The Safe Drinking Water and Toxic Enforcement Act of 1986 (Health & Saf. Code, §§ 25249.5-25249.13),1 adopted by the people at the November 4, 1986, General Election as Proposition 65 (hereinafter referred to as the Act or Proposition 65), provides in material part that “[n]o person in the course of doing business shall knowingly discharge or release a chemical known to the state to cause cancer or reproductive toxicity into water or onto or into land where such chemical passes or probably will pass into any source of drinking water . . . .” (§ 25249.5.) This case requires us to define what is meant by the phrase “source of drinking water.” The Attorney General, who brought this action to enforce the Act, contends that the phrase includes the water that is stored in or run through water faucets, and so defendant faucet manufacturers, whose products allegedly leach toxic chemicals into drinking water, may be sued for violations of the Act. The faucet manufacturers contend the contrary.
I. Statement of Facts and Procedural History
The Attorney General filed an action in superior court seeking injunctive relief and civil penalties against defendants, 16 manufacturers and distributors of drinking water faucets sold and used in California. The Attorney General alleged that the faucets they manufactured contained significant quantities of lead, which is among the substances listed by the state as known to cause cancer and reproductive toxicity within the meaning of the Act. (See § 25249.8; Cal. Code Regs., tit. 22, § 12000, subds. (b), (c).) The Attorney General further alleged that “[w]hen residential tapwater is stored in defendants’ faucets, lead is leached from the inside surface of the faucets into the drinking water. When tapwater is drawn from defendants’ faucets, lead is ingested by any person who ingests drinking water . . . .” He also alleged that “[T]he fact that lead leaches from faucets into tapwater has been known for a number of years.” Defendants have thereby violated section 25249.5 according to the Attorney General.
Defendants demurred to the first and second causes of action,3 arguing that faucet water was not a “source of drinking water” within the meaning of the statute and that the Attorney General had therefore failed to allege facts constituting a cause of action. The trial court sustained the demurrer to the first two causes of action, determining that “residential water faucets and the water within them are not ‘sources of drinking water.’ ” Although the trial court sustained the demurrer with leave to amend, the Attorney General concluded it was not possible to amend the complaint within the constraints of the trial court’s order. He instead sought a writ of mandate from the Court *300of Appeal. The Court of Appeal denied relief, concluding, for reasons discussed at greater length below, that the language of the statute could not support the Attorney General’s view that the water within household faucets is a “source of drinking water.” It further reasoned that, inasmuch as that term was ambiguous, the ambiguity should be resolved in favor of the defendants, since violators of the Act may be subject to civil penalties of up to $2,500 per day for each violation (see § 25249.7, subd. (b)), and “penal” statutes are traditionally construed strictly in favor of the defendant. We granted review to determine the proper construction of the phrase “source of drinking water.”4
In the time since we granted review, the Attorney General has entered into a settlement agreement, which provided for the dismissal of all defendants. Most of these settlements provide, among other things, that defendants will eventually sell in California only those faucets that leach lead below maximum levels set pursuant to the Act, and that compliance will be phased in over specified time periods. In the interim, warnings to consumers will be provided for those faucets that do not comply with the Act. The settlement with defendants Kohler and Sterling Plumbing Inc. (hereafter collectively Kohler), however, provides that Kohler will reduce lead content to levels that comply with the Act in 95 percent of its kitchen faucets by December 31, 1999, but that the companies will achieve 100 percent compliance if we decide this case in the Attorney General’s favor.
After conclusion of the settlement agreement Kohler filed a motion to dismiss, claiming that the issue was moot and without further legal importance. But as Kohler conceded, there are an undetermined number of other faucet manufacturers and plumbing suppliers which may yet be affected by our decision. Because of the continuing legal importance of the questions on which we granted review, as well as the contingent nature of some of the settlements entered into by the parties, we denied the motion to dismiss. (See State of Cal. ex rel. State Lands Com. v. Superior Court (1995) 11 Cal.4th 50, 60-62 [44 Cal.Rptr.2d 399, 900 P.2d 648].)
II. Does “Source of Drinking Water” Include Faucet Water?
“Our only task in reviewing a ruling on a demurrer is to determine whether the complaint states a cause of action. Accordingly we assume that the complaint’s properly pleaded material allegations are true and give the complaint a reasonable interpretation by reading it as a whole and all its parts in their context. [Citations.] We do not, however, assume the truth of *301contentions, deductions, or conclusions of fact or law.” (Moore v. Regents of University of California (1990) 51 Cal.3d 120, 125 [271 Cal.Rptr. 146, 793 P.2d 479, 16 A.L.R.5th 903].) In the present case there is no dispute that lead is a carcinogenic and reproductively toxic chemical within the meaning of the Act. (See Cal. Code Regs., tit. 22, § 12000, subds. (b), (c).) The sole question the parties raise is the proper construction of section 25249.5, specifically the meaning of the term “source of drinking water,” and whether it includes household faucets and faucet water.
