I respectfully dissent. Like Justice Panelli, I am concerned that the majority’s holding will lead to burdensome demands for discovery in situations not contemplated by the Legislature. Fortunately, the Legislature retains the authority to review this decision and take appropriate responsive action.
PANELLI, J.I respectfully disagree with the interpretation of the good cause requirement of Evidence Code section 1043, subdivision (b) offered by the majority opinion. Unlike the majority, I strongly believe that, in enacting Senate Bill No. 1436, 1977-1978 Regular Session (Stats. 1978, ch. 630, §§ 1-7, pp. 2082-2084; Pen. Code, §§ 832.5, 832.7 and 832.8; and Evid. Code, §§ 1043-1046), the Legislature meant to disavow the Caldwell court’s approval of Pitchess affidavits based solely on information and belief (Caldwell v. Municipal Court (1976) 58 Cal.App.3d 377 [129 Cal.Rptr. 834]) and to instead require traditional affidavits setting out the “materiality” of the information sought by the defendant to his prosecution and “reasonable belief” that the government agency has the information sought.
Pitchess and Caldwell have had a significant impact on the ease with which criminal defendants can obtain confidential police files. Filing a Pitchess motion has become a routine practice in prosecutions for resisting arrest or battery on a peace officer, regardless of whether the defendant or his attorney has any reason to believe the officers have actually used excessive force in the past. (Cal. Highway Patrol, Bill Rep. [on Sen. Bill No. 1436] (Sept. 5, 1978) p. 2.)
Though defendants certainly have the right to discover relevant information, my concern is the ease with which defense counsel may now obtain information from police officers’ confidential personnel files. In this case, Kennedy’s attorney merely submitted a form declaration in support of her Pitchess motion, with the names of the defendant and the police officers typed in the blank spaces. Based on this pro forma declaration, the trial court ordered an in camera inspection of the officers’ employment files, after which it would release “information regarding witnesses, complaining citizens . . ., only name and address, date of incident, telephone number.” In essence, this process was no more than “discovery upon demand,” a practice we clearly disapproved of in Pitchess (Pitchess v. Superior Court (1974) 11 Cal.3d 531, 536-537 [113 Cal.Rptr. 897, 522 P.2d 305]), and one the Legislature, by enacting Senate Bill No. 1436, clearly sought to stop. (Sen. Com. on Judiciary, Citizens’ Complaints & Peace Officer Personnel Records (Apr. 3, 1978) pp. 1, 7.)
*96The archive records reveal that Senate Bill No. 1436 grew out of dissatisfaction with the proliferation of Pitchess discovery.1 The Attorney General, author of the bill, informed the Governor that, “Senate Bill 1436 was drafted by my office in order to assist law enforcement officers . . . and represents a substantial step forward in protecting the[ir] rights . . . .” (Undated letter from Evelle Younger, Attorney General, to Edmund Brown, Jr., Governor.) The Senate Committee on the Judiciary described the purpose of the bill as, “to protect peace officer personnel records from discovery in civil or criminal proceedings” (Sen. Com. on Judiciary, Citizens’ Complaints & Peace Officer Personnel Records, op. cit. supra, at p. 1), stating also: “This bill is an attempt to cope with alleged law enforcement reaction (of shredding records to prevent discovery) to the California Supreme Court holding in Pitchess . . . .” (Id. at p. 7.) The Assembly Committee on Criminal Justice echoed this view of the purpose of the bill, stating that, “The thrust of this bill is to give the peace officer and his or her employing agency the right to refuse to disclose any information concerning the officer or complaints or investigations . . . .” (Assem. Com. on Crim. Justice, Bill Analysis [of Sen. Bill. No. 1436] (Aug. 7, 1978) p. 2.)
Those groups that considered the bill, unlike the majority here, understood that it constituted a significant restriction on defendants’ access to police personnel files. The State Public Defender, the California Attorneys for Criminal Justice, and the American Civil Liberties Union (ACLU) all opposed the bill, the ACLU stating: “In our view, the effect of SB 1436 is to preserve such records at the price of sealing them off forever from any public review.” (Mar. 30, 1978, letter from Brent Barnhart, legislative advocate for the ACLU, to Sen. Dennis Carpenter, the sponsor of the bill.) The California District Attorneys Association, the Peace Officers Research Association of California and the Attorney General all supported passage of the bill. Though these political advocates may be guilty of some hyperbole, criminal justice organizations, law enforcement groups, and Senate and Assembly committees readily gleaned that Senate Bill No. 1436 was intended to restrict access to police personnel files, rather than to grant access to these records on demand, as occurred in Caldwell, supra, 58 Cal.App.3d 377, and in the instant case.
