I concur with majority in its holdings regarding counts II and III of appellant’s complaint on the ground it specifies, namely, that appellant sued the wrong party in these counts.
*1373I respectfully dissent, however, from the majority’s disposition of count I, the administrative mandamus count. In the first place, I believe the trial court made a fundamental error of law in the course of adjudicating count I. Second, I disagree with the majority’s dicta implying that administrative mandamus does not lie because the agency was exercising a “quasi-legislative” function. Third, I would hold that respondent San Francisco Civil Service Commission (CSC) abused its discretion by canceling the promotional tests when, as here, (1) that cancellation adversely impacted another protected minority, (2) the staff report it commissioned and ended up approving (hereafter Brady report) declared so unequivocally that the tests were “job-related,” and (3) the cancellation was so patently the result of pressures by the union representing the status quo.
I will deal with these issues in the order just noted.
A. The Trial Court Misapplied the Law Concerning Count I
First of all, I believe the record makes clear that the trial court made an error of law in disposing of count I. This is a recognized basis for reversal of a trial court’s decision in a Code of Civil Procedure section 1094.5 (hereafter section 1094.5) action. (See Magit v. Board of Medical Examiners (1961) 57 Cal.2d 74, 87 [17 Cal.Rptr. 488, 366 P.2d 816]; Webb v. Miller (1986) 187 Cal.App.3d 619, 624 [232 Cal.Rptr. 50]; see also Cal. Administrative Mandamus 2d (Cont.Ed.Bar 1989) § 14.24, pp. 460-461.)
The error of law that is, I believe, plain from the record is that the superior court judge, who heard the administrative mandamus petition, quite apparently thought the sum total of the issue before the CSC at its hearing on April 3, 1995, was whether the tests had an adverse impact on African-Americans. This was incorrect. There were clearly two issues: (a) was there an adverse impact upon a cognizable class of persons, and (b) if so, were the examinations nonetheless “job-related”? That this is a two-prong test is made clear in the applicable California regulations. Thus, under California Code of Regulations, title 2, section 7287.4, subdivision (a), a showing of an adverse impact makes the practice “unlawful unless the policy or practice is job-related . . . .” And under title 2, section 7287.4, subdivision (e), if the “testing device” has an “adverse impact,” it is then “permissible only upon a showing that the selection practice is sufficiently related to an essential function of the job in question to warrant its use.” (Italics added.)
It is, I suggest, patent from these passages that adverse impact and job relatedness are two separate and distinct issues. This conclusion is fortified by a decision from Division Five of this district, City and County of San *1374Francisco v. Fair Employment & Housing Com. (1987) 191 Cal.App.3d 976 [236 Cal.Rptr. 716]. There, citing two United States Supreme Court cases, Justice Haning wrote: “Once the racially adverse impact of the examination was established, the burden of proof shifted to the City to prove the examination job-related.” (Id. at p. 989.)1
But that, clearly, is not how the trial court understood matters during the September 1997 hearing in front of it. At several different points during oral argument by appellant’s counsel, the judge made plain that, in his view, all that was necessary to sustain the action of the CSC in canceling the tests was the (admitted) fact that the written part of the test adversely impacted African-Americans. Thus, at one point, after appellant’s counsel had pointed out that the CSC staff had made an “undisputed admission that the test was job related,” the judge asked: “How can it have adverse impact and be solely job related?” At another point, after the same counsel argued that one “can’t say well, because a test had adverse impact, therefore, probably it wasn’t job related,” the judge asked: “Why can’t you do that?”
It is thus obvious that the judge thought he was dealing with a single-factor test, i.e., was there an adverse impact on a protected class? Once that question was answered affirmatively (as it clearly was here), he apparently concluded he could not find that the administrative agency abused its discretion. In fact, as we have seen, the regulations and cases make clear that what is involved is a two-prong test. The trial court should have gone on to examine whether there was anything in the record to establish whether or not the tests were job-related.2 Because it did not do so, I conclude that it erred on an issue of law, requiring us to reverse.