We begin our inquiry into the meaning of the phrase “source of drinking water” with an examination of the language of the Act itself. “Absent ambiguity, we presume that the voters intend the meaning apparent on the face of an initiative measure [citation] and the court may not add to the statute or rewrite it to conform to an assumed intent that is not apparent in its language.” (Lesher Communications, Inc. v. City of Walnut Creek (1990) 52 Cal.3d 531, 543 [277 Cal.Rptr. 1, 802 P.2d 317].) Of course, in construing the statute, “[t]he words . . . must be read in context, considering the nature and purpose of the statutory enactment.” (Nahrstedt v. Lakeside Village Condominium Assn. (1994) 8 Cal.4th 361, 378-379 [33 Cal.Rptr.2d 63, 878 P.2d 1275].)
The Act defines “source of drinking water” at section 25249.11, subdivision (d), as follows: “ ‘Source of drinking water’ means either a present source of drinking water or water which is identified or designated in a water quality control plan adopted by a regional board as being suitable for domestic or municipal uses.” Though this language does not fully resolve the issue before us, it does do much to illuminate the meaning of the phrase.
The latter part of the definition is reasonably clear. It refers to regional water quality control plans, which are required under the Porter-Cologne Water Quality Control Act. (Wat. Code, § 13000 et seq. (Porter-Cologne Act).) Pursuant to the Porter-Cologne Act, the state is divided into nine regions. (Id., § 13200.) Each region has a regional water quality control board responsible for formulating and implementing a plan to promote the quality of the bodies of water within its jurisdiction. (Id., § 13241.) As part of the regional water quality control plans, the boards must designate the various “beneficial uses” of each body of water, including “domestic, municipal, agricultural, and industrial supply; power generation; recreation; aesthetic enjoyment; navigation; and preservation and enhancement of fish, wildlife, and other aquatic resources or preserves.” (Id., § 13050, subd. (f); id., subd. (j).) The regional water quality control board’s jurisdiction encompasses both natural lakes, rivers and creeks and other bodies of water, as well as artificially created bodies such as reservoirs, canals, and dams. (See *302 id. § 13050, subd. (e) [“waters of the state” subject to water quality control plan “means any surface water or groundwater within the boundaries of the state”]; see also, e.g., California Water Quality Control Board, Central Valley Region, Central Regional Water Quality Control Plan (Basin Plan) for the Sacramento River Basin and the San Joaquin River Basin (3d ed. 1994) table II-1 (hereafter Central Regional Water Quality Control Plan). According to the plain language of section 25249.11, subdivision (d), therefore, waters of all types that are considered “suitable” for “domestic” or “municipal” use in these regional water quality control plans are “sources of drinking water” within the meaning of the Act. It is undisputed that regional water quality control districts do not regulate faucet water and that faucet water is not found in a water quality control plan.
Section 25249.11, subdivision (d), does not further define the first part of its definition, the term “present source of drinking water,” which is also considered a “source of drinking water” within the meaning of the Act. But two reasonable inferences can be made about the meaning of that phrase. First, since it is used in the alternative with “water which is identified . . . in a water quality control plan ... as being suitable for domestic or municipal uses,” it is presumably something other than the water sources referred to in this latter part of the definition. (Ibid.) Statutes, whether enacted by the people or the Legislature, will be construed so as to eliminate surplusage. (See Delaney v. Superior Court (1990) 50 Cal.3d 785, 798-799 [268 Cal.Rptr. 753, 789 P.2d 934].) Second, because the term is not further defined, it can be assumed to refer not to any special term of art, but rather to a meaning that would be commonly understood by the electorate.
Turning to the dictionary, we observe that the word “source” has several related but distinct meanings. One such meaning is “the point of origin of a stream of water.” (Webster’s Third New Intemat. Diet., supra, p. 2177.) The dictionary uses by way of example the following: “followed it from its source high in the mountains to its calmer reaches.” (Ibid.) This is the meaning of “source” that defendants assert is intended by the Act, that is, the ultimate origin of drinking water. At first blush, that assertion may appear credible, because this definition refers specifically to a source of water. On further reflection, however, this definition cannot be the sole and exclusive meaning of that term because, as even defendants concede, the term “source of drinking water” includes all waters suitable for municipal or domestic use in a regional water quality control plan, including reservoirs, channels, and bodies of water in urban areas. Accordingly, the term “source of drinking water” cannot be taken to mean the ultimate origins of drinking water in this narrow sense.
