Opinion

This case involves a challenge under the California Environmental Quality Act (Pub. Resources Code, § 21000 et seq.) (CEQA) to a decision by the City of Ceres (city) to grant approvals necessary to build a shopping center anchored by a Wal-Mart store. The challenger, Citizens for Ceres, has petitioned this court for writ relief from the trial court’s order upholding claims by the city and the developer that hundreds of documents be excluded from the administrative record because they are protected by the attorney-client privilege or the attorney work-product doctrine.

The dispute over these documents arose when the challenger pointed out that the administrative record prepared and certified by the city included no communications between the city and the developer. The city responded that the project had “the potential to be controversial”; that such communications therefore “were always made by and between legal counsel” for the city and the developer; and consequently all the communications were privileged. The challenger filed a motion to augment the administrative record by compelling the city to include the assertedly privileged communications. The trial court denied the motion, leading to these writ proceedings.

We reject the challenger’s argument that CEQA’s provisions defining the administrative record abrogate the attorney-client privilege and the attorney work-product doctrine. Those CEQA provisions do not reflect an intent on the part of the Legislature to eliminate privileges wholesale.

We conclude, however, that the common-interest doctrine, which is designed to preserve privileges from waiver by disclosure under some circumstances, does not protect otherwise privileged communications disclosed by the developer to the city or by the city to the developer prior to approval of the project. This is because, when environmental review is in progress, the interests of the lead agency and a project applicant are fundamentally divergent. While the applicant seeks the agency’s approval on the most favorable, least burdensome terms possible, the agency is dutybound to analyze the project’s environmental impacts objectively. An agency must require feasible mitigation measures for all significant impacts and consider seriously and without bias whether the project should be rejected if mitigation is infeasible or approved in light of overriding considerations.

The applicant and agency cannot be considered to be advancing any shared interest when they share legal advice at the preapproval stage. Under established principles, this means that the common-interest doctrine does not apply. After approval, by contrast, the agency and applicant have a united interest in defending the project as approved, and privileges are not waived by disclosures between them from that time onward. In making this distinction between preapproval and postapproval disclosures, we potentially disagree with California Oak Foundation v. County of Tehama (2009) 174 Cal.App.4th 1217, 1222-1223 [ 94 Cal.Rptr.3d 902 ] (California Oak), in which the court found the common-interest doctrine to be applicable to postapproval disclosures between an applicant and a lead agency and perhaps also to preapproval disclosures between them. We will grant writ relief to allow the trial court to apply the rule we have stated.

In the unpublished portion of the opinion, we consider several additional topics. First, we discuss the showing necessary to establish the common interest doctrine’s protection for any postapproval communications for which it may be claimed. Second, there are many other assertedly privileged documents that were not disclosed between the city and the developer. It will still be necessary for the trial court to reexamine those privilege claims because the court applied an incorrect standard in upholding them. In upholding all the challenged privilege claims without exception, the court expressed the view that the party asserting a claim of privilege need only assert it to obtain protection. In reality, the party asserting the privilege is required to make a showing of preliminary facts supporting the privilege. The court made no findings of these preliminary facts, and there is no substantial evidence in the record that would have supported those findings for any document. The city will be permitted to amend its submissions to make the necessary showings.

Next, some of the assertedly privileged documents are also claimed by the city to be excludable from the administrative record because they are “drafts” within the meaning of section 21167.6, subdivision (e)(10). The parties have a dispute over the scope of this exclusion. More broadly, the city argues that, even if none of the documents at issue are protected by privileges, they all belong to a phase of the environmental review that is excluded from the administrative record. This argument is based on a reading of section 21167.6, subdivision (e)(10), which, we conclude, it is unnecessary for us to rule upon in these writ proceedings. The trial court has not yet made any ruling on the subject and should do so in the first instance if necessary.

Finally, we reject four arguments for denying writ relief which are based on the allegations that (1) the challenger forfeited most of its challenges to the privilege claims by not presenting them properly in the trial court; (2) the challenger has not made a showing of prejudice; (3) the challenger failed to exhaust administrative remedies; and (4) the writ petition in this court is defective in form.

We issue a writ of mandate requiring the trial court to reconsider the claims of privilege in light of the holdings in this opinion.

FACTUAL AND PROCEDURAL HISTORIES

Real parties in interest Wal-Mart Stores, Inc., and Wal-Mart Real Estate Trust (the developer) applied to the city for land use approvals necessary to build a 300,000-square-foot shopping center anchored by a 200,000-square-foot Wal-Mart store. On September 12, 2011, the city certified an environmental impact report (EIR) and approved the project. The challenger initiated proceedings in the superior court, claiming the city failed to comply with CEQA.

After the city prepared a draft index for the administrative record, the challenger sent a letter to the city, stating: “The index . . . does not appear to include a single informal communication (such as [an] email or memo) between the agency and its consultants or the applicant. In my experience representing applicants as well as my experience with CEQA administrative records, there are typically lengthy communications between the applicant and the agency in this form and these are appropriately included in the record .... Yet the index is completely devoid of such communications or notes. In fact, it does not appear the agency staff/consultant e-mail accounts were reviewed for Communications related to this matter. Please explain whether this is an oversight that will be corrected or a deliberate omission, [f] If the latter, please explain the basis for the omission and if claimed for reasons of privilege, please provide a privilege log or similar device . . . .”

Counsel for the city answered in a letter stating that the omissions of communications between the city and the developer were deliberate and based on privilege. The city had deliberately structured all communications to be privileged because it anticipated that the project would be controversial and could lead to litigation, and that the city had no intention of providing any information at all about the withheld documents: “From the very earliest stages of the City’s consideration of this project, it was clear that the project had the potential to be controversial and that there was a relatively high risk of litigation. Thus, from the very earliest stages of the City’s consideration of the project, both the City and the project applicant retained legal counsel to assist with, and oversee compliance with CEQA and all other relevant laws and regulations. Communications, therefore were always made by and between legal counsel. These communications are protected from disclosure by the attorney-client privilege, the [attorney work-product] doctrine, the legislative privilege, the joint defense privilege, and, potentially, other privileges and protections. CEQA does not require the City to include any such privileged or protected documents in the administrative record or to waive any of these protections and privileges in preparing an administrative record. CEQA also does not require the preparation of a privilege log, as you have requested, and the City will not provide any such privilege log.” On December 19, 2011, the city certified the administrative record without including any of these communications.

