How California lives with two legislatures



California has two legislative bodies: the electorate and the legislature. Practical experience and separation-of-powers theory teach that two political actors simultaneously wielding the same governmental power is a recipe for disaster. Conflict is inevitable, and the greatest risk is a problem known as cycling: when two actors share a power, policy issues can cycle repeatedly between the actors and never be resolved. In this article we examine how the legislative powers of the California electorate and the legislature interact, and use a current initiative proposal as a practical example to show how the restrictions on the legislature’s ability to amend initiative measures are the key to prevent cycling.


California has two plenary legislative bodies

The California legislature has plenary legislative power except as specifically limited by the California constitution.[1] The relevant limitation here is the electorate’s direct democracy powers: the state’s entire lawmaking authority is vested in the legislature, except the electorate’s initiative and referendum powers.[2]

The state’s electorate also has plenary legislative power. The state constitution does not limit the subject matter of initiative statutes.[3] This sometimes raises questions about which actor has “more” legislative power — and the answer depends on whether you look at each alone or relative to each other.

From one viewpoint the legislature’s powers are broader than the electorate’s. The California Supreme Court has explained that the Progressive-era reforms that introduced direct democracy in 1911 restored to the electorate only a shared piece of the whole legislative power that the 1849 state constitution gave to the legislature.[4] But it is more accurate to say that the electorate’s initiative power is “coextensive” with the legislature’s;[5] or that the initiative is the same as the legislature’s power;[6] or that the electorate cannot enact a statute that the legislature itself could not enact.[7] And that only describes the creative power of either the legislature or the electorate alone — it does not compare their power relative to each other.

Cycling is prevented because the electorate holds the trump card

These legislative bodies act in concert, and cycling is prevented, because the electorate has the final say. That finality flows from a constitutional provision that gives the electorate control over amendments to its initiatives: Article II, section 10(c) of the California constitution provides that the legislature “may amend or repeal an initiative statute by another statute that becomes effective only when approved by the electors unless the initiative statute permits amendment or repeal without their approval.”[8] That section prevents cycling by giving the electorate the trump card. This suggests that the state’s two legislative bodies are not, in fact, equal in power. The electorate is the final word on California legislative policy decisions because it can both set the terms of and end a policy debate.

Consequently, between the two actors the electorate holds greater power because when the legislature and the electorate conflict, the electorate prevails.[9] A fundamental aspect of the initiative power is its ability to override the state legislature, making the electorate’s policy decision the final one. And most limits on the legislature’s lawmaking power do not apply to the electorate.[10] So when comparing the two legislative powers, the California Supreme Court held that “[t]he people’s reserved power of initiative is greater than the power of the legislative body.”[11] Article II, section 10(c) effectively gives the electorate a veto (perhaps even the equivalent of a line-item veto) over legislative amendments. And the referendum power obviously is a veto over legislative acts. The electorate can use this power to end policy debates between the legislature and the courts, or between the electorate and the courts (even by removing judicial officers).[12] No one has a veto over initiative acts.[13] These electorate powers, and the relative absence of countervailing powers held by other branches, make the electorate the ultimate authority in policy matters.

The standard of review for legislative amendments

Reviewing legislative amendments to initiatives requires balancing the interaction of two plenary policymakers. Because the initiative process “occupies an important and favored status in the California constitutional scheme,” the California Supreme Court has “consistently deemed it our duty to guard the people’s right to exercise the initiative power.”[14] Yet separation-of-powers concerns similarly compel the courts to respect the legislature’s powers, and so the legislature’s acts have a strong presumption of correctness.[15] The controlling principle is that the legislature’s powers are limited only by the state constitution — here, by the electorate’s initiative power. When the electorate exercises its power to limit the legislature’s ability to amend an initiative statute, that is a constitutional limitation. Accordingly, the legislature may only amend an initiative statute without subsequent voter approval when the initiative itself permits such amendment, “and then only upon whatever conditions the voters attached to the Legislature’s amendatory powers.”[16]