Another definition of “source” is “a point of origin or procurement.” (Webster’s Third New Intemat. Diet., supra, p. 2177.) Yet another possible *303meaning of “source” is “point of emanation,” as in “it is desirable to have the light source accurately located.” (Ibid.) Faucets and faucet water can reasonably be understood to be a “source” of drinking water in this sense, i.e., the point of procurement or emanation of drinking water. The Attorney General argues that these meanings are also included within the term “source of drinking water.”
Defendants insist, and the Court of Appeal held, that the electorate could not have understood the phrase in the manner urged by the Attorney General, and that it must have signified to them a more remote source. But they advance no persuasive reason why the above quoted alternate definitions of the word “source” could not have been incorporated into the Act. Indeed, the very fact that the phrase “remote source” is not a redundancy, nor “immediate source” an oxymoron, calls defendants’ position into question. The comprehensiveness of the definition of “source of drinking water” in section 25249.11, subdivision (d), which covers all “present” sources, as well as waters “suitable” as drinking water sources, suggests that the drafters of the Act, and the electorate that adopted it, had both immediate and remote sources in mind.
Defendants contend, and the Court of Appeal concluded, that the Attorney General’s interpretation of “source of drinking water” would excise the word “source” from the Act, because faucet water is not a source of drinking water but rather is simply drinking water. On the contrary, the Attorney General quite reasonably argues that a “source of drinking water” is any water that is part of the water supply and delivery system prior to coming out of the tap, from the mountain stream to the faucet, whereas “drinking water” is water after it comes out of the tap and is in the consumer’s possession. Thus, as the Attorney General concedes, the leaching of toxic chemicals into water once it has been drawn from the tap, as from glasses or cups, would not be considered a discharge into a “source of drinking water” under section 25249.S.5
Defendants also contend that clues as to the meaning of “source of drinking water” can be derived from use of the term “release” in section 25249.5, because that term “has been consistently defined to refer to events *304in the outside environment.” The statutes they cite do not support their position. For example, section 25501, subdivision (r), the hazardous materials law, defines “release” as any “spilling, leaking, pumping, pouring, emitting, emptying, discharging, injecting, escaping, leaching, dumping, or disposing into the environment.” (Id., subd. (r), italics added.) Section 25249.5, on the other hand, pertains to the “release,” including presumably “leaching,” of toxic chemicals into water or into or onto land when such chemical passes or probably will pass into a “source of drinking water,” rather than into the “environment.” Section 25501, subdivision (r), in other words, does not support the assertion that the term “release” limits the meaning of “source of drinking water” in the manner advocated by defendants; it does underscore that “release” includes “leaching.”
Defendants argue as well that the fact the statute specifies a prohibition of release of toxic chemicals “into water or into or onto land where such chemical passes or probably will pass into a source of drinking water” must exclude faucet water, since the toxic contamination of faucet water comes directly from lead pipes, not from water or land, and the statute would thereby be rendered irrational. Yet it is evident that the phrase “into water” in the first part of section 25249.5 refers both to drinking water sources themselves and to “nonsource” water that will convey toxic chemicals into a source of drinking water. Thus, for example, a discharge directly into a drinking water reservoir—clearly a drinking water source—would be considered a discharge “into water . . . where such chemical passes . . . into a source of drinking water.” (Ibid.) So too, there is nothing illogical about reading the statute to conclude that lead pipe contamination of faucet water is also a prohibited release of a toxic chemical directly into a “source of drinking water.”
The Attorney General and defendants also disagree about whether defendants’ interpretation of the phrase “present source of drinking water” creates surplusage. As noted, the phrase “present source of drinking water” must be interpreted to refer to something other than the “water . . . designated in a water quality control plan ... as being suitable for domestic or municipal uses.” (§ 25249.11, subd. (d).) Defendants argue that the distinction between the two prongs of the definitions “is only temporal—the first addresses those bodies of water that are presently designated as sources [of] drinking water and the second addresses those bodies of water that may in the future be designated as sources of drinking water.” Thus, “present source of drinking water” refers, according to defendants, “to those bodies of water, both natural and man-made, that are a current source of public, domestic and municipal water, e.g., Hetch Hetchy and the American River.”
The Attorney General argues, on the other hand, that waters designated as being suitable for “domestic” or “municipal” uses in a regional water quality *305control plan include both existing and potential sources. (See, e.g., Central Valley Region Water Quality Control Plan, supra, table II-1 [designating both existing and potential domestic and municipal uses].) He asserts therefore that under defendants' construction the phrase “present source of drinking water” would be wholly surplusage, because all the bodies of water they proposed for this category are designated in a water quality control plan as being suitable for domestic and municipal use. He contends accordingly that the definition of the phrase “present source of drinking water” is not confined to bodies of water that appear in water quality control plans.