The challenger filed an objection to the certification of the record because of the omission of the communications. Later, the challenger filed a motion asking the trial court to order the city to augment the administrative record to include them. The motion argued that communications between the city and the developer, as well as the city’s internal communications, were required to be included in the administrative record by section 21167.6, subdivision (e). The challenger further argued that, because section 21167.6 states that it applies “ ‘notwithstanding any other provision of law,’ ” no privileges applied.

In its opposition to the motion, the city informed the court that it had agreed to provide a privilege log, although it continued to maintain that it was under no obligation to do so. The log, as later supplemented, listed 3,311 documents. An overwhelming majority of the log entries indicated that the city was claiming the attorney-client privilege, the protection of the attorney work-product doctrine, or both. Many entries also indicated that, although the documents were disclosed between the city and the developer, waiver of privileges was prevented by the common-interest doctrine. The log actually refers to a joint-defense privilege, but, as we will explain, California has no joint-defense privilege. The city’s intention was to refer to the nonwaiver effect of the common-interest doctrine.

The city explained that there also were two groups of documents withheld for reasons other than privilege. One group was “administrative draft documents or documents not otherwise released to the public.” The other included documents related to a development agreement that was expected to be part of the project at an earlier stage but that had since been abandoned.

As far as we can tell from the record, the city provided little information on the basis of which it would be possible to determine whether any of the claimed privileges or protections applied. A total of about three dozen names appear in the privilege log as the names of people by or to whom documents were sent. With a few exceptions, however, neither the log nor any declaration supporting it provides any information identifying these people, stating which of them are attorneys or clients, or explaining which parties they represented or worked for. From the record as a whole, we have been able to identify six of them as attorneys for the city or the developer. There was, however, no straightforward way to identify the other 30 or so individuals listed. Further, although the record contains four declarations related to the assertedly privileged documents, none of these state the declarants’ personal knowledge that any of the documents were communications made in the course of an attorney-client relationship or were the work product of an attorney, with the exception of four items said to be attorney work product. Two declarations stated that the city and the developer sometimes disclosed privileged communications to each other and did so in pursuit of their common interests and with the expectation that the communications would remain confidential. The declarations did not, however, state that this was true, to the declarants’ personal knowledge, regarding any or all of the common-interest documents listed in the log except for four documents. For about 650 other documents where the protection of the common-interest doctrine is claimed in the log, there are no supporting facts.

The court held its first hearing on the motion to augment on April 20, 2012. The parties had formal meet-and-confer discussions in a jury room, and the hearing was continued. After the April 20 hearing, the challenger provided the city with a list of 2,275 privilege claims that it was disputing. On May 18, 2012, the city sent a chart indicating its responses regarding these disputed items. The responses indicated that the city had decided to disclose, and had already disclosed, a significant number of the documents. For the majority, however, the city adhered to its privilege claims—at least provisionally.

In a letter to the court dated May 23, 2012, describing the “issues that still remain between the parties,” the challenger reserved its right to maintain its challenges to the 2,275 items about which the city had supplied responses: “[Ojn May 18, 2012 the City provided responses ... to the Initial and Supplemental log lists of requested documents provided by Petitioner on April 20th and April 25th, respectively. Because these responses were not provided in conjunction with documents produced on May 7th, Petitioner has not had sufficient time to review these responses yet to determine their adequacy and/or whether they further answer issues raised herein. Thus, Petitioner reserves the right to object to any changes to the privilege logs or the City Responses.”

In a letter to the court dated May 24, 2012, the city stated that its review of the 2,275 challenges “has provided the opportunity to make necessary changes to the privilege log,” and it would submit an amended log “when the City is sure that no further changes will need to be made.” This implied the city was uncertain which of its privilege claims actually were valid.

In his letter to.the court dated May 23, 2012, the challenger’s counsel described the city’s production of “thousands of documents” after the April 20 hearing. These are documents conceded to be within CEQA’s description of the administrative record in section 21167.6, subdivision (e), but omitted from the administrative record that was certified by the city on December 19, 2011. Instead of promptly submitting these documents to the court to be included in the administrative record, the city apparently expected the challenger to review them and determine which ones it wanted to include. The city’s position was that it did not object to the inclusion of any of them. In a letter to opposing counsel dated May 7, 2012, counsel for the city remarked that, although the privilege log “references a large number of administrative draft sections of the EIR,” he found that it was “nearly impossible to match these documents to specific entries on the privilege log” his office had prepared months before. Further, the privilege log itself includes many entries for which privileges are claimed but no individuals are named as those participating in the communication. Ninety-three of these entries still had not been corrected by the time the parties submitted their final briefs preceding the July 6, 2012, hearing. This indicates that, at the time the city certified the administrative record and for seven months afterward, it never made a final determination of the documents it believed should be included in the administrative record or of the documents it wanted to claim it could withhold and why.

The court held another hearing on May 25, 2012. The court and parties discussed four categories of documents and agreed there were no other categories. The court and parties also expressed an expectation that the list of disputed documents would be narrowed by the time of the next hearing, so that the city and the developer could give a general indication of the substance and purpose of the documents, the challenger could make arguments based on that additional information, and the court could rule. The court stated that, after the parties had determined the set of documents remaining in dispute, they would submit simultaneous briefs arguing for and against the privilege claims on those documents. The court and parties did not refer to any specific limitation on the number of documents that could or would ultimately remain in contention.

When the simultaneous briefs were filed on June 26, 2012, it was obvious that the parties had not reached any agreement about the number of documents that remained in contention. The challenger attached to its brief a list of more than 500 documents and asked the court to order their disclosure. The city’s brief discussed 25 documents, implying that only these remained in dispute, while the developer’s brief asserted that the parties “have boiled the disputed issues down to” 19 documents and redactions in 13 other documents.