Courts will review legislation that affects the subject of an initiative on substantive and procedural grounds. If a legislative act is challenged for violating the initiative-amendment rule, the act will be evaluated based on two substantive questions:

  • Is the act unrelated to the initiative’s subject? If so, the legislature is free to act at will.[17]
  • If it is related to the initiative’s subject, does it change that subject? If so, then the court must evaluate whether the legislature acted within the initiative’s limits.[18]

Courts will also apply any procedural requirements imposed by the initiative. For example, in People v. DeLeon (which concerned criminal resentencing under Proposition 36’s reformation of the state’s Three Strikes law) the initiative permitted the legislature to provide additional rights for crime victims by a majority vote, but required a three-fourths legislative majority to amend the initiative by reducing those rights.[19]

In general, courts review statutes enacted by the legislature and the electorate by the same rules and canons of statutory construction.[20] Courts first consider the act’s language, giving the words their ordinary meaning and construing them in the context of the act as a whole.[21] If the language is not ambiguous, courts presume the voters intended the meaning apparent from that language.[22] If the language is ambiguous, courts consider ballot summaries and arguments in determining the voters’ intent and understanding of a ballot measure.[23]

A court’s primary task in interpreting provisions adopted by initiative “is to determine and give effect to the intent of the voters.”[24] And the same presumption of validity applies to all legislative acts, whether carried out by the electorate or the legislature.[25] Finally, it is common for an initiative measure to include a provision invoking Article II, section 10(c) and authorizing the legislature to amend the initiative without voter approval only if the amendment furthers the initiative’s purpose.[26] Consistent with the general rule for other limits on the legislature’s power, courts construe this limitation strictly.[27] The bottom line is that “the voters should get what they enacted, not more and not less.”[28]

A current-events example

As an example, we will apply the standard described above to a pending initiative measure. The California Privacy Rights and Enforcement Act (Proposition 24) is one of 12 initiatives qualified for the November 3, 2020 general election ballot.[29] (Download a PDF of the measure.) Its stated intent is “to further protect consumers’ rights, including the constitutional right of privacy.”[30] It warns that “the Legislature considered many bills in 2019 to amend the [California Consumer Privacy Act of 2018], some of which would have significantly weakened it. Unless California voters take action, the hard-fought rights consumers have won could be undermined by future legislation.”[31] The measure declares that “Rather than diluting privacy rights, California should strengthen them over time.”[32]

Like many initiatives, Proposition 24 contains a section that permits legislative amendments and defines the scope of permitted amendments:

The provisions of this Act may be amended after its approval by the voters by a statute that is passed by a vote of a majority of the members of each house of the Legislature and signed by the Governor, provided that such amendments are consistent with and further the purpose and intent of this Act as set forth in Section 3 . . . .[33]

That restriction applies to “all statutes amended or reenacted as part of this Act, and all provisions of such statutes, regardless of whether this Act makes any substantive change thereto.”[34] Finally, the amendment restriction applies even to statutes unaffected by Proposition 24: “Any amendments to this Act or any legislation that conflicts with any provision of this Act shall be null and void upon passage of this Act by the voters, regardless of the code in which it appears. Legislation shall be considered ‘conflicting’ for purposes of this subdivision, unless the legislation is consistent with and furthers the purpose and intent of this Act . . . .”[35]

This statement of intent unambiguously sets the floor and creates a one-way ratchet: the legislature can only amend this initiative by increasing privacy protections. Some opponents of Proposition 24 apparently argue that a legislative amendment of the law must satisfy every section of the law for it to be valid. That misstates the standard described above, and there is little danger a future court will use that approach, for three reasons.