B. Administrative Mandamus Was Appropriate
In two brief paragraphs in its opinion (maj. opn. ante, at pp. 363-364) the majority suggests although it admittedly does not decide that count I may be deficient because it attacks administrative action which is essentially “quasi-legislative” as opposed to “quasi-judicial.” I disagree with the majority’s dicta.
First of all, it is noteworthy that the respondents did not raise this issue in support of the outcome below. Nowhere in their briefs is their the slightest *1375mention of the agonizingly arcane dichotomy between “quasi-legislative” and “quasi-judicial” actions of administrative agencies. Nor was the point raised in oral argument by respondents.
But, second, and mindful of the refrain that appellate courts may affirm “on any basis presented by the record,” I disagree with the suggestion of the majority as to the proper characterization of this administrative action. The difference between “quasi-legislative” and “quasi-judicial” functions of an administrative agency is explored thoroughly in two leading texts. (See 8 Witkin, Cal. Procedure (4th ed. 1997) Extraordinary Writs, §§ 268, 269, pp. 1067-1071; Cal. Administrative Mandamus, supra, § 1.17, pp. 18-20 and § 3.11, pp. 81-82.) The appellate decision which most helpfully discusses the distinction is Pacifica Corp. v. City of Camarillo (1983) 149 Cal.App.3d 168, 176-178 [196 Cal.Rptr. 670]. That case stands for the proposition that when the administrative proceedings implicate the “application of specific standards to existing facts” (id. at p. 176), they are quasi-judicial. Indeed, a review of the authorities characterizing action as “quasi-legislative” reveals almost all of them to be cases involving an agency’s adoption of rules, regulations, guidelines, etc.
By contrast, what was involved here was an appeal by five individuals who complained that the cancellation of the tests violated the California Fair Employment and Housing Act (FEHA) (Gov. Code, § 12900 et seq.) and the regulations issued under it, and hence discriminated against them as members of an under-represented group. Indeed, the Brady report to the CSC summarized the issue before that agency as follows: “Five candidates, four of whom placed at or near the top of the two eligible lists, appealed the decision of the Director to cancel the examinations to the Civil Service Commission: . . . HQ . . . Two of the appellants [including this appellant] contend that they are members of under-represented groups in the management level of MUNI and they would be further adversely affected by nullifying the examination.”
I respectfully submit there is no reasonable way the resolution of such an issue can possibly be characterized as “quasi-legislative.”
C. The Agency Abused Its Discretion by Canceling the Promotional Tests
I do not understand the majority to contest the proposition that if, as and when an administrative agency abuses its discretion in a quasi-adjudicative setting, relief under section 1094.5 is available. (See Quackenbush v. Mission Ins. Co. (1996) 46 Cal.App.4th 458, 466 [54 Cal.Rptr.2d 112]; Saleeby v. State Bar (1985) 39 Cal.3d 547, 562 [216 Cal.Rptr. 367, 702 P.2d 525]; see *1376generally, 9 Witkin, Cal. Procedure, supra, Administrative Proceedings, § 114, pp. 1158-1159.) Nor, in turn, do I contest the fact that the record made before this administrative agency demonstrated that the examinations in question, particularly the written part, had an adverse impact upon African-Americans.3
As noted at the beginning, I part company with the majority because of three circumstances which I believe they do not adequately , weigh: (1) the adverse impact of the cancellation upon another protected minority, women, (2) the effectively-conceded “job relatedness” of the tests in question, and (3) the fact that, when confronted by union pressure to cancel the examinations, the CSC effectively eschewed any exercise of discretion.
