LAVORATO, Justice.

Hoa Thi Blood, a woman of Vietnamese origin, was employed by Hy-Vee Stores, Inc., as a checker in one of the company's grocery stores. The Iowa Civil Rights Commission found that Hy-Vee had discriminated against Blood on the basis of sex and national origin when it failed to promote her or give her more hours of work. Hy-Vee sought judicial review in the district court of the commission's decision. The district court ruled that substantial evidence supported the commission's finding of discrimination and most of the relief the commission had granted. Hy-Vee appealed and the commission cross-appealed. We affirm on the appeal and affirm in part and reverse in part on the cross-appeal.

On February 9, 1984, Blood filed a complaint against Hy-Vee with the Iowa Civil Rights Commission and the Cedar Rapids Civil Rights Commission. See Iowa Code § 601A.15(1) (1983). She alleged that Hy-Vee had discriminated against her on the basis of national origin and in retaliation for her complaint against Hy-Vee in 1977. See Iowa Code §§ 601A.6(1)(a); 601A.11(2). She later amended her complaint, alleging that Hy-Vee also discriminated against her on the basis of sex. See Iowa Code § 601A.6(1)(a).

The Cedar Rapids Civil Rights Commission staff investigated the complaint and recommended a determination that probable cause existed for the complaint. See Iowa Code § 601A.15(3)(a). Both the city and state commissions adopted this recommendation. Adoption of this recommendation triggered conciliation efforts between Blood and Hy-Vee. See Iowa Code § 601A.15(3)(c), (d). These efforts, however, were unsuccessful. The state commission then set the matter for hearing as a contested case. See Iowa Code § 601A.15(5), (6), (7).

Following the hearing, the hearing officer filed proposed findings of fact and conclusions of law. See Iowa Code § 601A.15(8). The hearing officer found that Hy-Vee had not discriminated against Blood on the basis of retaliation but had discriminated against her on the basis of sex and national origin. The hearing officer proposed certain class relief against Hy-Vee including an affirmative action plan. As to Blood the hearing officer proposed a promotion, back pay with full-time status to 1977, additional hours, front pay, and back pay at an increased rate to 1983. The hearing officer also proposed that Blood be denied punitive and emotional distress damages.

Hy-Vee appealed to the state commission. See Iowa Code § 17A.15(3). The commission adopted the hearing officer's proposed decision with two exceptions: it allowed $15,000 in punitive damages and $10,000 in compensatory damages for emotional distress. See Iowa Code §§ 17A.15(2), (3); 601A.15(8)(a)(7). Hy-Vee requested a rehearing, which the commission denied. See Iowa Code § 17A.16(2).

Hy-Vee then filed a petition for judicial review in the district court. See Iowa Code §§ 601A.17(1); 17A.19. The district court vacated the award of punitive damages; this portion of the court's ruling is not challenged on appeal. The court also vacated the award of additional hours, front pay, and back pay at an increased rate to June 1983. The court failed to consider the award of back pay at full time status to 1977. Finally, the court found that Blood filed her complaint in time and that the commission's violation of its rules on oral argument was harmless error. The court affirmed the commission's decision in all other respects.

Hy-Vee appealed and the commission cross-appealed. See Iowa Code § 17A.20. We first consider the issues raised in Hy-Vee's appeal and then the issues raised in the commission's cross-appeal.

Judicial review of the commission's findings is governed exclusively by the Iowa Administrative Procedure Act. Iowa Code §§ 17A.19; 601A.17(1). Our review of the district court's ruling is at law. Id. at § 17A.20; Iowa Civil Rights Comm'n v. Woodbury County Community Action Agency, 304 N.W.2d 443, 446 (Iowa App. 1981).

Under section 17A.19, the district court acted in an appellate capacity to correct errors of law specified for contested cases in section 17A.19(8)(f). Our review then is limited to determining whether the district court correctly applied the law. If we find the court did correctly apply the law, we affirm. Otherwise, we reverse. Iowa Civil Rights Comm'n v. Woodbury County, 304 N.W.2d at 446.

In deciding whether the district court correctly applied the law, we ask whether the commission's findings are supported by substantial evidence in the record as a whole. Iowa Code § 17A.19(8)(f); Iowa Civil Rights Comm'n v. Woodbury County, 304 N.W.2d at 446. Evidence is substantial if a reasonable person would find it adequate to reach the given conclusion, even if we might draw a contrary inference. Chauffeurs, Teamsters Helpers, Local 238 v. Iowa Civil Rights Comm'n, 394 N.W.2d 375, 379 (Iowa 1986).

The district court found that there was substantial evidence to support the commission's finding of employment discrimination based on sex and national origin. The district court also found that there was substantial evidence to support the commission's award for emotional distress. On appeal Hy-Vee challenges these findings. Hy-Vee also challenges the district court's findings that Blood filed her complaint in time and that the commission's violation of its rules on oral argument was harmless error.

I. Employment Discrimination Theories.

Blood based her employment discrimination claims on alleged violations of Iowa Code section 601A.6(1)(a). Among other things, this statute prohibits any person from discriminating in employment against any employee because of sex or national origin.

The United States Supreme Court recognizes two theories to prove employment discrimination under Title VII of the Civil Rights Act of 1964. One is the disparate treatment theory and the other is the disparate impact theory. Holdeman, Watson v. Ft. Worth Bank Trust: The Changing Face of Disparate Impact, 66 Den.U.L.Rev. 179, 180 (1989). Because the commission used the two theories, we discuss both.

The two theories are defined in International Brotherhood of Teamsters v. United States:

"Disparate treatment" . . . is the most easily understood type of discrimination. The employer simply treats some people less favorably than others because of their race, color, religion, sex, or national origin. Proof of discriminatory motive is critical, although it can in some situations be inferred from the mere fact of differences in treatment.

. . . .

Claims of disparate treatment may be distinguished from claims that stress "disparate impact." The latter involves employment practices that are facially neutral in their treatment of different groups but that in fact fall more harshly on one group than another and cannot be justified by business necessity. Proof of discriminatory motive, we have held, is not required under a disparate-impact theory. Either theory may, of course, be applied to a particular set of facts.

431 U.S. 324, 335, n. 15, 97 S.Ct. 1843, 1854, n. 15, 52 L.Ed.2d 396, 415, n. 15 (1977) (citations omitted). In short, the disparate treatment theory focuses on the employer's motivation; the disparate impact theory focuses on the consequences of the employer's conduct.

A. Disparate treatment. We recently summarized the analytical framework for allocating the burden and order of presentation of proof under chapter 601A in Hamilton v. First Baptist Elderly Hous. Found., 436 N.W.2d 336 (Iowa 1989). We borrowed this framework of analysis from two United States Supreme Court decisions, which applied it to Title VII cases: McDonnell Douglas Corp. v. Green, 411 U.S. 792, 93 S.Ct. 1817, 36 L.Ed.2d 668 (1973) and Texas Dep't of Community Affairs v. Burdine, 450 U.S. 248, 101 S.Ct. 1089, 67 L.Ed.2d 207 (1981). See Hamilton, 436 N.W.2d at 338-39.

