Opinion
CHIN, J.
In 1996, Proposition 2181 limited local government’s ability to impose real property assessments in two significant ways. An assessment can be imposed only for a “special benefit” conferred on real property (art. XIIID, § 2, subd. (b)), and the assessment on any parcel must be in proportion to the special benefit conferred on the particular parcel (art. XIII D, § 4, subd. (a)).
In 2001, the Santa Clara County Open Space Authority (OSA) imposed a countywide assessment to fund a program to acquire, improve, and maintain unspecified open space lands in the county. Plaintiffs sued, challenging that assessment on the grounds that it fails to satisfy the special benefit and proportionality requirements of Proposition 218. To decide whether OSA’s 2001 assessment violates article XIII D, we must first determine the appropriate standard of judicial review of a local governmental agency’s assessment determination. We conclude that Proposition 218 requires courts to make an independent review of local agency decisions that are governed by express constitutional provisions, as in this case, and that OSA’s assessment does not comply with the special benefit and proportionality requirements of article XIII D.
I. Factual and Procedural History
A. The Creation of OSA and the 1994 Special Assessment District
In 1992, the Santa Clara County Open-Space Authority Act (Pub. Resources Code, § 35100 et seq.) created OSA, with the express purpose of acquiring and preserving open space within the county to counter the conversion of land to urban uses, to preserve quality of life, and to encourage agricultural activities. (Pub. Resources Code, § 35101, subd. (a).) The act provides no particular method to fund open space acquisitions, but it authorizes OSA to levy special assessments under the Streets and Highways Code. (Pub. Resources Code, § 35173.) OSA’s jurisdiction included all Santa Clara [438]*438County lands except those already within the boundaries of the Midpeninsula Regional Open Space District.
In 1994, OSA formed an original assessment district under the authority of the Landscape and Lighting Act of 1972 (LLA).2 (Sts. & Hy. Code, § 22500 et seq.) OSA levied an annual special assessment on the district’s property owners to acquire and preserve open space land under the LLA’s procedures. Certain taxpayers challenged the 1994 assessment, but the Court of Appeal upheld it. The 1994 assessment raised approximately $4 million annually and allowed OSA to purchase thousands of acres of open space lands.3
B. The Creation of the 2001 Assessment District and the Passage of Proposition 218
In 2000, OSA determined that it needed additional annual funding to purchase open space. To raise these additional funds, OSA considered forming an additional assessment district. However, in 1996, California voters had passed Proposition 218 to “significantly tighten the kind of benefit assessments” an agency can levy on real property (Ballot Pamp., Gen. Elec. (Nov. 5, 1996) argument in favor of Prop. 218, p. 76) and to “ ‘protect^ taxpayers by limiting the methods by which local governments exact revenue from taxpayers without their consent.’ ” (Ballot Pamp., Gen. Elec., supra, text of Prop. 218, § 2, p. 108, reprinted in Historical Notes, 2A West’s Ann. Const. (2008 supp.) foil. Cal. Const., art. XIII C, p. 85 (Historical Notes).)
To achieve these goals, Proposition 218 tightened assessment requirements and definitions, imposed stricter procedures on agencies, and shifted traditional presumptions that had favored assessment validity. (Art. XIII D, §§ 2, subd. (i), 4.) Under Proposition 218’s procedures, local agencies must give the record owners of all assessed parcels written notice of the proposed assessment, a voting ballot, and a statement disclosing that a majority protest will prevent the assessment’s passage. (Art. XIII D, § 4, subds. (c), (d).) The proposed assessment must be “supported by a detailed engineer’s report.” (Art. XIII D, § 4, subd. (b).) At a noticed public hearing, the agencies must consider all protests, and they “shall not impose an assessment if there is a majority protest.” (Art. XIH D, § 4, subd. (e).) Voting must be weighted “according to the proportional financial obligation of the affected property.” (Ibid.)4
[439]*439OSA explored the possibility of creating a second assessment district that would comply with the new provisions of Proposition 218. As a first step, the OSA Board of Directors (OSA Board) authorized a poll of Santa Clara County property owners to determine whether they would support an assessment to fund the purchase of additional open space. The poll showed that approximately 55 percent of property owners would likely support up to a $20 per year property tax increase for acquiring and maintaining open space lands.
The OSA Board hired Shilts Consultants, Inc. (SCI), to prepare the engineer’s report. That report stated that the assessment would fund the “[acquisition, installation, maintenance and servicing” of open space lands for recreation, conservation, watersheds, easements, and similar purposes. Although the SCI report identified areas OSA was considering for potential acquisition and improvement and outlined general considerations OSA would use to identify and acquire open space lands, it identified no particular parcels to be acquired and no particular areas to be prioritized.
The proposed 2001 assessment district included all Santa Clara County lands that were in the 1994 assessment district. The proposed assessment district included approximately 314,000 parcels and over 800 square miles containing over 1 million people. The SCI engineer’s report identified the special benefits that would accrue to the assessed parcels, estimated the proportion of all the benefits that could be considered special, set the assessment for a single-family home at $20 per year, and provided a formula for estimating the proportionate special benefit that other property on the tax rolls would receive. Using the $20 property tax increase per single-family home, the SCI engineer’s report calculated that the assessment would produce an approximately $8 million increase in OSA’s budget.
The OSA Board accepted and filed the engineer’s preliminary report and authorized an assessment ballot proceeding. On September 1, 2001, OSA mailed an informational pamphlet to all of the approximately 314,000 property owners within the proposed district. The pamphlet described the assessment district and OSA’s goal of raising about $8 million annually to acquire open space lands within the county.
On September 14, 2001, OSA mailed a notice of the proposed assessment and an official ballot to all affected property owners. On October 25, 2001, OSA conducted an informational meeting, at which OSA’s general manager [440]*440and special counsel and a representative from SCI responded to numerous questions from the public. The formal public hearing was held on November 8, 2001.