The belief of the city and the developer that a drastic reduction in the scope of the dispute had taken place appears to be based on the challenger’s letter of June 14, 2012, in which the challenger discussed 50 documents. That letter did not, however, state that it contained the challenger’s list of the privilege claims being challenged. Instead, it explained that it contained the challenger’s remarks on certain documents the city and developer had already produced, some of which the challenger was arguing should be included in the administrative record. It also stated the challenger’s responses to a new set of privilege claims asserted by the city in a recent letter. The letter did not contain any agreement to abandon the challenges to any of the privilege claims.

On June 29, 2012, counsel for the city and the developer wrote to counsel for the challenger to express their “outrage” at the fact that a large number of documents remained in dispute; they claimed there was an “extreme disconnect” between the challenger’s statements at the May 25 hearing and its list of disputed documents in its June 26 brief. Counsel for the challenger wrote back that the challenger “has repeatedly stated that [it] does not waive any rights or claims to documents” and “has never agreed to limit its request” for documents in the manner the city and the developer assumed.

The final hearing on the challenger’s motion to augment the administrative record took place on July 9, 2012. The court expressed surprise that a large number of documents were still in dispute. “I got the impression we were down to 30-some odd documents,” it said. This impression appears to have been derived from the city and developer’s submissions alone, since the challenger’s brief included a list of more than 500 disputed documents. The city’s attorney said there were “[c]lose to 700,” the developer’s attorney agreed, and the court often referred to that figure, but we have found no basis for it in the record.

The court indicated that if there had been a small number of documents, and if the city had been willing to provide them for in camera review, there would have been no problem. On the other hand, with the large number of documents that the city was unwilling to produce, the court did not know how to cope with the task of ruling on the privilege claims: “I thought we were down to 32 or 35, something manageable based upon these letters that I got ... so I came in here with the idea today, hey, I got 30-some odd documents I have to look at. Hopefully the city attorney will give me some of these things that may be attomey/client privilege, so I can look at that small universe of documents, and then I can make a decision, [f] Whichever way I can make a decision, I disclose some of those 30-odd documents or some of them I would or maybe I wouldn’t disclose any, but then I’d have a nice packet. And I’d seal it up and say, here, do what you want with it if somebody wants to take a writ. [][] But what am I supposed to do with 700 of these things? How do I make a determination if 700 documents I’m not being shown qualifies as attomey/client privilege? [1] Anyone?”

The court, later said it felt “blindsided” by the fact that the challenger was still challenging several hundred of the city’s original 3,311 privilege claims. “I went from 32 documents, and now I’m supposed to do 700,” it added, again apparently relying on the city’s and the developer’s representations about only a few documents being in dispute.

“That’s not what I had in mind.” A moment later, it said, “And I’m not going to look at 700, so give me an example.” Still later, it said, “I thought okay, well, 32, I can do 32. No problem. I can’t do 700 . . . .”

Several times, the court stated its view that a party asserting a privilege had no burden beyond the mere assertion itself, while the party opposing the privilege claim had a burden of proving the privilege was inapplicable:

Finally, saying, “I don’t know what else to do with this,” and “I’m getting to the point where we need to get this thing resolved one way or the other,” the court made a blanket ruling upholding all the privilege claims on the ground that the city and the developer had asserted the claims and the challenger had not disproved their applicability: “I’m making the finding you have as officer of the court, they have the right to say attomey/client work product. They have done so ... . You can’t make your requirement ... to show me that it’s not carried, and it’s not protected by attomey/client or work product . . . . [][] I’m not going to order any further review on these attomey/client [or] work product.” After discussing some other matters, the court set a hearing on October 5, 2012, for determination of the merits, with the challenger’s opening brief due on August 24.

The challenger filed its petition for a writ of mandate in this court on September 7, 2012, seeking relief from the trial court’s order. The petition argues that the allegedly privileged documents should be ordered included in the administrative record because section 21167.6 renders all privileges inapplicable. Alternatively, the petition argues that several hundred of these documents should be ordered included in the administrative record because respondents never made the necessary showing of preliminary facts to establish that the privileges apply to the documents for which they are claimed.

We issued a stay order on September 17, 2012. On the same day, the city filed a “Preliminary Opposition” to the writ On September 18, 2012, Wal-Mart filed an “Informal Opposition.”

The preliminary and informal opposition briefs make six arguments: (1) The matter is not ripe for review because the trial court has not filed a written order embodying its ruling from the bench, and Court of Appeal, Fifth District, Local Rules, rule 3(b) (rule 3(b)), requires “a copy of the order or judgment from which relief is sought” to be attached to the writ petition. (2) The city and developer “are currently asking the trial court to directly address” their purported failure to establish the preliminary facts necessary to show that privileges apply and “intend to ask the Superior Court to defer entry of an order pending preparation of a new privilege log by the City.” (3) The challenger forfeited its arguments because it blindsided the city and developer in the trial court by stating in its final brief before the hearing that several hundred documents were at issue. (4) The challenger has not shown that it will be prejudiced at trial by its lack of access to the withheld documents. (5) Having no access to the withheld documents, the challenger was not able to raise any issue that might be found in them during the administrative proceedings, so the challenger has failed to exhaust its administrative remedies for any issue that might be found in them. (6) Section 21167.6 does not supersede the attorney-client privilege or the attorney work product doctrine.

On October 3, 2012, we issued an order to show cause why relief should not be granted. Included in the order to show cause was a briefing order, stating:

“The parties’ submissions should include, but need not be limited to, responses to the following questions:

The city and developer filed returns in which they expanded on their earlier arguments and responded to the court’s questions. The challenger filed a traverse, responding to the court’s questions and stating that it intended to continue to challenge the “draft” designations. The city filed a reply to the traverse. We granted leave to amici curiae California State Association of Counties and League of California Cities to file a joint brief in support of the city and the developer.

