Opinion

Southern California Edison Company (SCE) sued the Commissioners of the California Public Utilities Commission (PUC) in the United States District Court for the Central District of California, claiming PUC’s regulation of electricity rates violated federal law in several respects. The parties later reached an agreement settling the action, which became the basis for a stipulated judgment proposed to the district court. The. Utility Reform Network (TURN), which had intervened in the action, opposed the stipulated judgment, claiming, among other things, that PUC’s agreement to *787the settlement violated California law. The district court approved the settlement as fair and entered the stipulated judgment over these objections.

On appeal, the United States Court of Appeals for the Ninth Circuit discerned “a serious question” whether the agreement violated California law in several respects, both substantive and procedural. (Southern California Edison Co. v. Lynch (9th Cir. 2002) 307 F.3d 794, 809.) Because “as a matter of federal law, state officials cannot enter into a federally sanctioned con sent decree beyond their authority under state law,” the federal court of appeals believed the resolution of state law issues was essential to determining the validity of the stipulated judgment. (Id. at p. 812.) The court of appeals therefore certified to this court (Cal. Rules of Court, former rule 29.5; see id., rule 29.8) three questions of California law. We accepted the certification request, modifying one of the questions slightly. As accepted, the questions to be answered are:

1. Did the Commissioners of the California Public Utilities Commission have the authority to propose the stipulated judgment in light of the provisions of Assembly Bill No. 1890 (1995-1996 Reg. Sess.) codified in Public Utilities Code sections 330-398.5 (Stats. 1996, ch. 854)?

2. Did the procedures employed in entering the stipulated judgment violate the Bagley-Keene Open Meeting Act (Gov. Code, §§ 11120-11132.5)?

3. Does the stipulated judgment violate section 454 of the Public Utilities Code by altering utility rates without a public hearing and issuance of findings?

Having analyzed these questions, we conclude the settlement did not violate California law in any of these three respects.

BACKGROUND

The essential background of this case lies in California’s attempt, beginning in 1996, to move the system for provision of electrical power from a regulated to a competitive market, the crisis caused in mid-2000 to early 2001 by soaring prices for electricity on the wholesale market, and the urgency legislation enacted in January 2001 in response to that crisis.

Assembly Bill No. 1890 (1995-1996 Reg. Sess.) (hereafter Assembly Bill 1890), which became law in 1996 (Stats. 1996, ch. 854), was intended to provide the legislative foundation for “California’s transition to a more competitive electricity market structure.” (Assem. Bill 1890, § 1, subd. (a).) The new market structure included the creation of the California Power *788Exchange (CalPX), which was to run an “efficient, competitive auction” among electricity producers, and the Independent System Operator, which would control the statewide transmission grid. (Id., § 1, subd. (c).) The state’s main investor-owned electric utility companies (SCE, Pacific Gas and Electric Company (PG&E), and San Diego Gas and Electric Company (SDG&E); hereafter the utilities) were expected to divest themselves of substantial parts of their generating assets, while retaining others at least during the period of transition. (Id., § 10, adding Pub. Util. Code, former § 377.) Under the Assembly Bill 1890 scheme as implemented, all generators, including the utilities, sold their power through the CalPX; the utilities also bought, through that exchange, the electricity they needed to supply their retail customers. (Cal. Exchange Corp. v. FERC (In re Cal. Power Exchange Corp.) (9th Cir. 2001) 245 F.3d 1110, 1114-1115.)

Because this competition among producers was expected to bring down wholesale prices, the utilities believed that some of their generating assets, which they had built or improved with PUC approval, would become “uneconomic,” in that the costs of generation (and of certain long-term contracts between the utilities and other generators) would be higher than prevailing wholesale rates would support. The costs associated with these potentially uneconomic assets are also known as “stranded costs” or “transition costs.” The Legislature, in Assembly Bill 1890, intended to allow for “[a]ccelerated, equitable, nonbypassable recovery of transition costs” (Stats. 1996, ch. 854, § 1, subd. (b)(1)) and thereby to “provide the investors in these electrical corporations with a fair opportunity to fully recover the costs associated with commission approved generation-related assets and obligations” (Pub. Util. Code, § 330, subd. (t)). The legislative scheme for doing so without subjecting consumers to increased rates was complex, but consisted in its essentials of the following:

Under Public Utilities Code section 367,1 PUC was to identify and quantify potentially uneconomic costs (i.e., the PUC-approved costs that “may become uneconomic as a result of a competitive generating market”). The identified costs were to be recoverable through rates that would not exceed “the levels in effect on June 10, 1996,” and the recovery was not to “extend beyond December 31, 2001.” (§ 367, subd. (a).) The component of rates dedicated to recovery of transition costs was nonbypassable, i.e., it had to be paid to the utility whether the consumer bought power from the utility, from a generator in a single direct transaction, or from a generator in an aggregated direct transaction with other consumers. (§§ 365, subd. (b), 366, 370.)

*789Section 368 required each utility to propose, and PUC to approve, a “cost recovery plan” for the costs identified in section 367 that would set rates at June 10, 1996, levels, with a 10 percent reduction for residential and small commercial customers. Section 368, subdivision (a) continues; “These rate levels . . . shall remain in effect until the earlier of March 31, 2002, or the date on which the commission-authorized costs for utility generation-related assets and obligations have been fully recovered. The electrical corporation shall be at risk for those costs not recovered during that time period.”

PUC implemented this cost-recovery scheme in part by creating, for each electric utility, a transition cost balancing account (sometimes herein referred to as a TCBA), in which the PUC-identified stranded costs were tracked. Transition costs were not to be forecast, but rather entered in the transition cost balancing account as the PUC determined them. Costs associated with utility-retained generating assets were to be determined by comparing the book value of the assets with their market valuations, a process to be completed by the end of 2001. These uneconomic generating costs were to be netted against the benefits of any economic generating assets (those having higher market than book value). The difference between rate revenue and the utility’s other (nongeneration-related) costs was designated the utility’s “headroom” and was to be credited against the stranded costs in the transition cost balancing account. The portion of each rate serving as headroom was designated the competition transition charge. (In re Pacific Gas & Electric Co. (1997) 76 Cal.P.U.C.2d 627, 646-653, 740-744.)

In the first few years of the transition period, the utilities recovered much of their stranded costs. SDG&E was found to have recovered all its transition costs, ending the rate freeze for that utility under section 368. SCE and PG&E, however, were still subject to the rate freeze when, in the summer of 2000, power procurement prices, and particularly prices on the CalPX spot market, rose drastically. They incurred huge debts buying electricity through the CalPX. (Cal. Exchange Corp. v. FERC (In re Cal. Power Exchange Corp.), supra, 245 F.3d at p. 1115.)

In November 2000, as the wholesale price and supply problems continued, SCE brought its federal action against PUC, the subsequent settlement of which is the subject of this decision. In essence, SCE claimed the rate freeze imposed by Assembly Bill 1890 was now depriving SCE of its right, under federal law, to recover the costs of purchasing electricity for its customers. More particularly, SCE claimed the freeze rates had become unconstitutionally confiscatory and violated the federal “filed rate” rule, which assertedly allows a utility to recover in state-regulated retail rates the costs of purchases made under federally approved tariffs.

*790PUC granted SCE and the other utilities emergency rate relief in early 2001. Deeming the crisis one “that involves not only utility solvency but the very liquidity of the system,” PUC in January 2001 authorized a temporary surcharge of 1 cent per kilowatt-hour. (Application of Southern California Edison Co. (2001) Cal.P.U.C. Dec. No. 01-01-018, pp. 1-4, 2001 Cal. PUC LEXIS 44.) Two months later, still finding that “SCE’s and PG&E’s continued financial viability and ability to serve their customers has been seriously compromised by the dramatic escalation in wholesale prices,” PUC made the January increase permanent and authorized an additional 3 cents per kilowatt-hour increase. (Application of Southern California Edison Co. (2001) Cal.P.U.C. Dec. No. 01-03-082, pp. 2-4, 2001 Cal.PUC LEXIS 217.) PUC refers to these increases collectively as the “four cent surcharge,” a usage we adopt. (According to SCE, the surcharge amounted to an average increase of 40 percent in retail rates.) PUC’s March 2001 decision, while authorizing an increase to pay for ongoing power purchases, did “not address recovery of past power purchase costs and other costs claimed by the utilities.” (Id. at p. 2.)

The Legislature also took action in January 2001, in an extraordinary session called to address the power crisis. In that session’s Assembly Bill No. 1 (Stats. 2001, 1st Ex. Sess., ch. 4; hereafter Assembly Bill 1X), the Legislature authorized the state Department of Water Resources to begin buying power for customers of SCE and PG&E. (Id., § 4, adding Wat. Code, §§ 80100-80122.) In Assembly Bill No. 6 of that session (Stats. 2001, 1st Ex. Sess., ch. 2; hereafter Assembly Bill 6X), the Legislature amended several provisions of Assembly Bill 1890, halting at least temporarily the transition to a competitive electricity market. In particular, Public Utilities Code section 377, as first enacted by Assembly Bill 1890, had provided that PUC would continue regulating the utilities’ retained nonnuclear generating assets “until those assets have been subject to market valuation,” after which they would be sold off unless the utility convinced the PUC their retention was in the public interest. (Stats. 1996, ch. 854, § 10.) As amended by Assembly Bill 6X, section 377 provides that all the remaining generating assets are subject to PUC regulation and may not be sold until January 1, 2006, at the earliest. (Assem. Bill 6X, § 3.) Similarly, as enacted by Assembly Bill 1890, Public Utilities Code section 330, subdivision (1)(2) had provided that the generating assets “should be transitioned from regulated status to unregulated status through means of commission-approved market valuation mechanisms.” (Stats. 1996, ch. 854, § 10.) Assembly Bill 6X deleted this language, leaving only the general statement that “[generation of electricity should be open to competition.” (Id., § 2.) PUC subsequently issued decisions, based on Assembly Bill 6X, reestablishing cost-based rate regulation of SCE’s retained generating assets and modifying restrictions on the use of the four cent surcharge. (E.g., Application of Southern California Edison Co. (2002) *791Cal.P.U.C. Dec. No. 02-04-016, p. 2 [2000 Cal.PUC LEXIS 1110]; Application of Southern California Edison Co. (2002) Cal.P.U.C. Dec. No. 02-11-026, pp. 11-16, 2002 Cal.PUC LEXIS 720.)