On December 13, 2001, OSA reported the results of the balloting at a public hearing. Of the approximately 314,000 official ballots mailed, OSA received only 48,100 responses, a return of approximately 15 percent. Of those responses, 32,127 (66.8 percent) voted in favor of the assessment, while the rest voted “no” (33.2 percent). The returned ballots were weighted in proportion to the amount each parcel was to be assessed, mating the final tally 50.9 percent in favor and 49.1 percent opposed. Plaintiff Silicon Valley Taxpayers’ Association, Inc. (SVTA), objected to the results on procedural grounds no longer relevant to the issues raised here. The final engineer’s report, which was before OSA at the December meeting, contained some changes from the draft report filed in September. In particular, the final report emphasized that the “overriding” and “most important” criterion for OSA to use in acquiring open space was that the acquired lands be distributed throughout OSA’s jurisdiction. At the conclusion of the December hearing, the OSA Board approved the results, accepted the final engineer’s report, and established the new assessment district.
A year and a half later, the OSA Board renewed the assessment for 2003-2004 and added a cost-of-living increase of $0.34 per parcel.
C. Procedural History
SVTA, Howard Jarvis Taxpayers Association, and several individual taxpayers (collectively plaintiffs) filed this action for a writ of mandate, declaratory relief, and an injunction, seeking to invalidate the 2001 assessment. Plaintiffs’ second amended complaint contains two causes of action: the first alleges that OSA’s notice and balloting procedures did not comport with Proposition 218 and the Government Code; the second challenges the substantive validity of the assessment under Proposition 218 and the Landscaping and Lighting Act.
The parties filed cross-motions for summary judgment or, in the alternative, summary adjudication. The court issued an order granting summary adjudication in favor of OSA on the second cause of action.
After the OSA Board renewed its assessment for the 2003-2004 fiscal year, plaintiffs filed a second lawsuit challenging that assessment. The new complaint contained allegations similar to those in the original lawsuit and added claims contesting the increase in the new assessment. The two cases were then consolidated. The court issued an order granting summary adjudication [441]*441in OSA’s favor on the remaining causes of action. Based on that order and the previous order in the first lawsuit, the court entered judgment in favor of OSA.
In a two-to-one decision, the Court of Appeal affirmed the trial court’s judgment. The majority held that Proposition 218 had altered the traditionally deferential standard of review by eliminating the presumption that an assessment was valid. Nevertheless, the majority held that courts should still accord the final legislative determination substantial deference, as long as the agency had followed Proposition 218’s procedural requirements in levying the challenged assessment, and as long as substantial evidence in the administrative record supported the agency’s finding that the benefits were special. Using this limited scope of review, the majority determined that the engineer’s report supported OSA’s determination of special benefits and proportionality.
In her dissent, Justice Bamattre-Manoukian disagreed with the majority regarding the standard of review. In her view, the drafters of Proposition 218 had specifically targeted the deferential standard of review for change. Because the validity of a post-Proposition-218 assessment is now a constitutional question, she asserted that courts should exercise independent judgment in determining whether an assessment complies with article XIII D’s procedural and substantive requirements. The dissent independently analyzed the engineer’s findings concerning special benefits and proportionality and concluded that the identified benefits did not comply with Proposition 218’s legal requirements.
We granted plaintiffs’ petition for review.
II. Discussion
Plaintiffs contend that because state constitutional provisions now govern assessments, courts should apply an independent standard of judicial review to determine their validity. They claim that, in this case, the $20 flat-rate levy is an invalid assessment because it fails to satisfy several provisions of article XIII D, section 4, and that the levy is in essence a “special tax.” They argue further that, because OSA neither sought nor obtained the mandatory two-thirds voter approval for a special tax as required by Proposition 13, the $20 flat-rate levy violates both Propositions 13 (Cal. Const., art. XIII A) and 218. In discussing these claims, we first discuss the nature of special assessments before the enactment of Proposition 218, their relationship to Proposition 13 taxes, and how Proposition 218 changed the law governing assessments. As explained below, we agree with plaintiffs’ contentions.
We explained the nature of a special assessment in Knox v. City of Orland (1992) 4 Cal.4th 132 [14 Cal.Rptr.2d 159, 841 P.2d 144] (Knox), a [442]*442pre-Proposition 218 case. A special assessment is a “ ‘ “ ‘compulsory charge placed by the. state upon real property within a pre-determined district, made under express legislative authority for defraying in whole or in part the expense of a permanent public improvement therein . . .’ ” [Citation.]’ [Citation.] In this regard, a special assessment is ‘levied against real property particularly and directly benefited by a local improvement in order to pay the cost of that improvement.’ [Citation.] ‘The rationale of special assessments] is that the assessed property has received a special benefit over and above that received by the general public. The general public should not be required to pay for special benefits for the few, and the few specially benefited should not be subsidized by the general public. [Citation.]’ [Citation.] . . .
“A tax, on the other hand, is very different. Unlike a special assessment, a tax can be levied ‘ “without reference to peculiar benefits to particular individuals or property.” ’ [Citation.] Indeed, ‘[n]othing is more familiar in taxation than the imposition of a tax upon a class or upon individuals who enjoy no direct benefit from its expenditure, and who are not responsible for the condition to be remedied.’ [Citations.] . . .
“Therefore, while a special assessment may, like a special tax, be viewed in a sense as having been levied for a specific purpose, a critical distinction between the two public financing mechanisms is that a special assessment must confer a special benefit upon the property assessed beyond that conferred generally.” (Knox, supra, 4 Cal.4th at pp. 141-142.)
We explained the history of Proposition 218 in Apartment Assn. of Los Angeles County, Inc. v. City of Los Angeles (2001) 24 Cal.4th 830 [102 Cal.Rptr.2d 719, 14 P.3d 930] (Apartment Assn.): “ ‘Proposition 218 can best be understood against its historical background, which begins in 1978 with the adoption of Proposition 13. “The purpose of Proposition 13 was to cut local property taxes. [Citation.]” [Citation.] Its principal provisions limited ad valorem property taxes to 1 percent of a property’s assessed valuation and limited increases in the assessed valuation to 2 percent per year unless and until the property changed hands. (Cal. Const., art. XIII A, §§ 1, 2.)