DISCUSSION

I. CEQA does not abrogate privileges generally

The challenger’s most sweeping argument is that section 21167.6, subdivi- ' sion (e), supersedes evidentiary privileges because it requires materials to be included in the administrative record “[notwithstanding any other provision of law.” If correct, this would necessitate an order directing the trial court to reject all the city’s and developer’s privilege claims and require all documents within CEQA’s definitions to be included in the administrative record. We reject this argument.

Section 21167.6 provides:

Recently in Madera Oversight Coalition, Inc. v. County of Madera (2011) 199 Cal.App.4th 48 [ 131 Cal.Rptr.3d 626 ] (Madera Oversight), we made several observations about the contents of the administrative record as defined by these provisions. First, the language is mandatory: The administrative record shall include the listed items. Second, the list is nonexclusive; the administrative record’s contents include, but are not limited to, the listed items. (Id. at pp. 63-64.) Next, the administrative record as defined is very expansive. We quoted language that originated in one Court of Appeal case and was subsequently quoted in another: Section 21167.6 “ ‘ “contemplates that the administrative record will include pretty much everything that ever came near a proposed development or to the agency’s compliance with CEQA in responding to that development.” ’ ” (Madera Oversight, supra, at p. 64.) Fourth, the Court of Appeal does not directly review the agency’s decisions about what to include in the administrative record. Instead, it reviews the trial court’s decision on a party’s motion relating to the administrative record. It reviews the trial court’s findings of fact for substantial evidence and its conclusions of law de novo. (Id. at p. 65.) An appellant must affirmatively demonstrate error. Where the record is silent, the appellate court accepts all presumptions that support the trial court’s decision, including the presumption that the trial court made any necessary implied findings, so long as those findings are supported by substantial evidence. (Id. at p. 66.)

Several provisions of section 21167.6, subdivision (e), are of significance in this case, as they would require inclusion in the administrative record of types of documents the city apparently withheld, unless privileges are applicable and have not been waived by disclosure. Subdivision (e)(7)—requiring inclusion of all “written evidence or correspondence submitted to, or transferred from, the respondent public agency with respect to compliance with [CEQA] or with respect to the project”—encompasses correspondence between the applicant and agency pertaining to the project. Subdivision (e)(10)' includes “all internal agency communications, including staff notes and memoranda related to the project or to compliance with [CEQA].” Any materials for which the city claimed a privilege because they were communications among city staff in which the city’s counsel was included would fall within this provision, but for the privilege. Other provisions could be at issue as well. For instance, if an attorney was involved in the communication, the city might have withheld “staff reports and related documents prepared by the respondent public agency” that are relevant to CEQA compliance, the agency’s action on the project, or the agency’s findings, and that would belong in the administrative record if privileges are not applicable. (§ 21167.6, subd. (e)(2), (3).)

The challenger contends that, because section 21167.6, subdivision (e), applies “[notwithstanding any other provision of law,” its plain meaning requires inclusion in the administrative record even of documents that fall within the attorney-client privilege and the attorney work product doctrine. The challenger relies on the fact that, generally speaking, the expression “ ‘notwithstanding any other provision of law’ ” is “a ‘ “term of art” ’. . . that declares the legislative intent to override all contrary law.” (Klajic v. Castaic Lake Water Agency (2004) 121 Cal.App.4th 5, 13 [ 16 Cal.Rptr.3d 746 ], citation omitted.) The city and developer rely on California Oak, supra, 174 Cal.App.4th at page 1221 , which held that “[s]ection 21167.6 is not an abrogation of the attorney-client privilege or work product” doctrine. As the challenger points out, however, the California Oak opinion does not discuss the effect of the phrase “notwithstanding any other provision of law.”

We begin with a brief review of the attorney-client privilege, the attorney work product doctrine, and the purposes of both. A lawyer’s client has a privilege to refuse to disclose a confidential communication between the lawyer and the client made in the course of the lawyer-client relationship. (Evid. Code, §§ 952, 954.) “The party claiming the privilege has the burden of establishing the preliminary facts necessary to support its exercise, i.e., a communication made in the course of an attorney-client relationship. [Citations.] Once that party establishes facts necessary to support a prima facie claim of privilege, the communication is presumed to have been made in confidence and the opponent of the claim of privilege has the burden of proof to establish the communication was not confidential or that the privilege does not for other reasons apply. [Citations.]” (Costco Wholesale Corp. v. Superior Court (2009) 47 Cal.4th 725, 733 [ 101 Cal.Rptr.3d 758 , 219 P.3d 736 ] (Costco).)

The party claiming the privilege usually makes the preliminary showing via declarations. (Weil & Brown, Cal. Practice Guide: Civil Procedure Before Trial (The Rutter Group 2012) f 8:192, p. 8C-52 (rev. # 1, 2012).) In general, the court cannot require disclosure for in camera review of materials assertedly protected by attorney-client privilege. (Evid. Code, § 915; Costco, supra, 47 Cal.4th at pp. 736-737.)

The attorney work-product doctrine provides two levels of protection for attorney work product—absolute protection and qualified protection:

“(b) The work product of an attorney, other than a writing described in subdivision (a), is not discoverable unless the court determines that denial of discovery will unfairly prejudice the party seeking discovery in preparing that party’s claim or defense or will result in an injustice.” (Code Civ. Proc., § 2018.030.) Work produced by an attorney’s agents and consultants, as well as the attorney’s own work product, is protected by the attorney work product doctrine. (Scotsman Mfg. Co. v. Superior Court (1966) 242 Cal.App.2d 527, 531 [ 51 Cal.Rptr. 511 ].)