In October 2001, SCE and PUC reached a settlement of SCE’s federal rate action. In the settlement’s recitals, the parties agreed that during the period of very high wholesale power prices, SCE accumulated procurement-related liabilities and indebtedness of about $6,355 billion, creating severe liquidity and credit problems for the company. Conditions in 2001, including the four cent surcharge, had allowed SCE to collect retail revenues in excess of current costs. The settlement was intended to use the opportunity thus provided to restore SCE’s creditworthiness and avoid further instability and uncertainty for the company and consumers. (Settlement, Recitals D-F.)

PUC’s principal substantive concession in the settlement was its agreement to permit SCE to recover its past procurement-related costs by maintaining the existing rates until the end of 2003, if necessary. The parties agreed to establish a procurement-related obligations account (sometimes herein referred to as the PROACT), the initial balance of which was SCE’s procurement-related liabilities less its cash on hand. (The parties estimated the initial balance at about $3.3 billion.) (Settlement, § 2.1(a).) SCE agreed to apply all its surplus (its revenue in excess of defined recoverable costs), with some exceptions, to the account, gradually reducing its balance. (Settlement, § 2.1(b).) The PUC agreed to maintain the rates in effect on the settlement date (with some adjustments) during the “rate repayment period,” which was defined to end when the PROACT was paid down to zero or on December 31, 2003, whichever came first. (Settlement, §§ l.l(p), l.l(w), 2.2(a).) A potentially longer “recovery period” was defined as ending when the account was completely paid down or on December 31, 2005, whichever came first. (Settlement, § l.l(q).) The parties agreed that, if necessary, any obligations left in the PROACT at the end of the rate repayment period (i.e., at the end of 2003) would “be amortized in retail rates ratably during all or a portion of the remainder of the Recovery Period.” (Settlement, § 2.2(b).)2

Over TURN’S objection, the federal district court entered a stipulated judgment incorporating the terms of the settlement agreement, finding the agreement “adequate and fair.” On TURN’S appeal, the court of appeals resolved the federal law issues in favor of SCE and PUC (Southern California Edison Co. v. Lynch, supra, 307 F.3d at pp. 802-809), but found *792that the settlement and stipulated judgment appeared to violate California law in certain respects and, following “principles of comity” (id. at p. 812), tendered those state law questions to this court and stayed further proceedings pending our response (id. at p. 813). We proceed to decide the certified questions.

Question 1: Did the Commissioners of the California Public Utilities Commission have the authority to propose the stipulated judgment in light of the provisions of Assembly Bill No. 1890 (1995-1996 Reg. Sess.) codified in Public Utilities Code sections 330-398.5 (Stats. 1996, ch. 854)?

Answer: Yes.

PUC’s authority derives not only from statute but from the California Constitution, which creates the agency and expressly gives it the power to fix rates for public utilities. (Cal. Const., art. XII, §§ 1, 6.) Statutorily, PUC is authorized to supervise and regulate public utilities and to “do all things . . . which are necessary and convenient in the exercise of such power and jurisdiction” (§ 701); this includes the authority to determine and fix “just, reasonable [and] sufficient rates” (§ 728) to be charged by the utilities. Adverting to these provisions, we have described PUC as “ ‘a state agency of constitutional origin with far-reaching duties, functions and powers’ ” whose “ ‘power to fix rates [and] establish rules’ ” has been “ ‘liberally construed.’ ” (San Diego Gas & Electric Co. v. Superior Court (1996) 13 Cal.4th 893, 914-915 [55 Cal.Rptr.2d 724, 920 P.2d 669], quoting Consumers Lobby Against Monopolies v. Public Utilities Com. (1979) 25 Cal.3d 891, 905 [160 Cal.Rptr. 124, 603 P.2d 41].) If PUC lacked substantive authority to propose and enter into the rate settlement agreement at issue here, it was not for lack of inherent authority, but because this rate agreement was barred by some specific statutory limit on PUC’s power to set rates. (See Assembly v. Public Utilities Com. (1995) 12 Cal.4th 87, 103 [48 Cal.Rptr.2d 54, 906 P.2d 1209].)

TURN contends, first, that PUC’s agreement to the settlement violated a legislative directive in section 368, enacted as part of Assembly Bill 1890, which froze rates at June 1996 levels during the transition period. In particular, TURN argues the settlement illegally “extended]” the freeze period. But TURN errs in assuming that section 368 requires that rates be reduced at the end of the freeze period. In this respect, section 368 provides only that the freeze rates “shall remain in effect until the earlier of March 31, 2002, or the date on which the commission-authorized costs for utility generation-related assets and obligations have been fully recovered.” Section 368 does not dictate that rates be reduced, or changed in any way, at the end of the freeze period. True, section 330, subdivision (a) recites *793the Legislature’s “intent . . . that a cumulative rate reduction of at least 20 percent be achieved” by April 1, 2002, but section 330 consists of findings and declarations providing “guidance in carrying out” the provisions of Assembly Bill 1890, not binding limitations on PUC authority. While the Legislature certainly intended its Assembly Bill 1890 scheme to bring down retail rates through wholesale competition, progress toward that result was delayed, to say the least, by the unanticipated 2000-2001 rise in wholesale rates.

With more force, TURN contends the settlement allowed SCE to recover in the postfreeze period, in violation of section 368, costs incurred during the freeze period. TURN relies on the third sentence of section 368, subdivision (a), which provides: “The electrical corporation shall be at risk for those costs not recovered during that time period,” i.e., the freeze period ending March 31, 2002. After careful consideration, we conclude, contrary to TURN’S contentions, that after the enactment of Assembly Bill 6X in 2001, which required electrical utilities to retain their generating plants until at least 2006 and returned retained generating-asset rates to cost-based regulation, PUC was authorized to approve rates allowing SCE to recover the costs covered by the settlement agreement. While Assembly Bill 6X did not repeal section 368 or reverse all aspects of electricity deregulation, it constituted a major retrenchment from the competitive price-reduction approach of Assembly Bill 1890, reemphasizing instead PUC’s duty and authority to guarantee that the electric utilities would have the capacity and financial viability to provide power to California consumers.

We first consider whether, the effect of Assembly Bill 6X aside, the costs slated for recovery by the settlement agreement are uneconomic generating-asset costs (i.e., stranded or transition costs) restricted by section 368. On their face they are not: the settlement agreement expressly provides for postfreeze recovery of energy procurement, rather than generation, costs. Under the settlement, PUC agrees to maintain the rates in force at the time of the settlement until the earlier of December 2003 or the date the obligations recorded in the procurement-related obligations account have been recovered. SCE’s procurement-related liabilities are tallied in schedule 1.1 of the settlement, and include SCE’s debts to banks, electricity generators, the CalPX and Independent System Operator, and the Department of Water Resources, which in January 2001 began purchasing electricity for SCE customers. These liabilities resulted from “wholesale electricity procurement costs” (Settlement, Recital D) rather than from the “costs for generation-related assets and obligations” referred to in section 367, the recovery of which sections 367 and 368 restrict to the freeze period.

PUC nevertheless maintains that the costs to be recovered in retail sales under the settlement are not procurement costs but rather SCE’s “generation-related costs . . . which were previously called stranded costs.” PUC bases *794this characterization on an accounting change it made in March 2001, at TURN’S suggestion, by which SCE’s accumulated procurement liabilities were transferred into its transition cost balancing account, which had previously been used to track recovery only of stranded costs and which was, according to SCE, “overcollected” (i.e., in the black) at that time.3 As a result of this change, PUC later determined that SCE had, at the time of the settlement, a substantial amount of unrecovered transition costs. (Application of Southern California Edison Co., supra, Cal.P.U.C. Dec. No. 01-03-082, pp. 26-32; see Cal. P.U.C. Res. No. E-3765 (Jan. 23, 2002) p. 13.) Thus, PUC explains, “the effect of the Settlement is to recover the large stranded cost balance in the TCBA.” (Cal. P.U.C. Res. No. E-3765, supra, p. 13.)

Although PUC’s position is consistent with its earlier determination (in proceedings unrelated to this case) that costs are fungible for purposes of Assembly Bill 1890’s restrictions on cost recovery,4 we do not fully accept for present purposes PUC’s equation of SCE’s procurement liabilities accumulated during the wholesale rate crisis with its unrecovered transition costs. As discussed below, SCE’s true unrecovered transition costs appear indeterminable in light of PUC’s failure, following the changes wrought by Assembly Bill 6X, to complete the planned transition by assigning market values to SCE’s generating assets, a step that would have reduced the transition cost balancing account balance by an unknown but potentially significant amount. But whatever the amount of SCE’s unrecovered transition costs, there is no reason to assume it was exactly equivalent to the amount of the utility’s unrecovered procurement costs. Even assuming that when the March 2001 accounting change was made, some amount of transition costs should have remained in SCE’s transition cost balancing account, neither PUC nor TURN endeavors to explain why that amount would necessarily have been equal to the amount of SCE’s procurement costs (about $6,355 billion, according to the settlement) and hence, why the settlement recovery figure of $3.3 billion, calculated from SCE’s outstanding procurement liabilities, should be deemed to represent instead the exact amount of transition costs unrecovered at the time of the settlement. While the March 2001 accounting change may have been properly used to determine that the Assembly Bill 1890 rate freeze had not then ended (see fn. 3, ante), it should not bind us to a counterfactual characterization of all the procurement costs at issue here.

