Opinion
CHIN, J.The California Fair Employment and Housing Act (FEHA) (Gov. Code, § 12900 et seq.)1 generally prohibits employers from practicing some kinds of discrimination. We must decide whether persons claiming discrimination may sue their supervisors individually and hold them liable for damages if they prove their allegations. We conclude that the FEHA, like similar federal statutes, allows persons to sue and hold liable their employers, but not individuals. Our conclusion also applies to common law actions for wrongful discharge. Accordingly, we reverse the Court of Appeal judgment, which held that individual employees may be sued and held liable, and approve the contrary holding of Janken v. GM Hughes Electronics (1996) 46 Cal.App.4th 55 [53 Cal.Rptr.2d 741] (Janken).
. I. Procedural History
Plaintiff Kimberly Reno sued several defendants for various causes of action. Two of them are at issue here: (1) employment discrimination based on medical condition in violation of the FEHA, and (2) discharge in violation of public policy. Some of the defendants were business entities; others, including Marijo Baird, were individuals. As relevant here, the complaint alleged that the business entity defendants hired plaintiff as a registered nurse; that these businesses were employers as defined in the FEHA; that the *644individual defendants, including Baird, “acted as agents ... of [the business defendants] in violating the FEHA and were therefore also employers” as defined in the act; and that the defendants “discriminated against plaintiff on the basis of her medical condition, cancer, and discharged plaintiff because of her medical condition, cancer, in violation of’ the FEHA and public policy.
Baird moved for summary judgment, arguing that she could not be held individually liable for employment discrimination. The superior court granted the motion. Reno appealed. The Court of Appeal reversed. It held that, under the FEHA, “supervisory agents” who committed the alleged unlawful discrimination, as well as the employer, may be sued and held liable for that discrimination. It expressly disagreed with the contrary conclusion of Janken, supra, 46 Cal.App.4th 55.
We granted Baird’s petition for review. In addition to the parties, five amici curiae have filed briefs in this court. The California Employment Lawyers Association and a plaintiff in a similar, but separate, action support plaintiff Reno. The Employers Group, the California Employment Law Council, and the Attorney General support defendant Baird.
II. Discussion
A. Introduction
Two causes of action are at issue here: one under the FEHA and one for wrongful discharge in violation of public policy. (See Tameny v. Atlantic Richfield Co. (1980) 27 Cal.3d 167 [164 Cal.Rptr. 839, 610 P.2d 1330, 9 A.L.R.4th 314].) The FEHA prohibits various forms of discrimination in employment. (See Caldwell v. Montoya (1995) 10 Cal.4th 972, 978, fn. 3 [42 Cal.Rptr.2d 842, 897 P.2d 1320].) “There is no doubt that aggrieved persons may, after exhausting their administrative remedies, sue under the statute for civil damages.” (Ibid.) Certainly aggrieved persons may sue their employers, but may they also sue individual supervisors?
Although the FEHA prohibits harassment as well as discrimination, it treats them differently. It prohibits “an employer ... or any other person” from harassing an employee. (§ 12940, subd. (h)(1), italics added.) It defines a “person” as including “one or more individuals, partnerships, associations, corporations, limited liability companies, legal representatives, trustees, trustees in bankruptcy, and receivers or other fiduciaries.” (§ 12925, subd. (d).) The FEHA, however, prohibits only “an employer” from engaging in improper discrimination. (§ 12940, subd. (a).) In this connection, it defines *645an “employer” as including “any person regularly employing five or more persons, or any person acting as an agent of an employer, directly or indirectly . . . (§ 12926, subd. (d).) With regard to harassment, it defines an “employer” as “any person regularly employing one or more persons, or any person acting as an agent of an employer, directly or indirectly . . . (§ 12940, subd. (h)(3)(A).)
Recently, we noted that, although a number of cases “have involved individual defendants, with no argument made that they could not be personally liable,” “no prior published California decision has directly considered whether FEHA imposes personal liability on an individual employee or manager who causes or assists a covered ‘employer’ to violate the statute’s prohibitions against discriminatory hiring, firing, and personnel practices.” (Caldwell v. Montoya, supra, 10 Cal.4th at p. 978, fn. 3.) We expressly declined to address “that broad and difficult question . . . .” (Id. at p. 979, fn. 3.) Later, in scholarly decisions, two Courts of Appeal considered the question and reached opposite conclusions. The first, Janken, supra, 46 Cal.App.4th 55, which Justice Zebrowski authored, concluded that only the employer, and not individual supervisors, may be sued and held liable.2 The second, the Court of Appeal decision in this case, which Justice Lambden authored, concluded that individual supervisors also may be sued. We agree with Janken.
B. The Janken Decision
1. Distinction Between Discrimination and Harassment
The Janken court noted the FEHA’s differing treatment of harassment and discrimination. It “conclude[d] that the Legislature’s differential treatment of harassment and discrimination is based on the fundamental distinction between harassment as a type of conduct not necessary to a supervisor’s job performance, and business or personnel management decisions—which might later be considered discriminatory—as inherently necessary to performance of a supervisor’s job.” (Janken, supra, 46 Cal.App.4th at pp. 62-63.) The court noted that “harassment consists of a type of conduct not necessary for performance of a supervisory job. Instead, harassment *646consists of conduct outside the scope of necessary job performance, conduct presumably engaged in for personal gratification, because of meanness or bigotry, or for other personal motives. Harassment is not conduct of a type necessary for management of the employer’s business or performance of the supervisory employee’s job. (Cf. Lisa M. v. Henry Mayo Newhall Memorial Hospital (1995) 12 Cal.4th 291, 301 [48 Cal.Rptr.2d 510, 907 P.2d 358] [sexual assault not motivated by desire to serve employer’s interest]; Farmers Ins. Group v. County of Santa Clara (1995) 11 Cal.4th 992 [47 Cal.Rptr.2d 478, 906 P.2d 440] [sexual harassment by deputy sheriff not within scope of employment].)
“Discrimination claims, by contrast, arise out of the performance of nécessary personnel management duties. While harassment is not a type of conduct necessary to personnel management, making decisions is a type of conduct essential to personnel management. While it is possible to avoid making personnel decisions on a prohibited discriminatory basis, it is not possible either to avoid making personnel decisions or to prevent the claim that those decisions were discriminatory.