1. Adverse Impact via Test Cancellation
The appellant here claims, and claimed before the trial court in her mandamus hearing, that she was adversely impacted by the cancellation of the tests. Although the point is well buried in two federal cases decided under title VII, I think it is nevertheless clear that employment-related action such as canceling promotional tests or “freezing” promotional lists may be the sort of acts which trigger that federal statute and, a fortiori, the FEHA. (Cf. U.S. v. City and County of San Francisco (N.D.Cal. 1987) 656 F.Supp. 276, 283-285; Giles v. Ireland (11th Cir. 1984) 742 F.2d 1366, 1376-1378.)
Clearly, women are a protected class under the FEHA. (See Gov. Code, §§ 12920, 12940.) More importantly, I believe that appellant, in her mandamus papers to the trial court, made out a prima facie case of at least adverse impact caused by the cancellation of the examinations. This prima facie case rested principally on the fact that women were and are underrepresented in the management ranks of San Francisco’s Municipal Railway (Muni).
As already noted, the Brady report squarely presented appellant’s “underrepresented group” issue to the CSC. As I have also suggested, I believe the record shows that her contention had at least a prima facie basis. She *1377presented to the superior court4 evidence showing that the present Muni management at the pertinent levels is composed of only 6 percent women (2 out of 33) as compared with a universe of the total “available labor pool” in the San Francisco Bay Area of approximately 46 percent. And, according to more general figures supplied by the Muni itself, as of 1994 it had “unrealized goals” of between 15 percent and 47 percent of women in overall management, professional and supervisory roles. Finally, appellant also produced evidence that, prior to the cancellation of the tests in December 1994, the next most recent promotional lists for the position for which she tested expired as of 1989 and 1990. Thus, the cancellation of the tests on which appellant scored so well allowed Muni management to fill the positions she sought on a “temporary appointment” basis. Which, of course, is exactly what happened: after the cancellation, all of them men and two of them African-American, were appointed to fill the vacancies which were supposed to be filled via the canceled tests.
I submit that, on this record, appellant had established at least a prima facie case of adverse impact by the nullification of the tests.
2. Job Relatedness
It is on this issue that I principally part company with the majority’s reasoning. My colleagues seem to feel that the absence of a formal “validation study” regarding the test means that the test cannot be “job-related.” I profoundly disagree, and for several reasons.
First of all, the whole “validation study” business is strictly a creature of federal regulations adopted under title VII. Some of part 1607 of title 29 of the Code of Federal Regulations (1997) posits a long and involved process by which employment tests and the like may be “validated” and hence be absolutely, positively, etc., found to be job related. But no California case, no California statute, and not even any California regulation says that an employment test has to undergo anything like the tortuous, bureaucratically created process envisioned by section 1607. To be sure, California Code of Regulations title 2, section 7287.4, subdivision (a) says that an employment policy or practice (including, of course, a test) which has an adverse impact on a protected group is unlawful “unless the policy or practice is job-related, as defined in Section 7287.4(e). The [Fair Employment and Housing] Commission herein adopts the Uniform Guidelines on Employee Selection Procedures promulgated by various federal agencies, including the EEOC and Department of Labor.”
*1378I do not understand the Fair Employment and Housing Commission’s (FEHC) “adoption” of the federal guidelines to mean that the latter’s long and complex “validation” study process is thus incorporated into California law, much less incorporated as the exclusive means of establishing “job-relatedness.” Note that California Code of Regulations title 2, section 7287.4, subdivision (a) says that the term “job-related” is as defined in section 7287.4, subdivision (e). When one looks at that section, one finds the definition to be:“. . . that the selection practice is sufficiently related to an essential function of the job in question to warrant its use.” (Ibid.) I think it is very clear that if the latter is shown, i.e., that the test being challenged is “sufficiently related to an essential function of the job,” it is enough to mean, at least prima facie anyway, that it is job related.