There are three stages in this analysis. In the first stage the employee must establish a prima facie case of intentional discrimination by a preponderance of the evidence. Hamilton, 436 N.W.2d at 338. An employee establishes a prima facie case by showing that (1) the employee belonged to a protected group, (2) the employee made application and was qualified for a job for which the employer was seeking applicants, (3) the employee was rejected, (4) the position remained open and the employer continued to seek applicants with similar qualifications. Once the prima facie case is established, a presumption arises that the employer discriminated against the employee. Id.

In the second stage the employer must go forward with evidence to rebut the presumption of discrimination. The employer does so by producing evidence that the employee was not hired or promoted for a legitimate, nondiscriminatory reason. Id.

In the third stage the employee has the burden to show that the employer's proffered reason was not the true reason for the employment decision. The employee may meet this burden in two ways. The employee may directly persuade the fact finder that a discriminatory reason more likely motivated the employer or may indirectly show the employer's proffered reason is not worthy of belief. Id. at 339.

In all three stages the burden of persuasion rests with the employee. The only burden the employer shoulders is the burden to go forward with the evidence to rebut the presumption of discrimination. Id. at 338.

McDonnell Douglas and Burdine involved "pretext" cases — the employer's stated reasons for the employment decision are false. Until Price Waterhouse v. Hopkins, 490 U.S. ___, 109 S.Ct. 1775, 104 L.Ed.2d 268 (1989), there was no Supreme Court decision on "mixed motive" cases — those cases in which both legitimate and illegitimate considerations played a part in the employment decision.

A plurality of the Court in Price Waterhouse fashioned a different framework of analysis for the "mixed motive" cases. Under this analysis, the employee must first establish by a preponderance of the evidence that a discriminatory reason played a part in the employment decision. Price Waterhouse, 490 U.S. at ___, 109 S.Ct. at 1787, 104 L.Ed.2d at 285. The employer's burden at this point "is most appropriately deemed an affirmative defense" rather than a shift in the burden of proof. If the employer wishes to prevail, it must prove by a preponderance of the evidence that "its legitimate reason, standing alone, would have induced it to make the same decision." Price Waterhouse, 490 U.S. at ___, 109 S.Ct. at 1792, 104 L.Ed.2d at 289. Stated another way,

[w]hen a plaintiff . . . proves that [a discriminatory reason] played a motivating part in an employment decision, the defendant may avoid a finding of liability only by proving by a preponderance of the evidence that it would have made the same decision even if it had not taken the [discriminatory reason] into account.

Price Waterhouse, 490 U.S. at ___, 109 S.Ct. at 1795, 104 L.Ed.2d at 293. Such an approach is justified in mixed motivation cases because "the employer has created uncertainty as to causation by knowingly giving substantial weight to an impermissible criterion." Price Waterhouse, 490 U.S. at ___, 109 S.Ct. at 1796, 104 L.Ed.2d at 295 (O'Connor, J., concurring).

B. Disparate impact. The United States Supreme Court first applied the disparate impact theory to a Title VII claim in Griggs v. Duke Power Co., 401 U.S. 424, 91 S.Ct. 849, 28 L.Ed.2d 158 (1971). Later in Albemarle Paper Co. v. Moody, 422 U.S. 405, 95 S.Ct. 2362, 45 L.Ed.2d 280 (1975), the Court more fully developed the elements of proof under this theory.

Under the disparate impact theory as it developed under Griggs and Albemarle, a disparate impact case — like a disparate treatment case — proceeds through three stages. In the first stage the employee must show that a particular employment practice has an adverse impact on a protected group in "marked disproportion to its impact on employees outside that group." Holdeman, 66 Den.U.L.Rev. at 182. This stage depends almost entirely on statistical evidence. According to one commentator, this stage

often requires voluminous discovery, thorough and detailed analysis of the employer's total organization and operation, and expert testimony by statisticians, industrial psychologists, and personnel managers. The statistical comparisons must be valid in terms of significance (based on a sample large enough to yield reliable results), scope (covering an appropriate category of employees), and time (covering an appropriate length of time).

Id.

Once the employee has established a prima facie case, the case proceeds to the second stage. At this point, the burden of persuasion shifts to the employer to show the business necessity of the challenged employment practice. Albemarle, 422 U.S. at 425, 95 S.Ct. at 2375, 45 L.Ed.2d at 301; Griggs, 401 U.S. at 432, 91 S.Ct. at 849, 28 L.Ed.2d at 158; Holdeman, 66 Den.U.L.Rev. at 182. The quality of proof here under Albemarle and Griggs is important:

The employer must show that the employment [practice] is manifestly job related or necessary to [the employer's] business purpose, despite its disparate effect upon protected groups.

Holdeman, 66 Den.U.L.Rev. at 182.

If the employer succeeds in showing the business necessity of the employment practice, the case enters the third stage. In this stage, the employee must show there are other reasonable alternatives that would have less adverse impact. Albemarle, 422 U.S. at 425, 95 S.Ct. at 2375, 45 L.Ed.2d at 301; Holdeman, 66 Den.U.L.Rev. at 183.

The next development in the disparate impact theory came in Watson v. Fort Worth Bank Trust, 487 U.S. 977, 108 S.Ct. 2777, 101 L.Ed.2d 827 (1988). Before Watson Supreme Court cases on the disparate impact theory had only dealt with employment decisions based on objective criteria. See, e.g., Connecticut v. Teal, 457 U.S. 440, 443, 102 S.Ct. 2525, 2528, 73 L.Ed.2d 130, 134-35 (1982) (written examination); Dothard v. Rawlinson, 433 U.S. 321, 323-24, 97 S.Ct. 2720, 2724, 53 L.Ed.2d 786, 794 (1977) (height-weight requirement); Albemarle, 422 U.S. at 411, 95 S.Ct. at 2368, 45 L.Ed.2d at 293 (intelligence test); Griggs, 401 U.S. at 428, 91 S.Ct. at 852, 28 L.Ed.2d at 162 (intelligence test). The federal circuit courts were split on the question whether the theory could be applied to employment decisions based on subjective criteria. So the Supreme Court granted certiorari in Watson to resolve the question. Holdeman, 66 Den.U.L.Rev. at 189.

In Watson all the justices, except Justice Kennedy who took no part, agreed that the disparate impact theory applies to employment decisions based on subjective criteria. Watson, 487 U.S. at 980, 108 S.Ct. at 2781, 101 L.Ed.2d at 849. The plurality opinion, written by Justice O'Connor and joined by three other justices, undertook to reexamine the allocation of the burden of proof in disparate impact cases. This re-examination became the basis of the majority opinion in the next case on the subject: Wards Cove Packing Co., Inc. v. Atonio, ___ U.S. ___, 109 S.Ct. 2115, 104 L.Ed.2d 733 (decided June 5, 1989) (5-4 decision).

Although Watson expanded the disparate impact theory to include subjective criteria, Wards tightened the proof requirements for disparate impact cases. Id. at ___, 109 S.Ct. at 2115, 104 L.Ed.2d at 733. Wards significantly affects all three stages of proof. Now in the first stage of proof,

"the [employee's] burden in establishing a prima facie case goes beyond the need to show that there are statistical disparities in the employer's work force. The [employee] must begin by identifying the specific employment practice that is challenged. . . . Especially in cases where an employee combines subjective criteria with the use of more rigid standardized rules or tests, the [employee] is . . . responsible for isolating and identifying the specific employment practices that are allegedly responsible for any observed statistical disparities."