“ ‘To prevent local governments from subverting its limitations, Proposition 13 also prohibited counties, cities, and special districts from enacting any special tax without a two-thirds vote of the electorate. (Cal. Const., art. XIII A, § 4; Rider v. County of San Diego (1991) 1 Cal.4th 1, 6-7 [2 Cal.Rptr.2d 490, 820 P.2d 1000].) It has been held, however, that a special assessment is not a special tax within the meaning of Proposition 13. (Knox v. City of Orland[, supra,) 4 Cal.4th [at p.] 141 . . . , and cases cited.) Accordingly, a special assessment could be imposed without a two-thirds vote.
[443]*443“ ‘In November 1996, in part to change this rule, the electorate adopted Proposition 218, which added articles XIII C and XIII D to the California Constitution. Proposition 218 allows only four types of local property taxes: (1) an ad valorem property tax; (2) a special tax; (3) an assessment; and (4) a fee or charge. (Cal. Const., art. XIII D, § 3, subd. (a)(l)-(4); see also [id.], § 2, subd. (a).) It buttresses Proposition 13’s limitations on ad valorem property taxes and special taxes by placing analogous restrictions on assessments, fees, and charges.’ (Howard Jarvis [Taxpayers Assn. v. City of Riverside (1999)] 73 Cal.App.4th 679, 681-682 [86 Cal.Rptr.2d 592].)” (Apartment Assn., supra, 24 Cal.4th at pp. 836-837.)
Proposition 218 restricts government’s ability to impose assessments in several important ways. First, it tightens the definition of the two key findings necessary to support an assessment: special benefit and proportionality. An assessment can be imposed only for a “special benefit” conferred on a particular property. (Art. XIII D, §§ 2, subd. (b), 4, subd. (a).) A special benefit is “a particular and distinct benefit over and above general benefits conferred on real property located in the district or to the public at large.” (Art. XIII D, § 2, subd. (i).) The definition specifically provides that “[general enhancement of property value does not constitute ‘special benefit.’ ” (Ibid.) Further, an assessment on any given parcel must be in proportion to the special benefit conferred on that parcel: “No assessment shall be imposed on any parcel which exceeds the reasonable cost of the proportional special benefit conferred bn that parcel.” (Art. XIII D, § 4, subd. (a).) “The proportionate special benefit derived by each identified parcel shall be determined in relationship to the entirety of the capital cost of a public improvement, the maintenance and operation expenses of a public improvement, or the cost of the property-related service being provided.” (Ibid.) Because only special benefits are assessable, and public improvements often provide both general benefits to the community and special benefits to a particular property, the assessing agency must first “separate the general benefits from the special benefits conferred on a parcel” and impose the assessment only for the special benefits. (Art. XIII D, § 4, subd. (a).)
Second, as described above, Proposition 218 established strict procedural requirements for the imposition of a lawful assessment. (Ante, at p. 438.)
A. Standard of Review
Before Proposition 218 was passed, courts reviewed quasi-legislative acts of local governmental agencies, such as the formation of an assessment district, under a deferential abuse of discretion standard. (Knox, supra, 4 Cal.4th at pp. 145-149; Dawson v. Town of Los Altos Hills (1976) 16 Cal.3d [444]*444676, 684-685 [129 Cal.Rptr. 97, 547 P.2d 1377] (Dawson).) Because it was recognized that “the establishment of a special assessment district takes place as a result of a peculiarly legislative process grounded in the taxing power of the sovereign,” the scope of judicial review of such actions was “quite narrow.” (Dawson, supra, at pp. 683-684; see id. at p. 684 [“ ‘The board of supervisors is the ultimate authority which is empowered to finally determine what lands are benefited and what amount of benefits shall be assessed against the several parcels benefited . . . .’”].)
Accordingly, the standard of review was as follows: “A special assessment finally confirmed by a local legislative body in accordance with applicable law will not be set aside by the courts unless it clearly appears on the face of the record before [the legislative] body, or from facts which may be judicially noticed, that the assessment as finally confirmed is not proportional to the benefits to be bestowed on the properties to be assessed or that no benefits will accrue to such properties.” (Dawson, supra, 16 Cal.3d at p. 685; see also Knox, supra, 4 Cal.4th at p. 146.) Under the Dawson/Knox standard of review, courts presumed an assessment was valid, and a plaintiff challenging it had to show that the record before the legislative body “clearly” did not support the underlying determinations of benefit and proportionality. (See also Lent v. Tillson (1887) 72 Cal. 404, 429 [14 P. 71] [judicial interference is warranted only “when the courts can plainly see that the legislature has not really exercised this judgment at all, or that manifestly and certainly no such benefit can or could reasonably have been expected to result”].)
The drafters of Proposition 218 specifically targeted this deferential standard of review for change. Article XIII D, section 4, subdivision (f), provides: “In any legal action contesting the validity of any assessment, the burden shall be on the agency to demonstrate that the property or properties in question receive a special benefit over and above the benefits conferred on the public at large and that the amount of any contested assessment is proportional to, and no greater than, the benefits conferred on the property or properties in question.”
In determining the effect of article XIII D, section 4, subdivision (f), we apply the familiar principles of constitutional interpretation, the aim of which is to “determine and effectuate the intent of those who enacted the constitutional provision at issue.” (Richmond v. Shasta Community Services Dist. (2004) 32 Cal.4th 409, 418 [9 Cal.Rptr.3d 121, 83 P.3d 518].) “The principles of constitutional interpretation are similar to those governing statutory construction.” (Thompson v. Department of Corrections (2001) 25 Cal.4th 117, 122 [105 Cal.Rptr.2d 46, 18 P.3d 1198].) If the language is clear and unambiguous, the plain meaning governs. (People v. Lopez (2003) 31 Cal.4th. 1051, 1056 [6 Cal.Rptr.3d 432, 79 P.3d 548].) But if the language is [445]*445ambiguous, we consider extrinsic evidence in determining voter intent, including the Legislative Analyst’s analysis and ballot arguments for and against the initiative. (People v. Canty (2004) 32 Cal.4th 1266, 1281 [14 Cal.Rptr.3d 1, 90 P.3d 1168]; People v. Rizo (2000) 22 Cal.4th 681, 685 [94 Cal.Rptr.2d 375, 996 P.2d 27].)