First, the cases do not so hold. For example, in Gardner v. Schwarzenegger, the court rejected the legislature’s attempt to allow incarceration for drug-related probation violations — because that contravened two express purposes of the initiative at issue (to enhance public safety by freeing jail cells for violent criminals and save money by affording treatment in lieu of incarceration) and because “Proposition 36 specifically limits a court’s ability to order incarceration following a first or second drug-related probation violation.”[36] And in Foundation for Taxpayer & Consumer Rights v. Garamendi, the court rejected the legislative amendment because one of the proposition’s fundamental purposes was to eliminate discrimination against previously uninsured drivers and specifically provided that the absence of prior automobile insurance coverage could not be a criterion for insurance eligibility — yet the legislature’s amendment facially conflicted with that purpose and provision by permitting insurers to use persistency of insurance coverage as a rating factor.[37] In both cases the courts applied the standard described above and properly focused on the fact that the legislature directly contradicted an initiative’s stated purpose.

And those cases describe the opposite of a legislative amendment that would be permitted because it furthers an initiative’s purpose. In both cases the legislature erred with attempts to overturn primary features of the law, not acts in furtherance of its purpose. Those cases correctly applied the “major purpose” standard described above to prevent the legislature from diluting an initiative’s intent, and neither case stands for the proposition that courts should make legislative amendments more difficult than an initiative itself makes them. The standard described above expressly permits the legislature to amend measures consistent with their stated intent — here, “to further protect consumers’ rights.”[38] For example, the legislature could add an opt-in regime, an expanded private right of action, or expand the anti-discrimination features of the law. All of those amendments would increase the degree of consumer protection, and so would be permitted by Proposition 24 because they are consistent with that measure’s intent.

Finally, the point of the look-at-the-whole standard is to determine its “major purposes,” not to find technical issues for the legislature to stumble over.[39] For example, in applying that standard, one Court of Appeal decision described the task as seeking the “fundamental purpose or primary mandate of an initiative” when determining whether an amendment furthers its purpose.[40] Consequently, it is a mistake to define a measure’s purpose at “a level so granular as to equate that intent with each of the specific provisions in the initiative.”[41] On the contrary, the standard only requires a reviewing court to articulate the measure’s purpose at a “high level” — otherwise any subsequent legislative amendments to a measure’s provisions would necessarily change its intent, because “any amendment changes an initiative’s scope or effect, whether by addition, omission, or substitution of provisions.”[42] If the standard meant that any amendment to an initiative’s provisions is inconsistent with the initiative’s intent, then no initiative could ever be amended. Such an approach would render superfluous language expressly allowing (as does Proposition 24) legislative amendments that further the measure’s intent. And that would contradict the principle of statutory construction that interpretations that render any part of a statute superfluous are to be avoided.[43]

When a court reviews Proposition 24 as a whole to ascertain the electorate’s intent,[44] it will find ample evidence of a plain intent to protect consumer privacy, to strengthen existing provisions, and to permit legislative amendments that provide even greater protection. To the extent that plain statement is unclear, a court will next examine the ballot materials.[45] Consistent with the measure’s text, the proponents’ ballot arguments state a plain intent that the legislature may amend the law only to strengthen the protections it provides: Proposition 24 will “MAKE IT MUCH HARDER TO WEAKEN PRIVACY in California in the future, by preventing special interests and politicians from undermining Californians’ privacy rights, while allowing the Legislature to amend the law to further the primary goal of strengthening consumer privacy to better protect you and your children, such as opt-in for use of data, further protections for uniquely vulnerable minors, and greater power for individuals to hold violators accountable.”[46] A reviewing court will understand the measure’s intent to permit only those legislative amendments that increase privacy protections. As Justice Chin wrote, “The substance, not the form, is what matters.”[47] Divining the intent in Proposition 24 to only permit amendments that increase privacy protections is a simple task.