Id. at ___, 109 S.Ct. at 2124, 104 L.Ed.2d at 751 (quoting the plurality in Watson) (emphasis added). In short, the employee must show a causal link between the challenged employment practice and the disparate impact.

To establish this link, an employee may not simply rely on statistics that show an imbalance of a protected class in the work force. By rejecting such statistical comparisons in the work force, the Court significantly limited the role of statistics in the first stage. According to the Court, the proper comparison now is between the

[protected class] composition of [the at-issue jobs] and the [protected class] composition of the qualified . . . population in the relevant labor market. It is such a comparison — between the [protected class] composition of the qualified persons in the labor market and the persons holding at issue jobs — that generally form the proper basis for the initial inquiry in a disparate impact case.

Id. at ___, 109 S.Ct. at 2121, 104 L.Ed.2d at 747 (citations omitted).

Hazelwood School Dist. v. United States, 433 U.S. 299, 97 S.Ct. 2736, 53 L.Ed.2d 768 (1977) provides an example of such a comparison. There the Court found that the proper comparison was between the "racial composition of the [employer's] teaching staff and the racial composition of the qualified public teacher population in the relevant labor market." Id. at 308, 97 S.Ct. at 2742, 53 L.Ed.2d at 777.

If the employer's practices deter or create barriers to the members of the protected class, however, "the [statistical] analysis would be different." Id. ___ U.S. at ___, 109 S.Ct. at 2122 n. 7, 104 L.Ed.2d at 748, n. 7. The Court does not specify what that analysis would be. But the implication of the Court's statement seems to be that internal work force comparisons would then apply. This is so because the employer's deterrence skews the proper comparison. Cf. id. at ___, 109 S.Ct. at 2136, 104 L.Ed.2d at 754 (Blackmun, J., dissenting) (the structure of an industry may render meaningless any statistical comparison other than an internal work force comparison).

Under the Court's revamped disparate impact theory in Wards, the dispositive issue in the second stage is

whether a challenged practice serves, in a significant way, the legitimate employment goals of the employer. The touchstone of this inquiry is a reasoned review of the employer's justification for his use of the challenged practice. A mere insubstantial justification in this regard will not suffice, because such a low standard of review would permit discrimination to be practiced through the use of spurious, seemingly neutral employment practices. At the same time, though, there is no requirement that the challenged practice be "essential" or "indispensable" to the employer's business for it to pass muster: this degree of scrutiny would be almost impossible for most employers to meet, and would result in a host of evils. . . .

Id. at ___, 109 S.Ct. at 2125-26, 104 L.Ed.2d at 752-53 (citations omitted).

The employer's burden in the second stage is now clearly one of producing evidence of a business justification and not a burden of persuasion. Id. at ___, 109 S.Ct. at 2126, 104 L.Ed.2d at 753. The Court acknowledges that some of its earlier decisions can be read as suggesting the burden of persuasion in this stage was on the employer. Id.

In the third stage of proof, the employee can still prevail even though the fact finder believes the employer had a business justification for its employment decision. To prevail, the employee must convince the fact finder that (1) the employer has alternative hiring practices that could reduce the disparate impact and (2) the employer refuses to adopt these alternatives. Id. at ___, 109 S.Ct. at 2126, 104 L.Ed.2d at 753-54.

Like the disparate treatment theory in pretext cases, the burden of persuasion in a disparate impact case now rests with the employee in all three stages of proof. Id. at ___, 109 S.Ct. at 2119, 104 L.Ed.2d at 733. In our view this is the most significant effect of the Court's redefinition in Wards of the burden of proof under the disparate impact theory.

II. Issues on Appeal.

A. Substantial evidence to support employment discrimination based on sex. On our review of the record as a whole, we find that there is substantial evidence to support the following findings of fact. Blood was employed as a checker at Hy-Vee Store No. 5 in Cedar Rapids beginning on August 5, 1975. She was still employed as a checker at the time of the hearing before the commission in 1985.

At the time of the hearing Hy-Vee owned 159 stores — 115 were in Iowa and the remainder were in five other states. Store No. 5, where Blood worked, had four departments: grocery/general, meat, produce, and frozen. Each department had its own manager who reported to the store manager. The grocery/general department was supervised by the assistant store manager.

The line of authority was the same in most Hy-Vee stores. The store manager was the highest authority in a store, followed by the assistant store manager, shift managers (formerly "swing men"), aisle coordinators (formerly "aislemen"), stocker/checkers, stockers and checkers, and, last, courtesy/carryout clerks. Stockers/checkers apparently performed a combination of stocking and checking activities. Stockers and checkers were viewed as being on the same level of authority.

"Full-time" employees worked forty or more hours per week, "regular-time" employees worked thirty to forty hours per week, and "part-time" employees worked less than thirty hours per week. Full-time compensation included all benefits and store bonuses, regular-time included only the benefits, and part-time included only the reduced benefits of vacation/holiday pay.

Hy-Vee usually hired employees as courtesy clerks, but occasionally it would hire a new employee, like Blood, as a checker. The store managers made promotions beyond these levels on an ad hoc basis, without posting any notices and without considering any written qualifications.

An employee who wanted to be promoted to an aisle coordinator position generally had to have stocking and checking experience. An employee gained this experience in either the stocker or stocker/checker positions. Stockers spent 75% of their time stocking and 25% checking. Checkers, on the other hand, spent 75% of their time checking and only 25% stocking.

In promoting or hiring, however, Hy-Vee gave only men employees or applicants the stocker positions, while it gave women the checker positions. Don Sapp, the manager of Store No. 5, admitted that this was the policy being used in his store:

Q. . . . my question to you was, males have always had a better opportunity to be promoted to aisle coordinator, isn't that correct? A. Yes.

Q. And that decision to make . . . [an] employee a stocker or a checker, that's based just on sex, isn't it? A. No.

Q. It's not? Well, you just told me all the females are made checkers, all the males are made stockers isn't that right? A. Yes.

Q. The only difference between the checkers and the stockers was their sex, isn't that right? A. Yes.

Q. And you also told me that there were no males that were checkers and no females that were stockers? A. Yes.

Data from Store No. 5 showed that in 1980 and 1985, all regular-time and full-time stockers were men. In 1980 all regular-time and full-time checkers were women.

Data from the ten Cedar Rapids-area stores showed that in 1980 all 170 part-time stockers were men; in 1985, 96 were men and 2 were women. The same data showed that in these stores in 1980, 165 part-time checkers were women and 8 were men; in 1985, 159 were women and 18 were men.

In a company bulletin dated July 31, 1974, the president of Hy-Vee explained the company's promotion policy, which implicitly included not assigning women to the stocker positions:

1. The physical work factor eliminates many of [the women]. For years there were laws limiting the weight which a woman could be required to lift. Some more sturdy women could qualify but few of them chose to.