Article XIII D, section 4, subdivision (f), states that the agency has the burden of demonstrating special benefit and proportionality in any legal action contesting the validity of any assessment. Although it is clear that the voters intended to reverse the usual deference accorded governmental action and to reverse the presumption of validity by placing the “burden” on the agency, the provision does not specify the scope of that burden. Because the language imposing a “burden” on the agency is somewhat imprecise, we look to the ballot materials as further indicia of voter intent.
The Legislative Analyst explained to the voters that Proposition 218 was designed to “constrain local governments’ ability to impose . . . assessments . . .” and to “place extensive requirements on local governments charging assessments.” (Ballot Pamp., Gen. Elec., supra, analysis of Prop. 218 by Legis. Analyst, p. 73.) Addressing the burden of demonstration language of proposed article XIII D, section 4, subdivision (f), the Legislative Analyst explained: “Currently, the courts allow local governments significant flexibility in determining fee and assessment amounts. In lawsuits challenging property fees and assessments, the taxpayer generally has the ‘burden of proof’ to show that they are not legal. This measure shifts the burden of proof in these lawsuits to local government. As a result, it would be easier for taxpayers to win lawsuits, resulting in reduced or repealed fees and assessments.” (Ballot Pamp., Gen. Elec., supra, analysis of Prop. 218 by Legis. Analyst, p. 74.) Or stated another way, Proposition 218 was intended to make it more difficult for an assessment to be validated in a court proceeding.
As the dissent below points out, a provision in Proposition 218 shifting the burden of demonstration was included in reaction to our opinion in Knox. The drafters of Proposition 218 were clearly aware of Knox and the deferential standard it applied based on Dawson, supra, 16 Cal.3d 676. The argument in favor of Proposition 218 referred to a “growing list of assessments imposed without voter approval” after Proposition 13 that are in fact special taxes. (Ballot Pamp., Gen. Elec., supra, argument in favor of Prop. 218, p. 76.) As one example of several named abuses of the assessment process, it specified that “[i]n Northern California, taxpayers 27 miles away from a park are assessed because their property supposedly benefits from that park.” (Ibid.) The reference to 27 miles was based on the facts of Knox, which involved an assessment to raise funds to maintain five existing parks serving four school districts. We upheld the assessment, deferring to the City of Orland’s [446]*446determination that the property owners were “uniquely benefited by the proximity of these facilities to their properties” (Knox, supra, 4 Cal.4th at p. 149), although the assessment district contained 42,300 acres of land and geographically consisted of the entire city and portions of outlying areas in Glenn County (id. at p. 137, fn. 5).
Also, in Knox, we declined a request to reevaluate the Dawson deferential standard of review for special assessments, finding “no basis” for requiring the assessing agency to bear the burden of proof “in the context of benefit assessments.” (Knox, supra, 4 Cal.4th at p. 147.) The Knox plaintiffs argued that, as in Beaumont Investors v. Beaumont-Cherry Valley Water Dist. (1985) 165 Cal.App.3d 227, 235 [211 Cal.Rptr. 567], the local agency should bear the burden of proof in establishing the validity of a special assessment, and we should reassess the traditional standard of review that we reaffirmed in Dawson. (Knox, supra, 4 Cal.4th at pp. 146-147.) In rejecting the argument, we distinguished benefit assessments from the development fees in Beaumont, noted the different statutory contexts, and refused to change the deferential standard of review. (Ibid.) Thus, it appears that the inclusion of the burden of demonstration language was intended to supply the “basis” found lacking in Knox, and that the drafters of Proposition 218 particularly targeted Knox.
As further evidence that the voters sought to curtail local agency discretion in raising funds, Proposition 218’s preamble includes an express statement of purpose: “The people of the State of California hereby find and declare that Proposition 13 was intended to provide effective tax relief and to require voter approval of tax increases. However, local governments have subjected taxpayers to excessive tax, assessment, fee and charge increases that not only frustrate the purposes of voter approval for tax increases, but also threaten the economic security of all Californians and the California economy itself. This measure protects taxpayers by limiting the methods by which local governments exact revenue from taxpayers without their consent.” (Ballot Pamp., supra, text of Prop. 218, § 2, p. 108; see Historical Notes, supra, at p. 85; People v. Canty, supra, 32 Cal.4th at p. 1280 [“In considering the purpose of legislation, statements of the intent of the enacting body contained in a preamble, while not conclusive, are entitled to consideration.”].) In passing Proposition 218, the voters clearly sought to limit local government’s ability to exact revenue under the rubric of special assessments.
The Court of Appeal majority below recognized that the voters intended to change the deferential standard of review: “[B]y placing the burden to demonstrate special benefit and proportionality on the agency the new law must now require that which Lent held was not necessary, i.e., that the record contain affirmative evidence of the two substantive bases for the assessment.” [447]*447Nevertheless, the majority maintained that courts should continue to give deference to the local agency’s assessment decision (an act of a legislative body) for two reasons. First, “the constitutional separation of powers demands that we give it deference. (Cal. Const., art. Ill, § 3; [citations].)” Second, if the challenged assessment was levied according to Proposition 218’s procedural requirements, courts will continue to accord the final legislative determination substantial deference. Otherwise, “invalidating an assessment that received the support of a majority of the property owners would frustrate the will of those property owners.” The majority concluded that the scope of judicial review was “limited.”