*795The passage of Assembly Bill 6X in January 2001 introduced additional grounds against deeming recovery of procurement liabilities to be limited by section 368. Assembly Bill 6X eliminated Assembly Bill 1890’s requirement for market valuation of utility-retained generating assets, required SCE to keep its remaining generating assets until 2006, and allowed PUC to regulate the rates for power so generated pursuant to ordinary “cost-of-service” ratemaking. PUC was thus authorized to permit SCE such recovery of past costs as necessary to render the utility financially viable and to ensure SCE would be able to continue serving its customers through electricity generated in its retained plants. In a technical sense, moreover, Assembly Bill 6X largely eliminated the category of “uneconomic” generating asset costs, the only costs whose recovery is limited under section 368. Since “uneconomic” costs are those that “may become uneconomic as a result of a competitive generating market” in that they “may not be recoverable in market prices in a competitive market” (§ 367), and under Assembly Bill 6X the retained assets will not be included in a competitive generating market until at least 2006, section 368, as PUC argues, “no longer applies to the generation-related costs of the utilities.”

TURN concedes that Assembly Bill 6X returned to cost-of-service regulation those generating assets SCE still owned when the law was enacted.5 The January 2001 measure, TURN further concedes, meant that PUC would be able to ensure, by rate regulation, that SCE “would be given an opportunity, in the future, to earn a return on [its] investment in those plants” and that, consequently, “[t]he remaining book-value of utility-retained generation is not any part of the unrecovered stranded costs.” But TURN argues Assembly Bill 6X did not affect section 368’s mandate that SCE be “at risk” for what TURN calls “the stranded costs represented by the plants it did sell, or by the depreciation expense already recorded on its retained plants during the rate freeze.” The stranded costs accumulated by January 2001 were, TURN argues, unrecoverable under section 368 after the freeze period ended in March 2002.

We are not persuaded the settlement violates section 368 as TURN claims, even if the settlement is regarded as permitting recovery of some generation-related costs rather than, as indicated on its face, only procurement costs. SCE’s transition cost balancing account, designed to track its transition costs, was overcollected at the beginning of 2001. Even if, as TURN claims, the *796January 2001 account balance understated SCE’s transition costs, SCE persuasively argues it also overstated such costs because it did not reflect the increased market value of SCE’s retained generating assets in an environment of higher wholesale prices.

The process of selling, appraising, or otherwise placing a market value on the utilities’ generation assets, to be completed by December 31, 2001 (§ 367, subd. (b)), was essential to the Assembly Bill 1890 scheme, since the true stranded cost of an asset depended in part on its market value. (See In re Pacific Gas & Electric Co., supra, 76 Cal.P.U.C.2d at pp. 674—675.) Each utility’s transition cost balancing account tracked not only its generating assets’ amortized book values and its competition transition charge revenues, but also any “market valuation credits.” (Id. at p. 674.) The utilities thus expected to “adjust the transition cost balancing account when assets are sold or market-valued.” (Id. at p. 675.) But according to SCE, PUC “had never rendered a decision assigning market values to SCE’s generation assets— assets that became extremely valuable in a market in which prices had risen dramatically.” The passage of Assembly Bill 6X, which ended the sale of generating assets and returned them to traditional PUC rate regulation, removed the rationale and opportunity for market valuation, thus preventing the transition cost balancing account from serving as a complete or accurate record of transition costs.

Finally, we note that the Legislature, in section 367, gave PUC the authority to identify uneconomic costs. Pursuant to that authority the agency has determined that after passage of Assembly Bill 6X, generation-related costs are, for the reasons already stated, no longer “uneconomic” within the meaning of section 367. (Application of Southern California Edison Co., supra, Cal. P.U.C. Dec. No. 02-11-026, p. 14.) Cognizant of the principle that PUC’s interpretation of the Public Utility Code “should not be disturbed unless it fails to bear a reasonable relation to statutory purposes and language” (Greyhound Lines, Inc. v. Public Utilities Com. (1968) 68 Cal.2d 406, 410-411 [67 Cal.Rptr. 97, 438 P.2d 801]), we are unable to reach a different conclusion here.

Whether we regard the costs to be recovered under the PUC-SCE settlement as procurement costs or as generation-related costs, therefore, they were not uneconomic costs restricted in recovery by section 368. Consequently, PUC’s agreement to the settlement was not in violation of Assembly Bill 1890.

*797 Question 2: Did the procedures employed in entering the stipulated judgment violate the Bagley-Keene Open Meeting Act (Gov. Code, §§ 11120-11132.5)?

Answer: No.

The Bagley-Keene Open Meeting Act (the Bagley-Keene Act) applies to most state boards and commissions, including PUC. (See Gov. Code, §§ 11121, 11121.1, 11126, subd. (d).) The purpose of the law, stated in Government Code section 11120, is to ensure that “actions of state agencies be taken openly and that their deliberation be conducted openly.” The Bagley-Keene Act implements this policy by mandating that “[a]ll meetings of a state body shall be open and public . . .” (Gov. Code, § 11123), by requiring advance public notice of meetings (id., § 11125), by authorizing legal actions to prevent threatened violations of the act or declare its applicability to past or threatened future “actions” of a body (id., § 11130), and to declare null and void an “action taken” in violation of Government Code sections 11123 or 11125 (id., § 11130.3). “Action taken” is defined broadly to include “a collective decision” of the members and “a collective commitment or promise ... to make a positive or negative decision.” (Id., § 11122.)

The October 2001 settlement, which the subsequent stipulated judgment implemented, was signed by the five PUC commissioners and was both a collective decision of the commissioners and a collective commitment or promise to take further actions. It was thus an “action taken” by PUC and subject, under the above provisions, to the Bagley-Keene Act. The parties, moreover, agree this action was taken in a meeting, to wit, the closed or executive session of the regularly scheduled PUC meeting of October 2, 2001. The published agenda for the October 2 meeting listed, in the closed session section, this item: “FEX-2: Conference with Legal Counsel—Existing . . . Litigation. Case name unspecified. (Disclosure of case name would jeopardize existing settlement negotiations.) (Gov. Code Sec. 11126(e)(2)(A).)” According to PUC, the commissioners unanimously approved the settlement during the closed session, then reconvened in public session to announce their action. This is confirmed by the published PUC results sheet for the October 2 meeting, which lists this item: “FEX-2: SCE Settlement. Approved Staff Recommendation as Modified 5-0. Reconvened in Public Session at 1:30 p.m. to announce the action taken in Executive Session.”

PUC contends taking this action in closed session did not violate the Bagley-Keene Act, but, rather, was permitted under an exception to the law’s open-meeting requirement, Government Code section 11126, subdivision (e)(1), which provides as follows: “Nothing in this article shall be *798construed to prevent a state body, based on the advice of its legal counsel, from holding a closed session to confer with, or receive advice from, its legal counsel regarding pending litigation when discussion in open session concerning those matters would prejudice the position of the state body in the litigation.” We agree. On its face, subdivision (e)(1) permits a body only to “confer with” and “receive advice from” its attorney regarding litigation. But subdivision (e)(1) must be read in light of its purposes and in consonance with a closely related provision of the Bagley-Keene Act, Government Code section 11126.3, subdivision (a), which allows a body to withhold the identity of litigation to be considered in closed session if to identify it would “jeopardize its ability to conclude existing settlement negotiations to its advantage.” (Italics added.) Read in light of its purposes and in that statutory context, Government Code section 11126, subdivision (e)(1) was, as will be seen below, clearly intended to permit the body not only to deliberate with counsel regarding a settlement, but actually to settle the litigation in a closed session when closure is deemed necessary to avoid prejudice to a favorable settlement.

Settlement discussions with counsel are obviously an aspect of litigation particularly vulnerable to prejudice through public exposure and are thus one of the areas Government Code section 11126, subdivision (e)(1) was centrally intended to shelter from public revelation. In Sacramento Newspaper Guild v. Sacramento County Board of Supervisors (1968) 263 Cal.App.2d 41 [69 Cal.Rptr. 480], the court held that the enactment of the Ralph M. Brown Act (Gov. Code, §§ 54950-54962; hereafter Brown Act), the open meeting law applicable to local public entities, was not intended to remove protection of the attorney-client privilege from local government bodies’ deliberations with their attorneys concerning litigation. Public entities have as great a need for confidential counsel from their attorneys as private litigants and should not be put at a disadvantage in litigation by depriving them of that essential assistance. (Sacramento Newspaper Guild, supra, at p. 55.) In particular, the court explained, a public entity’s discussion with counsel about possible settlement must occur in private, for such conferences require a frank evaluation of the case’s strengths and weaknesses, and “[i]f the public’s ‘right to know’ compelled admission of an audience, the ringside seats would be occupied by the government’s adversary, delighted to capitalize on every revelation of weakness.” (Id. at p. 56; accord, Roberts v. City of Palmdale (1993) 5 Cal.4th 363, 373-374 [20 Cal.Rptr.2d 330, 853 P.2d 496].) The Legislature subsequently added protective provisions to both the Bagley-Keene and Brown Acts, vindicating the view expounded in Sacramento Newspaper Guild. Both new provisions were phrased in the language of current Government Code section 11126, subdivision (e)(1). (See Stats. 1981, ch. 968, § 12, p. 3690, adding former subd. (q) to Gov. Code, § 11126; Stats. 1984, ch. 1126, § 3, p. 3802, adding Gov. Code, § 54956.9.)

*799In 1992, the California Attorney General’s Office construed Government Code section 54956.9, the Brown Act provision paralleling Government Code section 11126, subdivision (e)(1), as authorizing a public entity to act on a settlement proposal, as well as deliberate on it, in closed session with its counsel. (75 Ops.Cal.Atty.Gen. 14 (1992).) The Attorney General noted, first, that the Brown Act’s “personnel exception” (Gov. Code, § 54957) has been construed to permit closed-session action on appointments and dismissals (see Lucas v. Board of Trustees (1971) 18 Cal.App.3d 988, 991 [96 Cal.Rptr. 431]), even though on its face the statute authorizes only a closed session to “consider” such personnel matters. “The parallel between section 54957 (‘to consider’) and section 54956.9 (‘to confer’) warrants similar treatment.” (75 Ops.Cal.Atty.Gen., supra, at p. 19.)