At first glance one might wonder how a government with two legislatures could possibly function. Indeed, California’s government design is sometimes attacked as dysfunctional, and the initiative in particular is scorned as “out of control.”[48] We have shown that many critiques of the initiative are overblown, and that the initiative in California largely serves its intended purpose of empowering the voters to solve intractable problems and break political stalemates.[49] Rather than viewing the state as having two equivalent legislative bodies, those actors are more accurately described as having a master-servant relationship. As in the case of measures like Proposition 24 (with its explicit instructions for courts to follow when reviewing attempted legislative amendments) the legislature is hobbled by the restraints imposed by the electorate. That makes the legislature the electorate’s servant on any matter covered by an initiative, and the electorate controls future changes to those laws.


Senior research fellow Stephen M. Duvernay contributed to this article.

[1] Marine Forests Soc. v. Cal. Coastal Com. (2005) at 43–44, 47–48; Fitts v. Superior Court (1936) at 234 (“[T]he Legislature is vested with the whole of the legislative power of the state.”); People v. Tilton (1869) 37 Cal. 614, 626 (state constitutions are not grants of legislative power).

[2] Methodist Hosp. of Sacramento v. Saylor (1971) at 691.

[3] Santa Clara County Local Trans. Authority v. Guardino (1995) at 246; DeVita v. County of Napa (1995) at 776 (explaining that to the extent that the initiative is the constitutional power of the electors to propose statutes and to adopt or reject them, it is generally coextensive with the legislature’s power to enact statutes).

[4] Howard Jarvis Taxpayers Assn. v. Padilla (2016) at 516 (“When they adopted the initiative power in 1911, they restored to themselves only a shared piece of that power.”).

[5] Kennedy Wholesale, Inc. v. State Bd. of Equalization (1991) at 1364; see also DeVita at 776 (explaining that to the extent that the initiative is the constitutional power of the electors to propose statutes and to adopt or reject them, it is generally coextensive with the legislature’s power to enact statutes).

[6] Legislature v. Deukmejian (1983) at 673 (“the reserved power to enact statutes by initiative is a legislative power, one that would otherwise reside in the Legislature. It has heretofore been considered to be no greater with respect to the nature and attributes of the statutes that may be enacted than that of the Legislature”), 674–75 (“the power of the people through the statutory initiative is coextensive with the power of the Legislature”) and 676.

[7] Bldg. Indus. Assn. v. City of Camarillo (1986) at 821; Deukmejian at 675 (initiative statutes are subject to the same constitutional limitations and rules of construction as legislative statutes); State Comp. Ins. Fund. v. State Bd. of Equalization (1993) 14 Cal.App.4th 1295, 1300 (“Apart from procedural differences, the electorate’s lawmaking powers are identical to the Legislature’’s.”).

[8] See also County of San Diego v. Com. on State Mandates (2018) at 206 (legislature is prohibited from amending an initiative statute unless the initiative itself permits amendment.).

[9] Carrillo, Duvernay & Stracener, California Constitutional Law: Popular Sovereignty (2017) 68 Hastings L.J. 731, 746 (“[T]he electorate has a veto over acts of the legislature through the referendum power. As a result, arguably more than half of the state’s legislative power rests with the voters.”) (italics omitted).

[10] California Cannabis Coalition v. City of Upland (2017) at 935 (when voters exercise the initiative power “they do so subject to precious few limits on that power” and procedural requirements imposed on the legislature “do not similarly constrain the electorate’s initiative power without evidence that such was their intended purpose”); see also Brown v. Superior Court (2016) at 354.

[11] Rossi v. Brown (1995) at 715 (original italics).

[12] Carrillo, Duvernay, Gevercer & Fenzel, California Constitutional Law: Direct Democracy (2019) 92 So.Cal. L. Rev. 557, 611 (“[I]n California the ultimate check on judicial authority lies with the electorate, which has used its power to remove state high court justices.”).

[13] “Neither the Governor, the Attorney General, nor any other executive or legislative official has the authority to veto or invalidate an initiative measure that has been approved by the voters.” Perry v. Brown (2011) at 1126.

[14] Brown at 351.