“Courts have employed the concept of delegable authority as a test to distinguish conduct actionable as discrimination from conduct actionable as harassment. We adopt this approach to find that the exercise of personnel management authority properly delegated by an employer to a supervisory employee might result in discrimination, but not in harassment. (See Birkbeck v. Marvel Lighting Corp. (4th Cir. 1994) 30 F.3d 507, 510 and fn. 1 [distinguishing ‘personnel decisions of a plainly delegable character’ from harassment]; and Stephens v. Kay Management Co., Inc. (E.D.Va. 1995) 907 F.Supp. 169, 171, 173 [no personal liability of individual supervisors for ‘employment-related decisions’ or ‘ “personnel decisions of a plainly delegable character’”].) Making a personnel decision is conduct of a type fundamentally different from the type of conduct that constitutes harassment. Harassment claims are based on a type of conduct that is avoidable and unnecessary to job performance. No supervisory employee needs to use slurs or derogatory drawings, to physically interfere with freedom of movement, to engage in unwanted sexual advances, etc., in order to carry out the legitimate objectives of personnel management. Every supervisory employee can insulate himself or herself from claims of harassment by refraining from such conduct. An individual supervisory employee cannot, however, refrain from engaging in the type of conduct which could later give rise to a discrimination claim. Mating personnel decisions is an inherent and unavoidable part of the supervisory function. Without mating personnel decisions, a supervisory employee simply cannot perform his or her job duties.
“We conclude, therefore, that the Legislature intended that commonly necessary personnel management actions such as hiring and firing, job or *647project assignments, office or work station assignments, promotion or demotion, performance evaluations, the provision of support, the assignment or nonassignment of supervisory functions, deciding who will and who will not attend meetings, deciding who will be laid off, and the like, do not come within the meaning of harassment. These are actions of a type necessary to carry out the duties of business and personnel management. These actions may retrospectively be found discriminatory if based on improper motives, but in that event the remedies provided by the FEHA are those for discrimination, not harassment. Harassment, by contrast, consists of actions outside the scope of job duties which are not of a type necessary to business and personnel management. This significant distinction underlies the differential treatment of harassment and discrimination in the FEHA.” {Janken, supra, 46 Cal.App.4th at pp. 63-65, fns. omitted.)
2. The Statutory “Agent” Language
The Janken court confronted the plaintiffs’ argument that an individual supervisor acts “as an agent of an employer” within the meaning of “employer” in section 12926, subdivision (d), and is therefore personally liable as an employer. It found two possible constructions of the “agent” language. “One construction is that argued for by plaintiffs here: that by this ■ language the Legislature intended to define every supervisory employee in California as an ‘employer,’ and hence place each at risk of personal liability whenever he or she makes a personnel decision which could later be considered discriminatory. The other construction is the one widely accepted around the country: that by the inclusion of the ‘agent’ language the Legislature intended only to ensure that employers will be held liable if their supervisory employees take actions later found discriminatory, and that employers cannot avoid liability by arguing that a supervisor failed to follow instructions or deviated from the employer’s policy.” {Janken, supra, 46 Cal.App.4th at pp. 65-66, original italics.) The court adopted the latter construction for several reasons.
a. Cases Interpreting Federal Statutory Counterparts
The Janken court relied in part on decisions interpreting similar language in federal statutes. “Because the antidiscrimination objectives and relevant wording of title VII of the Civil Rights Act of 1964 (Title VII) [(42 U.S.C. § 2000e et seq.)], the Age Discrimination in Employment Act (ADEA) [(29 U.S.C. § 621 et seq.)] and the Americans with Disabilities Act (ADA) [(42 U.S.C. § 12111 et seq.)] are similar to those of the FEHA, California courts often look to federal decisions interpreting these statutes for assistance in interpreting the FEHA. (See, e.g., Los Angeles County Dept. *648of Parks & Recreation v. Civil Service Com. (1992) 8 Cal.App.4th 273, 280 [10 Cal.Rptr.2d 150]; Fisher v. San Pedro Peninsula Hospital (1989) 214 Cal.App.3d 590, 606 [262 Cal.Rptr. 842]; 8 Witkin, Summary of Cal. Law (9th ed. 1988) Constitutional Law, § 759, pp. 255-259.)
“Title VII defines employer as ‘a person . . . who has fifteen or more employees . . . and any agent of such a person.’ [(42 U.S.C. § 2000e(b).)] The ADEA defines employer as ‘a person . . . who has twenty or more employees’ including ‘any agent of such a person.’ [(29 U.S.C. § 630(b).)] The ADA defines employer as ‘a person . . . who has 15 or more employees . . . and any agent of such person.’ [(42 U.S.C. § 12111(5)(A).)] These three federal statutes thus contain definitions of employer identical in all relevant respects to the definition of employer contained in the FEHA: specifically, all the definitions of employer in these statutes are worded to cover the ‘agent’ of the employer.” (Janken, supra, 46 Cal.App.4th at p. 66, fns. omitted.)
The court cited the “clear and growing consensus” (Stephens v. Kay Management Co., Inc. (E.D.Va. 1995) 907 F.Supp. 169, 173) of courts interpreting the federal statutes that supervisors cannot be held personally liable for employment discrimination. It noted that, since 1993, eight federal circuits have either (1) held that the “agent” language does not create individual liability for discrimination, or (2) found that, although individuals can be sued in their official or representative capacity, they may not be sued in their individual capacity and have no personal liability, or (3) interpreted similar language in a state statute as not creating individual liability. (Jan-ken, supra, 46 Cal.App.4th at p. 67.) The court discussed these circuit decisions individually.
“Tomka v. Seiler Corp. (2d Cir. 1995) 66 F.3d 1295, 1313-1314 was a Title VII case. In Tomka, the Second Circuit rejected ‘a narrow, literal reading of the agent clause’ which implied ‘that an employer’s agent is a statutory employer for purposes of liability.’ Instead, on ‘a broader consideration’ of the purposes of Title VII, Tomka found that the ‘agent’ language was intended as a simple expression of respondeat superior liability. (66 F.3d 1295, 1314.)
“Birkbeck v. Marvel Lighting Corp. [(4th Cir. 1994)] 30 F.3d 507 was an ADEA case. In Birkbeck, the Fourth Circuit noted that ‘[t]he few courts that have found individual liability under the ADEA tend to seize on the use of the word “agent” ’ in the definition of employer. Birkbeck found that rationale ‘unpersuasive,’ finding instead that the ‘agent’ reference was ‘an unremarkable expression of respondeat superior—that discriminatory personnel actions taken by an employer’s agent may create liability for the employer.’ (30 F.3d 507, 510.) *649“Grant v. Lone Star Co. (5th Cir. 1994) 21 F.3d 649, 651-653 was a Title VII case. In Grant, the Fifth Circuit stated that neither public nor private sector employees are subject to individual liability on the basis of the ‘agent’ language, instead agreeing that the purpose of the ‘agent’ provision was to incorporate respondeat superior liability into Title VII. (See also Harvey v. Blake (5th Cir. 1990) 913 F.2d 226, 227-228 [public officials may be liable in their official capacities, but not in their individual capacities].)