The same parallel may be drawn between the corresponding provisions of the Bagley-Keene Act. Subdivision (a)(1) of Government Code section 11126 permits closed sessions “to consider” personnel matters. Though case law has not yet addressed the point, we note that the immediately following provision, subdivision (a)(2), refers to “any disciplinary or other action taken against any employee at the closed session,” indicating that the Legislature intended, in the Bagley-Keene Act as (according to the Attorney General) in the Brown Act, that the government body could not only deliberate, but act, in closed session. The language used in Government Code section 11126, subdivision (e)(1), permitting a body “to confer” with counsel on settlement of pending litigation, is not so dissimilar to that in subdivision (a)(1) (“to consider”) as to warrant a different interpretation.

Interpreting the Brown Act counsel provision, the Attorney General also reasoned that consultation with counsel in the course of litigation often focuses on possible action—e.g., whether to file a suit or countersuit, what claims and defenses to plead, what parties to join. Conferring with counsel on these matters necessarily includes deciding on a course of action and instructing or authorizing counsel to pursue it. The same applies to settlement discussions. “Unless a local agency is to be a ‘second class citizen’ with its opponents ‘filling the ringside seats’ (Sacramento Newspaper Guild v. Sacramento County Bd. of Suprs., supra, 263 Cal.App.2d at p. 56), it must be able to confer with its attorney and then decide in private such matters as the upper and lower limits with respect to settlement, whether to accept a settlement or make a counter offer, or even whether to settle at all. These are matters which will depend upon the strength and weakness of the individual case as developed from conferring with counsel. A local agency of necessity must be able to decide and instruct its counsel with respect to these matters in private.” (75 Ops.Cal.Atty.Gen., supra, at pp. 19-20.)

This reasoning is equally applicable to state bodies governed by the Bagley-Keene Act. In providing (in Gov. Code, § 11126, subd. (e)(1)) for *800private conferences with counsel regarding pending litigation, the Legislature must have intended the scope of privacy to be broad enough to include the bodies’ instructions to their attorneys as to how to proceed, including whether and with what limits to negotiate settlement. The legislative purpose of placing public agencies on a roughly equal footing with private parties in litigation would otherwise be defeated.

In this case, of course, PUC went beyond instructing counsel in the closed meeting of October 2, 2001; it actually concluded the settlement, unanimously voting to accept the proposed agreement with SCE, and reconvened in public session only to announce the action taken. Theoretically, the PUC commissioners could instead have deliberated in private on this step, then reconvened in public session (at the same or a later meeting) to actually vote. But such a procedure could serve the purposes of the Bagley-Keene Act only if the body announced, before the public session, the identity of the litigation proposed for settlement, for only then could the public possibly be informed of, and comment on, the substance of the proposed action. (See Gov. Code, § 11125.7, subds. (a), (g) [state bodies, specifically including PUC, to provide opportunity for public comment during consideration of each agenda item]; id., subd. (d) [requirement does not apply to closed session items].) To convene publicly simply to vote on an unidentified and undescribed litigation proposal, without the opportunity for meaningful public comment, would be an empty gesture, which we will not assume the Legislature intended to require.

The question, then, is whether the Bagley-Keene Act requires a state body, after deliberating on a proposed settlement in closed session pursuant to Government Code section 11126, subdivision (e)(1), to announce its proposed decision in public session—identifying the litigation involved—and accept public comment on the proposed settlement before voting on it. To this question we think Government Code section 11126.3, subdivision (a) dictates a negative answer.

Government Code section 11126.3 sets forth the required procedures for closed sessions. Subdivision (a) mandates public disclosure of the “general nature” of each closed session item by, for example, a listing on the public agenda. The last sentence of subdivision (a) provides: “If the session is closed pursuant to subparagraph (A) of paragraph (2) of subdivision (e) of Section 11126 [pending litigation or administrative adjudications], the state body shall state the title of, or otherwise specifically identify, the litigation to be discussed unless the body states that to do so would jeopardize the body’s ability to effectuate service of process upon one or more unserved parties, or that to do so would jeopardize its ability to conclude existing settlement negotiations to its advantage.” (Italics added.)

*801Under the quoted provision, a body may decline to identify the litigation under discussion in closed session if the body states that to identify it would jeopardize the conclusion of an advantageous settlement. Were the body required, after its closed-session deliberations but before actually concluding the settlement, to announce publicly the proposed settlement and the name of the litigation, the protective purpose of Government Code section 11126.3, subdivision (a) would be defeated. The Legislature clearly intended, in enacting Government Code section 11126.3, subdivision (a), that a state body be able to keep private the identity of litigation it is considering settling until it has “conclude[d]” the settlement (assuming the body believes privacy is strategically necessary to the settlement negotiations). Construing the closely related provisions of Government Code section 11126, subdivision (e)(1) to require a public identification of the proposed settlement, before it has been concluded would defeat that purpose and could not have been the legislative intent.6

In this case PUC strictly followed the procedure mandated in Government Code section 11126.3, subdivision (a), noting in the closed-session agenda item, “Case name unspecified. (Disclosure of case name would jeopardize existing settlement negotiations.) (Gov. Code Sec. 11126(e)(2)(A).)” To require PUC, under Government Code section 11126, subdivision (e)(1), to reconvene in open session and publicly announce it was considering settling SCE’s federal litigation would subject PUC to the potential loss of the very negotiating equality that Government Code section 11126.3, subdivision (a) was designed to preserve. Reading Government Code section 11126, subdivision (e)(1) in statutory context, therefore, we conclude it authorized PUC not only to discuss, but also to conclude the settlement in closed session.

TURN contends that even if Government Code section 11126, subdivision (e)(1) generally permits state bodies to take action on a settlement in closed session, another provision of that section, subdivision (d)(1), specifically required this settlement agreement to be acted on in public session because the settlement raised electricity rates. Government Code section 11126, subdivision (d)(1) provides: “Notwithstanding any other provision of *802law, any meeting of the Public Utilities Commission at which the rates of entities under the commission’s jurisdiction are changed shall be open and public.”

We agree, however, with PUC and SCE that by agreeing to the settlement PUC did not “change[]” the “rates” (Gov. Code, § 11126, subd. (d)(1)) that SCE could charge for electricity. The central commitment PUC made in the settlement was to maintain the then existing rates for an agreed period. (Settlement, § 2.2(a).)

According to TURN, the settlement agreement changed rates by making regulatory concessions to SCE that (TURN asserts) will lead to future rates higher than would otherwise obtain. Government Code section 11126, subdivision (d)(1), TURN argues, applies to “any PUC decision that results in customers paying higher rates than they would absent the action.” We reject this interpretation of the statute as establishing a standard impossible to apply, because it depends on the unknowable course of future events under hypothetical conditions. Had PUC not settled SCE’s federal lawsuit, SCE might have won its case and rates might have been raised even higher or been kept in place longer; had SCE lost or continued in protracted litigation, it might have gone into bankruptcy and the bankruptcy court might have approved higher rates. This court, moreover, cannot know whether at some time in the future PUC would or would not have ordered rate reductions, customer refunds, or cost-accounting changes that might indirectly have resulted in lower rates. Under TURN’S interpretation, virtually any regulatory action would be a change in rates because PUC could have taken some other action potentially leading to lower rates in the future. (See Toward Utility Rate Normalization v. Public Utilities Com. (1988) 44 Cal.3d 870, 873 [245 Cal.Rptr. 8, 750 P.2d 787] [PUC “authorized no changes in rates” in adopting accounting procedure that prevented otherwise automatic rate reduction].)

In this case, for example, TURN asserts the settlement agreement deprived customers of refunds they would otherwise have been entitled to receive for overcollection of the four cent surcharge, that is, for any collections of the surcharge not needed for current power purchases, as was the originally stated purpose of the surcharge. But TURN cannot show PUC would, absent the settlement, have ordered refund of surcharge revenues used to pay procurement debts incurred during the crisis. While the March 2001 PUC decision allowing the surcharge stated that the surcharge’s purpose was to pay for ongoing power purchases and that surcharge revenues were “subject to refund if, at a later date, we determine that the utilities failed to use the funds to pay for future power purchases,” that decision explicitly did “not address recovery of past power purchase costs and other costs claimed by the utilities.” (Application of Southern California Edison Co., supra, Cal.P.U.C. Dec. No. 01-03-082, pp. 2, 60.) Later, the PUC explicitly permitted use of the *803surcharge revenues to pay utilities’ past power purchase debts. (Application of Southern California Edison Co., supra, Cal.P.U.C. Dec. No. 02-11-026, pp. 2, 4, 15-16.)7 Given PUC’s stated concern with restoring SCE to “reasonable financial health” in order that it be able to reliably provide electric power to its customers (Application of Southern California Edison Co., supra, at p. 4), we cannot assume PUC would have taken a different regulatory course absent the settlement.8

TURN also cites legislative history documents indicating that the 1975 amendment adding the language in Government Code section 11126, subdivision (d)(1) was intended to require that meetings for any PUC “deliberation on rate proceedings” or “at which rates of entities under PUC jurisdiction are considered” be open and public. (Legis. Analyst, analysis of Sen. Bill No. 1 (1975-1976 Reg. Sess.) as amended Apr. 24, 1975, p. 1; Assem. Ways & Means Com., staff analysis of Sen. Bill No. 1 (1975-1976 Reg. Sess.) as amended June 24, 1975, p. 2.) The cited history, however, does not indicate an intent to apply Government Code section 11126, subdivision (d)(1) to regulatory decisions other than rate changes. As the language of subdivision (d)(1) reflects, the primary legislative concern was with decisions to change rates; decisions leaving rates unchanged, but taking regulatory action that results in rates remaining at current levels longer than they otherwise might—the most that can be said about the settlement agreement’s effect—are not clearly within the legislative purpose. For this reason, and *804because we have earlier found a clear legislative intent, expressed in Government Code sections 11126, subdivision (e)(1) and 11126.3, subdivision (a), to allow settlement of pending litigation without a public meeting, we conclude PUC’s agreement to the settlement was not a meeting “at which the rates of entities under the commission’s jurisdiction are changed” for purposes of Government Code section 11126, subdivision (d)(1).