2. The transfer requirement eliminates several more. In every case assistant managers (and often swing men) are required to work in several stores and towns to secure the broad experience necessary to qualify them for further promotion to store management. Most women are or will be married and must live where their husband is employed. They thus are not willing to accept a transfer.

3. There is one subjective factor which is not easy to define but still does exist. In our experience many men and some women find it difficult to accept supervision from women. Thus the matter of personality and temperament becomes even more important than normal (and it is higher on Hy-Vee's list of management criteria than in most supermarket companies because of our autonomous operating philosophies). Obviously some women would so qualify but ingrained prejudices of employees themselves (including women) make the selection process more difficult.

This policy statement — characterized by the district court as direct evidence of sexual stereotyping — was in effect at the times in question here. It has apparently been replaced by a very general policy statement entitled "Discrimination and Sexual Harassment."

The result of the 1974 promotion policy is reflected in the sexual percentage breakdown for managerial positions. In the ten Cedar Rapids stores in 1980, 98.9% of the women employees were in nonmanagerial positions; in 1985, 93.2% of them were still in such positions. By comparison, 48.2% of the men employees were in managerial positions in 1980, and 51.5% of them were in such positions in 1985.

Women employees were not represented at all in the ranks of store managers, assistant store managers, or shift managers in either 1980 or 1985. No women were aisle coordinators in 1980, and only one was in 1985. Among department managers, thirty-nine were men and one was a woman in 1980; forty-two were men and six were women in 1985.

Frank B. Martin, an assistant professor of statistics at the University of Minnesota, appeared as an expert witness. Based on his statistical analysis he concluded that the number of men who hold full-time positions is disproportionately high to the overall proportion of men in the Hy-Vee system. The deviation from chance of the proportion of men and women in managerial positions, Martin also concluded, was highly statistically significant: close to six standard deviations.

When Blood amended her complaint to allege employment discrimination based on sex, she said:

[Hy-Vee] routinely hires and retains women in the position of "checker" with no opportunity for promotion. Males hired by [Hy-Vee] spend the majority of their time as "stockers," and thus have the experience regarded as necessary for promotion. During the last ten years, I have been classified as a "checker" and despite repeated requests for promotion and a continuing excellent work record, I have never been promoted. During that time period only one woman has been promoted to the position of aisleman/coordinator or to a supervisor position. The line of progression established and implemented in Hy-Vee Store No. 5, Cedar Rapids, Iowa denies women the same promotional opportunities that are given to men. I hereby charge Hy-Vee Food Stores and its representatives with discrimination on the basis of sex in violation of the Municipal Code of Cedar Rapids, Iowa and Iowa Code chapter 601A. . . .

Applying the disparate impact theory, the commission found that Blood had established a prima facie case of employment discrimination based on sex in violation of section 601A.6(1)(a). The district court concluded there was substantial evidence to support this finding. We agree.

Blood is a woman. Women are a protected class under Iowa Code section 601A.6(1)(a).

Blood identified the particular discriminatory practice responsible for the disparate impact on promotion of women: the sexual segregation of the stocker and checker positions. This segregation is very explicit.

Blood did use internal work force comparisons to establish the causal link between the discriminatory practice and the disparate impact. Although the Supreme Court generally disapproved of such a comparison in Wards, we think the comparison is nevertheless proper here.

Hy-Vee was explicitly discriminatory by the segregation, giving its women employees the clear message that they would never be considered to have the proper stocking experience necessary for promotion. Hy-Vee also had in effect a policy statement outlining the reasons for not promoting women.

So Hy-Vee was deterring its women employees from applying for promotions. As the Court said in Wards, the analysis must be different when the employer's barriers are deterring potential applicants of a suspect class.

Martin, the statistics expert, translated the 1985 statistical data into standard deviation units. He testified that when the difference between the actual number of women employed at the managerial level and the expected number is greater than two or three standard deviations, the likelihood that this results from chance is remote.

Statisticians agree that with two standard deviations, there is only a 5% possibility that the observed incident is caused by chance. With a standard deviation of more than three, there is only a .27% possibility that the observed incident is caused by chance. D.W. Barnes and J.M. Conley, Statistical Evidence in Litigation 135 (1986).

The United States Supreme Court has accepted a two or three standard deviation as statistically significant in discrimination cases. See Hazelwood, 433 U.S. at 308-12, 97 S.Ct. at 2741-43, 53 L.Ed.2d at 778-79 (1976) (a "pattern or practice" of employment discrimination case brought by black teachers under Title VII), Castaneda v. Partida, 430 U.S. 482, 494-97, 97 S.Ct. 1272, 1280-81, 51 L.Ed.2d 498, 510-12 (1977) (jury discrimination case brought as an equal protection claim). In Hazelwood, the deviation was more than six standard deviations in one year and more than five in another year. The Court held that

as a general rule for such large samples, if the difference between the expected value and the observed number is greater than two or three standard deviations then the hypothesis that teachers were hired without regard to race would be suspect.

Hazelwood, at 308 n. 14, 97 S.Ct. at 2742 n. 14, 53 L.Ed.2d at 778 n. 14.

The Court concluded that the demonstrated differences (63 expected versus 16 observed in 1972-73 and 70 expected versus 22 observed in 1973-74) were substantial on their face and enough to establish a prima facie case. Id. at 305, 97 S.Ct. at 2740, 53 L.Ed.2d at 776. Several federal courts have since used the two or three standard deviation test in discrimination cases. See, e.g., Board of Educ. of City School Dist. of N Y v. Califano, 584 F.2d 576, 584 n. 29 (2d Cir. 1978) aff'd sub. nom. Board of Educ. of City School Dist. of N.Y. v. Harris, 444 U.S. 130, 100 S.Ct. 363, 62 L.Ed.2d 275 (1979); Battle v. Anderson, 457 F. Supp. 719, 732 (E.D.Okla. 1978).

In 1985 there were 112 managerial positions in the ten Hy-Vee stores located in Cedar Rapids. During that same year there were 294 employees; 206 were men and 88 were women. According to Martin, based on those figures one would expect to see 30 women in managerial positions. There were only six. This represented a standard deviation of six. The standard deviation in 1980 was even greater. The chance of seeing a deviation of six in Martin's opinion is "miniscule." He concluded that it was "extremely unlikely" that the distribution of men and women in managerial positions was based on chance.

The commission relied in part on this statistical data in finding that Blood had established a prima facie case of discrimination. In its analysis the commission applied the "two or three standard deviation" test recognized in the federal cases.

In our view the statistical disparity shown here is so dramatic that it alone establishes prima facie proof of purposeful discrimination. Although Wards limits the use of statistics in discrimination cases, it still recognizes that statistical proof can alone make out a prima facie case. Wards, ___ U.S. at ___, 109 S.Ct. at 2118, 104 L.Ed.2d at 747 (citing International Bhd. of Teamsters, 431 U.S. at 339, 97 S.Ct. at 1856, 52 L.Ed.2d at 417-18, and Hazelwood, 433 U.S. at 307-08, 97 S.Ct. at 2741, 53 L.Ed.2d at 777). This is especially true when, as here, an employer does not offer meaningful statistical evidence to attack or rebut the statistical conclusion of employment discrimination.