[15] There is a “strong presumption of the constitutionality of an act of the Legislature.” Delaney v. Lowery (1944) at 569. “In considering the constitutionality of a legislative act we presume its validity, resolving all doubts in favor of the Act. Unless conflict with a provision of the state or federal Constitution is clear and unquestionable, we must uphold the Act.” Amwest Surety Ins. Co. v. Wilson (1995) at 1252 (citations and quotation omitted). “[J]udicial decisions abound with declarations to the effect that all presumptions and intendments favor the validity of statutes; that mere doubt by the judicial branch of the government as to the validity of a statute will not afford a sufficient reason for a judicial declaration of its invalidity, but that statutes must be upheld as constitutional unless their invalidity clearly, positively, and unmistakably appears.” People v. Superior Court (1937) at 298.

[16] People v. Superior Court (Pearson) (2010) at 568.

[17] The legislature remains free to address a related but distinct area or a matter that an initiative measure does not specifically authorize or prohibit. People v. Kelly (2010) at 1025–26; see also Com. on State Mandates at 214 (balancing the electorate’s “precious” initiative power against the legislature’s ability to change statutory provisions outside the scope of an initiative’s provisions by holding that the legislature in most cases retains the power to amend the restated provision through the ordinary legislative process. “This conclusion applies unless the provision is integral to accomplishing the electorate’s goals in enacting the initiative . . . .”).

[18] Because an amendment is “a legislative act designed to change an existing initiative statute by adding or taking from it some particular provision,” to determine whether a legislative act amends an initiative, “we simply need to ask whether it prohibits what the initiative authorizes, or authorizes what the initiative prohibits.” Superior Court (Pearson) at 571.

[19] See People v. DeLeon (2017).

[20] Evangelatos v. Superior Ct. (1988) at 1212 (“[I]nitiative measures are subject to the ordinary rules and canons of statutory construction.”).

[21] Superior Court (Pearson) at 571.

[22] Ibid.

[23] Ibid.

[24] Bowens v. Super. Ct. (1991) at 46; Mosk v. Super. Ct. (1979) at 495 (superseded by Adams v. Com. on Judicial Performance(1994) at 650–51).

[25] Saylor at 691–92 (discussing act by legislature); Brosnahan v. Brown (1982) at 240–41 (discussing act by electorate).

[26] Amwest at 1251.

[27] Id. at 1255–56.

[28] Hodges v. Superior Court (1999) at 888.

[29] California 2020 Ballot Propositions (2020),

[30] Proposition 24 (2020) section 3.

[31] Proposition 24 (2020) section 2(D).

[32] Proposition 24 (2020) section 2(E).

[33] Proposition 24 (2020) section 25(a).

[34] Proposition 24 (2020) section 25(c).

[35] Proposition 24 (2020) section 25(d).

[36] Gardner v. Schwarzenegger (2009) at 1378.

[37] Foundation for Taxpayer & Consumer Rights v. Garamendi (2005) at 362.

[38] Proposition 24 (2020) section 3.

[39] Amwest at 1259.

[40] Howard Jarvis Taxpayers Assn. v. Newsom (2019) at 174.

[41] People v. Superior Court (Alexander C.) (2019) at 1003.

[42] Ibid. (citations and quotations omitted).

[43] Wells v. One2One Learning Foundation (2006) at 135.

[44] Amwest at 1251.

[45] Superior Court (Pearson) at 571.

[46] Proposition 24 (2020) Proponent’s Argument in Favor.

[47] Superior Court (Pearson) at 573.

[48] See, e.g., Diaz, Opinion: California Initiative Process is Out of Control, S.F. Chron. (Sept. 7, 2018).

[49] Carrillo, Duvernay, Gevercer & Fenzel, supra note 12, at 590. Yes, the electorate does sometimes err. See, e.g., Raven v. Deukmejian (1990) (partly invalidating Proposition 115) and Obergefell v. Hodges (2015) (abrogating the California Supreme Court decision in Strauss v. Horton (2009) that upheld Proposition 8).