Question 3: Does the stipulated judgment violate section 454 of the Public Utilities Code by altering utility rates without a public hearing and issuance of findings?

Answer: No.

Section 454, subdivision (a) provides that “no public utility shall change any rate or so alter any classification, contract, practice, or rule as to result in any new rate, except upon a showing before the commission and a finding by the commission that the new rate is justified.” Contrary to the premise of the certified question, section 454 does not require PUC to hold a “public hearing” before allowing a change in rates. Indeed, the statute provides that PUC may adopt rules governing “the nature of the showing required” and “the form and manner of the presentation of the showing, with or without a hearing.” (§ 454, subd. (b), italics added; see also Wood v. Public Utilities Commission (1971) 4 Cal.3d 288, 292 [93 Cal.Rptr. 455, 481 P.2d 823] [“The Public Utilities Code does not require public hearings before rate increases or rule changes resulting in rate increases may be authorized”].)

Section 454 contemplates an “application” for a rate change by the utility and requires a “showing” in support of the application and a “finding” by PUC that the change is justified. How the statutory requirements of a showing and finding might be applied to a settlement agreement, rather than an application, is unclear. But the problem in applying section 454 is more fundamental still: SCE submitted no application for a change in rates. If, as appears from section 454, a major rate change may be made only by application (see Pacific Bell v. Public Utilities Com. (2000) 79 Cal.App.4th 269, 274 [93 Cal.Rptr.2d 910] [under § 454 and PUC’s implementing regulations, a utility may raise rates significantly only through the process of a formal application to PUC]), then a settlement agreement like that in dispute here, which resolved a federal court suit rather than an application before the commission, could not include a change in rates. We need not decide whether such a reading of section 454 is correct, however, because the settlement agreement effected no rate change subject to section 454.9

*805As discussed in relation to the second certified question, PUC agreed, in the settlement, to maintain SCE’s approved rates for a specified period, rather than to change them; nor did the other regulatory actions promised in the agreement change rates. TURN suggests that by allowing the current rates, including the surcharge, to be used for past procurement debts, the settlement established a “new rate” within the meaning of section 454. The premise of TURN’S argument is that, absent the settlement, rates would have been reduced when the freeze ended in March 2002 and wholesale prices dropped, making the surcharge unnecessary for current power purchases. In answering the first certified question, however, we explained that nothing in Assembly Bill 1890 required freeze rates to be changed after March 2002, and in answering the second question we noted that the original restriction on use of the surcharge revenue could have been, and was, eventually removed on grounds independent of the settlement. Again, to assert PUC would have reduced rates at any particular time, if not bound by the settlement to maintain them, would be to speculate. TURN’S effort to transmute a continuing rate into a new rate therefore fails.

Section 454, moreover, contemplates a formal application for a “change” in rates or for alteration of some condition of service so as to create a “new rate.” That a utility would formally apply merely to maintain a rate appears not within the statute’s contemplation. The setting of a future rate to be the same as the present rate, as here, is thus not within the purview of section 454, which focuses more narrowly on changed or new rates, which must be pursued by application.

For these reasons, we conclude PUC’s agreement to the settlement did not violate section 454’s requirement that a rate change or new rate be justified by a showing and finding.

CONCLUSION

In response to the court of appeals’ certified questions, we conclude that PUC’s agreement to the settlement and stipulated judgment did not violate the provisions of Assembly Bill 1890 and that the procedures employed in entering the stipulated judgment did not violate either the Bagley-Keene Act or Public Utilities Code section 454.

*806George, C. J., Kennard, J., Brown, J., Moreno, J., and Rushing, J.,* concurred.

I accept the majority’s conclusion that the terms of the settlement between the Public Utilities Commission (PUC or Commission) and Southern California Edison Company (SCE), which became the basis for a stipulated judgment in federal district court, did not exceed the Commission’s authority under Assembly Bill No. 1890 (1995-1996 Reg. Sess.) (Assembly Bill No. 1890), as codified in Public Utilities Code sections 330-396. (Stats. 1996, ch. 854, § 10.) Unlike the majority, however, I believe the process by which the PUC entered the settlement violated two other important statutes.

First, the PUC contravened the Bagley-Keene Open Meeting Act (Act or Bagley-Keene Act; Gov. Code, § 11120 et seq.). The Commission misused an exception in the Act, intended to permit closed meetings to “confer with, or receive advice from, . . . counsel” about pending litigation (id, § 11126, subd. (e)(1)), to approve in secret a legal settlement in which it guaranteed SCE billions of dollars in past and prospective rate relief, and thus “changed” the rates to be paid by SCE’s customers (id, § 11126, subd. (d)(1)).

Second, the PUC acted illegally under the Public Utilities Code, by so “changing]” SCE’s rates, through a secretly approved settlement, without any “showing before the [Cjommission and a finding by the [Cjommission that the new rate [was] justified.” (Pub. Util. Code, § 454, subd. (a), italics added.)

The Bagley-Keene Act was adopted to require state agencies to exercise their essential regulatory authority through public deliberations and decisions, subject to direct scrutiny and comment from the citizens whose daily lives these decisions affect. Because the PUC’s power over utility rates is especially crucial, the Legislature added specific provisions, in both the Bagley-Keene Act and the Public Utilities Code, requiring the Commission to make its rate decisions openly, and to follow formalities designed to ensure its determination that the approved rates are in the public interest.

The majority’s holding that the PUC could bypass these protections if it did so to settle litigation opens the door to a widespread danger of secret “government by lawsuit,” in which state agencies conduct their most important regulatory business in private, through the device of settling litigation between themselves and the entities they regulate. By the same device, the *807majority allow the PUC to engage in significant ratemaking decisions without showings or findings that the rates thereby set are just and reasonable. I cannot accept such a conclusion. I therefore respectfully dissent from the majority’s answers to questions 2 and 3 certified by the Ninth Circuit Court of Appeals.

I address the two critical statutes in turn.

A. Bagley-Keene Act.

The majority correctly note the general outlines of the Bagley-Keene Act, adopted in 1967 (Stats. 1967, ch. 1656, § 122, p. 4026) and often amended thereafter. The Act states “[i]t is the public policy of this state that . . . the proceedings of [covered state] agencies be conducted openly,” and declares the intent of the statute to be “that actions of state agencies be taken openly and that their deliberation be conducted openly.” (Gov. Code, § 11120, italics added.)

Accordingly, the Act mandates that, except as otherwise specifically provided, a covered “state body” (Gov. Code, § 11121.1) must (1) conduct its “meetings . . . open[ly] and public[ly]” (id., § 11123, subd. (a)), (2) provide advance public notice and an agenda for each such meeting (id., § 11125); (3) allow members of the public to address the body on each agenda item (id., § 11125.7, subd. (a)); and (4) permit public criticism of the body’s policies or actions (id., § 11125.7, subd. (c)). The Attorney General, a district attorney, or an interested person may sue to prevent future violations of the Act, or to determine the applicability of the Act to past or threatened future conduct by a state body. (Id., § 11130, subd. (a).) An interested person may also sue to obtain a judicial determination that an “action taken” in violation of the open-meeting requirements is null and void. (Id., § 11130.3, subd. (a).) Any member of a state body who attends a meeting of that body in violation of the Act, with intent to deprive the public of information to which the member knows or has reason to know the public is entitled under the Act, is guilty of a misdemeanor. (Id., § 11130.7.) “Except as expressly provided by [the Act], no closed session may be held by any state body.” (Id., § 11132.)

A “meeting” includes “any congregation of a majority of the members of a state body at the same time and place to hear, discuss, or deliberate upon any item that is within the subject matter jurisdiction of the state body to which it pertains.” (Gov. Code, § 11122.5, subd. (a), italics added.) An “action taken” includes “a collective decision” of the members and any “collective commitment or promise ... to make a positive or negative decision.” (Id., § 11122, italics added.)

*808Thus, except as otherwise specified, the Act (1) directly prohibits closed or secret meetings of state bodies to discuss or deliberate on public business, and (2) separately provides for nullification of the actions and decisions taken at such illegal meetings.

As the majority indicate, all agree that the PUC’s decision to approve the SCE settlement was an “action taken” at a “meeting” that did not conform to the open and public requirements of the Bagley-Keene Act. The PUC’s published agenda for the regularly scheduled commissioners’ meeting of October 2, 2001, listed “Conference with Legal Counsel—Existing . . . Litigation. Case name unspecified” as a matter to be discussed in closed session. As suggested by the official minutes of the October 2 meeting, the commissioners unanimously approved the “SCE [settlement” during the closed discussion, then reconvened in public session to announce their action.

To validate the “action taken” at this closed meeting, the PUC, SCE, and the majority invoke an exception to the open-meeting requirements, set forth in subdivision (e)(1) of Government Code section 11126. Subdivision (e)(1) states that “[n]othing in [the Act] shall be construed to prevent a state body, based on the advice of its legal counsel, from holding a closed session to confer with, or receive advice from, its legal counsel regarding pending litigation when discussion in open session concerning these matters would prejudice the position of the state body in the litigation.” (Italics added.) But neither the plain meaning of this language—whether read in isolation or in the overall statutory context—nor its legislative history supports the majority’s conclusion that the limited right to confer with counsel in closed session connotes the additional right to take final action in secret on the matter discussed.