More than complicated statistics, simple observation reveals sex discrimination here. There is the explicitly discriminatory policy statement by Hy-Vee's president. There is the store manager's admission that he carried out this policy. And, finally, there is the "inexorable zero:" the "glaring absence" of women from any but the lower level supervisory positions. See International Bhd. of Teamsters, 431 U.S. at 342 n. 23, 97 S.Ct. at 1858 n. 23, 52 L.Ed.2d at 419 n. 23.

Hy-Vee's attempt to discount the statistics by showing store No. 5 promoted four women other than Blood misses the mark. The promotion of four women in a company-wide system that shows, statistically, very little chance of consistent advancement for women does not in our view overcome Blood's prima facie case of employment discrimination. Finally, this rebuttal evidence loses all probative value when we consider it in the light of Hy-Vee's policy reasons for segregating stockers and checkers on the basis of sex.

Hy-Vee came forward with no evidence to justify the use of the discriminatory practice. Nor does Hy-Vee even argue any business justification for the practice. The reason is obvious: the practice is so explicitly discriminatory that it cannot be justified. We think the commission correctly determined liability at this stage of the case. Cf. Hamilton, 436 N.W.2d at 338 (judgment in favor of employee is warranted when employer offers no evidence to rebut employee's prima facie case of employment discrimination).

B. Substantial evidence to support employment discrimination based on national origin. As to this phase of Blood's case, the record evidence establishes the following facts. In 1972 Blood came to the United States from Vietnam with her three children and husband, a United States citizen. She began working for Hy-Vee in August 1975.

In October 1977 Blood and Hy-Vee entered into a conciliation agreement in settlement of an employment discrimination complaint. As part of that agreement, Hy-Vee agreed to promote Blood to a thirty hour week and regular-time status. Shortly after that agreement was reached, Blood repeatedly asked Sapp, the store manager, for additional hours. Sapp told her no additional hours were available.

In 1980 two white employees, Laura Melick and Sue Zahari, asked Sapp for more hours. They received an additional six or seven hours. Before this, Melick and Zahari had been working thirty hours per week as regular-time checkers. Though Blood had seniority over both women, Sapp did not offer her additional hours.

In the summer of 1981 Blood again asked Sapp for more hours. She followed up this request in late 1981 and early 1982. Finally, in February 1982 Sapp gave her an additional two hours, bringing her hours to a total of 32 per week. From the summer of 1982 until the hearing before the commission, Blood asked for, but did not receive any, additional hours.

In June 1981 Blood asked Sapp to promote her. Sapp told her that her chances of promotion would be greater if she would transfer and gain some multi-store experience. Blood then informed Sapp in writing that she would be willing to transfer within the Cedar Rapids area. Thereafter Blood checked back with Sapp about promotions about every six months.

In February 1982 Blood asked Sapp to recommend her for the position of deli manager at a new store in Cedar Rapids. Sapp told her he could not recommend her because she had no prior deli experience in a Hy-Vee store. He also told her that he could only recommend her for a deli trainee position, which would require her to take a four or five dollar cut in pay.

About a year later, Sapp promoted three white women employees: Melick, Zahari, and Kathy Wisehart. Sapp promoted Melick from stocker-checker and part-time bookkeeper to full-time bookkeeper in October 1982. Blood then asked Sapp to promote her to the part-time bookkeeping position that was vacant. But he gave that position to Wisehart. Wisehart had been a checker who had no prior bookkeeping experience and who had less seniority than Blood. Blood, on the other hand, had attended commercial college for two years where she took courses in accounting and business management. She also had five years work experience as a bookkeeper before taking the job at Hy-Vee.

Sapp promoted Zahari to aisle coordinator in June 1983. The following December Sapp also gave her supervisory front-end duties in the cashier section of the store. Blood, however, was not offered either position. Several months later Blood filed her discrimination complaint.

In this instance, the commission applied the disparate treatment analysis. It found that Blood had also established a prima facie case of employment discrimination based on national origin. The district court concluded there was substantial evidence to support this finding. For reasons that follow we also agree with this conclusion.

Blood is Vietnamese. Because of her national origin she is a member of a protected class.

As the district court noted, it is difficult to determine whether Blood applied for and was qualified for any particular job. Hy-Vee had no formal policy about promotions and had no job descriptions informing employees what qualifications were necessary for promotion. Promotions were based on the subjective decision of the store manager. In addition, employees had to rely on word of mouth or their own awareness about job openings because vacancies were not posted.

The record is clear, however, that Blood made known to Sapp her desire for promotion and more working hours. The record is also clear that when more hours and promotions were available, Sapp simply ignored Blood's requests. Instead, Sapp gave those hours and promotions to non-Vietnamese employees who had less qualifications and seniority than Blood.

At this stage of the case the burden was on Hy-Vee to go forward with evidence to rebut the presumption of discrimination. It could do so by producing evidence that Blood was not promoted or given more hours because of a legitimate, nondiscriminatory reason.

In proffering reasons why Blood was not given more hours or promoted, Hy-Vee triggered the third stage of the case. In this stage Blood had the burden of persuading the commission that (1) a discriminatory reason more likely motivated Hy-Vee, or (2) the proffered reasons were not worthy of belief.

One reason Sapp gave for denying Blood more hours was that no hours were available. Sapp testified that Blood did not ask for additional hours until 1981 after Melick and Zahari had received extra hours. By then, no extra hours were available. He did eventually give Blood two additional hours in February 1982.

The commission found that Sapp's reason was not true. Like the district court we think there is substantial evidence to support this finding. Blood testified she began asking Sapp for more hours in late 1977, shortly after she and Hy-Vee entered the conciliation agreement. She continued to ask for more hours but Sapp always told her none were available. Yet he gave Zahari and Melick more hours in 1980 after they asked. Based on this evidence, reasonable minds could conclude more hours were available but Sapp never offered them to Blood.

Sapp's reasons for rejecting Blood for the part-time bookkeeping job in 1983 were two-fold. He thought Wisehart — the employee who did get the position — "had the mentality to go in and do the job." And he thought Blood, because of her children, could not work an early morning shift.

The commission found Sapp's reasons were "clearly subjective and lacking in credence." The district court found there was substantial evidence to support this finding and so do we. Blood's business education and experience clearly qualified her for the job. In addition, Sapp never asked Blood if she would be available to work the early morning shift — he simply assumed she would not be able to do so. Blood, on the other hand, testified she would have been willing to work this shift so she could have more bookkeeping hours.

Sapp also gave two reasons why he gave Zahari rather than Blood the aisle-coordinator promotion and the front-end supervisory duties in 1985. At the time, he believed Zahari would get along better with the other employees. In addition, these duties would have required Blood to carry customer's groceries outside the store. He thought Blood, because of her nationality, would have trouble with the cold weather.

The commission refused to consider either reason legitimate and nondiscriminatory. As to the first, it found that Zahari had been disciplined three times for flareups with other employees. Blood, on the other hand, had only one blemish on her record. She had been disciplined for using store coupons, an incident Sapp himself said he would not consider in promotions. In addition, Sapp and his assistant store manager described Blood as a good or very good employee.