“U.S. E.E.O.C. v. AIC Security Investigations, Ltd. (7th Cir. 1995) 55 F.3d 1276 was an ADA case. In U.S. E.E.O.C., the Seventh Circuit joined ‘analogous decisions of our sister Circuits in holding that individuals who do not independently meet the ADA’s definition of “employer” cannot be held liable under the ADA.’ (55 F.3d 1276,1279.) The Seventh Circuit stated that the EEOC was ‘fighting against the weight of authority’ in urging individual liability on the basis of the ‘agent’ wording in the employer definition. While finding the EEOC’s ‘ “plain language” ’ argument to have ‘surface appeal,’ the Seventh Circuit found ‘upon closer examination that appeal is really an illusion.’ (55 F.3d 1276, 1281.) The court ruled that the reason for the inclusion of the ‘agent’ language in the definition of employer ‘was to ensure that the courts would impose respondeat superior liability upon employers for the acts of their agents. [Citations.]’ ([Ibid.], original italics.)
“Miller v. Maxwell’s Intern. Inc. (9th Cir. 1993) 991 F.2d 583 was a Title VII and ADEA case. In Miller, the Ninth Circuit noted that, because of the use of the word ‘agent’ in the definition of employer, ‘some courts have reasoned that supervisory personnel ... are themselves employers for purposes of liability.’ (Miller, supra, 991 F.2d 583, 587.) Although the Ninth Circuit found that this statutory construction argument was ‘not without merit,’ the court found the ‘better rule’ to be that the purpose of the agent provision was to incorporate respondeat superior liability into the statute. ([Ibid.].) [To the same effect, see also Greenlaw v. Garrett (9th Cir. 1995) 59 F.3d 994, 1001.]
“Sauers v. Salt Lake County (10th Cir. 1993) 1 F.3d 1122 was a Title VII case. In Sauers the Tenth Circuit found that a supervisor with significant control over hiring, firing or conditions of employment acts as an ‘agent’ of the employer, and that the employer is therefore liable for a hostile work environment created by the supervisor whether or not the employer knew of the supervisor’s conduct. (1 F.3d 1122, 1125.) The supervisor himself, however, was not the employer and could be sued only in his official, and not in his individual, capacity. The ‘agent’ language in the definition of employer did not render the individual liable. (But see Brownlee v. Lear Siegler Management Services Corp. (10th Cir. 1994) 15 F.3d 976, 978 [suggesting individual liability under Title VII in dictum].)
*650“The Eleventh Circuit has ruled that individual supervisory employees cannot be held personally liable under either Title VII or the ADEA, because they are not employers. Instead, the individual nonemployers may be sued in their official or agency capacity only. (Cross v. State of Ala. (11th Cir. 1995) 49 F.3d 1490, 1504 [individuals may be sued under Title VII in their official capacities only]; Smith v. Lomax (11th Cir. 1995) 45 F.3d 402, 403, fn. 4 [under Title VII and the ADEA, individual supervisors are not employers and cannot be held personally liable]; Busby v. City of Orlando (11th Cir. 1991) 931 F.2d 764, 772 [supervisory employees may be named in their agency capacity, but are not personally liable].)” {Janken, supra, 46 Cal.App.4th at pp. 67-69, fn. omitted.)
The last federal case the Janken court surveyed was Lenhardt v. Basic Institute of Technology, Inc. (8th Cir. 1995) 55 F.3d 377, in which the Eighth Circuit Court of Appeals construed similar language in a Missouri statute as not imposing individual liability on supervisors. “Reviewing federal law, the Eighth Circuit found a ‘clear consensus’ that supervisors cannot be held liable in their individual capacity, and that instead supervisors could be named only in their ‘official’ capacities, finding that ‘[e]very circuit that has considered the issue ultimately has concluded that an employee, even one possessing supervisory authority, is not an employer upon whom liability can be imposed under Title VII.’ (55 F.3d 377, 381.)” {Janken, supra, 46 Cal.App.4th at pp. 69-70, fn. omitted.)
b. Incongruity Between Exempting Small Employers and Holding Individual Nonemployers Liable
The Janken court also noted that the FEHA exempts small employers from liability for discrimination. “Section 12926, subdivision (d) defines ‘employer’ as including ‘any person regularly employing five or more persons.’ A person who regularly employs less than five other persons is not an ‘employer’ for purposes of FEHA prohibitions on discrimination, and hence cannot be sued for discrimination. {Jennings v. Marralle (1994) 8 Cal.4th 121 [32 Cal.Rptr.2d 275, 876 P.2d 1074].) For purposes of harassment, however, ‘employer’ is specially defined in section 12940, subdivision (h)(3)(A) to include any person regularly employing one or more persons. Section 12940, subdivision (h)(4) and (5) make clear that this special definition of ‘employer’ as someone employing only one other person applies only to harassment claims, and that discrimination claims continue to be covered by the ‘five or more’ definition in section 12926, subdivision (d). The Legislature thus made a clear distinction in California in the treatment of harassment claims versus the treatment of discrimination claims: small employers can be sued for harassment, but they cannot be sued for discrimination.