The majority concede that “[o]n its face, subdivision (e)(1) permits a [state] body only to ‘confer with’ and ‘receive advice from’ its attorney regarding litigation.” (Maj. opn., ante, at p. 798, italics added.) Indeed, subdivision (e)(1) uses more restrictive language in this regard than the several other open-meeting exceptions contained in Government Code section 11126. These variously allow the state body, meeting in closed session, to “consider,” “discuss,” “deliberate on,” or even “give instructions” concerning the subject matter addressed. (See, e.g., id., subds. (a)(1) [state body may “consider” employee personnel matters], (c)(3) [state body may “deliberate on” quasi-judicial decision under Administrative Procedure Act], (4) [state body may “consider[]” term, parole, or release of prisoner if public disclosure of subject matter is prohibited by statute], (5) [state body may “consider” conferring of honorary degrees, or gifts, donations, or bequests, where donor desires confidentiality], (7)(A) [state body may “give instructions to” negotiator regarding purchase, sale, exchange, or lease of real *809property], (7)(E) [state body may “discuss[ ]” eminent domain proceedings], (8) [California Postsecondary Education Commission may “consider” appointment or termination of commission’s director], (9) [Council for Private Postsecondary and Vocational Education may “consider” appointment or termination of council’s executive director], (10) [Franchise Tax Board may “consider[ ]” appointment or termination of board’s executive officer], (16) [appropriate state body may “consider[ ]” investment decisions for retirement or pension funds], (17) [state body may hold closed sessions when “discharging responsibilities” with regard to labor negotiations], (18) [state body may “consider” matters posing criminal or terrorist threats to its personnel or property], (d)(2) [PUC may “deliberate” on disciplinary actions against any person or entity under its jurisdiction].)

Moreover, Government Code section 11126, subdivision (e) makes clear that the “confer with counsel” exception is not intended to grant state bodies a general license to decide in secret whether to enter settlements. Instead, the purpose of subdivision (e) is merely to preserve for a state agency, in the context of actual, threatened, or proposed litigation (see id., subd. (e)(2)(A)-(C)), a limited form of the privilege available to private litigants for confidential communications between lawyer and client. Subdivision (e)(2) specifies that “[f]or purposes of [the Act], all expressions of the lawyer-client privilege other than those provided in this subdivision are hereby abrogated. This subdivision is the exclusive expression of the lawyer-client privilege for purposes of conducting closed-session meetings pursuant to [the Act].” (Italics added.)

The legislative history of subdivision (e) of Government Code section 11126 confirms that a state body’s privilege to confer privately with counsel about pending litigation should be construed narrowly, to cover only lawyer-client consultation and advice. As originally adopted, neither the Bagley-Keene Act nor its local-agency counterpart, the Ralph M. Brown Act (Brown Act; Gov. Code, § 54950 et seq.), included any reference to an agency’s right to meet in private to consult with its counsel or discuss litigation. A subsequent Court of Appeal decision, Sacramento Newspaper Guild v. Sacramento County Bd. of Supervisors (1968) 263 Cal.App.2d 41 [69 Cal.Rptr. 480] (Sacramento Newspaper Guild), addressed whether the public-meeting provision of the Brown Act “abrogates by implication the statutory policy [of Evidence Code sections 950-952] assuring opportunity for private legal consultation by public agency clients.” (Sacramento Newspaper Guild, supra, at p. 55, italics added.) The Court of Appeal concluded that public agencies involved in actual or pending litigation, facing the same stakes and realities as private litigants, should have the same privilege as their private counterparts to “ ‘the effective aid of legal counsel,’ ” and thus to the “ ‘opportunity for confidential legal advice.’ ” (Id. at p. 56, italics added.)

*810The court reasoned that “[sjettlement and avoidance of litigation are particularly sensitive activities, whose conduct would be grossly confounded, often made impossible, by undiscriminating insistence on open lawyer-client conferences. In settlement advice, the attorney’s professional task is to provide his client a frank appraisal of strength and weakness, gains and risks, hopes and fears. If the public’s ‘right to know’ compelled admission of an audience, the ringside seats would be occupied by the government’s adversary, delighted to capitalize on every revelation of weakness.” (Sacramento Newspaper Guild, supra, 263 Cal.App.2d 41, 56, italics added, fn. omitted.)

The Legislature later codified this principle in the Bagley-Keene Act by adding former subdivision (q) to section 11126. (Stats. 1981, ch. 968, § 12, p. 3690.) As adopted in 1981, former subdivision (q) simply provided that “[n]othing in [the Act] shall be construed to prevent a state body from holding a closed session to confer with legal counsel regarding pending litigation when discussion in open session concerning those matters would adversely affect or be detrimental to the public interest.”

In 1987, however, the Legislature tightened and refined the Act’s provision for private conferences with counsel concerning pending litigation. (Stats. 1987, ch. 1320, § 2, p. 4762.) At that time, former subdivision (q) of Government Code section 11126 was rewritten in language roughly equivalent to that of current subdivision (e). The 1987 amendment removed permission for state bodies to meet with counsel in closed session about pending litigation whenever public discussion would adversely affect the “public interest.” Under the amendment, a closed-session consultation was allowed only when public discussion “would prejudice the position of the state body in the litigation.” (Gov. Code, § 11126, subd. (e)(1); see id., former subd. (q).) The amendment added the further proviso that, for purposes of the Act, the section is the “exclusive” expression of the lawyer-client privilege, which is otherwise “abrogated.” (Ibid.)

The 1987 amendment also included the extensive discussion, now contained in Government Code section 11126, subdivision (e)(2), of when “litigation shall be considered pending” for purposes of the privilege to confer in private with counsel. This requires that (1) an adjudicatory proceeding before a court, an administrative body, a hearing officer, or an arbitrator, to which proceeding the state body is a party, has already been initiated; (2) existing facts and circumstances have persuaded the state body, based on counsel’s advice, that it faces significant exposure to litigation; or (3) based on existing facts and circumstances, the state body has decided to initiate, or is deciding whether to initiate, litigation. (Gov. Code, § 11126, subd. (e)(2)(A), (B)(i), (C)(i); see id., former subd. (q)(1), (2)(A), (3).)

*811Finally, the 1987 amendment added the requirement, still in effect, that counsel prepare and submit to the state body, prior to the closed session if possible, but in no event more than a week thereafter, a memorandum stating “the specific reasons and legal authority for the closed session.” This memorandum must include, in the case of litigation not yet formally initiated, “the facts and circumstances” justifying a belief that the body faces a significant exposure to litigation or should decide whether to initiate such proceedings. (Gov. Code, § 11126, subd. (e)(2)(C)(ii); see id., former subd. (q).)

The source of the 1987 legislation was Senate Bill No. 200 (1987-1988 Reg. Sess.) (Senate Bill No. 200). The bill was described as “codif[ying] the exclusive use of the attorney-client privilege for the purpose of conducting closed sessions,” and as allowing “[c]losed sessions ... to seek the advice of legal counsel with regard to ‘pending litigation’ if discussion with legal counsel in open session would ‘prejudice the position’ of the public entity.” (Assem. Subcom. on Admin. of Justice, Analysis of Sen. Bill No. 200 (1987-1988 Reg. Sess.) as amended May 4, 1987, p. 1, italics added.)

Urging passage of Senate Bill No. 200 in the Assembly, the California Attorney General explained the concerns that had prompted the proposed legislation (which made conforming amendments to both the Brown and Bagley-Keene Acts): “Briefly, the problem is this: Several years ago the courts ruled that, absent specific legislation to the contrary, the Brown Act will not be construed to limit the availability of the traditional attorney-client privilege to local governmental bodies. [Citations.] This leaves . . . public agencies with broad freedom to go into executive session for confidential attorney-client discussion of virtually any issue which may involve ‘pending litigation’ or the ‘avoidance of litigation.’ [f] [Senate Bill No.] 200 will eliminate this loophole by placing clear and reasonable limitations upon when . . . governmental agencies may hold closed meetings to discuss legal issues. It protects the legitimate need of public officials to obtain confidential legal advice on issues which may end up in litigation but does not sacrifice the public’s right to open meetings. In short, the bill strikes an appropriate balance between two important but often conflicting principles of public policy.” (Atty. Gen. John K. Van de Kamp, letter to Assembly Member Lloyd G. Connelly, re Sen. Bill No. 200 (1987-1988 Reg. Sess.) July 10, 1987, italics added.)

The 1987 amendments, and the accompanying analyses and comments, do not directly address whether, during a closed lawyer-client litigation conference, a state body may make its final decision on how to resolve the pending proceeding. However, the amendments do confirm these general principles: First, Government Code section 11126, subdivision (e) defines, and strictly limits, a state agency’s exercise of its attorney-client privilege under the *812Bagley-Keene Act. (See Roberts v. City of Palmdale (1993) 5 Cal.4th 363, 373-381 [20 Cal.Rptr.2d 330, 853 P.2d 496] [construing parallel provisions of the Brown Act].) Second, this privilege has been statutorily narrowed over time, and is not coextensive with a private party’s rights to maintain secrecy in litigation matters. Third, the scope of the privilege has been carefully calibrated to allow the state body to conduct necessary private consultations with its counsel about pending litigation, while still maintaining, to the maximum possible extent, the Act’s overall requirement of public deliberation and decision. All these circumstances suggest that the privilege must be narrowly, not broadly, construed, where final decisionmaking by the agency is at stake.

The majority’s expansive interpretation of the privilege contravenes these tenets. The majority imply, on the basis of obsolete authority, that the public and private attorney-client privileges are coextensive. (Maj. opn., ante, at p. 798, citing Sacramento Newspaper Guild, supra, 263 Cal.App.2d 41, 55.) More importantly, the majority’s construction far exceeds a public agency’s need for attorney-client confidentiality, while unduly restricting the right of the people to public decisionmaking by state agencies.