As to the second, the commission found that Sapp's testimony was direct evidence that Blood's nationality was "an admitted factor in not allowing her to work the front-end." In its review, the district court noted that Sapp never asked Blood whether cold weather would keep her from doing her job. It also noted that Blood testified she would have been willing to take the front-end position even though occasionally she would have to go out into the cold.

The district court concluded there was substantial evidence to support the commission's findings that these reasons were pretextual. We agree.

The commission also found that the need for multi-store experience, although generally true at upper level management, was not a necessity for lower level in-store promotions. And it found that Hy-Vee employees did not have to take $4 to $5 pay cuts in order to transfer to another store. Sapp also gave Blood these reasons when she asked to be promoted. We think there is substantial evidence to support the commission's refusal to give credence to these reasons.

The commission concluded that Blood had proven all of Hy-Vee's proffered reasons were unworthy of belief. It concluded, therefore, that Blood had proven her allegations of employment discrimination on the basis of national origin. The district court thought there was substantial evidence to support these conclusions, and so do we.

C. Substantial evidence to support an award for emotional distress. The hearing officer rejected Blood's claim for emotional distress. The commission reversed this part of the hearing officer's decision and awarded Blood $10,000 for this claim. This was $5000 more than she requested. Hy-Vee contended in district court, as it does here, that the award is not supported by substantial evidence. In support of this contention, Hy-Vee argues that the emotional distress was not the result of outrageous conduct, it was not severe enough, and it was not caused by anything Hy-Vee did.

We recently held that damages for emotional distress are recoverable under our civil rights statute. Chauffeurs, 394 N.W.2d at 383. However, we left open the question whether a civil rights complainant must show outrageous conduct. Id.

Following Chauffeurs we held that emotional distress is an element of damages for retaliatory discharge. We also held that the distress need not be severe to be compensable. See Niblo v. Parr Mfg., Inc., 445 N.W.2d 351, 357 (Iowa 1989) (jury found that plaintiff was fired because she was going to file a workers' compensation action). Our survey of the cases led us to conclude that we had been lenient in allowing incidental damages without a showing of physical injury in tort cases involving intentional or illegal conduct. Id. at 354-55. We said that the "wrongful or retaliatory discharge in violation of public policy is an intentional injury committed by the employer against an employee who chooses to exercise some substantial right." Id. at 355.

For this reason, "[w]e believe[d] that public policy also require[d] us to allow a wrongfully discharged employee a remedy for his or her complete injury." Id. We then cited with approval the majority of jurisdictions that were in accord. Id. at 355-56. For these same reasons, we rejected the contention that the distress had to be severe to be compensable. Id. at 355.

We think our reasoning in Niblo applies with equal force here. Hy-Vee's acts were intentional and violated not only a statute but a strong public policy underlying that statute. In Chauffeurs we recognized that our civil rights statute is to be liberally construed to eliminate unfair and discriminatory acts and practices. Chauffeurs, 394 N.W.2d at 382; Iowa Code § 601A.18.

We therefore hold that a civil rights complainant may recover compensable damages for emotional distress without a showing of physical injury, severe distress, or outrageous conduct. Other courts interpreting civil rights statutes have reached the same result. See, e.g., Amos v. Prom, Inc., 115 F. Supp. 127, 132 (N.D. Iowa 1953) (interpreting an Iowa civil rights statute on public accommodation); Buckley Nursing Home v. Commission Against Discrimination, 20 Mass. App. 172, 182, 478 N.E.2d 1292, 1299 (1985); Catalina Beach Club v. State Div. of H.R., 95 A.D.2d 766, ___, 463 N.Y.S.2d 244, ___ (1983); William v. Joyce, 4 Or. App. 482, 483, 479 P.2d 513, 514 (1971); Gray v. Serruto Builders, Inc., 110 N.J. Super. 297, 314, 265 A.2d 404, 414 (1970).

Blood suffers from psoriasis, a hereditary skin disease with no known cure. Symptoms include red patches of skin that often itch, deformed and thickened nails, and a buildup of the outer layer of skin. Blood first showed signs of the disease in 1973.

According to a dermatologist who began treating Blood in February 1975, psoriasis is an unpredictable disease in which spontaneous remissions, and exacerbations are very typical. He also testified that stress plays a part in aggravating the disease and that removal of stress helps a psoriatic get better. In his opinion Blood is a typical psoriatic whose condition might be made worse by stress.

Blood testified how she felt when others were promoted and she was not. She was upset, felt bad, and had headaches. At one point she described her reaction to the stress she was experiencing: "I would like to have a bag to beat it up, punch it in." She described how her psoriasis would flare up and cause her to itch. She attributed the flareups to this stress. She came to this conclusion following a layoff from work for three months because of an injury. During those three months, the psoriasis cleared up. Once she returned to work, the psoriasis returned.

The district court concluded this evidence was sufficient

to allow a reasonable mind to conclude that Blood suffered emotional distress in the form of anxiety, headaches, and flareups of her psoriasis and that the distress was proximately caused by the discriminatory policies and acts of Hy-Vee and its representatives.

We agree.

Hy-Vee raises two additional issues about the award for emotional distress. It contends the award for emotional distress is excessive because the commission awarded Blood $5000 more than she requested. Hy-Vee also contends Blood cannot recover for occupational injury or disease under chapter 601A because her remedy is limited to an action under workers' compensation law.

In an answer to an interrogatory posed by Hy-Vee, Blood stated she sustained $12,500 in emotional distress damages. Later Blood filed exceptions to the proposed decision of the hearing officer who had denied Blood damages for emotional distress. In the exceptions Blood requested $5000 for emotional distress. The commission responded by awarding her $10,000.

Hy-Vee did not raise the excessive damage award issue in its petition for judicial review nor in its extensive briefs filed with the district court. Apparently this accounts for the district court judge's failure to address the issue in his detailed and well-reasoned decision. Because Hy-Vee did not raise the issue before the district court, it did not preserve error for appellate review on this issue. See Bonds v. State, 447 N.W.2d 135, 136 (Iowa 1989).

We also note that Hy-Vee did not raise its contention that the award is preempted by workers' compensation in its application for rehearing to the commission. See Iowa Code § 17A.16(2) ("any party may file an application for rehearing, stating the specific grounds therefore and the relief sought. . . ."). To preserve error for appeal, a party must raise the issue before the agency. Chauffeurs, 394 N.W.2d at 382. We think Hy-Vee also failed to preserve error on this issue.

D. Timeliness of the complaint. In its exception to the hearing officer's proposed decision, Hy-Vee contended that Blood's complaint was not filed within the time prescribed by Iowa Code section 601A.15(12). In support of its exception, Hy-Vee argued there was no discriminatory activity within the time prescribed by section 601A.15(12). The commission did not rule on this issue; however, the district court did.

We determined that the district court made findings of fact and conclusions of law on the statute of limitations issue that should properly have been made by the commission. So we ordered a limited remand to allow the commission to address the issue. It did and found that Hy-Vee's national origin discrimination in the assignment of working hours and promotion were both of a continuing nature because such discriminatory practices constituted a series of related acts. The commission then found that the last day of each discriminatory practice — discrimination in the assignment of work hours and promotion — fell within the time prescribed by section 601A.15(12).