Accordingly, the majority stated the new standard of review as follows: “A special assessment finally confirmed by a local legislative body in accordance with applicable law will not be set aside by the courts so long as the local legislative body demonstrates, by reference to the face of the record before that body, that the property or properties in question will receive a special benefit over and above the benefits conferred on the public at large and that the amount of any contested assessment is proportional to, and no greater than, the benefits conferred on the property or properties in question. In all other respects, such an assessment shall not be set aside by the courts unless it clearly appears on the face of the record before the legislative body, or from facts which may be judicially noticed, that the assessment constitutes a manifest abuse of discretion.”
Under the majority’s standard, an assessing agency’s determinations regarding whether benefits are special and proportional under the state Constitution must be affirmed if substantial evidence supports them. Although the substantial evidence standard is less deferential than the Dawson/Knox standard of review, it nevertheless is still highly deferential. (Crawford v. Southern Pacific Co. (1935) 3 Cal.2d 427, 429 [45 P.2d 183] [power of appellate court begins and ends with a determination as to whether there is any substantial evidence, contradicted or uncontradicted, to support conclusions below]; Jessup Farms v. Baldwin (1983) 33 Cal.3d 639, 660 [190 Cal.Rptr. 355, 660 P.2d 813] [reviewing court views the evidence in the light most favorable to the prevailing party, giving it the benefit of every reasonable inference and resolving all conflicts in its favor].) The majority’s choice of the deferential substantial evidence standard comported with its emphasis on the constitutional separation of powers doctrine, the legislative character of the assessment determinations at issue, and the consent of the weighted majority of property owners in the district.
However, a valid assessment under Proposition 218 must not only be approved by a weighted majority of owners under the procedural requirements in article XIII D, section 4, subdivisions (c), (d), and (e), but must also [448]*448satisfy the substantive requirements in section 4, subdivision (a). (Art. XIIID, § 4, subds. (a), (c)-(e).) These substantive requirements are contained in constitutional provisions of dignity at least equal to the constitutional separation of powers provision. (Cal. Const., art. Ill, § 3.) Before Proposition 218 became law, special assessment laws were generally statutory, and the constitutional separation of powers doctrine served as a foundation for a more deferential standard of review by the courts. But after Proposition 218 passed, an assessment’s validity, including the substantive requirements, is now a constitutional question. “There is a clear limitation, however, upon the power of the Legislature to regulate the exercise of a constitutional right.” (Hale v. Bohannon (1952) 38 Cal.2d 458, 471 [241 P.2d 4].) “ ‘[A]ll such legislation must be subordinate to the constitutional provision, and in furtherance of its purpose, and must not in any particular attempt to narrow or embarrass it.’ ” (Ibid.) Thus, a local agency acting in a legislative capacity has no authority to exercise its discretion in a way that violates constitutional provisions or undermines their effect.
We “ ' “must enforce the provisions of our Constitution and ‘may not lightly disregard or blink at... a clear constitutional mandate.’ ” ’ ” (State Personnel Bd. v. Department of Personnel Admin. (2005) 37 Cal.4th 512, 523 [36 Cal.Rptr.3d 142, 123 P.3d 169].) In so doing, we are obligated to construe constitutional amendments in a manner that effectuates the voters’ purpose in adopting the law. (Howard Jarvis Taxpayers Assn. v. City of Satinas (2002) 98 Cal.App.4th 1351, 1355 [121 Cal.Rptr.2d 228].)
Proposition 218 specifically states that “[t]he provisions of this act shall be liberally construed to effectuate its purposes of limiting local government revenue and enhancing taxpayer consent.” (Ballot Pamp., supra, text of Prop. 218, § 5, p. 109; see Historical Notes, supra, at p. 85.) Also, as discussed above, the ballot materials explained to the voters that Proposition 218 was designed to: constrain local governments’ ability to impose assessments; place extensive requirements on local governments charging assessments; shift the burden of demonstrating assessments’ legality to local government; make it easier for taxpayers to win lawsuits; and limit the methods by which local governments exact revenue from taxpayers without their consent. Because Proposition 218’s underlying purpose was to limit government’s power to exact revenue and to curtail the deference that had been traditionally accorded legislative enactments on fees, assessments, and charges, a more rigorous standard of review is warranted. We construe article XIII D, section 4, subdivision (f)—the “burden ... to demonstrate” provision—liberally in light of the proposition’s other provisions, and conclude that courts should exercise their independent judgment in reviewing local agency decisions that have determined whether benefits are special and whether assessments are proportional to special benefits within the meaning of Proposition 218. (Redevelopment Agency v. County of Los Angeles (1999) [449]*44975 Cal.App.4th 68, 74 [89 Cal.Rptr.2d 10] [courts exercise independent judgment in matters involving constitutional interpretation]; see People v. Cromer (2001) 24 Cal.4th 889, 894 [103 Cal.Rptr.2d 23, 15 P.3d 243] [courts use independent, de novo review for mixed questions of fact and law that implicate constitutional rights].)
Defendants argue that because a weighted majority of property owners approved the assessment, it furthers Proposition 218’s emphasis on voter consent, and we should accord deference to those voting owners’ wishes. However, voter consent cannot convert an unconstitutional legislative assessment into a constitutional one. Under Proposition 218, all valid assessments must both clear the substantive hurdles in article XIII D, section 4, subdivision (a) and be approved by a weighted majority of owners under section 4, subdivisions (c), (d), and (e). Moreover, Proposition 218 was designed to prevent a local legislative body from imposing a special tax disguised as an assessment. (Apartment Assn., supra, 24 Cal.4th at p. 839 [“The ballot arguments identify what was perhaps the drafter’s main concern: tax increases disguised via euphemistic relabeling as ‘fees,’ ‘charges,’ or ‘assessments.’ ”].)5 The judicial invalidation of an assessment does not thwart the objective of taxpayer consent; under Proposition 13, two-thirds of the voters must still approve the proposed revenue source (i.e., a special tax). (Cal. Const., art. XIII A, § 4; art. XIII D, § 3, subd. (a)(2).) Neither the separation of powers nor property owner consent justifies allowing a local legislative body or property owners (both bound by the state Constitution) to usurp the judicial function of interpreting and applying the constitutional provisions that now govern assessments.