Second, the United States Supreme Court has itself made clear that a formal “validation study” is not, even under title VII and section 1607 (29 C.F.R. (1997)) of the federal regulations, at all mandatory to a finding of job relatedness. As the majority concedes, in Watson v. Fort Worth Bank & Trust (1988) 487 U.S. 977, 998-999 [108 S.Ct. 2777, 2791, 101 L.Ed.2d 827], a plurality of the court speaking through Justice O’Connor held exactly this, stating: “Our cases make it clear that employers are not required, even when defending standardized or objective tests, to introduce formal ‘validation studies’ showing that particular criteria predict actual on-the-job performance. In [New York City Transit Authority v.] Beazer [(1979) 440 U.S. 568 [99 S.Ct. 1355, 59 L.Ed.2d 587]], for example, the Court considered it obvious that ‘legitimate employment goals of safety and efficiency’ permitted the exclusion of methadone users from employment with the New York City Transit Authority; the Court indicated that the ‘manifest relationship’ test was satisfied, even with respect to non-safety-sensitive jobs, because those legitimate goals were ‘significantly served by’ the exclusionary rule at issue in that case even though the rule was not required by those goals. [Citation.] Similarly, in Washington v. Davis [(1976) 426 U.S. 229 [96 S.Ct. 2040, 48 L.Ed.2d 597]], the Court held that the ‘job relatedness’ requirement was satisfied when the employer demonstrated that a written test was related to success at a police training academy ‘wholly aside from [the test’s] possible relationship to actual performance as a police officer.’ [Citations.] (‘[A]s a matter of law, it is permissible for the police department to use a test for the purpose of predicting ability to master a training program even if the test does not otherwise predict ability to perform on the job’).”
Citing Watson, two federal district courts have ruled that a formal validation study is not a prerequisite to a finding of job relatedness. (See Garner v. Runyon (N.D.Ala. 1991) 769 F.Supp. 357, 362; Rudder v. District of Columbia (D.D.C. 1995) 890 F.Supp. 23, 45, fn. 14.) If, notwithstanding section *13791607 (29 C.F.R. (1997)) of the federal regulations, a “validation study” is not necessary to establish “job-relatedness,” a fortiori it is not in California under the FEHA and the FEHC’s regulations.5
The third and final reason why job relatedness should have, at least, been further explored by the CSC, and perhaps even found on the record before it, derives from the refreshingly candid Brady report presented to that agency in advance of its hearing on the complaint of appellant and her several colleagues. That report was prepared by one Terri Brady, who also testified at the beginning of the CSC hearing on April 3, 1995. In a six-page, single-spaced report to the CSC, Brady set out the background of the controversy coming before that body. In describing the tests that had been given, she preliminarily observed that, although “time-consuming and costly to develop and administer,” they nonetheless “tested the required knowledge areas, abilities and skills needed to perform the essential functions of the positions.”
That bit of prose might, to a layman, seem to concede job relatedness, but there is, as the majority itself notes, much more. A couple of pages later, under the heading of “Standards,” and after noting in considerable statistical detail how the written part of the tests adversely impacted African-American applicants, the Brady report took up the job relatedness issue: “The second burden of proof cited in the [federal] Uniform Guidelines pertains to the job-relatedness of the examinations. . . . Both the subject matter expert who developed the examinations, and the candidates who participated in the examinations stated that the written exercises were job-related and accurately measured the knowledge, abilities and skills necessary to perform to essential functions of the job. Thus, the second burden of proof as provided in the Guidelines was satisfied.'’'’ (Italics added.)
Citing a portion of the federal guidelines, the majority dismisses this concession of job relatedness as not binding because of the sources relied upon.6 But the critical point is not whether that statement is a binding legal admission, but whether CSC’s utter disregard of its essence constitutes an *1380abuse of discretion. I conclude it does. Put another way, I maintain the Brady report’s statements mean, at the minimum, that the examinations were presumptively job related. I would hold that, in light of them, the CSC was at least required to take the job relatedness issue to the next level of inquiry, i.e., to (a) undertake a formal “validation study” (or its equivalent) and (b) then make a definitive determination as to whether the tests were job related.