The commission also found that Hy-Vee's sex discrimination promotion policies and practices were of a continuing nature, constituting a systematic policy of discrimination in effect within the time prescribed by section 601A.15(12).

Iowa Code section 601A.15(12) provides that

[a] claim under [chapter 601A] shall not be maintained unless a complaint is filed with the commission within one hundred eighty days after the alleged discriminatory or unfair practice occurred.

In connection with this statutory limitation a commission rule provides that

[i]f the alleged unlawful discriminatory practice or act is of a continuing nature, the date of the occurrence of the alleged unlawful practice shall be deemed to be any date subsequent to the commencement of the alleged unlawful practice up to and including the date upon which the unlawful practice has ceased.

161 Iowa Admin. Code 3.3(2) (1989) (formerly 240 Iowa Admin. Code 1.3(3) "b" (1983)). The commission rule codifies what is known in the federal job discrimination cases as the "continuing violation" doctrine. So under the rule if the alleged discriminatory act is of a "continuing nature," then the act is considered to have occurred as of the last date of the act.

In Annear v. State, 419 N.W.2d 377, 379 (Iowa 1988) we recognized that Iowa Code chapter 601A is patterned after Title VII. We also noted that a similar 180-day limitation is found in that statute at 42 U.S.C. § 2000e-5(e). Because of these similarities we found federal cases on the "continuing violation" doctrine instructive. Annear, 419 N.W.2d at 379.

In general the "continuing violation" doctrine does not excuse compliance with the time limits for filing a charge. But if a violation is continuing, the time does not begin to run when the discrimination first happens. Instead the action is considered filed in time if there are discriminatory acts within the limitations period. 45A Am.Jur.2d Job Discrimination § 1231 at 1024 (1986).

Because the rationale for the doctrine is "to provide a remedy for past actions that operate to discriminate" now, the alleged discrimination must be ongoing. Such discrimination may not be "merely the consequence of a now time-barred event." Id.

The federal courts recognize two types of continuing violations. The first type consists of "a series of acts with one independent discriminatory act occurring within the charge-filing period." Schlei Grossman, Employment Discrimination Law 1047 (1983), accord Milton v. Weinberger, 645 F.2d 1070, 1075 (D.D.C. 1981); 45A Am.Jur.2d § 1231 at 1025. The second type is directed toward the "maintenance of a system or policy which discriminates." Schlei Grossman, 1050; accord Milton, 645 F.2d at 1074; 45A Am.Jur.2d § 1231 at 1025.

Four elements make up the "series of acts" type of continuing violation. These include the following: (1) the alleged discrimination pervades the series of events, (2) there is a present violation of the statute, (3) the present acts of alleged discrimination are related to the time-barred events, and (4) the charge covering the present violation is filed within the limitations period. 45A Am.Jur.2d 1232, p. 1026. In Milton, the court made it clear that the first and third elements refer to the same thing: the acts in the series must be shown to be related, and not isolated. Milton, 645 F.2d at 1076.

Another federal case delineates several factors to help courts determine

when alleged discriminatory acts are related closely enough to constitute a continuing violation and when they are merely discrete, isolated, and completed acts which must be regarded as individual violations.

Berry v. Board of Supervisors of L.S.U., 715 F.2d 971, 981 (5th Cir. 1983). In other words these factors help a fact finder to determine whether elements one and three — the pervasiveness of the alleged discrimination and the relation of present acts to time-barred events — have been established.

The factors include subject matter, frequency, and permanence. Id. As to subject matter, the relevant question is whether "the alleged acts involve the same type of discrimination tending to connect them in a continuing violation." Id. Frequency deals with the question whether the alleged acts are of a recurring nature or more in the nature of an isolated employment decision. Id.

The last factor — permanence — deals with the question whether an employee should or should not realize the employer is discriminating. Id. For example, the discriminatory acts may be so persistent and longstanding that the employee should realize the employer is discriminating. Id. At that point the statute of limitations is triggered. The limitations period, however, is not triggered when the consequences of the discriminatory acts is something the employee might reasonably expect without suspecting discrimination. Id. For example, an employee would probably not suspect that the employer is discriminating when the employer's reasons for the acts are pretextual and seemingly legitimate.

As we said, the commission applied the "series of acts" continuing violation theory to Blood's charge of national origin discrimination and determined that her claim was not time-barred. In making this determination the commission made a number of findings, applying the four elements of the theory in light of the three factors on pervasiveness and relatedness set out in Berry.

The commission found that Hy-Vee's assignment of working hours involved the same type of recurring discrimination: a consistent, repetitive practice of denying Blood's requests for more hours while increasing the hours of similarly situated non-Vietnamese employees. It made a similar finding with regard to promotions. Finally, it found that these acts continued into the 180-day limitation of Iowa Code section 601A.15(12).

The commission further found that there was no degree of permanency evident in any one denial of Blood's requests for more working hours or for promotion. As to the requests for more working hours, the commission focused on two facts. First, Sapp repeatedly told Blood that no additional hours were available. Second, Sapp did give her two additional working hours. Because of these circumstances, the commission concluded that Blood "would not know for some period of time that the denials were either discriminatory or permanent."

As to the requests for promotion, the commission relied in part on the recognition in some federal cases that a discriminatory failure to promote represents an actionable continuing violation. See, e.g., Trevino v. Celanese Corp., 701 F.2d 397, 402 (5th Cir. 1983); Rich v. Martin Marietta Corp., 522 F.2d 333, 348 (10th Cir. 1975) (failure to promote, unlike hiring or discharge or other events often found as not constituting continuing discrimination, does not take place on a particular day, "it invariably arises during a lengthy period of time."); see also Annear, 419 N.W.2d at 379 ("Failure to promote is viewed as ongoing, while failure to hire is viewed as an [isolated] act").

In addition, the commission believed that Blood, a capable and ambitious employee, may not have fully realized that Hy-Vee was engaging in a consistent discriminatory practice not to promote her until Zahari's two promotions. The first occurred in June 1983 and the second, the following December. Consistent with the last fact is the circumstance of the previous complaint that resulted in more working hours for Blood. The settlement, the commission found, could reasonably lead Blood to expect that Hy-Vee would not continue to discriminate against her in the future.

As to the sex discrimination charge, the commission applied the continuing violation doctrine based on the theory that Hy-Vee maintained a discriminatory policy or system into the limitations period. Under this theory of the continued violation doctrine,

a systematic policy of discrimination is actionable even if some or all of the events evidencing its inception occurred prior to the limitations period. The reason is that the continuing system of discrimination operates against employees and violates his or her rights right up to a point in time that falls within the applicable limitations period. Such continuing violations are most likely to occur in the matter of placements or promotions. A minority employee who is not promoted in 1973, for example, and is subject to a continuing policy against promotion of minorities, may file a timely charge in 1976, because the policy against promoting him or her continued to violate the employee's rights right up to the time the charge was filed.