As major support for their conclusion that the closed-session provision extends beyond the limited privilege to confer with counsel, and encompasses a final decision by the state body to accept a proposed settlement, the majority cite another provision of the Bagley-Keene Act, Government Code section 11126.3, subdivision (a). Under this provision, a state body that intends to consult its counsel in closed session about already existing litigation must publicly identify, by name or other specific means, the litigation to be discussed “unless the body states that to do so would jeopardize ... its ability to conclude existing settlement negotiations to its advantage.” (Ibid., italics added.) The PUC availed itself of the privilege not to identify the SCE settlement as the subject of its closed session on October 2, 2001, stating in its public agenda that “ ([d]isclosure of case name would jeopardize existing settlement negotiations).”

Focusing on the single word “conclude” in Government Code section 11126.3, subdivision (a), the majority reason broadly that this must mean the state body can use the cloak of confidentiality, not only to discuss the pros and cons of settlement with its counsel, but also to “conclude” the settlement. But this is a thin reed for the majority to grasp. Just as the inability to confer with counsel in private might compromise the agency’s strategy and jeopardize its ability to “conclude” a settlement to its advantage, a requirement that the agency prematurely identify the matter to be discussed in such a conference may also do so. But however confidential such preliminary legal discussions and negotiations may be, nothing in section 11126.3, subdivision (a) states or implies that the agency may actually resolve pending *813litigation in a regulatory matter without warning that a settlement of the particular case is imminent, explaining in public the proposed settlement terms, and allowing public response, at a public meeting, before making its final decision.

Certainly a state body may frankly discuss with its counsel, in private, the progress of ongoing settlement negotiations, including a candid assessment of the agency’s negotiating strategy, the “strength and weakness” of the agency’s position, and the “gains and risks, hopes and fears” a settlement entails, without affording its opponents “ringside seats” at these preliminary discussions. (Sacramento Newspaper Guild, supra, 263 Cal.App.2d 41, 56.) No doubt the agency may privately instruct its counsel, negotiating on its behalf, concerning terms it is inclined to accept.1 Moreover, it may well be that in subsequent public consideration of the matter, the state body need not fully disclose the litigation-related concerns that it discussed privately with its counsel under cover of the attorney-agency privilege, even if this means the public is less than fully informed about all the reasons the agency is tentatively prepared to accept a resolution on particular terms.

But none of this implies that a final regulatory decision, framed as the settlement of a pending lawsuit, itself can be undertaken without any public scrutiny or input, as occurred here. Government Code sections 11126, subdivision (e) and 11126.3, subdivision (a) were not intended to provide state agencies conducting the public’s business with the same right private litigants may have to resolve their disputes entirely away from the public’s prying eyes. Where significant regulatory decisions are at stake, parties involved in litigation with a state agency must understand that this is so. Whatever incidental litigation disadvantage this may impose on state agencies in the conduct of their regulatory business, as opposed to individuals and organizations in the conduct of their private affairs, is a necessary corollary to the express statutory policy of public decisionmaking.

To this extent, I am not persuaded by the California Attorney General’s construction of Government Code section 54956.9, the Brown Act analog to section 11126, subdivision (e)(1). (75 Ops.Cal.Atty.Gen. 14 (1992).) The Attorney General reasoned that the language of section 54956.9 (a local governmental body may meet in closed session to “confer with, or receive *814advice from, its legal counsel regarding pending litigation”) allows a local government body not simply to consult and confer in private on litigation matters, but also to take final action to settle a lawsuit.

For this conclusion, the Attorney General first cited language in the Brown Act’s “personnel” exception, now contained in subdivision (b)(1) of Government Code section 54957, which permits closed meetings “to consider the appointment, employment, evaluation of performance, discipline, or dismissal of a public employee . . . .” (Italics added; see Gov. Code, § 11126, subd. (a)(1) [parallel Bagley-Keene Act personnel exception].) Noting that several Court of Appeal decisions had construed the personnel exception to permit not only deliberation, but final action, the Attorney General asserted that the operative word “consider” in the personnel exception, and the operative word “confer” in the pending-litigation exception, were enough alike to dictate a similar interpretation of the latter provision. (75 Ops.Cal.Atty.Gen., supra, at p. 19.)

The Attorney General also reasoned that a public agency’s right to confer with counsel in secret about confidential litigation and settlement strategy necessarily implies the further right to decide, in confidence, what course to take. Otherwise, the Attorney General cautioned, an agency’s litigation adversaries would have “ ‘ringside seats’ ” for its decisions, and public litigants would be “ ‘second-class citizen[s],’ ” at a disadvantage compared to their private counterparts. (75 Ops.Cal.Atty.Gen., supra, at p. 19, quoting Sacramento Newspaper Guild, supra, 263 Cal.App.2d 41, 56.)

For reasons I have already discussed at length, I believe these conclusions are flawed. The words of the pending-litigation and personnel exceptions, respectively, are materially different. The former permits the public entity only to “confer with, or receive advice from, its counsel” in private (Gov. Code, § 11126, subd. (e)(1)), while the latter, by its use of the broader word “consider” (id., subd. (a)(1)), specifically allows active deliberation on the issue under discussion. In light of the 1987 narrowing of the pending-litigation privilege, and given the overall statutory policy of open deliberations and actions, these linguistic distinctions should not be conflated to allow a broad right of agencies to settle regulatory litigation in private.

Moreover, as we have seen, the current pending-litigation exception is not intended to afford public agencies litigation privacy entirely equivalent to that enjoyed by private parties. Instead, the statutory exception seeks to balance competing policies by providing a limited, and exclusive, form of attorney-client confidentiality for public agencies, while interfering as little as possible with the fundamental requirement that the collective actions of such agencies be taken in public. Thus, the pending-litigation exception does not imply a *815loophole allowing agencies covered by the Bagley-Keene Act to meet in secret to make final decisions on matters of significant regulatory interest.

In this age of high-stakes litigation, the majority’s contrary conclusion opens the door to secret “government by lawsuit,” allowing governmental bodies to exercise significant portions of their regulatory authority in private by the device of settling lawsuits between themselves and the entities they regulate. I cannot accept such a weakening of the clear purposes of the Bagley-Keene Act. I conclude that subdivision (e) of Government Code section 11126 did not permit the PUC to act in secret to finally approve a settlement of its litigation with SCE.

But even if Government Code section 11126, subdivision (e) generally allowed state bodies to approve regulatory settlements in closed session, respondent The Utility Reform Network (TURN) correctly urges that the PUC’s approval of this particular settlement nonetheless violated the Bagley-Keene Act. TURN points to another portion of section 11126—subdivision (d)(1)—that deals specifically with this agency and the subject matter of this settlement. Section 11126, subdivision (d)(1) provides that “[njotwithstanding any other provision of law, any meeting of the Public Utilities Commission at which the rates of entities under the [Cjommission’s jurisdiction are changed shall be open and public.” (Italics added.) By any common understanding, the PUC’s agreement to entry of the stipulated judgment in SCE’s federal action constituted the Commission’s commitment to “change[ ]” the electricity rates SCE could charge.

As the majority indicate, the parties to the October 2001 settlement agreed that, because of falling wholesale electricity prices during 2001, SCE’s existing rates “had allowed SCE to collect retail revenues in excess of current costs.” (Maj. opn., ante, at p. 791.) Among other components, these rates included emergency surcharges, totaling 4 cents per kilowatt-hour, which the PUC had granted to SCE, and to Pacific Gas and Electric Company (PG&E), in early 2001, at a time of very high wholesale power prices. When the PUC granted these surcharges, it had restricted their application to future power purchases, not past liabilities, and had made surcharge revenues refundable to ratepayers to the extent not used for this limited purpose. (Application of Southern California Edison Co. (2001) Cal.P.U.C. Dec. No. 01-01-018, pp. 2-3, 10-17, 24, 2001 Cal.PUC LEXIS 44; Application of Southern California Edison Co. (2001) Cal.P.U.C. Dec. No. 01-03-082, pp. 16-18, 60-61, 2001 Cal.PUC LEXIS 217.)

In the settlement, however, the PUC agreed, as its “principal substantive concession” (maj. opn., ante, at p. 791), to permit SCE to recover certain past costs by (1) applying the overcollections already in SCE’s coffers—i.e., *816its “cash on hand” (ibid.)—to this purpose and (2) “maintaining [SCE’s] existing rates until the end of 2003, if necessary,” to allow further such recovery (ibid., italics added). The settlement called for the establishment of a tracking account, known as PROACT, that would record SCE’s progress toward recouping these costs. (Ibid.) PROACT’s initial balance would be the gross amount of SCE’s accumulated past liabilities subject to recovery—an amount the parties agreed to be about $6,355 billion—less the surplus SCE had already collected. (Ibid.) Under this formula, the initial PROACT balance was estimated at some $3.3 billion. (Ibid.) The settlement rates would remain in effect until “the PROACT [account] was paid down to zero or . . . December 31, 2003, whichever came first. [Citation.]” (Ibid.)

The settlement thus authorized three fundamental “change[s]” in SCE’s rates. First, it either provided, or extended, a guaranteed duration to the rates in existence at the time of the settlement.2 Second, it cancelled, nunc pro tunc, ratepayers’ rights to refunds of amounts SCE had already collected under the 2001 surcharges, but had not used for ongoing power purchases as the terms of those surcharges originally required. Third, it removed, for the future, the original limitation on SCE’s use of the surcharges. That allowed SCE to continue to assess the surcharges, and to retain the revenues therefrom, under circumstances not permitted by the original terms and conditions of these special rates.

The majority insist the PUC did not agree in the settlement to “ ‘change[ ]’ ” SCE’s rates, but only made a “commitment ... to maintain the then existing rates for an agreed period.” (Maj. opn., ante, at p. 802.) This pinched and hypertechnical analysis ascribes too narrow a meaning to the broad statutory phrase “rates ... are changed.” (Gov. Code, § 11126, subd. (d)(1).) Surely it does not comport with the legislative purpose to ensure that the PUC’s core ratemaking decisions be open and public.

