general shall, as directed by general law, request the opinion of the justices of the supreme court as to the validity of any initiative petition circulated pursuant to Section 3 of Article XI.
[4] Section 16.061(1), Florida Statutes (1997), provides in pertinent part:
The Attorney General shall, within 30 days after receipt of a proposed revision or amendment to the State Constitution by initiative petition from the Secretary of State, petition the Supreme Court, requesting an advisory opinion regarding the compliance of the text of the proposed amendment or revision with s. 3, Art. XI of the State Constitution and the compliance of the proposed ballot title and substance with s. 101.161.
[5] The official vote, as certified by the Secretary of State, was 3,625,500 in favor of the amendment, and 1,097,127 opposed. (Defendant's ex. # 1.)
[6] In their complaint, appellants addressed their arguments only to the portions of the amendment regarding state legislators. At oral argument, appellants agreed that if their rationale prevailed, the entire amendment must be stricken, including the portions addressing Florida's Lieutenant Governor and cabinet.
[7] We authorized the filing of briefs by amici curiae in support of the positions of the respective parties.
[8] See § 101.171, Fla. Stat. (1991).
[9] The petitioners in that case alleged that the proposed amendment to article VII, section 4 would trigger the "repealer" clause of article VII section 6, which the ballot title and summary did not mention. Therefore, the petitioners argued, the proposed amendment must be stricken from the ballot. We ultimately disagreed. See Florida League of Cities v. Smith, 607 So.2d 397, 401 (Fla.1992).
[10] We also reject appellants' argument that the amendment disparately affects voters in odd-numbered districts because senators serving in odd numbered districts will be denied the right to run for reelection in the year 2000, while senators currently serving in even-numbered districts, who are not be subject to reelection until 2002, will be allowed to serve until that time. This argument was not raised in the complaint below and was not considered by the trial court. We therefore do not address it in this case.
[11] At various times, this Court has characterized the single-subject requirement as "a rule of restraint that protects against unbridled cataclysmic changes in Florida's organic law," see, e.g., Advisory Opinion to the Attorney General re People's Property Rights Amendments, 699 So.2d 1304, 1307 (Fla.1997), and guards against "log-rolling," defined as aggregating "dissimilar provisions in one law in order to attract the support of diverse groups to assure its passage." Fine v. Firestone, 448 So.2d 984, 988 (Fla.1984). Surely those sensible rationales underlying the single-subject requirement were ignored here in what can only be described as a dramatic change to "Florida's organic law" and a veritable tutorial in logrolling.
[12] This Court has wisely heeded Justice Kogan's admonition in similar cases. See, e.g., Advisory Opinion to the Attorney General re Right of Citizens to Choose Health Care Providers, 705 So.2d 563 (Fla.1998). In finding the proposed amendment violative of the single-subject requirement, this Court determined:
The proposed amendment combines two distinct subjects by banning limitations on health care provider choices imposed by law and by prohibiting private parties from entering into contracts that would limit health care provider choice. The amendment forces the voter who may favor or oppose one aspect of the ballot initiative to vote on the health care provider issue in an "all or nothing" manner. Thus, the proposed amendment has a prohibited logrolling effect and fails the single-subject requirement.
Id. at 566 (emphasis added); see also Advisory Opinion to the Attorney General-Save Our Everglades, 636 So.2d 1336, 1341 (Fla.1994) (finding single-subject violation where "[m]any voters sympathetic to restoring the Everglades might be antithetical to forcing the sugar industry to pay for the cleanup by itself, and yet those voters would be compelled to choose all or nothing") (emphasis added). Because the voters here were presented with precisely the same "all or nothing" choice, how is it that the same "prohibited logrolling effect" did not result in a single-subject requirement violation? No rational distinction supports such contradictory outcomes. Moreover, this Court also found a single-subject violation "because [the proposed amendment] would significantly affect the legislative and executive branches of government...." Health Care Providers, 705 So.2d at 566 n. 1. Again, this case presents exactly the same scenario in that the amendment's broad swath included not only state and federal legislators, but numerous executive branch officers, among them the Lieutenant Governor and "any office of the Florida cabinet." Limited Political Terms, 592 So.2d at 226.