We find the Attorney General’s interpretation of section 25249.11, subdivision (d) to be more plausible. If defendants were correct, then it would be expected that the statute would refer to water “designated in a water quality control plan” in both parts of its definition of “source of drinking water.” In other words, the statute would read something like: “source of drinking water is all water designated in a water quality control plan ... as being suitable for domestic or municipal use, whether or not presently so used.” But the fact that section 25249.11, subdivision (d) refers to water quality control plans only in the second part of its definition implies that the first part of the definition—present source of drinking water—was not intended to refer exclusively to bodies of water included in water quality control plans.
The fact that the Attorney General’s interpretation of the phrase “source of drinking water” is consistent with a plausible reading of section 25249.11, subdivision (d), while defendants’ construction is not, does not necessarily mean that the Attorney General’s interpretation of the critical phrase is correct. Yet this fact weighs undeniably in his favor. Moreover, when we consider the purpose of the Act and the consequence of adopting each of the two proffered interpretations, we conclude that the Attorney General offers the only reasonable construction of “source of drinking water.”
“ ‘[W]here the language of a statutory provision is susceptible of two constructions, one of which, in application, will render it reasonable, fair and harmonious with its manifest purpose, and another which would be productive of absurd consequences, the former construction will be adopted.’ [Citation.] . . . Stated differently, ‘Where uncertainty exists consideration should be given to the consequences that will flow from a particular interpretation.’ [Citation.] A court should not adopt a statutory construction that will lead to results contrary to the Legislature’s apparent purpose.” (Western Oil & Gas Assn. v. Monterey Bay Unified Air Pollution Control Dist. (1989) 49 Cal.3d 408, 425 [261 Cal.Rptr. 384, 111 P.2d 157] (Western Oil & Gas Assn.).) At issue in Western Oil & Gas Assn, was whether the Tanner Act, *306which provides a comprehensive statutory scheme for the identification and control of air pollutants on a statewide level by the State Air Resources Board, prohibited local air quality control boards from regulating emissions before they have been identified under that Act’s extensive administrative procedures. In concluding that no such prohibition should be implied, we looked to the considerable authority that the local boards possessed prior to the Act, to the purpose of the Act, and to the practical consequences that would flow from a contrary construction. We found that “[i]t would ... be unreasonable to conclude that the Legislature intended to repeal the districts’ long-standing power, leaving emissions totally unregulated until the board acts. That drastic result would be contrary to the Legislature’s declaration that the Tanner Act was necessary ‘to achieve the earliest practicable control of toxic contaminants.’ [Citations.] Moreover, the Legislature’s obvious purpose in passing the act was to improve and strengthen air pollution regulation. The result sought by the association would be inimical to that purpose. ‘ “[T]he objective sought to be achieved by a statute as well as the evil to be prevented is of prime consideration in its interpretation.” ’ ” (Id. at p. 426, italics in original.)
The purposes of Proposition 65 are stated in the preamble to the statute, section 1, which declares in pertinent part: “The people of California find that hazardous chemicals pose a serious potential threat to their health and well-being, that state government agencies have failed to provide them with adequate protection, and that these failures have been serious enough to lead to investigations by federal agencies of the administration of California’s toxic protection programs. The people therefore declare their rights: [^] (a) to protect themselves and the water they drink against chemicals that cause cancer, birth defects, or other reproductive harm.” (Ballot Pamp., Proposed Stats. with arguments to voters, Gen. Elec. (Nov. 4, 1986) p. 53 (hereafter Ballot Pamphlet).)
Further evidence of the Act’s purpose and intent can be gleaned from the ballot materials. “[W]hen ... the enactment follows voter approval, the ballot summary and arguments and analysis presented to the electorate in connection with a particular measure may be helpful in determining the probable meaning of uncertain language.” (Amador Valley Joint Union High Sch. Dist. v. State Bd. of Equalization (1978) 22 Cal.3d 208, 245-246 [149 Cal.Rptr. 239, 583 P.2d 1281].) An examination of the ballot summary and arguments in this case makes clear that the primary focus of the Act, aside from its warning provisions, was the protection of drinking water, i.e., the water that comes from the tap, from contamination. First, the “Official Title and Summary Prepared by the Attorney General” states in part, “Restrictions on Toxic Discharges Into Drinking Water; Requirement of Notice of Persons’ Exposures to Toxics. Initiative *307Statute. Provides persons doing business shall neither expose individuals to chemicals known to cause cancer or reproductive toxicity without first giving clear and reasonable warning, nor discharge such chemicals into drinking water." (Ballot Pamp., supra, at p. 52, italics added.)