*651“In Title VII and the ADA, employer is defined as a person employing 15 or more. [(42 U.S.C. §§ 2000e(b), 12111(5)(A).)] In the ADEA, employer is defined as a person employing 20 or more. [(29 U.S.C. § 630(b).)] These definitions exempt small employers from federal employment discrimination litigation because ‘Congress did not want to burden small entities with the costs associated with litigating discrimination claims.’ (Miller v. Maxwell’s Intern. Inc., supra, 991 F.2d 583, 587; see also, e.g., Birkbeck v. Marvel Lighting Corp., supra, 30 F.3d 507, 510 [purpose of exempting small employers is to reduce burden on small businesses].) Many of the federal cases which found no personal liability against individual supervisory employees based their decisions in part on the incongruity that would exist if small employers were exempt from liability while individual nonemployer supervisors were at risk of personal liability. (Tomka v. Seiler Corp., supra, 66 F.3d 1295, 1314 [inconceivable that a Congress concerned with protecting small employers would simultaneously impose liability on individual employees]; Birkbeck v. Marvel Lighting Corp., supra, 30 F.3d 507, 510 [incongruous to impose personal liability on individual nonemployer when small employers are exempted]; Grant v. Lone Star Co., supra, 21 F.3d 649, 652 [inconceivable that Congress intended to protect small entities, yet allow liability against individuals—‘the smallest of legal entities’]; U.S. E.E.O.C. v. AIC Security Investigations, Ltd., supra, 55 F.3d 1276, 1281 [limitation protecting small employers balances goal of stamping out discrimination with protection of small entities from hardships of litigation, individual liability for employees upsets that balance]; Miller v. Maxwell’s Intern. Inc., supra, 991 F.2d 583, 587 [if Congress intended to protect small entities with limited resources, inconceivable that Congress intended to allow liability against individual employees].)
“The same reasoning applies to our task of construing the employer definition in the FEHA. The Legislature clearly intended to protect employers of less than five from the burdens of litigating discrimination claims. (Jennings v. Marralle, supra, 8 Cal.4th 121.) We agree that it is ‘inconceivable’ that the Legislature simultaneously intended to subject individual nonemployers to the burdens of litigating such claims. To so construe the statute would be ‘incongruous’ and would ‘upset the balance’ struck by the Legislature.” {Janken, supra, 46 Cal.App.4th at pp. 71-72, original italics, fns. omitted.)
c. Policy to Avoid Conflicts of Interest and Chilling of Effective Management3
The Janken court stated that “imposing liability on individual supervisory employees would do little to enhance the ability of victims of discrimination *652to recover monetary damages, while it can reasonably be expected to severely impair the exercise of supervisory judgment. The minimal potential for benefit to an alleged victim juxtaposed with the potentially severe adverse effects of imposing personal liability on individual supervisory employees is an additional reason for our conclusion that this is not the result intended by the Legislature.
“Many courts have noted the importance of maintaining the conditions in which impartial judgment can be exercised by officials performing duties in the public sector. (See, e.g., Caldwell v. Montoya, supra, 10 Cal.4th 972, 982 [[Njoting ‘ “vital public interest in securing free and independent judgment” ’ of public sector employees in ‘dealing with personnel problems.’]; Lipman v. Brisbane Elementary Sch. Dist. (1961) 55 Cal.2d 224, 230 [11 Cal.Rptr. 97, 359 P.2d 465] [‘There is a vital public interest in securing free and independent judgment of school trustees in dealing with personnel problems, ...’];...) Judge Learned Hand made a similar point long ago in Gregoire v. Biddle (2d Cir. 1949) 177 F.2d 579, 581, stating: ‘It does indeed go without saying that an official, who is in fact guilty of using his powers to vent his spleen upon others, or for any other personal motive not connected with the public good, should not escape liability for the injuries he may so cause; and, if it were possible in practice to confine such complaints to the guilty, it would be monstrous to deny recovery. The justification for doing so is that it is impossible to know whether the claim is well founded until the case has been tried, and that to submit all officials, the innocent as well as the guilty, to the burden of trial and to the inevitable danger of its outcome, would dampen the ardor of all but the most resolute, or the most irresponsible, in the unflinching discharge of their duties ... the answer must be found in a balance between the evils inevitable in either alternative. . . .’ rather than in subjecting ‘those who try to do their duty to the constant dread of retaliation.’ (See also Tietz v. Los Angeles Unified Sch. Dis. (1965) 238 Cal.App.2d 905, 910 [48 Cal.Rptr. 245], quoting Judge Learned Hand.)
“No one could reasonably doubt that effective and efficient management of industrial enterprises and other economic organizations is also important to the public welfare. The societal interest in effective private sector personnel management may be less direct, but only marginally (if at all) less compelling . . . . Yet it is manifest that if every personnel manager risked losing his or her home, retirement savings, hope of children’s college education, etc., whenever he or she made a personnel management decision, management of industrial enterprises and other economic organizations *653would be seriously affected. (See, e.g., Birkbeck v. Marvel Lighting Corp., supra, 30 F.3d 507, 510 [‘personal liability would place a heavy burden on those who routinely make personnel decisions’].)
“Plaintiffs’ theory would place a supervisory employee in a direct conflict of interest- with his or her employer every time that supervisory employee was faced with a personnel decision. . . . [It] would coerce the supervisory employee not to make the optimum lawful decision for the employer. Instead, the supervisory employee would be pressed to make whatever decision was least likely to lead to a claim of discrimination against the supervisory employee personally, or likely to lead only to that discrimination claim which could most easily be defended. The employee would thus be placed in the position of choosing between loyalty to the employer’s lawful interests at severe risk to his or her own interests and family, versus abandoning the employer’s lawful interests and protecting his or her own personal interests. The insidious pressures of such a conflict present sobering implications for the effective management of our industrial enterprises and other organizations of public concern. We believe that if the Legislature intended to place all supervisory employees in California in such a conflict of interest, the Legislature would have done so by language much clearer than that used here.
“Moreover, imposing personal liability against individual supervisory employees adds little to an alleged victim’s legitimate prospects for monetary recovery. The plaintiff-employee’s primary target remains the employer. Adding individual supervisors personally as defendants adds mostly an in terrorem quality to the litigation, threatening individual supervisory employees with the spectre of financial ruin for themselves and their families and correspondingly enhancing a plaintiff’s possibility of extracting a settlement on a basis other than the merits. Enhancing the prospects for obtaining a settlement on a basis other than the merits is hardly a worthy legislative objective ....
“The only marginally legitimate purpose for adding individual supervisory employees as defendants is the possibility that the employer will file for bankruptcy protection and be unable to respond fully in damages. This prospect cuts both ways, however. Employment with an employer about to go bankrupt is in jeopardy in the first instance, and hence is of inherently lesser value than employment with a healthy and solvent employer. Loss or impairment of employment with an employer on the brink of bankruptcy would thus cause correspondingly less compensable damage. At the same *654time, it is often when companies are struggling for their financial survival that the most difficult personnel decisions must be made. A legal regime in which individual supervisory employees were personally liable after bankruptcy of the employer (a bankruptcy which would potentially result in unemployment for the supervisory employees as well) would mean that salvage of a troubled business would be all that less likely, as managers declined to make decisions necessary to save the enterprise. . . .