Courts are familiar with the process of determining the constitutionality of the taxes, fees, and assessments that local governments impose. (See Richmond v. Shasta Community Services Dist., supra, 32 Cal.4th at pp. 418-428 [determination whether charge that water district imposed violated art. XIII D restrictions required de novo review]; Howard Jarvis Taxpayers Assn. v. City of Roseville (2002) 97 Cal.App.4th 637, 647-650 [119 Cal.Rptr.2d 91] [court found in-lieu fee that city imposed was unconstitutional under art. XIII D]; Howard Jarvis Taxpayers Assn. v. City of Riverside, supra, 73 Cal.App.4th 679, 684-690 [question whether existing streetlight assessment was subject to Prop. 218 limitations involved court’s de novo interpretation of the Constitution and voters’ intent]; Howard Jarvis [450]*450Taxpayers Assn. v. City of Salinas, supra, 98 Cal.App.4th at pp. 1354-1359 [court independently interprets constitutional amendments contained in art. Xni D to determine whether water fee was a property-related fee requiring property owners’ vote]; Graber v. City of Upland (2002) 99 Cal.App.4th 424, 429 [121 Cal.Rptr.2d 649] [question whether local ordinance violated constitutional provisions relating to tax increment financing was subject to de novo review].)
Accordingly, courts should exercise their independent judgment in reviewing whether assessments that local agencies impose violate article XIII D.6
B. The 2001 Special Assessment
We apply this standard of review to the special assessment in this case to determine whether OSA met its burden of demonstrating that the assessed properties received a special benefit and that the assessment is proportional to that special benefit.
1. Special Benefits
“Under Proposition 218, only special benefits are assessable. (Cal. Const., art. XIIID, § 4, subd. (a).) Local governments may not impose assessments to pay for the cost of providing a general benefit to the community . . . .” (City of Saratoga v. Hinz (2004) 115 Cal.App.4th 1202, 1223 [9 Cal.Rptr.3d 791].) If a proposed project will provide both general benefits to the community and special benefits to particular properties, the agency can impose an assessment based only on the special benefits. It must separate the general benefits from the special benefits and must secure other funding for the general benefits. (Art. XIII D, § 4, subd. (a); Hinz, supra, 115 Cal.App.4th at p. 1223.)
Both before and after Proposition 218 passed, special assessments were distinguished from special taxes through the concept of special benefits. (Knox, supra, 4 Cal.4th at p. 142; Ventura Group Ventures, Inc. v. Ventura Port Dist. (2001) 24 Cal.4th 1089, 1106 [104 Cal.Rptr.2d 53, 16 P.3d 717] (Ventura Group Ventures).) In Knox, we referred to a special benefit as a benefit “ ‘over and above that received by the general public.’ ” (Knox, supra, 4 Cal.4th at p. 142.) There, we presumed (in the absence of evidence to the [451]*451contrary) that the presence of well-maintained open parkland contributed to the district’s attractiveness and thus was a special benefit because it enhanced the desirability of the residential properties in that district. (Knox, supra, 4 Cal.4th at p. 149.)
Proposition 218 made several changes to the definition of special benefits. First, Proposition 218 defines a special benefit as “a particular and distinct benefit over and above general benefits conferred on real property located in the district or to the public at large,” with the additional italicized requirement. (Art. XIII D, § 2, subd. (i), italics added.) Correspondingly, it emphasizes that “[g]eneral enhancement of property value does not constitute ‘special benefit.’ ” {Ibid.) Since the “[g]eneral enhancement of property value” is a “general benefit[] conferred on real property located in the district” {ibid.), Proposition 218 clearly mandates that a special benefit cannot be synonymous with general enhancement of property value. Thus, Proposition 218 tightened the definition of special benefits and broadened the definition of general benefits to include benefits conferred generally “on real property located in the district.” (Art. XIII D, § 2, subd. (i).)7
Relying on Harrison v. Board of Supervisors (1975) 44 Cal.App.3d 852 [118 Cal.Rptr. 828] (Harrison), the Court of Appeal majority below commented that “[i]f there is a significant difference between the two definitions [of special benefits before and after Proposition 218], we do not detect it.” Harrison simply held that an increase in property value alone did not amount to a special benefit. (Harrison, supra, 44 Cal.App.3d at pp. 858-859.) This holding did not preclude a determination of special benefit based in part on the general enhancement of property value.
Moreover, while pre-Proposition 218 case law makes clear that assessments may not be levied for purposes of conferring purely general benefits, courts did not invalidate assessments simply because they provided general benefits to the public in addition to the requisite special benefits, and did not demand a strict separation of special and general benefits. (See, e.g., Knox, [452]*452supra, 4 Cal.4th at pp. 137, 149 [upheld validity of assessment for park maintenance despite fact city did not separate general benefits to people outside area and to community at large from special benefits to residential parcels]; Allen v. City of Los Angeles (1930) 210 Cal. 235, 238 [291 P. 393] [“It would be well within the power of the city council to make the cost of the entire proceeding rest upon the shoulders of the property owners of a given district especially benefited thereby.”]; Federal Construction Co. v. Ensign (1922) 59 Cal.App. 200, 210 [210 P. 536] (Ensign) [“To invalidate the assessment the general public benefit must be the only result of the improvement”; 100 percent of cost of new sewage treatment plant fully assessable notwithstanding general benefits]; 51 Cal.Jur.3d (2003) Public Improvements, § 19, p. 900 [“For an assessment to be invalid because it confers a general public benefit, the general benefit must be the only result of the assessment.”].)
Consequently, the pre-Proposition 218 cases on which the Court of Appeal majority below and OSA relied are not instructive in determining whether a benefit is special under Proposition 218. Instead, under the plain language of article XIII D, a special benefit must affect the assessed property in a way that is particular and distinct from its effect on other parcels and that real property in general and the public at large do not share.8 (Art. XIIID, § 2, subd. (i).)