The attorney is the holder of this privilege. (Lasky, Hass, Cohler & Munter v. Superior Court (1985) 172 Cal.App.3d 264, 271 [ 218 Cal.Rptr. 205 ].) A party asserting the privilege must “prove the preliminary facts to show that the privilege applies.” (Mize v. Atchison, T. & S. F. Ry. Co. (1975) 46 Cal.App.3d 436, 447 [ 120 Cal.Rptr. 787 ].) When a party asserts the absolute privilege, the court cannot require the material to be produced for in camera review to evaluate the claim of privilege, but the court can require production for in camera review when the party asserts only the qualified privilege. (Evid. Code, § 915.) An opposing party seeking to overcome a claim of qualified privilege has the burden of establishing prejudice. (Coito v. Superior Court (2012) 54 Cal.4th 480, 499 [ 142 Cal.Rptr.3d 607 , 278 P.3d 860 ].)

The purpose of the attorney-client privilege is to enhance the effectiveness of our adversarial legal system by encouraging full and candid communication between lawyers and clients. (See, e.g., City & County of S. F. v. Superior Court (1951) 37 Cal.2d 227, 235 [ 231 P.2d 26 ].) The purposes of the work product doctrine are to “[preserve the rights of attorneys to prepare cases for trial with that degree of privacy necessary to encourage them to prepare their cases thoroughly and to investigate not only the favorable but the unfavorable aspects of those cases,” and to “[p]revent attorneys from taking undue advantage of their adversary’s industry and efforts.” (Code Civ. Proc., § 2018.020.)

The question presented here is difficult. California Oak does not explain why “any other provision of law” does not include the provisions establishing privileges. Roberts v. City of Palmdale (1993) 5 Cal.4th 363 [ 20 Cal.Rptr.2d 330 , 853 P.2d 496 ] (Roberts), on which the city and developer also rely, does not present an analogous situation. There, our Supreme Court held that the California Public Records Act (Gov. Code, § 6250 et seq.) and the Ralph M. Brown Act (Gov. Code, § 54950 et seq.) did not abrogate the attorney-client privilege and require disclosure of an attorney’s letter communicating legal advice to a city council. Both the Public Records Act and the Brown Act, however, have provisions expressly making exceptions to disclosure for privileged communications. (Roberts, supra, at pp. 370, 379 .) CEQA does not.

Despite the lack of controlling authority, we are persuaded that section 21167.6 does not mean agencies must disregard all privileges when assembling CEQA administrative records. Courts are required to go cautiously when interpreting statutes that might either expand or limit privileges, for we are forbidden to create privileges or establish exceptions to privileges through case-by-case decisionmaking. (Evid. Code, § 911, subd. (b) [no privileges exist except by statute]; Roberts, supra, 5 Cal.4th at p. 373 [courts may not find implied exceptions to privileges]; Dickerson v. Superior Court (1982) 135 Cal.App.3d 93, 99 [ 185 Cal.Rptr. 97 ] [same].)

The area of privilege is “one of the few instances where the Evidence Code precludes the courts from elaborating upon the statutory scheme.” (Cal. Law Revision Com. com., 29B pt. 3A West’s Ann. Evid. Code (2009 ed.) foll. § 911, p. 219.) Knowing this (see Voters for Responsible Retirement v. Board of Supervisors (1994) 8 Cal.4th 765, 779, fn. 3 [ 35 Cal.Rptr.2d 814 , 884 P.2d 645 ] [Legislature presumed aware of all existing law when it acts]; Bailey v. Superior Court (1977) 19 Cal.3d 970, 977, fn. 10 [ 140 Cal.Rptr. 669 , 568 P.2d 394 ] [same]), the Legislature did not likely intend to make CEQA administrative records a privilege-free zone by the indirect means of placing the phrase “[notwithstanding any other provision of law” at the beginning of section 21167.6, four subdivisions away from the administrative record provisions in subdivision (e). Our Supreme Court has explained that the policies behind the attorney-client privilege are just as applicable when the client is a public agency as in other contexts: “Open government is a constructive value in our democratic society. [Citations.] The attorney-client privilege, however, also has a strong basis in public policy and the administration of justice. The attorney-client privilege has a venerable pedigree that can be traced back 400 years .... It is no mere peripheral evidentiary rule, but is held vital to the effective administration of justice. [Citation.] The privilege promotes forthright legal advice and thus screens out meritless litigation that could occupy the courts at the public’s expense .... [f] A city council needs freedom to confer with its lawyers confidentially in order to obtain adequate advice, just as does a private citizen who seeks legal counsel .... The public interest is served by the privilege because it permits local government agencies to seek advice that may prevent the agency from becoming embroiled in litigation, and it may permit the agency to avoid unnecessary controversy with various members of the public.” ( Roberts, supra, 5 Cal.4th at pp. 380-381.) Similar considerations apply to the attorney work-product doctrine. In light of all this, we believe that if the Legislature had intended to abrogate all privileges for purposes of compiling CEQA administrative records, it would have said so clearly.

Our conclusion does not render meaningless the phrase “[n]otwithstanding any other provision of law” in section 21167.6. The phrase applies to the whole of section 21167.6, not just subdivision (e). One main effect the phrase has is to distinguish CEQA’s mandamus procedures from some of the procedures for mandamus actions generally, which are set out in the Code of Civil Procedure.

For all these reasons, we conclude that section 21167.6 does not abrogate the attorney-client privilege or the attorney work-product doctrine.

II. The common-interest doctrine does not protect agency-applicant disclosures made before project approval

The challenger argues that, because the interests of a lead agency and a project applicant diverge fundamentally while the project application is pending, the common-interest doctrine does not operate to prevent waiver of privileges when the agency and applicant disclose communications to each other during the application’s pendency. We agree.

The dispute over the administrative record in this case began when the challenger noticed that the record certified by the city contained no communications between the city and the developer. The city responded by saying that all these communications were privileged because they were made through counsel, and privileges were not waived by disclosure because of the common-interest doctrine. The doctrine therefore is central to the case.

The common-interest doctrine allows disclosure between parties, without waiver of privileges, of communications protected by the attorney-client privilege or the attorney work-product doctrine where the disclosure is necessary to accomplish the purpose for which the legal advice was sought. (Weil & Brown, Cal. Practice Guide: Civil Procedure Before Trial, supra, ¶ 8:199.10, pp. 8C-55 to 8C-56 (rev. # 1, 2012).) The doctrine is not an independent privilege but a doctrine specifying circumstances under which disclosure to a third party does not waive privileges. (OXY Resources California LLC v. Superior Court, supra, 115 Cal.App.4th at p. 889 (OXY Resources).) It does not mean there is “an expanded attorney-client relationship encompassing all parties and counsel who share a common interest.” (Ibid.)