The majority misinterprets editorial changes made to an early draft of Senate Bill No. 1436. Subdivision (b)(3) of Evidence Code section 1043 contains two prerequisites for Pitchess motions: “materiality” of the requested information to the defense and “reasonable belief” that the information exists in the police department’s files. The materiality requirement *97originally stated that a motion for the discovery of police personnel records must set forth “in full detail and upon personal knowledge the materiality thereof to the issues in litigation,” while the final draft dropped the requirement of full detail upon personal knowledge and stated that the motion must simply present “the materiality thereof to the subject matter involved in the pending litigation.” (Ante, p. 89.) Importantly, the deletion of personal knowledge language related only to materiality, not to the reasonable belief requirement. By deleting personal knowledge from the materiality requirement, the drafters approved of declarations by defense counsel regarding the legal relevance of prior wrongdoings of the officers. The drafters did not, however, alter the section’s requirement that the motion be based solely on affidavits.
The drafters also altered the language relating to reasonable belief, requiring that the motion describe the “type of records or information” rather than “the particular records or information sought.” (Evid. Code, § 1043, subd. (b)(2).) From this change, the majority infers that the Legislature did not intend to require the defendant to prove that prior complaints of excessive force exist in the officer’s file. (Ante, p. 92.) The majority ignores, however, the section’s later requirement that the motion state “upon reasonable belief that such governmental agency identified has such records or information from such records.” (§ 1043, subd. (b)(3).) Clearly, we must read these two subdivisions of section 1043 in tandem. Though the statute does not mandate the defendant to describe the contents of the file with particularity, it certainly demands a certain amount of prediscovery research into the officer’s past.
The majority claims that a requirement of prediscovery research would become an “insurmountable obstacle.” (Ante, p. 92, fn. 7; see generally ante, pp. 89-92.) Myriad defendants, however, have been able to establish the existence of other citizens’ complaints against the officers who arrested them, if not also able to locate the very individuals who filed those complaints. (See e.g., Jalilie v. Superior Court (1987) 195 Cal.App.3d 487, 489 [240 Cal.Rptr. 662]; cf. Craig v. Municipal Court (1979) 100 Cal.App.3d 69 [161 Cal.Rptr. 19].) The Pitchess defendant’s affidavit, based on the personal knowledge of the defendant rather than on information and belief, specifically listed the information sought, explained the unavailability of the citizens who lodged the complaints, and made an explicit showing of the information’s relevance to the claim of self-defense. (Pitchess, supra, 11 Cal.3d at p. 537.) The attorney in People v. Memro (1985) 38 Cal.3d 658, 674 [214 Cal.Rptr. 832, 700 P.2d 446], could also assert a “reasonable belief” that the police department had citizens’ complaints against the officers, because he was able to locate individuals who claimed they had *98been beaten and threatened by the same officers who had interrogated Memro.
Federal cases have reached a similar conclusion about the obligation of a criminal defendant to demonstrate the existence of relevant information in a police officer’s employment records, in the context of discovery under Brady v. Maryland (1963) 373 U.S. 83 [10 L.Ed.2d 215, 83 S.Ct. 1194], In United States v. Navarro (7th Cir. 1984) 737 F.2d 625, 630-632, the court disapproved of “[m]ere speculation” that a government file contains information relevant to the credibility of a witness as the basis for the file’s disclosure. (There, the defendant sought the Immigration and Naturalization Service file of a Drug Enforcement Agency (DEA) informant, who was a legal resident of the United States, surmising that the informant’s continued residency in the United States was conditioned on his cooperation with the DEA.) The court in United States v. Andrus (7th Cir. 1985) 775 F.2d 825, 842-843, relied on the holding of Navarro to prevent the discovery of police personnel files by a defendant accused of conspiracy to distribute cocaine. The defendant conceded he had no reason to believe the files actually contained information useful to impeach the officers at trial, but he sought their discovery nevertheless. “[W]ithout even a hint that impeaching material was contained therein,” the court ruled that the defendant was not entitled to the production of the files. {Id. at p. 843.)
Also, the majority deemphasizes the importance of section 1043’s requirement of an affidavit. (Ante, pp. 86-88.) The Legislature’s intent to disavow the holding of Caldwell and require a showing on personal knowledge of a “reasonable belief” that the agency in fact has the information sought is manifest from its inclusion of the affidavit requirement in section 1043. The drafters could have made the files of an officer automatically available to the defense; they could have required only an informal showing of relevance; they could have stated that a declaration on information and belief was adequate. Instead, they inserted the word “affidavit,” which carries with it an established legal meaning.