The complex and crucial task of ratemaking does not simply set the amount of money a utility may charge for a unit of service at any particular moment. It also necessarily establishes the terms and conditions attached to the authorized charge. When, as here, the PUC agrees to grant or extend a rate freeze, or alters the circumstances under which a rate may be charged or its revenues retained, it “change[s]” the rate.

*817The majority reject TURN’S argument that the settlement “changed” rates because it will lead to higher future rates than customers would otherwise have paid. This is an impossible standard to apply, the majority reason, because “it depends on the unknowable course of future events under hypothetical conditions.” (Maj. opn., ante, at p. 802.) In other words, the majority explain, for all we know, events other than the settlement, whether regulatory, legal, or economic, would have produced those same future rates.

This rationale misses the point. The settlement’s effect was to (1) provide, or extend, the guaranteed duration of a rate, (2) allow SCE to retain past and future surcharge revenues that would otherwise have been subject to refund, and (3) there by narrow the opportunity for electricity customers to obtain rebates, or to enjoy lower future rates, based on the conditions that might then prevail. In making these fundamental alterations in the structure of SCE’s existing rates, the settlement “changed” the rates. And the effect on SCE’s ratepayers was hardly minimal. The settlement’s avowed purpose was to enable SCE to secure from its customers billions of dollars in excess of current operational costs that were deemed necessary to restore SCE to financial health.3

The Legislature cannot have meant to allow ratemaking decisions with such significant effect on the public to escape the open-meeting requirement set forth in subdivision (d)(1) of Government Code section 11126. Indeed, this is confirmed by the legislative history of subdivision (d)(1), a factor discounted by the majority. The substance of present subdivision (d)(1) was adopted in 1975 as part of former subdivision (p). (Stats. 1975, ch. 959, § 5, p. 2238.) Legislative analyses of this provision, as enacted by Senate Bill No. 1 (1975-1976 Reg. Sess.) (Senate Bill No. 1), frequently described its effect as prohibiting the PUC from holding closed sessions for “deliberation on rate proceedings” (Legis. Analyst, analysis of Sen. Bill No. 1, as amended Apr. 24, 1975, p. 1; see also Assem. Off. of Research, 3d reading analysis of Sen. Bill No. 1, as amended Aug. 12, 1975, p. 1), and as requiring any PUC meeting where “rates ... are considered” to be open and public (Assem. Ways & Means Com., Staff Analysis of Sen. Bill No. 1, as amended June 24, 1975, p. 2).

The PUC urges that even if its initial acceptance of the settlement in closed session violated the Bagley-Keene Act’s open-meeting requirement, the Commission “cure[d]” or “correctfed]” the violation (Gov. Code, § 11130.3, subd. (a)) by two subsequent actions taken in public meetings. But the two *818actions the PUC cites merely implemented the terms of a settlement, and the resulting stipulated judgment, already reached in violation of the Act.

By P.U.C. Resolution No. E-3765 (2002), the Commission simply granted, with minor modifications, SCE’s request to establish the PROACT account called for by the settlement. (Cal. P.U.C. Res. No. E-3765 p. 37.) Indeed, in the resolution, the PUC rebuffed an argument by the California Manufacturers & Technology Association that granting SCE’s request would impermissibly change certain prior Commission decisions. According to the PUC, “this . . . issue arguefd] the legality of the [settlement, which [was] beyond the scope of’ the proceeding then before the Commission. (Id., at p. 35.)

The PUC also points to its Decision No. 02-11-026, 2002 Cal.PUC LEXIS 720, which modified Application of Southern California Edison, supra, Cal.P.U.C. Decision No. 01-03-082. As indicated above, Decision No. 01-03-082, issued in March 2001, had granted both PG&E and SCE a 3-cent surcharge (and had also made “permanent” an additional 1-cent surcharge granted to those utilities in January 2001), but had restricted use of these surcharges to ongoing power procurement and made them otherwise refundable. Decision No. 02-11-026 relaxed this restriction by allowing the 2001 surcharges to be used both for future power purchases and for the more general purpose of “returning each utility to financial health.” (Application of Southern California Edison Co. (2002) Cal.P.U.C. Dec. No. 02-11-026.)

But Decision No. 02-11-026 neither mentioned the SCE settlement nor reflected any effort to reconsider its terms. Moreover, as applied to SCE, repeal of prior restrictions on use of the 4-cent surcharge had already occurred by virtue of the settlement, and was required in any event by the resulting stipulated judgment in the federal action, entered October 5, 2001. Nothing in Decision No. 02-11-026 indicates it was a sincere and effectual attempt by the Commission to reassess the SCE settlement itself in an open and public meeting.4

*819For all these reasons, I would answer “yes” to the Ninth Circuit’s question whether the PUC’s approval of the SCE settlement in a closed session violated the Bagley-Keene Act.

B. Public Utilities Code section 454.

As indicated above, the Public Utilities Code separately provides, in pertinent part, that “no public utility shall change any rate or so alter any classification, contract, practice, or rule as to result in any new rate, except upon a showing before the [Commission and a finding by the [CJommission that the new rate is justified. . . .” (Pub. Util. Code, § 454, subd. (a) (section 454(a)), italics added.) The obvious purpose of the statute is to require a demonstration by the utility, and a resulting determination by the Commission, that the rate change is just and reasonable.

As discussed above, the settlement between the PUC and SCE constituted their agreement to a “change” in SCE’s rates. The settlement either froze rates or extended a freeze already in effect. It eliminated, for both past and future purposes, prior restrictions on SCE’s use of the 2001 surcharges. It cancelled ratepayers’ rights, both past and future, to refunds of surcharge amounts overcollected by SCE under the terms and conditions originally applicable to these rates. It thus afforded SCE the opportunity to recover from its ratepayers some $6,355 billion in liabilities already accrued by SCE, with approximately $3.3 billion of that amount to appear on their future electricity bills. Insofar as the PUC accepted this “change” without resort to the requirements of Public Utilities Code section 454(a), it acted illegally.5

SCE and the PUC argue that even if the settlement did authorize a change in SCE’s rates, Public Utilities Code section 454(a) has no relevance here. *820According to SCE and the PUC, the statute and its implementing regulations are concerned solely with the procedures by which a utility may seek the Commission’s approval to change the utility’s “tariff,” or published schedule of rates (see, e.g., Pub. Util. Code, §§ 489, subd. (a), 491; Cal. P.U.C. Gen. Order No. 96-A (1996) §§ I.B., III.C., VI; Pacific Bell v. Public Utilities Com. (2000) 79 Cal.App.4th 269, 274 [93 Cal.Rptr.2d 910])—procedures that simply do not pertain to a settlement and stipulated judgment in a lawsuit. (See also maj. opn., ante, at p. 804.)

Moreover, SCE and the PUC insist, a utility may change its rates only through the formal alteration of its tariff. (See Pub. Util. Code, §§ 489, subd. (a), 491, 532; Cal. P.U.C. Gen. Order No. 96-A, supra, § VI; see also Transmix Corp. v. Southern Pacific Co. (1960) 187 Cal.App.2d 257, 265 [9 Cal.Rptr. 714].) As SCE and the PUC observe, any modification of SCE’s tariff did not occur directly by virtue of the settlement and stipulated judgment, but only through implementing decisions, such as P.U.C. Resolution No. E-3765 (see discussion, ante), that resulted from formal Commission proceedings. Hence, these parties conclude, the settlement itself did not violate Public Utilities Code section 454(a).

Neither SCE nor the PUC cites authority holding that the Commission may authorize a utility rate change, by means of a legal settlement, without complying with the basic requirements of Public Utilities Code section 454(a). The majority in this case deliberately avoid that issue by concluding, erroneously in my view, that the instant settlement involved no “change” in SCE’s rates. (Maj. opn., ante, at p. 804.)

In any event, the technical arguments advanced by SCE and the PUC obscure the fundamental purpose of the scheme for public utility regulation. Such utilities are subject to control by the Legislature (Cal. Const., art. XII, § 3), which has mandated in the Public Utilities Act that their rates be “just and reasonable” (Pub. Util. Code, § 451). Regulatory authority over the rates of public utilities is vested in the Commission (Cal. Const., art. XII, §§ 1, 4, 6), which is responsible, through specified procedures, to assure that these rates meet the “just and reasonable” standard required by law (Pub. Util. Code, §§ 454(a), 728). Allowing the Commission to use a legal settlement to grant a significant change in a utility’s rates, without resort to a showing and finding that the change is just and reasonable, fundamentally undermines this regulatory structure. It invites such utility litigation as a means of “end-running” the established regulatory process.

As TURN suggests, the contention by SCE and the PUC that rates are “change[d],” for purposes of Public Utilities Code section 454(a), only upon completion of the tariff-setting process unduly elevates the ministerial act of *821implementing rate changes already mandated, in essential outline, by a prior Commission decision. Section 454(a) is superfluous unless it means that the fundamental determination whether a proposed change in rates is to be allowed at all can be made only upon a showing and finding, under normal regulatory procedures, that the change is just and reasonable.

I recognize that the regulatory process for approving rate changes is not readily compatible with the practicalities of settling lawsuits. My conclusion that Public Utilities Code section 454(a) nonetheless applies may well mean that the Commission simply cannot engage in significant ratemaking by such means. But any disadvantage this may ascribe to the Commission, or to a financially distressed utility, in a particular case is outweighed by the overarching regulatory policy of assuring that the rates paid by California’s utility customers are just and reasonable.

For all these reasons, I would answer “yes” to the Ninth Circuit’s question whether the settlement and stipulated judgment between SCE and the PUC violated Public Utilities Code section 454.

Appellant’s petition for a rehearing was denied October 22, 2003. Chin, J., and Brown, J., did not participate therein. Baxter, J., was of the opinion that the petition should be granted.