The doctrine is based on Evidence Code sections 912 and 952:

“As used in this article, ‘confidential communication between client and lawyer’ means information transmitted between a client and his or her lawyer in the course of that relationship and in confidence by a means which, so far as the client is aware, discloses the information to no third persons other than those who are present to further the interest of the client in the consultation or those to whom disclosure is reasonably necessary for the transmission of the information or the accomplishment of the purpose for which the lawyer is consulted, and includes a legal opinion formed and the advice given by the lawyer in the course of that relationship.” (Evid. Code, § 952.) Although these provisions deal specifically with the attorney-client privilege, the same considerations apply to waiver or nonwaiver of the work-product doctrine. (OXY Resources, supra, 115 Cal.App.4th at p. 891 .)

The OXY Resources court stated the elements of the common-interest doctrine: “Applying these waiver principles in the context of communications among parties with common interests, it is essential that participants in an exchange have a reasonable expectation that information disclosed will remain confidential. If a disclosing party does not have a reasonable expectation that a third party will preserve the confidentiality of the information, then any applicable privileges are waived. An expectation of confidentiality, however, is not enough to avoid waiver. In addition, disclosure of the information must be reasonably necessary for the accomplishment of the purpose for which the lawyer was consulted. [Citation.] Thus, ‘[f|or the common interest doctrine to attach, most courts seem to insist that the two parties have in common an interest in securing legal advice related to the same matter—and that the communications be made to advance their shared interest in securing legal advice on that common matter.’ ” (OXY Resources, supra, 115 Cal.App.4th at p. 891 .)

Applications of Evidence Code sections 912 and 952 can be divided into two categories. The first is where the third party has no interest of his or her own in the matter, but a litigant must disclose a confidential communication to the third party because the third party is an agent or assistant who will help to advance the litigant’s interests. This is the category the Law Revision Commission described in commenting on Evidence Code section 912, subdivision (d) (which deals with necessary disclosures for purposes of various privileges, not just the attorney-client privilege), as part of the report it issued when the Evidence Code was proposed for adoption in 1965: “Subdivision (d) is designed to maintain the confidentiality of communications in certain situations where the communications are disclosed to others in the course of accomplishing the purpose for which the lawyer, physician, or psychotherapist was consulted. For example, where a confidential communication from a client is related by his attorney to a physician, appraiser, or other expert in order to obtain that person’s assistance so that the attorney will better be able to advise his client, the disclosure is not a waiver of the privilege, even though the disclosure is made with the client’s knowledge and consent. Nor would a physician’s or psychotherapist’s keeping of confidential records necessary to diagnose or treat a patient, such as confidential hospital records, be a waiver of the privilege, even though other authorized persons have access to the records. Communications such as these, when made in confidence, should not operate to destroy the privilege even when they are made with the consent of the client or patient.” (Recommendation Proposing an Evidence Code (Jan. 1965) 7 Cal. Law Revision Com. Rep. (1965) p. 162.)

The second category is where the third party is not in any sense an agent of the litigant or attorney but is a person with interests of his or her own to advance in the matter, interests that are in some way aligned with those of the litigant. In its comment on Evidence Code section 952 (dealing with the attorney-client privilege specifically), the Law Revision Commission reiterated much of what it said about section 912, but added a remark about disclosures bearing upon a matter of “joint concern”: “Confidential communications also include those made to third parties—such as the lawyer’s secretary, a physician, or similar expert—for the purpose of transmitting such information to the lawyer because they are ‘reasonably necessary for the transmission of the information.’. . . [f] A lawyer at times may desire to have a client reveal information to an expert consultant in order that the lawyer may adequately advise his client. The inclusion of the words ‘or the accomplishment of the purpose for which the lawyer is consulted’ assures that these communications, too, are within the scope of the privilege . . . . [f] The words ‘other than those who are present to further the interest of the client in the consultation’ indicate that a communication to a lawyer is nonetheless confidential even though it is made in the presence of another person—such as a spouse, parent, business associate, or joint client—who is present to further the interest of the client in the consultation. These words refer, too, to another person and his attorney who may meet with the client and his attorney in regard to a matter of joint concern.” (Recommendation Proposing an Evidence Code (Jan. 1965) 7 Cal. Law Revision Com. Rep., supra, p. 172.)

It is this last notion, “joint concern,” that is the basis of the common-interest doctrine. Evidence Code sections 912 and 952, however, make no reference to common interests or joint concerns; they refer instead to a reasonable necessity of disclosure. Those two sections give rise to the common-interest doctrine. This is because, in limited situations, the alignment of the parties’ common interests may mean disclosures between them are reasonably necessary to accomplish the purposes for which they are consulting counsel.

This, consequently, is the limited manner in which California has adopted a rule preserving privileges when parties with common interests disclose privileged communications to each other. The privilege survives disclosure to a party with a common interest only if it is necessary to accomplish the privilege holder’s purpose in seeking legal advice. The doctrine extends no further than this because in California there is no independent statutory joint defense or common interest privilege, and California courts are not authorized to establish one. The “. . . Federal Rules of Evidence provide that principles of common law govern rules of privilege . . .” and, consequently, the Ninth Circuit, for example, “has recognized a ‘joint defense privilege’ as an ‘extension of the attorney-client privilege’ since at least 1964. [Citation.]” (OXY Resources, supra, 115 Cal.App.4th at p. 888 .) We, by contrast, are “not free to create new privileges as a matter of judicial policy and must apply only those which have been created by statute.” (Dickerson v. Superior Court, supra, 135 Cal.App.3d at p. 99 .)