“. . . The construction we choose avoids conflicts of interest by harmonizing the FEHA statutory provisions with the duties of supervisory employees, and does so without doing any significant harm to plaintiffs’ legitimate interests.” (Janken, supra, 46 Cal.App.4th at pp. 72-76, fns. omitted.)
d. Responding to the “Chamber of Horrors” Argument
The Janken court argued that its interpretation would not create a “chamber of horrors.” “In many of the federal cases which construed the ‘agent’ language not to create personal liability for individual supervisors, it was argued that such a construction would open the floodgates of discrimination, would give supervisors a ‘free pass’ to discriminate, would ‘liberate’ supervisors to discriminate ‘with impunity,’ and the like. All the recent cases have rejected such arguments. One case rejected this ‘parade of horribles’ as ‘Chicken Little-esque.’ (U.S. E.E.O.C. v. AIC Security Investigations, Ltd., supra, 55 F.3d 1276, 1282.) Another considered this ‘chamber of horrors’ argument and was ‘not persuaded.’ (Lenhardt v. Basic Institute of Technology, Inc., supra, 55 F.3d 377, 381.) Another found the argument ‘unsound.’ (Miller v. Maxwell’s Intern. Inc., supra, 991 F.2d 583, 588.)
“The cases have rejected such arguments because the employer remains liable. ‘An employer that has incurred civil damages because one of its employees believes he can violate Title VII with impunity will quickly correct that employee’s erroneous belief.’ (Miller v. Maxwell’s Intern. Inc., supra, 991 F.2d 583, 588.) ‘The employing entity is still liable, and that entity and its managers have the proper incentives to adequately discipline wayward employees, as well as to instruct and train employees to avoid actions that might impose liability.’ (U.S. E.E.O.C. v. AIC Security Investigations, Ltd., supra, 55 F.3d 1276, 1282.) ‘As a practical matter employees who unlawfully discriminate against their fellow employees, and who thereby expose their employer to liability, do not get anything like a “free pass” to continue their wrongdoing with impunity. By incorporating the principles of respondeat superior into Title VII, Congress has required employers to answer for prohibited acts of discrimination perpetrated by their employees. An employer subjected to well-founded claims of employment discrimination as a result of an employee’s intentional acts of discrimination is not likely to look favorably upon the offending employee. To the *655contrary, the employer, to protect its own interests and to avoid further liability, almost certainly will impose some form of discipline upon the offending employee. That discipline may include a “free pass” to the unemployment line, a result that would seem particularly likely if the employee engages in repeated acts of intentional discrimination against fellow employees.’ (Lenhardt v. Basic Institute of Technology, Inc., supra, 55 F.3d 377, 381; see also Birkbeck v. Marvel Lighting Corp., supra, 30 F.3d 507, 510 [‘Employer liability ensures that no employee can violate the civil rights laws with impunity’]; Stephens v. Kay Management Co. Inc., supra, 907 F.Supp. 169, 174 [supervisory employees are not liberated to discriminate with impunity because the employer remains liable].)
“The reasoning of these cases applies here. The fact that the employer is liable via the respondeat superior effect of the ‘agent’ language provides protection to employees even if individual supervisors are not personally liable. Hence we do not find this consideration to compel a conclusion that the Legislature must have intended to impose personal liability on individual supervisory employees.” (Janken, supra, 46 Cal.App.4th at pp. 76-77, fn. omitted.)
3. The “Aiding and Abetting” Claim
The Janken court considered and rejected another textual argument for personal liability under the FEHA. “Section 12940, subdivision (g) provides that it is an unlawful employment practice for ‘any person to aid, abet, incite, compel or coerce the doing of any of the acts forbidden’ by the FEHA, or ‘to attempt to do so.’ Plaintiffs contend that the supervisory employee defendants here can be held liable on the theory that, in making the challenged personnel decisions on behalf of [the employer], they ‘aided and abetted’ [the employer].
“Preliminarily, we can discern purposes for the ‘aid and abet’ language independent of any involving the liabilities of supervisory employees. This language makes it unlawful, for example, for third parties such as customers or suppliers to induce or coerce prohibited discrimination or harassment. Plaintiffs contend, however, that this ‘aid and abet’ language also places individual supervisory employees at risk of personal liability for personnel management decisions in which they participate. We must now decide whether the Legislature intended to accomplish a result so significant by a method so abstruse.
“Aiding and abetting occurs when one helps another commit a prohibited act. [Citation.] The concept of aiding and abetting involves two separate *656persons, one helping the other. Here we deal with individual employees of a corporate employer. A corporation can act only through its individual employees. [Citation.] Our question is whether it can properly be said that a corporate employee is ‘aiding and abetting’ his or her corporate employer when the corporate employee acts on behalf of the corporation by making a personnel decision or, more precisely, whether this is the meaning intended by the Legislature.
“. . . Linguistically, it is questionable whether it can properly be said that an employee who exercises delegated personnel management authority is ‘aiding and abetting’ his or her employer in managing personnel, and the stilted and unusual nature of such a usage alone casts doubt on plaintiffs’ construction. Had the Legislature intended to place all employees charged with the duty of making personnel decisions in California at risk of personal liability, we believe the Legislature would have done so by language more direct and less susceptible to doubt. . . .
“. . . For the reasons stated here plus those stated above in connection with the ‘agent’ issue, we conclude that the Legislature did not intend to impose personal liability upon individual supervisory employees by the roundabout method of ‘aiding and abetting’ language.” (Janken, supra, 46 Cal.App.4th at pp. 77-79.)
4. Conclusion
The Janken court concluded that because “only employers—and not individual supervisory employees—are at risk of liability for discrimination,” the trial court had correctly dismissed the individual supervisory employees. (Janken, supra, 46 Cal.App.4th at p. 80.)
C. Counterarguments
Plaintiff and the Court of Appeal in this case criticize Janken in several respects that we find, in sum, unpersuasive.
The Court of Appeal here noted that, “without analysis and discussion,” earlier Court of Appeal decisions allowed plaintiffs to sue agent employees as well as employers. We made essentially the same statement in Caldwell v. Montoya, supra, 10 Cal.4th at pages 978-979, footnote 3. We also observed that no decision had directly considered the question, and we expressly reserved it. (Ibid.) Janken was the first court to consider it. As we implied in *657Caldwell, the earlier decisions provide no authority for what they did not consider.