3. Union Pressure
And why didn’t it do so? Because, very probably, of political pusillanimity. As made manifest by Brady in both her report and her CSC testimony, the union representing the current occupants of Muni supervisory and management positions, Transport Workers Union, Local 200 (Local 200), wouldn’t hear of it.
After, as noted above, the Brady report came down hard and fast on the job relatedness point, it went on to conclude that such was not the end of the inquiry. It said that there was yet a “third burden of proof’ which could be “the determining factor when deciding to use or nullify these examinations,” namely whether there are “alternate tests or comparable selection devices that are job-related but do not evidence adverse impact.”7 It noted that the staff had considered this possibility and, indeed, during a meeting with Local 200 had “recommended either a broader certification rule or even [a] rule of the list certification procedures be employed to resolve the issue.” (The report explained this meant that a larger number of persons who took and passed the tests could be considered under this compromise solution.)
But the compromise went nowhere and the Brady report explains why: “This proposed resolution to the protest would have satisfied the third burden of proof as the Guidelines stipulate. However, Local 200 would not agree to broadening the certification rule and dismissed this proposed resolution out of hand.” (Italics added.)
The Brady report then explained once more why the “broadening of the certification rule” solution was fair and equitable, but then noted again: *1381“However, Local 200 stated they would not accept broadening the rule of certification and insisted the lists be nullified.”
That effectively ended the staff’s search for a resolution. The Brady report then concluded with this frank summary: “The Uniform Guidelines on Employee Selection Procedures issued by the Federal government dictate that selection procedures be job-related and free of adverse impact.™ Given the prima facie case of statistically significant differing rates of passage on the written components of both . . . examinations, and given the limitations placed on the use of alternate selection devices, the Director acted appropriately in nullifying these examinations.” (Italics added.) The italicized portion of this summary “says a mouthful.”
Ms. Brady appeared before the CSC on April 3, 1995, to formally present her report and this recommendation. She reiterated the attempted compromise she and her staff had tried to reach with Local 200 to try to effect “a mutually satisfactory resolution to the problem” by “broadening the rule of certification for these lists . . . .” But, she again noted: “Local 200 did not agree to this compromise and stated that the only solution that was acceptable to them was to have the examination list nullified.” With practically no discussion among its members, the CSC voted to accept this staff recommendation.
From all of this, I believe it is obvious that Local 200’s adamant opposition to any compromise regarding “comparable selection devices” was the dominant factor in the ultimate cancellation of the tests. Put another way, the CSC simply turned “toes up” in the face of the union’s pressure and never even considered exercising any discretion in this matter.8 9 Such a well-documented refusal of an administrative agency to exercise its discretion necessarily makes its ultimate action “arbitrary, i.e., unsupported by a *1382rational basis” and thus an abuse of discretion. (See In re Executive Life Ins. Co. (1995) 32 Cal.App.4th 344, 358 [38 Cal.Rptr.2d 453].)10
The majority holds that “[tjhere was no legal duty in this case to adopt the tests or to try to validate them.” (Maj. opn., ante, at p. 1368.) It sums up its holding as being that “a court cannot, for the benefit of one protected group, compel a validation study through administrative mandate where an agency has properly exercised its discretion to cancel an employment test shown to have adverse impacts on another protected group.” (Id. at p. 1370.) I would hold, to the contrary, that it was an abuse of discretion for this agency to decline to undertake a validation study inasmuch as (1) the cancellation of the tests had an apparent adverse impact upon another protected minority, (2) the tests were presumptively job related as per the agency’s staff’s own thorough investigation and analysis, and (3) the agency clearly declined to pursue a “comparable selection device” alternative because of union pressure.
For these reasons, I respectfully dissent from the majority’s affirmance of the trial court’s denial of appellant’s petition for a writ of administrative mandamus. I would reverse as to that count and remand it to the trial court for further proceedings consistent with the foregoing.