With this background in mind, we turn to whether a lead agency can share with the project applicant a preapproval interest in the creation of a legally defensible EIR that supports the applicant’s proposal. It is important to be clear at the outset that the common interest, if there is any, is in the creation of a legally defensible environmental document that supports the applicant’s proposal. There is no point in asking, as the city and developer in this case would have it, whether the applicant and agency have a common interest simply in the development of a legally defensible environmental document. This is because the developer has no interest in the development of an environmental document that does not support the developer’s proposal.

Before completion of environmental review and project approval, the law presumes the lead agency is neutral and objective and that its interest is in compliance with CEQA. It is this neutral role which could cause it to reject the project or certify an EIR supporting one of the project alternatives or calling for mitigation measures to which the applicant is opposed. The agency’s unbiased evaluation of the environmental impacts of the applicant’s proposal is the bedrock on which the rest of the CEQA process is based. In Save Tara v. City of West Hollywood (2008) 45 Cal.4th 116 [ 84 Cal.Rptr.3d 614 , 194 P.3d 344 ] (Save Tara), for example, our Supreme Court held that a city could not define “approval” of a project in such a way that its actual commitment to the project was already made before environmental review had been completed and the project formally approved. (Id. at p. 132.)

In Save Tara , the city entered into an agreement to develop property conditioned on subsequent environmental review and CEQA compliance. Before environmental review was completed, the city lent money to the developer for preparatory activities, announced publicly that it was determined to proceed with the project, and began relocating tenants whom the project would displace. (Save Tara, supra, 45 Cal.4th at pp. 140-142.) Our Supreme Court held that the city violated CEQA because it was committed to the project before going through environmental review. The defect in the proceedings was that the city’s actions tended “strongly to show that City’s commitment to the .. . project was not contingent on review of an EIR.” (Save Tara, supra, at p. 142 .) In other words, CEQA forbids an agency to be committed to accepting an applicant’s proposal before environmental review has been completed.

Save Tara applied a concept that is common in CEQA cases. This is the concept that a primary purpose and effect of CEQA is to require agencies to confront environmental impacts before deciding in favor of applicants’ projects. In Woodward Park Homeowners Assn., Inc. v. City of Fresno (2007) 150 Cal.App.4th 683 [ 58 Cal.Rptr.3d 102 ], for instance, we reiterated the point that “ ‘[t]here is a sort of grand design in CEQA: Projects which significantly affect the environment can go forward, but only after the elected decision makers have their noses rubbed in those environmental effects, and vote to go forward anyway.’ ” (Id. at p. 720, quoting Vedanta Society of So. California v. California Quartet, Ltd. (2000) 84 Cal.App.4th 517, 530 [ 100 Cal.Rptr.2d 889 ].) This means that the product of the agency’s efforts in conducting environmental review must reveal the true impacts of the proposed project, no matter how unattractive. The agency must unblinkingly include all significant impacts in the EIR and consider them with an open mind when deciding on project approval.

In our view, the lead agency’s obligation not to commit to the project in advance, but instead to carry out an environmental review process and create environmental documents that reveal the project’s impacts without fear or favor, and only then make up its mind about project approval, means the agency cannot have an interest, prior to project approval, in producing a legally defensible EIR or other environmental document that supports the applicant’s proposal. At the same time, of course, the applicant’s primary interest in the environmental review process is in having the agency produce a favorable EIR that will pass legal muster. These interests are fundamentally at odds.

The conflict between the agency’s interests and the applicant’s is far from being only theoretical. For many issues of the first degree of importance— whether an impact is significant or a mitigation measure is feasible, for instance—there may be substantial evidence on both sides. Either conclusion might survive judicial review. It is on issues like these that the preparers of environmental documents must make their most cmcial decisions. On these issues, the interests of the agency and applicant are opposed, even though they share an interest in producing a document that will be legally sufficient. The agency’s duty is to present the conclusion best supported by the facts, while the applicant’s interest is to present the conclusion most favorable to its proposal. These often will be opposite conclusions, although either would be legally sufficient.

In some cases, the common-interest doctrine may apply to parties that have interests that are partly common and partly opposed. (See OXY Resources, supra, 115 Cal.App.4th at p. 888 [concept of joint defense has expanded over the years to encompass, among other things, “ ‘parties who oppose one another in a case but are able to join forces on a particular issue of common interest’ ”]; STI Outdoor v. Superior Court (2001) 91 Cal.App.4th 334, 337-341 [ 109 Cal.Rptr.2d 865 ] [trial court erred in finding waiver by disclosure between public agency and private party negotiating for public contract even though contract was not yet finalized; disclosure of communications was reasonably necessary to further interests of both parties in finalizing it].) This, however, is not that kind of case. The relationship between a lead agency and project applicant is unique. Before project approval, the agency must objectively judge whether the project as proposed is environmentally acceptable and therefore must make a decision about whether it will align itself with the applicant in part, in whole, or not at all. Only after approving the proposal can the agency be said to join forces with the applicant. There may be, and typically are, extensive communications between them, but they cannot yet be said to be “on the same side.” Before project approval, therefore, the agency does not have even partially common interests with the applicant. The nature of its interest is held in abeyance until it decides whether to approve the project.

In saying this, we do not mean to imply that the members of an agency’s governing board are legally prevented from having a favorable opinion of a project from the outset. We also do not mean to imply that the agency and applicant should not work together on the EIR. Agencies and applicants routinely do so and CEQA contemplates that they will. The point is simply that the lead agency, as an agency, cannot have any commitment to the project as proposed until after environmental review is complete. This means its interests as it pursues the environmental review process are fundamentally not aligned with those of the applicant, and preapproval disclosure of communications by one to the other waives any privileges the communications may have had.

For similar reasons, the policies behind the attorney-client privilege and the attorney work-product doctrine do riot support the suspension of waiver principles when communications are disclosed between agency and applicant before project approval. The purpose of the attorney-client privilege is to enhance the effectiveness of the adversarial system by encouraging candid communication between lawyers and their clients. This purpose does not include encouraging strategizing between a private applicant and a government agency to meet a future challenge by members of the public to a decision in favor of the applicant if, at the time of the strategizing, the agency has not, and legitimately could not, have yet made that decision. The purpose of the attorney work-product doctrine is to allow attorneys to advise and prepare without risk of revealing their strategies to the other side or of giving the other side the benefit of their efforts. Before completion of environmental review, the agency cannot have as a legitimate goal the secret preparation, in collaboration with the applicant, of a legal defense of a project to which it must be still uncommitted.