The Court of Appeal found “the Janken court’s attempt to differentiate harassment from discrimination claims to be untenable.” It stated that any supervisor forced to interact with other employees risks harassment claims, just as any supervisor forced to make personnel decisions risks discrimination claims. It observed that employers cannot allow supervisors to discriminate unlawfully any more than to harass; they may only delegate the responsibility to make employment decisions based on lawful factors. However, the Legislature has already distinguished between discrimination and harassment. Whatever similarities there may be between the two, the employer ultimately does the former; coworkers and supervisors do the latter. Harassment claims are legitimately distinguished from discrimination claims because they are based on different types of conduct. Behavior that gives rise to a harassment claim is not related to performing one’s job duties except insofar as it occurs within the work environment. Behavior that gives rise to a discrimination claim, on the other hand, is often indistinguishable from performing one’s job duties. Thus, the Legislature properly tailored the FEHA in order to address these distinct claims.
Plaintiff claims the distinction between harassment and discrimination is “unworkable.” She argues that Janken, “in effect, [says] that an agent of an employer could be personally liable for hanging a pin-up calendar and offending an employee but could not be held liable if the same agent were to shout out loud T refuse to hire black people’ and then proceed to do just that.” Shouting out loud, however, as distinct from making personnel decisions, might be deemed actionable harassment. It is not the type of conduct an employer could properly delegate; making a supervisor liable for that conduct would not create a conflict of interest between the supervisor and the employer. For the reasons the Janken court stated, the Legislature reasonably could, and did, distinguish between harassment and discrimination claims.
The Court of Appeal interpreted Janken as concluding that the “agent” language merely incorporated respondeat superior principles. It criticized that conclusion on the basis that the FEHA would be found to incorporate those principles even without that language. “Therefore,” it said, “if the agent language merely articulated the rule of respondeat superior, it was unnecessary. Thus, under the Janken court’s interpretation, the definition of employer . . . was surplusage, violating the rule of statutory construction to ‘give meaning to every word of a statute if possible’ (Arnett [v. Dal Cielo (1996) 14 Cal.4th 4,] 22 [56 Cal.Rptr.2d 706, 923 P.2d 1]).” We disagree that the “agent” language is surplusage under any interpretation.
*658The issue in this case is individual liability for discrimination. Therefore, we express no opinion on the scope of employer liability under the FEHA for either discrimination or harassment. We specifically express no opinion on whether the “agent” language merely incorporates respondeat superior principles or has some other meaning. We need not because, whatever that language means precisely, it is not surplusage. The Legislature may reasonably have chosen to define the scope of employer liability expressly rather than to leave it to judicial interpretation. It is a maxim of statutory construction that “Courts should give meaning to every word of a statute if possible, and should avoid a construction making any word surplusage.” {Arnett v. Dal Cielo (1996) 14 Cal.4th 4, 22 [56 Cal.Rptr.2d 706, 923 P.2d 1].) But “surplusage” in this context means surplusage as to other statutory language, not as to some possible judicial interpretation. For example, in the case the Court of Appeal cited, the statutory language at issue was a reference to “discovery or subpoena.” {Ibid.) We held that an interpretation that “discovery” included “subpoena” should be avoided because that interpretation would render the word “subpoena” surplusage to the word “discovery.” {Ibid.)
The rule avoiding surplusage does not apply here. Although the Legislature would not need to refer to “discovery or subpoena” if the former included the latter, the Legislature may choose to state all applicable legal principles in a statute rather than leave some to even a predictable judicial decision. Express statutory language defining the scope of employer liability is not surplusage. Rather, it may eliminate potential confusion and avoid the need to research extraneous legal sources to understand the statute’s full meaning. Legislatures are free. to state legal principles in statutes, even if they repeat preexisting law, without fear the courts will find them unnecessary and, for that reason, imbued with broader meaning.
The Court of Appeal questioned the Janken court’s reliance on federal cases interpreting federal statutes in two respects. First, it stated, “The federal courts have not reached consensus regarding a supervisor’s individual liability, and the United States Supreme Court has yet to rule on this question.” It noted that the First Circuit Court of Appeals has not resolved the issue and cited two recent district court decisions criticizing the circuit court decisions. {Ruffino v. State Street Bank and Trust Co. (D.Mass. 1995) 908 F.Supp. 1019, 1047-1048; Ostrach v. Regents of the University of California (E.D.Cal. 1997) 957 F.Supp. 196, 199-200.) It also suggested that the Fifth Circuit’s rulings have been inconsistent by “comparing] Stults v. Conoco, Inc. (5th Cir. 1996) 76 F.3d 651, 654-655 [no individual liability] with Garcia v. Elf Atochem North America (5th Cir. 1994) 28 F.3d 446, 451 [supervisors are individually liable].” The two cases are not inconsistent. *659Garda v. Elf Atochem North America (5th Cir. 1994) 28 F.3d 446, effectively disapproved on other grounds in Oncale v. Sundowner Offshore Services, Inc. (1998) 523 U.S. 75, _ [118 S.Ct. 998, 1001, 140 L.Ed.2d 201], indicated that individuals may be sued for harassment in their official capacity as agents of the employer. However, citing Grant v. Lone Star Co. (5th Cir. 1994) 21 F.3d 649, it also said that “Title VII liability does not attach to individuals acting in their individual capacity.” (Garcia, supra, 28 F.3d at p. 451, fn. 2.)
Although the high court has not yet ruled on this question, the federal circuit courts support the Janken ruling overwhelmingly. Indeed, the Sixth Circuit Court of Appeals recently joined the “majority of our sister circuits that have addressed this issue [and] have held that an employee/supervisor, who does not otherwise qualify as an ‘employer,’ cannot be held individually liable under Title VII and similar statutory schemes.” (Wathen v. General Elec. Co. (6th Cir. 1997) 115 F.3d 400, 404; see id. at p. 405.) A clear consensus now exists. We find the cases concluding supervisory employees are not individually liable persuasive in both number and reasoning.
The Court of Appeal here also concluded the federal cases do not apply because of “the significant differences between the federal statutes and the FEHA . . . .” It relied on certain differences in remedy, the aiding and abetting language of section 12940, and language indicating the Fair Employment and Housing Commission (FEHC) may issue a written accusation against the accused “person [or] employer . . . .” (§ 12965, subd. (a), italics added; see also §§ 12965, subd. (b), 12970, subds. (a), (b).) None of these points convince us. In general, “The language, purpose and intent of California and federal antidiscrimination acts are virtually identical. Thus, in interpreting FEHA, California courts have adopted the methods and principles developed by federal courts in employment discrimination claims arising under” the federal acts. (Cummings v. Benco Building Services (1992) 11 Cal.App.4th 1383, 1386 [15 Cal.Rptr.2d 53].) The general rule applies to the issue in this case. The substance of the relevant language involved here—including the “agent” provision—is found in each of the analogous federal statutes. Any differences in remedy and the aiding and abetting language do not persuade us the Legislature intended a rule regarding individual liability unlike the federal rule.