Appellant’s petition for review by the Supreme Court was denied October 28, 1998.
Of course, under title VII of the Federal Civil Rights Act (42 U.S.C. § 2000e et seq. (hereafter title VII)) it has long been clear that there are two prongs to the test. (See, e.g., Griggs v. Duke Power Co. (1971) 401 U.S. 424, 436 [91 S.Ct. 849, 856, 28 L.Ed.2d 158], and Albemarle Paper Co. v. Moody (1975) 422 U.S. 405, 425 [95 S.Ct. 2362, 2375, 45 L.Ed.2d 280].)
As I will explain in more detail below, it should especially have done so here because the Brady report presented to the CSC concluded the tests were almost certainly “job-related.”
The majority suggests that, because a particular federal guideline (29 C.F.R. § 1607.4D (1997)) declares that an 80 percent or lower passage rate by a particular ethnic group “will generally be regarded by the Federal enforcement agencies as evidence of adverse impact,” this means there is then “a presumption of adverse impact.” (Maj. opn., ante, at p. 1368.) In the first place, that is not what the federal regulation says. Second, I question the need to look to federal regulations when there are two California regulations directly on point. (See Cal. Code Regs., tit. 2, § 7287.4, subds. (a) and (e).)
It is unclear to me, from the record as presented to us, whether these same statistics were before the CSC.
The majority states that “under this guideline [29 C.F.R. § 1607.9A (1997)], no test having an adverse impact can be used absent a successful validation study." (Maj. opn., ante, at p. 1367.) That is, I respectfully suggest, patently contrary to the unequivocal holding of Watson. It is not clear to me whether the majority is (a) interpreting that case as saying something different than is conveyed by its clear language or (b) laying down a stricter standard for FEHA purposes than obtains under title VII.
At another point, the majority deprecates the Brady report’s conclusion regarding job relatedness as a “first-impression opinion” which is a “foregone conclusion” because the staff is assessing “its own handiwork.” (Maj. opn., ante, at p. 1370.) I respectfully submit that these characterizations are quite unfair; such is apparent from the conscientiousness and candor which literally pervades the report. Also, and although not demonstrable from the report *1380itself, I strongly suspect that the final characterization is simply wrong for the reason that personnel experts are probably not entrusted with the substance of examinations to, e.g., test who would be a good San Francisco Muni supervisor.
Query whether such a “third factor” exists in California law. It is nowhere mentioned in the applicable regulation (Cal. Code Regs., tit. 2, § 7287.4), although it is in the federal guidelines. (See 29 C.F.R. § 1607.3B (1997).) The only California case I have found mentioning it suggests it applies only if the employment device under scrutiny is a “mere pretext.” (See City and County of San Francisco v. Fair Employment & Housing Com., supra, 191 Cal.App.3d at p. 985.)
Not, as we have seen, quite correct: The correct formulation is that, unless they are free of adverse impact, they must be job related.
To my dismay, the majority seemingly approves this exercise in timidity. It says that attempting validation “increased the likelihood of a lawsuit, from the union or others” (maj. opn., ante, at p. 1368) while crediting the test results and simultaneously undertaking a validity study would also “have run a serious gamble.” (Id. at p. 1369.) It concludes: “[D]eference to the proper exercise of administrative discretion dictates that an agency in the Commission’s position cannot be forced by mandate to risk the expense, time and potential legal liability of a validation study.” (Ibid.) I regret seeing such sentiments expressed as potential legal precedent. They seem to me to say that the “cut and run” exercise pursued here by the CSC is defendable because such bodies shouldn’t be forced to (a) decide what is right and (b) defend that decision in court. I had always thought that was generally the best public policy.
I note with regret that the majority nowhere deals with, or even refers to, the obviously severe impact of Local 200’s obstinacy in this controversy.