Our examination of the engineer’s report supporting the assessments reveals that OSA has failed to meet its burden of demonstrating that the assessment is based only on the special benefits conferred on the particular parcel and is in proportion to those benefits. Various studies supported the listed benefits in the engineer’s report. But, as discussed below, the report’s [453]*453designation of these listed benefits as “special” failed to satisfy the constitutional requirements for assessments that fund open space acquisitions.
The engineer’s report enumerates seven “special benefits” that the assessment will confer on all residents and property owners in the district: (1) enhanced recreational activities and expanded access to recreational areas; (2) protection of views, scenery, and other resources; (3) increased economic activity; (4) expanded employment opportunity; (5) reduced costs of law enforcement, health care, fire prevention, and natural disaster response; (6) enhanced quality of life and desirability of the area; and (7) improved water quality, pollution reduction, and flood prevention.
The report states that the benefit of “[e]nhanced recreational opportunities and expanded access to recreational areas” will be conferred on “all property owners, residents, employees and customers throughout the OSA” and that “[a]ll properties will benefit from the assessments . . . .” It explains that residential properties will benefit because “[t]hese improved open space areas will be available to residents and guests of property owners within the OSA, thereby making these properties more valuable,” and that nonresidential properties will benefit because additional recreation areas available to employees will “enhance an employer’s ability to attract and keep quality employees.” The “enhanced economic conditions benefit the [nonresidential] property by making it more valuable.” The report therefore acknowledges that all people in OSA’s territory will benefit broadly, generally, and directly from the assessment, resulting in all properties receiving a derivative, indirect benefit.
Similarly, the report describes the second listed “special benefit” as benefiting everyone in the district generally (“[protection of views, scenery and other resources values and environmental benefits enjoyed by residents, employees, customers and guests”). The report concludes that “[t]hese benefits ultimately accrue to properties because properties are more desirable in areas that offer environmental and economic benefits.” The report makes no attempt to tie this benefit to particular properties. Instead, it concludes that all properties throughout the district will receive this benefit equally.
“Increased economic activity” and “[e]xpanded employment opportunity” are also listed in the report as “special benefits.” Again, the report states that increased economic activity and expanded employment opportunity will result from the acquisition of additional open space because increased recreational opportunities will likely attract more people to the county. These people, in turn, will patronize county services and businesses, thereby fostering economic growth and “additional employment opportunities for OSA residents.” The report broadly concludes that the increased economic [454]*454activity in the area is “a benefit ultimately to residential, commercial, industrial and institutional property.” However, it simply assumes that the resultant increased economic activity will affect people and property throughout the county equally, but makes no direct connection to any particular properties.
The remaining listed “special benefits” do not satisfy the constitutional requirements either. Relying on various studies, the report claims that because open space and parks promote good health and reduce crime and vandalism, the county can expect a reduction in health care and law enforcement costs. It reasons that “[s]uch cost reduction frees public funds for other services that benefit properties,” and “[a]ll of these factors ultimately benefit property by making the community more desirable and property, in turn, more valuable.” The report also asserts that, because open space helps protect water quality and reduce flooding, the costs of public utility services for properties in the district will decrease.
Finally, the report emphasizes that open space areas will “enhance the overall quality of life and desirability of the area.” All the listed benefits are general benefits in this case, shared by everyone—all 1.2 million people— living within the district. The report does not even attempt to measure the benefits that accrue to particular parcels. Indeed, the report describes OSA’s mission, which is “[t]o preserve, protect and manage, for the use and enjoyment of all people, a well-balanced system of urban and non-urban areas of scenic recreational and agricultural importance.” (Italics added.) OSA is responsible, as the report explains, “for preserving and maintaining open space for approximately 1.2 million people residing within its boundaries, representing over two-thirds of the population within Santa Clara County.” Although it is reasonable to conclude that quality-of-life benefits to people living in, working in, and patronizing businesses in the district will, in turn, benefit property in the district, such derivative benefits are only “general benefits conferred on real property located in the district or to the public at large.” (Art. XIII D, § 2, subd. (i).) Moreover, to the extent that the value of property located in a desirable community is enhanced, this is a “[gjeneral enhancement of property value,” and is thus, by definition not a special benefit. (Ibid.)
In addition, the report’s description of general benefits fails to comport with the Constitution. The engineer’s report acknowledges that the acquisition, maintenance and preservation of open spaces “provide a degree of general benefit to the public at large.” But it then asserts that the ratio of general to special benefit that will be derived from OSA’s open space acquisition program will be 10 percent general benefit and 90 percent special benefit, based on its determination that general benefit is measured only as [455]*455the benefit conferred on “individuals who are not residents, employees, customers or property owners” (italics added) in the assessment district. This distinction finds no support in the Constitution.
Under article XIII D, general benefits are not restricted to benefits conferred only on persons and property outside the assessment district, but can include benefits both “conferred on real property located in the district or to the public at large.” (Art. XIII D, § 2, subd. (i), italics added.) “At large” means “[n]ot limited to any particular . . . person” or “[fjully; in detail; in an extended form.” (Black’s Law Dict. (8th ed. 2004) p. 136.) By its plain language, section 2, subdivision (i), does not permit OSA to choose one segment of the “public at large” to measure general benefit. The “public at large” thus means all members of the public—including those who live, work, and shop within the district—and not simply transient visitors. The report assumes that people and property within the district—an area covering over 800 square miles, with a population of approximately 1.2 million people—will receive no general benefit at all, only special benefits, from OSA’s acquisition of open space. But under these circumstances, “[i]f everything is special, then nothing is special.” (Ventura Group Ventures, supra, 24 Cal.4th at p. 1107.)