As the motion-practice equivalent of oral testimony, an affidavit must conform, as a rule, to the rules of testimony (McLellan v. McLellan (1972) 23 Cal.App.3d 343, 359 [100 Cal.Rptr. 258]; 6 Witkin, Cal. Procedure (3d ed. 1985) Proceedings Without Trial, § 29, pp. 345-346), including the requirement that it contain only facts personally known to the declarant (i.e., the rule against hearsay). (1 Jefferson, Cal. Evidence Benchbook (2d ed. 1982) § 18.8, p. 439; Evid. Code, § 1200.) The majority itself acknowledges that “an affidavit is normally presumed to state matters personally known to the affiant . . . .” (Ante, p. 87; see also City of Santa Cruz v. Superior Court (1987) 190 Cal.App.3d 1669, 1674 [236 Cal.Rptr. 155]; Star Motors *99Imports v. Superior Court (1979) 88 Cal.App.3d 201, 204 [151 Cal.Rptr. 721] [civil suit discussing affidavit in support of petition for writ of mandate].)
The majority opinion stresses that in other contexts, most commonly in regard to applications for search warrants, courts allow affidavits to be based on information and belief. (Ante, pp. 87-88.) As the majority points out, courts have long interpreted Penal Code section 1525 to permit magistrates to issue search warrants solely on the basis of hearsay allegations in police officer affidavits. (See e.g., Illinois v. Gates (1982) 462 U.S. 213 [76 L.Ed.2d 527, 103 S.Ct. 2317]; People v. Magana (1979) 95 Cal.App.3d 453 [157 Cal.Rptr. 173].) Courts have also demanded, however, much more than bald assertions by police officers that evidence of a crime exists at a particular location. Clearly, the assertions of Kennedy’s attorney in her declarations that “I am informed and believe based upon information contained in the police report as well as from defendant, that Officers Harris and Dickson . . . used excessive, unreasonable and unnecessary force to arrest defendant,” and that “I believe that other complaints of use of excessive force by the officer [sic] may have been filed by members of the public,” do not reach the level of reliability that we expect in an application for a search warrant. No magistrate could lawfully authorize the police to search a citizen’s home based on a form affidavit in which the officer had typed the suspect’s name and address and stated that he “is informed and believes” that the suspect is storing contraband there. Similarly, we should not misconstrue the legislative intent behind section 1043 as authorizing municipal courts to permit defense counsel to search police personnel files based on fill-in-the-blank affidavits containing boilerplate “information and belief” clauses.
The United States Supreme Court has formulated a “totality of circumstances” rule for determining probable cause to search a defendant’s home or person. (Illinois v. Gates, supra, 462 U.S. at p. 238 [76 L.Ed.2d at p. 548] [“The task of the issuing magistrate is simply to make a practical, commonsense decision whether, given all the circumstances set forth in the affidavit before him, including the ‘veracity’ and ‘basis of knowledge’ of persons supplying hearsay information, there is a fair probability that contraband or evidence of a crime will be found in a particular place.”].) This rule allows parties to submit affidavits containing hearsay, but requires affiants to prove the reliability of their informants. Moreover, the hearsay statements in such affidavits, included to show the state of mind of the affiant rather than for their own intrinsic truth, are admitted in order to demonstrate to the magistrate that probable cause to search exists. We should require no less in the Pitchess context.
*100I think a similar rule can work in the context of Pitchess motions. Municipal court judges could make a “commonsense decision” on whether defense counsel has made a good faith effort to locate prior complainants, witnesses, or any reliable information concerning the officer and whether the discovery motion stems from a real need for evidence in support of the defendant’s claim of self-defense or from a desire to freely peruse police files in order to create a claim of self-defense. (Cf. Craig, supra, 100 Cal.App.3d at p. 76.) By inserting the requirement that Pitchess motions include affidavits stating upon reasonable belief that the police department has records of prior complaints, the Legislature clearly intended defense counsel to undertake some amount of prediscovery research. Though I will accept the majority’s contention that “reasonable belief” is a less stringent standard than “with particularity,” I must argue that section 1043 requires more than mere conclusory allegations based upon “information and belief” without demonstrating the basis for such information and belief.
The majority opinion states several times that the use of force in Kennedy’s arrest, in and of itself, establishes good cause for perusal of the officers’ employment files. (Ante, pp. 80, 85-86, 93 [“While the recited facts do not, of course, establish that the force used was in fact excessive, it is altogether fair and rational, on the basis of such facts and averments, to entertain a ‘reasonable belief’ or inference that the same officers may have been accused of the use of excessive force in the past, and that such information may be found in their personnel records.”].) I disagree. We should not “entertain” the inference that the officers used excessive force in the past based on their use of force in the present arrest. Although the police report reveals that the officers had to use physical force to restrain Kennedy, the majority admits that it does not establish that the officers used excessive force in the arrest. (Ante, p. 93.) Moreover, the officers physically restrained Kennedy only after receiving a call that he had brandished a knife and threatened a passerby, and then only after he refused to submit to a pat-down search and took a swing at one of them. (Ante, pp. 78-79.) Objectively, would we not expect an officer to use some degree of physical force under these circumstances? Contrary to the majority’s view, the police report of Kennedy’s arrest (which Kennedy was required to append to his moving papers under Evid. Code, § 1046) affords no “reasonable belief” or cause to suspect that one of the arresting officers had ever been accused of acting violently in the past.