Moreover, the “Argument in Favor of Proposition 65” states in part: “There are certain chemicals that are scientifically known—not merely suspected, but known—to cause cancer and birth defects. Proposition 65 would: [‘JO Keep these chemicals out of our drinking water.” (Ballot Pamp., supra, at p. 54.) Further along in the ballot argument under the title “Safe Drinking Water” it states: “Proposition 65 singles out chemicals that are scientifically known to cause cancer or reproductive disorders (such as birth defects). Effectively it tells businesses: Don’t put these chemicals into our drinking water supplies.” (Ibid.)
Thus, one of the predominant purposes of the Act, as stated in the preamble, ballot summary, and arguments, was to protect drinking water from toxic contamination. In light of that purpose, the term “source of drinking water” was expansively defined to include any water currently destined to be used as drinking water, as well as any water officially designated as suitable for drinking water. The Act thereby creates a broad zone of protection for drinking water before it comes out of the tap, outlawing all toxic discharges that will have the probable consequences of contaminating such water. We must conclude that the electorate did not intend a breach in that protective zone by exempting discharges of lead into faucet water—where toxic contamination of drinking water is certain to occur—absent an express exclusion. Accordingly, we find the Attorney General’s interpretation of the statute and the phrase “source of drinking water” to be not only faithful to the statutory language, but also “reasonable, fair and harmonious with its manifest purpose,” whereas defendants’ construction is “productive of absurd consequences.” (Western Oil & Gas Assn., supra, 49 Cal.3d at p. 425.)
Defendants contend that this interpretation of the statute violates the statutory construction principle known as noscitur a sociis (it is known by its associates). “In accordance with this principle of construction, a court will adopt a restrictive meaning of a listed item if acceptance of a more expansive meaning would make other items in the list unnecessary or redundant, or would otherwise make the item markedly dissimilar to the other items in the list.” (Moore v. California State Bd. of Accountancy (1992) 2 Cal.4th 999, 1012 [9 Cal.Rptr.2d 358, 831 P.2d 798].) Defendants contend that a plumbing fixture “is markedly dissimilar from bodies of water which are or have the potential for being used as a source of drinking water,” and *308therefore should not be inferred to be included within the Act. But when the purpose of the Act is considered, the seeming dissimilarities between water in plumbing fixtures and natural bodies of water become less significant than their similarities. Both pertain to water that will be used for drinking purposes, and both require regulation in order to ensure safe drinking water. We therefore find defendants’ argument unpersuasive.
Defendants contend that the “Analysis by the Legislative Analyst” of the Act that appeared in the ballot pamphlet supports its position. We find no such support. The analysis begins with a review of various then-existing government programs and statutes designed to protect the people of the state against possible exposures to harmful chemicals, and the agencies that implement these statutes. The analysis then explains that “[t]hese regulatory agencies must make judgments about the amounts of harmful chemicals that can be released into the environment. In doing so, they try to balance what it costs to prevent the release of chemicals against the risks the chemicals pose to public health and safety.” (Ballot Pamp., supra, at p. 52.) The analysis then summarizes the provisions of the Act, concluding that “[bjecause these new requirements would result in more stringent standards, the practical effect of the requirements would be to impose new conditions for the issuance of permits for discharges into sources of drinking water. In order to implement the new requirements, state agencies that are responsible for issuing permits would be required to alter state regulations and develop new standards for the amount of chemicals that may be discharged into sources of drinking water.” (Ibid.)
Defendants contend that the reader of the analysis would have understood from the foregoing that the Act does not apply to plumbing faucets because of its reference to government-issued “permits.” As they state: “It is indisputable that such permits relate to discharges to the state’s surface and ground waters, and have never been issued for and have nothing to do with faucets.” Yet it is clear from the context that the Legislative Analyst, in explaining the “practical effect” of the Act’s new requirements, was attempting to put the Act in the context of existing environmental regulations, which the first part of the analysis was devoted to explaining. The Legislative Analyst did not suggest that all the effects and ramifications of the Act were being set forth in his brief summation. As discussed, the language of section 25249.11, subdivision (d), makes clear that Proposition 65 protects waters beyond the jurisdiction of the State Water Quality Control Board and the regional boards, and therefore beyond their permitting authority. In light of the explicit language and purpose of the statute, and the generality and brevity of the Legislative Analyst’s commentary, the latter cannot plausibly be viewed as implicitly limiting the scope of the statute in the manner advocated by defendants.
*309Defendants also contend that we should defer to certain contemporaneous constructions of the Act by the administrative agency charged with interpreting it. For reasons discussed below, we conclude that no administrative agency has in fact endorsed defendants’ position, and to the extent that some agency staff expressed views consistent with defendants’ position, they are neither controlling nor persuasive.