Further, we note the validity of this assessment would be questionable even under the pre-Proposition 218 cases on which OSA relies. (See, e.g., Knox, supra, 4 Cal.4th 132 [assessment valid for maintenance of five existing parks in four school districts in city]; City of San Diego v. Holodnak (1984) 157 Cal.App.3d 759 [203 Cal.Rptr. 797] [assessment valid to fund parks and other public facilities located in new development]; Ensign, supra, 59 Cal.App. 200 [assessment valid to fund new sewer system].) Unlike the assessment here, the assessments in the pre-Proposition 218 cases involved specific, identified improvements that directly benefited each assessed property and whose costs could be determined or estimated and then allocated to the properties assessed. Also, in Knox and Holodnak, the properties assessed received special benefits from the particular park because of their proximity to park facilities. (Knox, supra, 4 Cal.4th at p. 149; Holodnak, supra, 157 Cal.App.3d at p. 763.)
Here, with a district of 314,000 parcels, OSA shows no distinct benefits to particular properties above those which the general public using and enjoying the open space receives. The special benefits, if any, that may arise would likely result from factors such as proximity, expanded or improved access to the open space, or views of the open space. (See Ensign, supra, 59 Cal.App. at p. 217 [property which is specially benefited is “ ‘real property adjoining, or near the locality of the improvement’ ”].) But, because OSA has not identified any specific open space acquisition or planned acquisition, it cannot show any [456]*456specific benefits to assessed parcels through their direct relationship to the “locality of the improvement.” The improvement is only to OSA’s budget for open space acquisitions.
Based on the undisputed facts in OSA’s record (the engineer’s report), OSA has failed to demonstrate that the properties in the assessment district receive a particular and distinct special benefit not shared by the district’s property in general or by the public at large within the meaning of Proposition 218.
2. Proportionality
For an assessment to be valid, the properties must be assessed in proportion to the special benefits received: “No assessment shall be imposed on any parcel which exceeds the reasonable cost of the proportional special benefit conferred on that parcel.” (Art. XIII D, § 4, subd. (a).) “The proportionate special benefit derived by each identified parcel shall be determined in relationship to the entirety of the capital cost of a public improvement, the maintenance and operation expenses of a public improvement, or the cost of the property-related service being provided.” (Ibid., italics added.) Capital cost is defined as “the cost of acquisition, installation, construction, reconstruction, or replacement of a permanent public improvement by an agency.” (Art. XIII D, § 2, subd. (c), italics added.)
To satisfy the proportionality requirement, the engineer’s report assigned all single-family homes in the district one single-family equivalent (SFE) unit and assigned other types of property greater or lesser SFE’s, depending on the estimated number of people using those properties. Condominiums received a lesser SFE because the average number of people per unit was estimated to be fewer than in an average single-family residence. Commercial properties received a higher SFE than single-family residences because greater numbers of people use them. Each SFE corresponded to an annual assessment of $20, an amount a majority of property owners surveyed would be willing to pay.
Because all single-family homes were assessed the same $20 amount, the engineer’s report assumed that all single-family homes throughout the 800-square-mile district would receive an equal special benefit, regardless of their proximity to open space areas that might be acquired at some time in the future. The report contains no detailed analysis on how specific properties, blocks, school districts, or even cities would benefit from their proximity to open space. OSA contends that its assessment is nonetheless valid because it plans to acquire space equally throughout the district, and all properties will be equally close to and benefit from open space areas. The engineer’s report [457]*457lists 30 priority acquisition areas and identifies a number of other “potential acquisition and improvement areas.” This, OSA claims, is sufficient to satisfy Proposition 218’s proportionality requirement. We disagree.
The report’s proportionality analysis fails to satisfy Proposition 218 largely because the special assessment is based on OSA’s projected annual budget of $8 million for its open space program rather than on a calculation or estimation of the cost of the particular public improvement to be financed by the assessment. The figure of $8 million was derived from the additional $20 per year in property taxes multiplied by the number of properties on the tax rolls in the district. The $8 million collected for the assessment annually—with an automatic cost-of-living increase—provides a continuing source of revenue for OSA’s budget. However, the purpose of an assessment is to require the properties which have received a special benefit from a “public improvement” “to pay the cost of that improvement,” and not to fund an agency’s ongoing budget. (Ventura Group Ventures, supra, 24 Cal.4th at p. 1106, italics added; see Knox, supra, 4 Cal.4th at p. 142.)
The engineer’s report generally describes a program to acquire various properties throughout the county, as well as to provide maintenance and servicing of these public areas. Such future acquisitions include, but are not limited to, “greenbelts, hillsides, viewsheds and watersheds, baylands, riparian corridors, urban open space, parklands, agricultural lands, development rights on agricultural lands and other land-use types, conservation easements, other property rights, wetlands, utility right-of-ways, surplus school sites, [and] quarries.” OSA argues its goal is to acquire open space land that is evenly distributed throughout the district. Although the report lists 30 general priority acquisition areas, it further notes this list is not exclusive. The report identifies no particular parcels or specific areas within the district that OSA plans to acquire for open space or parks. Further, the engineer’s report notes that OSA “should” complete at least one acquisition of open land every five years. Notably, OSA is not required to do so.
Thus, the report fails to identify with sufficient specificity the “permanent public improvement” that the assessment will finance, fails to estimate or calculate the cost of any such improvement, and fails to directly connect any proportionate costs of and benefits received from the “permanent public improvement” to the specific assessed properties. As the dissent below observed, “an assessment calculation that works backward by starting with an amount taxpayers are likely to pay, and then determines an annual spending budget based thereon, does not comply with the law governing assessments, either before or after Proposition 218.”
As with its determination of special benefits, OSA has failed to demonstrate proportionality. Accordingly, we conclude that the assessment is invalid [458]*458for failing to meet the requirements of Proposition 218. In light of this disposition, we need not reach the other arguments plaintiffs raise.
III. Disposition
We reverse the judgment of the Court of Appeal and remand the matter to that court for further proceedings consistent with our opinion.
George, C. J., Kennard, J., Baxter, J., Werdegar, J., Moreno, J., and Corrigan, J., concurred.