One case is cited by the city and the developer that arguably contains a holding contrary to ours. In California Oak, supra, 174 Cal.App.4th 1217 , a county considering a project subject to CEQA received four documents from its outside counsel, which had been retained to advise it on CEQA compliance. The county disclosed the documents to counsel for the developer. (California Oak, supra, at p. 1221.) The county claimed the documents were protected by the attorney-client privilege, and the common-interest doctrine prevented waiver by disclosure to the developer. The challengers moved unsuccessfully for an order compelling the county to include the documents in the administrative record. On appeal, they argued that section 21167.6 superseded the attorney-client privilege, and the common-interest doctrine did not apply. (California Oak, supra, at pp. 1220-1221.) The Court of Appeal rejected these arguments. On the common-interest doctrine, the court stated: “[The challenger] argues that [the county’s] communication to [the developer] was not reasonably necessary for the accomplishment of the purpose for which [the county] took advice from the outside counsel. To wit: ‘[T]his purpose—to achieve compliance with CEQA—differed from [the developer’s] purpose, which was to defend their permits against a CEQA [lawsuit].’ [The challenger] takes too crabbed a view of [the county’s] purpose in considering the advice of the outside counsel, [ft] The purpose of achieving compliance with the CEQA law, reasonably viewed, entails a further purpose. It includes producing an EIR that will withstand a legal challenge for noncompliance. Thus, disclosing advice to a codefendant in the subsequent joint endeavor to defend the EIR in litigation can reasonably be said to constitute ‘ “involvement of third persons to whom disclosure is reasonably necessary to further the purpose of the [original] legal consultation.” ’ ” (California Oak, supra, 174 Cal.App.4th at pp. 1222-1223, italics added.)

The italicized phrase above is somewhat ambiguous. It arguably means the disclosure by the agency to the applicant took place after the project was approved and legal defense of the agency’s decision had commenced. The application of the doctrine to a disclosure taking place at that stage is straightforward. After environmental review is complete and the agency has certified the EIR and approved the project, there is no longer any conflict between the agency’s role as an ally of the developer and its role as an objective evaluator of the project. Nothing in the published portion of the opinion directly indicates when the disclosure took place.

The city and the developer argue that the court’s remarks imply that it intended to refer to all privileged communications between the agency and applicant, not just those taking place after project approval. They say the statement that the agency’s purposes include producing a legally sufficient EIR mean the court “must have been addressing communications relating to the production of a legally compliant EIR, not just defending an already approved EIR.” If so, we disagree for the reasons we have stated and decline to follow California Oak .

The developer also cites San Bernardino Valley Audubon Society, Inc. v. County of San Bernardino (1984) 155 Cal.App.3d 738 [ 202 Cal.Rptr. 423 ]. There, the Court of Appeal held that a project applicant was required to pay half of the attorney’s fees awarded to a plaintiff who successfully challenged an EIR, while the agency paid the other half. The court stated that the developer worked with the agency’s staff in developing the EIR, and the threat of a fee award would incentivize applicants to try to make sure EIR’s are legally adequate. (Id. at p. 756.) This case does not show that an agency and applicant have a common interest for purposes of the common interest doctrine. We agree that an EIR is often the work of both the agency and the applicant, and that the applicant has an interest in making sure the EIR is compliant with CEQA. Yet these facts do not show that the interests of the agency and applicant in the development of the EIR and completion of the environmental review process are the same. A fundamental tension between the agency’s interest in objective environmental analysis and the applicant’s interest in obtaining approval of the proposed project prevents those interests from being aligned before the project is approved. Both have an interest in producing a legally adequate EIR, but as we have said, the agency cannot share the applicant’s interest in an EIR that supports the project as proposed until the environmental review process is complete.

Contrary to city and developer’s contentions, our holding is not affected by the proposition that the applicability of the common interest doctrine does not depend on the commencement of litigation. The attorney-client privilege and the attorney work-product doctrine apply to many situations not yet involving litigation or never involving it. (See, e.g., Roberts, supra, 5 Cal.4th at p. 371 ; STI Outdoor v. Superior Court, supra, 91 Cal.App.4th at pp. 340-341.) The crucial point in time for our purposes here is project approval, not the commencement of litigation. That point is crucial for the reasons we have stated. The time of commencement of litigation has no significance.

For all these reasons, we conclude that the city and developer have waived the attorney-client privilege and the protection of the attorney-client work product doctrine for all communications they disclosed to each other before the city approved the project. Consequently, any such communications that fall within section 21167.6, subdivision (e), must be included in the administrative record.

As we have already indicated, the situation is different after project approval. Then the agency’s and applicant’s interests are aligned, assuming the approval has not left any dispute remaining between them. Both are legitimately committed to the same thing at that point—defending the project as approved. Under those circumstances, there is nothing about the agency-applicant relationship that would stand in the way of applying the common-interest doctrine, assuming its elements are satisfied with respect to the particular communications for which its protection is claimed.

III.-V.

DISPOSITION

Let a writ of mandate issue directing the superior court to do all of the following:

Costs are awarded to Citizens for Ceres.

This court’s order staying the proceedings in the superior court is vacated.

Cornell, J., and Franson, J., concurred.

Subsequent statutory references are to the Public Resources Code unless otherwise noted. In a number of places, the city and developer assert that the court refused to rule op all but a few of the city’s privilege claims because it felt the challenger had not made timely challenges to the others. As will be seen, the record does not support this view. The court never stated that the challenger forfeited any of its challenges or that any of the city’s privilege claims were being upheld through some kind of default. The record can reasonably be read only as showing that the court upheld all the privilege claims on their merits. See footnote, ante, page 889.