The Act, at section 25249.12, provides that “The Governor shall designate a lead agency and such other agencies as may be required to implement the provisions of the chapter including this section. Each agency so designated may adopt and modify regulations, standards, and permits as necessary to conform with and implement the provisions of this chapter and to further its purposes.” In 1987, pursuant to Governor’s Executive Order No. D-61-87, the Health and Welfare Agency (HWA) was designated as the “lead agency” under the Act. HWA’s interpretation of the Act is therefore of particular relevance. “Although not necessarily controlling, as where made without the authority of or repugnant to the provisions of a statute, the contemporaneous administrative construction of [an] enactment by those charged with its enforcement ... is entitled to great weight, and courts generally will not depart from such construction unless it is clearly erroneous or unauthorized.” (Coca-Cola Co. v. State Bd. of Equalization (1945) 25 Cal.2d 918, 921 [156 P.2d 1].)
The record shows that HWA never took a position on whether faucets are to be regulated under the Act. In response to comments regarding its promulgation of regulations to implement the Act, HWA declined the request of one commenter to further define “source of drinking water.” “This definition is not necessary because this term is already adequately defined in the Act at Health and Safety Code section 25249.11, subsection (d), and any further definition could cause unnecessary confusion. The designation in regional water quality control plans mentioned in the act provide sufficient guidance on which specific bodies of water are protected under the Act.” (HWA, amendment to final statement of reasons (1988) adopting Cal. Code Regs., tit. 22, § 12102 et seq.) Although defendants would have us read into this comment an implied endorsement of its position that only bodies of water mentioned in regional water quality control plans should be considered “sources of drinking water,” it is evident that HWA was not, by the comment, taking a position on the question before us. Moreover, as discussed above, the statute’s definition of “present source of drinking water” transcends the jurisdictional limits of the state or regional water quality control boards. Nor does HWA’s report, Safe Drinking Water and Toxic Enforcement Act of 1986: A Blueprint for Effective Implementation (Fiscal Year 1989-1990), support defendants’ position, as they contend, for this *310document also does not consider the question whether faucet water is a source of drinking water.
After these implementing regulations were promulgated, the Plumbing Manufacturers Institute (PMI), an industry group to which most of the defendants belong, made a request for administrative guidelines exempting plumbing products from section 25249.5. In a private, unpublished letter, dated December 1,1988, Steven A. Book, science adviser to the Secretary of HWA, responded “that the agency . . . cannot adopt such a position.” Further, Book rejected PMI’s central contention, namely that privately installed plumbing pipes were part of a “public water system” and therefore exempt from the Act pursuant to section 25249.11, subdivision (b).
The Book letter did go on to state that “[t]he Agency has never taken the position that the leaching of chemicals from a private plumbing facility into water within the facility constitutes a discharge into a ‘source of drinking water.’ The term ‘source of drinking water’ appears to refer to the geographic sources of water, whether used as a present source of drinking water, or simply designated by a regional water quality control board as suitable for domestic or municipal use.” Defendants would have us consider these comments as HWA’s definitive position on this issue, entitled to judicial deference. Yet it is clear that the Book letter cannot be taken as an implied agency endorsement of defendants’ position. This is evident not only from Book’s explicit refusal to adopt the position requested by PMI, but also from the conclusion of the letter, which advised, after noting that some plumbing suppliers were developing products that did not present lead exposure problems, that “avoiding the use of listed chemicals, or reducing their levels, is a good way to avoid application of the Act.”
Moreover, as the Office of Environmental Health Hazard Assessment (OEHHA), the successor lead agency for the implementation of the Act,6 pointed out in an amicus curiae brief filed with the trial court in this case, the Act’s regulations prescribe two methods by which private parties can obtain administrative interpretations of the Act: an “interpretive guideline” on a general matter (Cal. Code Regs., tit. 22, § 12102, subd. (b)), or a “safe use determination” which interprets and applies the Act to a specific set of facts (id., subd. (c)). Both of these types of administrative rulings require publication before their formal adoption. (See id., §§ 12103, 12104.) OEHHA correctly concluded that the science adviser’s letter was neither an interpretive guideline nor a safe use determination, and did not follow the public comment procedures provided for under the administrative regulations.