The majority also stresses that municipal court judges have generally refused to disclose verbatim reports from personnel files, ordering instead, as the judge in this case indicated he would order, the release of the names, addresses and telephone numbers of prior complainants and witnesses and the dates of the incidents in question. (Ante, p. 84; Kelvin v. Superior Court *101(1976) 62 Cal.App.3d 823, 828-829 [133 Cal.Rptr. 325].) We can not rely on the municipal courts to independently restrict discovery, however, when neither the Evidence Code nor case authority requires them to do so. In addition, the in camera review of an officer’s personnel file may reveal confidential personal information not relevant to the inquiry before the court. In cities the size of Santa Cruz, where the officer may frequently appear before the judge conducting the in camera review, the disclosure of such personal information would constitute an unnecessary loss of privacy and may have a deleterious effect on the officer’s relationship with the court.
Employment files by definition contain confidential information. Penal Code section 832.7 explicitly states that “Peace officer personnel records . . . are confidential,” and Penal Code section 832.8 delineates the contents of these records: “(a) Personal data, including marital status, family members, educational and employment history, or similar information; [¶] (b) Medical history; [¶] (c) Election of employee benefits; [¶] (d) Employee advancement, appraisal, or discipline; [¶] (e) Complaints, or investigations of complaints, concerning an event or transaction in which [the officer] participated, or which he perceived, and pertaining to the manner in which he performed his duties; or [¶] (f) Any other information the disclosure of which would constitute an unwarranted invasion of personal privacy.” The majority dwells on the importance of the discovery of complaints about the officer’s performance, while ignoring the loss of privacy which the officer suffers when anyone—whether judge, defense counsel or defendant—gains access to information regarding his life out of uniform. We should not give defendants the power to force such an invasion of privacy without a credible showing that the intrusion is absolutely necessary to a fair trial.
In short, several reasons exist compelling this court to reject the pro forma declarations used by Kennedy’s attorney in support of her Pitchess motion. We must first acknowledge the legislative intent to limit access to police personnel files. (Ante, pp. 95-96.) Second, we should respect the privacy rights of police officers and the confidentiality of their employment files. (Cal. Const., art. I, § 1; Pen. Code, § 832.7.) Third, as we expressly cautioned in Pitchess, the right of discovery created in that seminal case does not authorize defendants to embark on “fishing expeditions” into the confidential personnel files of law enforcement agencies. (Pitchess, supra, 11 Cal.3d at p. 538; People v. Cooper (1960) 53 Cal.2d 755, 770 [3 Cal.Rptr. 148, 349 P.2d 964].) Finally, defendants’ attempts to obtain the personnel files of police officers in order to buttress their claims of self-defense are merely a tertiary use of those records. Police departments are required to keep records of citizens’ complaints in order to discipline officers who use unnecessary excessive force (Pen. Code, § 832.5), not to aid the legal theories of future defendants. At the very least, we must give effect to our charge *102in Pitchess that defendants must demonstrate more than “ ‘a mere desire for the benefit of all information which has been obtained by the People in their investigation of the crime.’” (11 Cal.3d at p.537.)
The majority opinion has perhaps lost the forest for the trees, dwelling on the dissection of statutory language while ignoring the broader intent of Senate Bill No. 1436. Regardless of our interpretation of the terms “materiality,” “reasonable belief,” “personal knowledge” and “information and belief,” section 1043 instituted a “good cause” requirement for Pitchess discovery. Though we may argue over the exact parameters of that good cause requirement, we can not contest its existence. Good cause is certainly not satisfied by an unsupported statement that the police officers involved used excessive force in the past or by defense counsel’s reiteration of the defendant’s self-serving story that the officers used excessive force in his arrest. In enacting Senate Bill No. 1436, the Legislature clearly intended to discourage courts from using such paltry information in pro forma, boilerplate affidavits as the basis for granting access to the confidential employment files of peace officers. Today, the majority repudiates that legislative charge. The Legislature might do well to again turn its attention to Pitchess discovery, and to reconsider whether the relevant statutes, as construed today by a majority of this court, truly effectuate its intended goals and effects.
Eagleson, J., concurred.
All legislative history cited in the dissenting and majority opinions is found in the Governor’s Bill File for Senate Bill No. 1436.