The general language allowing the FEHC to issue an accusation against a “person” does not mean that all the FEHA provisions apply to persons as well as employers. The accusation provisions apply generally. Some of the FEHA provisions, such as those involving harassment, do, indeed, apply to *660individual persons as well as employers. (E.g., § 12940, subd. (h)(1) [prohibiting “any other person” from harassing an employee].) Therefore, in general, the FEHC may issue accusations against persons as well as employers. The discrimination provisions, however, apply only to employers. General language permitting the FEHC to issue accusations against persons in some situations does not mean it may issue them in all. The Court of Appeal specifically noted that in 1992, “the Legislature amended section 12965, subdivision (b), to include the following language: ‘If the person claiming to be aggrieved does not request a right-to-sue-notice,— ... [a] city, county, or district attorney in a location having an enforcement unit ... for the purpose of prosecuting AIDS/HIV discrimination claims, may also bring a civil action under this part against the person, employer, labor organization, or employment agency named in the notice.’ (Italics added [by the Court of Appeal].)” This particular language is not before us because no “enforcement unit” is seeking to bring an action against anyone for “AIDS/HIV discrimination”; accordingly, we need not decide whether the reference to “person” imposes personal liability on individuals not otherwise qualifying as employers or merely allows an enforcement unit to sue persons in their representative capacities. We do not, however, believe this narrow language referring to a civil action by an enforcement unit for a specific type of discrimination was intended to create individual liability under the FEHA for any discrimination claim brought by anyone.
The Court of Appeal also relied on the FEHC’s administrative decisions. We assign great weight to the interpretations an administrative agency like the FEHC gives to the statutes under which it operates, although ultimately statutory interpretation is a question of law the courts must resolve. (Robinson v. Fair Employment & Housing Com. (1992) 2 Cal.4th 226, 234-235 & fn. 6 [5 Cal.Rptr.2d 782, 825 P.2d 767] (Robinson); Carmona v. Division of Industrial Safety (1975) 13 Cal.3d 303, 310 [118 Cal.Rptr. 473, 530 P.2d 161].) The Court of Appeal found that the FEHC “has consistently held supervisors individually liable as agents of the employer for their acts of discrimination and harassment. (See, e.g., Dept. Fair Empl. & Hous. v. Bee Hive Answering Service (1984) No. 84-16, FEHC Precedential Decs. 1984-1985, CEB 8, p. 23 [employer and individual supervisor liable for sexual discrimination and harassment]; Dept. Fair Empl. & Hous. v. Del Mar Avionics (1985) No. 85-19, FEHC Precedential Decs. 1984-1985, CEB 16, p. 25 [FEHC found employer and supervisor liable for race and sex discrimination and harassment]; Dept. Fair Empl. & Hous. v. Madera County (1990) No. 90-03, FEHC Precedential Decs. 1990-1991, CEB 1, p. 26 [‘The Department asserts that respondent Gordon is personally liable [for sexual harassment] as an employer under the Act because he acted as an agent of Madera County.’].)”
*661Although these decisions do state what the Court of Appeal asserts, on examination, they do not provide strong support for plaintiff’s position. None considered the differing statutory treatment of harassment and discrimination. More importantly, they rested solely on now-outdated federal authority. The first decision finding individual liability, Dept. Fair Empl. & Hous. v. Bee Hive Answering Service (1984) No. 84-16, FEHC Precedential Decs. 1984-1985, CEB 8, pages 14-16, 23, did not itself examine the question of individual liability but only cited the then existing federal rule. “Since neither the [FEHA] nor our regulations define the term, ‘agent,’ and since Title VII also defines employer by extending coverage to any agent of the larger employing entity (and similarly fails to define ‘agent’), we have examined Title VII law on this point. Inasmuch as we find that Title VII precedent is applicable here, we have chosen to be guided by it.” (Id. at p. 14, italics added.) The decision then merely cited early federal district court decisions that found individual liability. (Id. at p. 15.) Later FEHC decisions simply cited earlier decisions. None ever considered the question on its own merits.
As discussed above, federal circuit court decisions now overwhelmingly find no individual liability. The older district court decisions which the FEHC cited are obsolete. The question arises, to what part of the FEHC interpretation should we give great weight? Its conclusion that title VII precedent applies? Or its reliance on now obsolete Title VII precedent? Under the circumstances, the FEHC conclusion that individuals are personally liable is entitled to but little consideration. We must decide the question ourselves.
In a similar vein, the Court of Appeal also found that legislative inaction in the face of the FEHC interpretation and earlier appellate court decisions indicates acquiescence. (See Robinson, supra, 2 Cal.4th at p. 235 & fn. 7.) As noted, appellate court decisions before Janken did not directly consider the question. The same question arises regarding the FEHC decisions. In what, exactly, is the Legislature supposed to have acquiesced? In the conclusion that federal precedent controls? Or in the reliance on now obsolete federal precedent? In this situation, especially, legislative inaction is a “ ‘ “ ‘weak reed upon which to lean.’ ” ’ ” (Harris v. Capital Growth Investors XIV (1991) 52 Cal.3d 1142, 1156 [278 Cal.Rptr. 614, 805 P.2d 873].)
The Court of Appeal here also disagreed with the Janken court’s discussion of the FEHA’s purposes. It was concerned that shielding supervisory employees who commit discriminatory acts from liability undermines the FEHA’s policy and does not ensure that the law bar all modes of discrimination. However, as the Janken court pointed out (Janken, supra, 46 *662Cal.App.4th at pp. 76-77), employers will not condone discriminatory acts by their supervisory employees, because the employers must ultimately pay. Furthermore, supervisors will not escape punishment. They will merely escape the cost of defending suit. Supervisors guilty of engaging in unlawful discrimination, and thus causing their employers to incur monetary liability, will often suffer demotion or unemployment. Their reputation with potential future employers will also be affected. Until the Legislature provides for punishing individual employees, we should leave that task to the employers.