*311Furthermore, as the Court of Appeal recently concluded, the views of an administrative agency that are “the product of a nonadversarial, ex parte process, conducted at the request of an organization that exclusively represents the interests” of a private industry group are entitled to less deference than administrative decisions made after formal proceedings in which adversarial views are aired. (Hudgins v. Neiman Marcus Group, Inc. (1995) 34 Cal.App.4th 1109, 1125-1126 [41 Cal.Rptr.2d 46].) In Hudgins, the Court of Appeal declined to defer to the Labor Commissioner’s private opinion letter to a retail industry group interpreting relevant statutes to conclude that the policy of deducting unidentified returned merchandise costs from employees’ sales commissions was lawful. (Ibid.) Deference to the pronouncement of an administrative agency is even less warranted in this case than in Hudgins, since the science adviser’s private letter explicitly declined to take a definitive position on the question posed by the PMI. For all of the foregoing, the science adviser’s comment does not qualify as an official pronouncement of a lead agency entitled to deference from this court.
Defendants also cite certain administrative interpretations of “source of drinking water” by the State Water Quality Control Board and by regional water quality control boards. But these agencies did not take a position on the question whether plumbing supplies are regulated within the purview of safe drinking water provisions of the Act. They were concerned only with the meaning of “source of drinking water” as it applied to bodies of water within their jurisdictions. As already discussed, the Act implicitly contemplates that there may be “present” sources of drinking water beyond the purview of the state and regional boards. The administrative agencies’ interpretation of the phrase “source of drinking water” cannot, therefore, be taken as a tacit adoption of defendants’ position.7
*312Finally defendants argue that the Act is a “penal” statute because, in addition to provisions for injunctive relief set forth in section 25249.7, subdivision (a), the Act also provides in section 25249.7, subdivision (b) a civil penalty of up to $2,500 per day for each violation of section 25249.5. Defendants further contend that because the Act is penal in nature, it should be strictly construed in their favor, and that therefore we should adopt the construction of the term “source of drinking water” favorable to their position. It is indeed the case that “[w]hen language which is susceptible of two constructions is used in a penal law, the policy of this state is to construe the statute as favorably to the defendant as its language and the circumstance of its application reasonably permit.” (People v. Overstreet (1986) 42 Cal.3d 891, 896 [231 Cal.Rptr. 213, 726 P.2d 1288].) But we find defendants argument unconvincing for two reasons. First, we do not believe, in light of the language and purpose of the Act discussed above, that it is “reasonably susceptible” to the construction defendants propose. The rule of strict construction of penal statutes “ ‘ “is not an inexorable command to override common sense and evident statutory purpose.” ’ ” (People v. Anderson (1987) 43 Cal.3d 1104, 1145-1146 [240 Cal.Rptr. 585, 742 P.2d 1306].)
Second the rule of strict construction of penal statutes has generally been applied in this state to criminal statutes, rather than statutes which prescribe only civil monetary penalties. (See, e.g., People v. Forbis (1996) 42 Cal.App.4th 599, 603-604 [49 Cal.Rptr.2d 836]; In re Rottanak K. (1995) 37 Cal.App.4th 260, 269 [43 Cal.Rptr.2d 543]; People v. Barraza (1988) 197 Cal.App.3d 613, 617 [243 Cal.Rptr. 98]; People v. Martin (1985) 168 Cal.App.3d 1111, 1123-1124 [214 Cal.Rptr. 873]; In re Fain (1983) 145 Cal.App.3d 540, 549 [193 Cal.Rptr. 483]; Keeler v. Superior Court (1970) 2 Cal.3d 619, 631 [87 Cal.Rptr. 481, 470 P.2d 617, 40 A.L.R.3d 420]; People v. Smith (1955) 44 Cal.2d 77, 79 [279 P.2d 33]; People v. Ralph (1944) 24 Cal.2d 575, 581 [150 P.2d 401]; see also 1 Witkin & Epstein, Cal. Criminal Law (2d ed. 1988) § 28, pp. 37-38, and cases cited therein.) As was first stated in Ex parte Rosenheim (1890) 83 Cal. 388, 391 [23 P. 372], a criminal case, and reiterated many times since, “[w]hile it is true, the rule of the common law that penal statutes are to be strictly construed has been abrogated by [Penal Code section 4], . . . it is also true that the defendant is entitled to the benefit of every reasonable doubt, whether it arise out of a question of fact, or as to the true interpretation of words or the construction of language used in a statute . . . .”
This rule of strict construction for criminal statutes, also referred to as the rule of “lenity,” was explained by the United States Supreme Court in *313somewhat different terms: “Application of the rule of lenity ensures that criminal statutes will provide fair warning concerning conduct rendered illegal and strikes the appropriate balance between the legislature, the prosecutor, and the court in defining criminal liability. [Citation.] (‘[B]ecause of the seriousness of criminal penalties, and because criminal punishment usually represents the moral condemnation of the community, legislatures and not courts should define criminal activity’).” (Liparota v. United States (1985) 471 U.S. 419, 427 [85 L.Ed.2d 434, 441, 105 S.Ct. 2084].) Thus, as both Rosenheim and Liparota recognize, criminal penalties, because they are particularly serious and opprobrious, merit heightened due process protections for those in jeopardy of being subject to them, including the strict construction of criminal statutes.