The Court of Appeal also noted the “reality that corporate entities are comprised of individuals” and can act only through individuals. The point is obviously correct but does not address the question of who is liable: the corporation, the individuals, or both. It is settled that the corporation can be held liable. The question here is whether to impose liability also on the individuals. Corporate decisions are often made collectively by a number of persons. Different individuals might have differing levels of awareness and participation in the decisions. When a collective decision is discriminatory, some participants might have acted innocently, others less so. Assessing individual blame might be difficult, in contrast to simply placing blame on the corporation, on whose behalf the individuals acted. Moreover, to make collective decisions possible, individuals often must rely on information or evaluations that others supply. Imposing individual liability for collective decisions might place the individuals in an adversarial position to each other (as well as to the corporation). Individuals might fear to act in reliance on input from others. Some might fear that a potentially controversial but, so far as they can know, correct and necessary collective decision might be misconstrued and give rise to a discrimination action. Out of caution, they might feel compelled to dissent from that decision, or attempt to disassociate themselves from it, merely to protect their pocketbooks. For these reasons, imposing liability on the corporate whole rather than each individual who participated in the corporate decision is sensible.
The Court of Appeal stated that, contrary to what the Janken court found, “Concern over litigation costs was not the principal reason for the small employer exemption in the FEHA.” The court relied on our discussion of the act’s purposes in Robinson, supra, 2 Cal.4th at page 240, which did not mention protecting small employers. The issue in Robinson was the reach of the FEHC’s jurisdiction under the FEHA. We viewed the issue from the perspective of the FEHC and discussed the act’s purposes solely in terms of how the FEHC could operate most effectively. Our discussion of the act’s purposes was not intended to be exhaustive or to preclude a finding that the act was also intended to protect small entities.
We find persuasive the Janken court’s discussion of the incongruity between exempting small employers and imposing liability on individual *663supervisors. (Janken, supra, 46 Cal.App.4th at pp. 71-72.) We do not decide merely whether individuals should be held liable for their wrongdoing, but whether all supervisors should be subjected to the ever-present threat of a lawsuit each time they make a personnel decision. Litigation is expensive, for the innocent as well as the wrongdoer. By limiting the threat of lawsuits to the employer itself, the entity ultimately responsible for discriminatory actions, the Legislature has drawn a balance between the goals of eliminating discrimination in the workplace and minimizing the debilitating burden of litigation on individuals.
The Court of Appeal was also concerned that, although ordinarily the employer will be the “deep pocket” defendant, sometimes a plaintiff will be unable to recover from the employer and will turn to the individual. Plaintiff states that the business employers she has sued are now defunct, and she must look to Baird personally for recovery. The Court of Appeal stated that, if that is the case, “it would be unfair to bar Reno from recovering against Baird.” We agree with the Janken court’s response to a similar concern. (Janken, supra, 46 Cal.App.4th at p. 75.) We do not believe the Legislature intended a “legal regime in which individual supervisory employees were personally liable after bankruptcy of the employer . . . .” (Ibid.)
For these reasons, we conclude that individuals who do not themselves qualify as employers may not be sued under the FEHA for alleged discriminatory acts.
D. Discharge in Violation of Public Policy
In addition to the cause of action under the FEHA, plaintiff alleged a cause of action for discharge in violation of public policy. The basis for this cause of action, and the public policy plaintiff cites, is the FEHA. (See Gantt v. Sentry Insurance (1992) 1 Cal.4th 1083, 1095 [4 Cal.Rptr.2d 874, 824 P.2d 680] [“[C]ourts in wrongful discharge actions may not declare public policy without a basis in either constitutional or statutory provisions.” (Original italics.)].) Our conclusion that individual supervisors may not be sued under the FEHA applies likewise to this cause of action.
We decided a similar issue in Jennings v. Marralle (1994) 8 Cal.4th 121 [32 Cal.Rptr.2d 275, 876 P.2d 1074]. There, the plaintiff attempted to sue for age discrimination someone too small to qualify as an “employer” under the FEHA. We held she could not do so because “the ‘public policy’ on which plaintiff relies is not applicable to defendant. He is not an ‘employer’ subject to the age discrimination provisions of the FEHA.” (Id. at p. 130; see also Stevenson v. Superior Court (1997) 16 Cal.4th 880, 894 [66 *664Cal.Rptr.2d 888, 941 P.2d 1157].) The same rationale applies here. It would be absurd to forbid a plaintiff to sue a supervisor under the FEHA, then allow essentially the same action under a different rubric. Because plaintiff may not sue Baird as an individual supervisor under the FEHA, she may not sue her individually for wrongful discharge in violation of public policy. (See also Jacobs v. Universal Development Corp. (1997) 53 Cal.App.4th 692, 704 [53 Cal.App.4th 692, 62 Cal.Rptr.2d 446]; Weinbaum v. Goldfarb, Whitman & Cohen (1996) 46 Cal.App.4th 1310, 1315 [54 Cal.Rptr.2d 462].) We express no opinion on the general question whether discrimination based on medical condition violates one of the “fundamental public policies” that will support a Tameny action. (Tameny v. Atlantic Richfield Co., supra, 27 Cal.3d at p. 176.)
III. Conclusion
We reverse the judgment of the Court of Appeal and remand the matter for'further proceedings consistent with this opinion.
George, C. J., Kennard, J., Baxter, J., and Corrigan, J.,* concurred.
All further statutory references are to the Government Code unless otherwise indicated.
Two recent Court of Appeal decisions have agreed with Janken but contain no independent analysis. (Acuna v. Regents of University of California (1997) 56 Cal.App.4th 639, 651 [65 Cal.Rptr.2d 388]; Melugin v. Zurich Canada (1996) 50 Cal.App.4th 658, 666 & fn. 4 [57 Cal.Rptr.2d 781].) Another cited Janken approvingly and relied on substantial portions of its analysis in discussing the scope of individual liability for harassment under the FEHA. (Fiol v. Doellstedt (1996) 50 Cal.App.4th 1318, 1326, 1328-1329, 1330-1331 [58 Cal.Rptr.2d 308].) (This case involves a claim for discrimination, not harassment. We express no opinion regarding individuals’ liability for harassment.)
Although setting policy is a legislative, not a judicial, function, “Where uncertainty exists consideration should be given to the consequences that will flow from a particular interpre*652tation.” (Dyna-Med, Inc. v. Fair Employment & Housing Com. (1987) 43 Cal.3d 1379, 1387 [241 Cal.Rptr. 67, 743 P.2d 1323].)
Associate Justice of the Court of Appeal, First Appellate District, Division Three, assigned by the Chief Justice pursuant to article VI, section 6 of the California Constitution.'