Hale v. Morgan (1978) 22 Cal.3d 388 [149 Cal.Rptr. 375, 584 P.2d 512] (Hale), cited by defendants, does not support their position that all statutes with civil monetary penalties should also be strictly construed. In that case, a landlord was found to be liable under Civil Code section 789.3 for terminating a tenant’s utilities with the intent to terminate his tenancy. The fact of the landlord’s liability under the statute was not at issue. Rather we addressed questions pertaining to the amount of civil penalty which could be assessed against him. We held that the statute in question violated the due process clause of the federal and California Constitutions because it did not permit the trial court any discretion in imposing the civil penalty—it did not allow the court, for example, to take into account the good faith motivation of the offending landlord. (22 Cal.3d at pp. 404-405.) We also construed the penalty portion of the statute, subdivision (b) of Civil Code section 789.3, determining when a tenant can be said to be “deprived” of utility services for purposes of quantifying the landlord’s liability. We concluded, after consulting the dictionary definition of “deprived,” that a tenant would not be considered to be deprived of utility services if “the tenant actually succeeds in restoring service, or, by reasonable effort, could have done so.” (22 Cal.3d at p. 406.) On our way to that conclusion, we stated in dictum that “[b]e-cause the statute is penal, we adopt the narrowest construction of its penalty clause to which it is reasonably susceptible in the light of its legislative purpose.” (Id. at p. 405, italics added.)
Hale did not purport to alter the general rule that civil statutes for the protection of the public are, generally, broadly construed in favor of that protective purpose. (See, e.g., Ford Dealers Assn. v. Department of Motor Vehicles (1982) 32 Cal.3d 347, 357 [185 Cal.Rptr. 453, 650 P.2d 328]; Kizer v. Waterman Convalescent Hospital, Inc. (1992) 10 Cal.App.4th Supp. 8, 15 [13 Cal.Rptr.2d 239]; Skyline Homes, Inc. v. Department of Industrial Relations (1985) 165 Cal.App.3d 239, 250 [211 Cal.Rptr. 792]; Montessori *314 Schoolhouse of Orange County, Inc. v. Department of Social Services (1981) 120 Cal.App.3d 248 [175 Cal.Rptr. 14]; California State Restaurant Assn. v. Whitlow (1976) 58 Cal.App.3d 340, 347 [129 Cal.Rptr. 824]; Buchwald v. Superior Court (1967) 254 Cal.App.2d 347, 354-358 [62 Cal.Rptr. 364].) The focus of Hale was rather on safeguarding against the excessive penalization of those found liable under a civil statute. We construed a portion of the statute that was concerned solely with the manner of calculating the amount of penalty—the defendant did not claim that the portion of the statute pertaining to the fact of his liability was ambiguous. We accordingly specified that we were narrowly construing the “penalty clause.” (Hale, supra, 22 Cal.3d at p. 405.)
In the present case we are concerned with a statute, section 25249.5, that defines the conduct proscribed by the Act, and the scope of the government’s authority to enjoin and prohibit that conduct, rather than the method of assessing the amount of penalty for transgressing the proscription. We do not consider, for example, which acts by faucet manufacturers or distributors constitute a discrete “violation” of the Act that would subject them to a separate monetary penalty under section 25249.7, subdivision (b) (see Hale, supra, 22 Cal.3d at pp. 401-402; People ex rel. Younger v. Superior Court (1976) 16 Cal.3d 30, 42-44 [127 Cal.Rptr. 122, 544 P.2d 1322]; People v. Superior Court (1973) 9 Cal.3d 283, 288-289 [107 Cal.Rptr. 192, 507 P.2d 1400, 55 A.L.R.3d 191]); nor do we consider whether their alleged good faith belief that they were not violating the statute should affect the penalty for which they are liable.8 Accordingly, Hale's dictum is inapposite. Although there may be circumstances in which civil statutes should be strictly construed, we find no reason to do so in this case.
In sum, we consider Proposition 65 to be a remedial statute intended to protect the public from, among other things, toxic contamination of its drinking water. We construe the statute broadly to accomplish that protective purpose. As discussed above, we conclude that the Attorney General’s construction of section 25249.5 to include faucet water as a source of drinking water is not only a reasonable reading of the language of the statute, *315but is also harmonious with its basic remedial objectives, i.e., protection of the public.9
For all the foregoing, the judgment of the Court of Appeal is reversed and the cause is remanded with directions to cause issuance of a peremptory writ of mandate as prayed.
George, C. J., Kennard, J., Werdegar, J., and Brown, J., concurred.