Williams v. Owens-Illinois, Inc., 665 F.2d 918, 924 (9th Cir. 1982); accord Higgins v. Oklahoma ex rel. Oklahoma Employment Sec. Comm'n, 642 F.2d 1199, 1200 n. 2 (10th Cir. 1981); Clark v. Olinkraft Inc., 556 F.2d 1219, 1221-22 (5th Cir. 1977), cert. denied, 434 U.S. 1069, 98 S.Ct. 1251, 55 L.Ed.2d 772 (1978); Rich v. Martin Marietta Corp., 522 F.2d 333, 348 n. 15 (10 Cir. 1975); Wetzel v. Liberty Mut. Ins. Co., 508 F.2d 239, 246 (3rd Cir. 1975), cert. denied, 421 U.S. 1011, 95 S.Ct. 2415, 44 L.Ed.2d 679 (1975); Schlei Grossman, at 1056.

Under this theory, the attack is upon the system rather than just its application to the complainant personally. So an employee may challenge the discriminatory policy even though the employee was not denied a particular job or employment benefit within the limitations period. Reed v. Lockheed Aircraft Corp., 613 F.2d 757, 759-61 (9th Cir. 1980); accord Roberts v. North Am. Rockwell Corp., 650 F.2d 823, 828 (6th Cir. 1981) (hiring system would not consider women); Rich, 522 F.2d at 348 ("Plaintiffs here challenge the entire system maintaining that it continually operated so as to hold them in lower echelons. Hence, the ninety day period prior to the filing of the EEOC charge looms inconsequential in this kind of case."); Wetzel, 508 F.2d at 246 (practices of assigning women to segregated jobs from which promotional opportunities were limited constituted "continuing violations of Title VII and would allow filing a charge at anytime by a present employee."); Stallings v. Container Corp. of America, 75 F.R.D. 511 (D.Del. 1977) (plaintiff may attack promotion practices even if he cannot show he was passed over during statutory period); Schlei Grossman, 1050.

The commission noted that Blood's complaint alleged that Hy-Vee had a

continuing practice to set aside stocker positions for males while simultaneously enforcing a policy requiring stocking experience for higher level positions in the line of progression.

The commission found this was true. It also found that the promotion practice constituted a system or policy of sex discrimination that continued until at least seven days before Blood filed her complaint.

We, like the district court, think the commission applied the correct legal analysis under section 601A.15(12) and that its findings of fact are supported by substantial evidence.

E. The commission's violation of its oral argument rules. Hy-Vee complains about the commission's failure to abide by its own rules as to oral argument on review of a hearing officer's recommended decision. We doubt whether Hy-Vee properly preserved error on this issue for two reasons. First, Hy-Vee failed to make a record on the issue before the commission. Second, Hy-Vee failed to raise the issue in its judicial review petition.

Nevertheless the district court did consider the issue. The court found that the commission had indeed violated its rules because it did not notify Hy-Vee's counsel that it would hear arguments. The court made this finding based on statements of counsel at the hearing on the judicial review petition.

But the court concluded that the error was harmless because Hy-Vee's counsel had notice from opposing counsel one day before the arguments and chose not to appear. The court determined Hy-Vee was entitled to no relief on this issue because it had not established prejudice of substantial rights as required by Iowa Code section 17A.19(8). We agree and so we choose not to disturb the district court's resolution of the issue.

We think, however, that we should impress upon the commission the importance of abiding by its rules, especially a rule pertaining to an important procedure such as oral argument.

III. Issues on the Cross-Appeal.

In the cross-appeal we must decide whether there was substantial evidence to support an award of back pay with full-time status to 1977, additional hours, front pay, and back pay at an increased rate to 1983.

A. Back pay at full-time status to 1977. The commission ordered Hy-Vee to pay Blood the difference between what Blood actually received beginning December 1, 1977, and what she would have earned as a full-time employee. Hy-Vee challenged this award in the district court on the ground that the award was not supported by substantial evidence.

The district court did not address this back-pay issue, thinking — incorrectly — that the commission had rejected the hearing officer's proposed award. We remanded the back pay issue along with the statute of limitations issue. On remand the district court concluded there was substantial evidence to support the award of back pay at full-time status to 1977.

Iowa Code section 601A.15(8) gives the commission authority "to take the necessary remedial action as in the judgment of the commission will carry out the purposes of chapter 601A." One type of remedial relief the commission may award includes "upgrading of employees with or without pay." Iowa Code § 601A.15(8)(a)(1). We viewed a similar provision under an older version of chapter 601A as "invest[ing] the commission with considerable discretion in fashioning an appropriate remedy." Foods, Inc. v. Iowa Civil Rights Comm'n, 318 N.W.2d 162, 171 (Iowa 1982). We think section 601A.15(8)(a)(1) gives the commission authority to award back pay at full-time status when warranted by the evidence.

In computing awards in discrimination cases federal courts follow two basic principles. First, an unrealistic exactitude is not required. Pettway v. American Cast Iron Pipe Co., 494 F.2d 211, 260-61 (5th Cir. 1974); 45B. Am.Jur.2d Job Discrimination § 2395 at 794 (1986). Second, uncertainties in determining what an employee would have earned before the discrimination should be resolved against the employer. Id. One federal court characterized these principles this way:

While the mechanics of computing back pay are difficult and alternative figures might have been used by the trial court in fashioning a remedy, whatever difficulty of ascertainment exists was due to the discriminatory wage structure maintained by the defendant. In such cases, it suffices for the trial court to determine the amount of back wages as a matter of just and reasonable inferences. Difficulty of ascertainment is no longer confused with right of recovery.

Brennan v. City Stores, Inc., 479 F.2d 235, 242 (5th Cir. 1973).

These principles are consistent with the ultimate goal of compensating an injured person: to place the injured person in the position the person would have been in had there been no injury. Foods, Inc., 318 N.W.2d at 171. They are also consistent with our obligation to construe chapter 601A liberally to effectuate its purpose. Chauffeurs, 394 N.W.2d 375, 382 (Iowa 1986).

In concluding the back pay award was supported by substantial evidence, the district court noted the following:

There is record evidence demonstrating that in the ten Cedar Rapids Hy-Vee stores in 1980, only 32 of 309 [10.35%] women employees had full-time status while 185 of 601 [30.78%] men employees worked full-time. In 1985, 41 of 473 [ 8.6%] women employees were full-time employees while 137 of 655 [ 26.4%] men had full-time status. There are no statistics for 1977 in the record. But there is evidence demonstrating that a white male employee hired five months before Blood in 1977 became full-time in July 1977, while Blood, despite repeated requests for more hours never did achieve full-time status. While the court might draw a contrary conclusion, the record evidence is sufficient to support a finding that Blood would have achieved full-time status in 1977 but for the sex discrimination practices of Hy-Vee. The evidence presents a judgment call which must be left to the sound discretion of the commission.

We agree with this assessment.

Moreover, we agree with the commission that the disparate treatment in full-time hours is magnified by the clear-cut evidence of Hy-Vee's discriminatory policy against promoting women. As applied to Blood this discrimination had a direct effect on her ability to achieve full-time status.

Federal courts in job discrimination cases commonly calculate awards on the experience of comparable employees outside of the claimant's protected class. See, e.g., Stallworth v. Shuler, 777 F.2d 1431, 1434-35 (11th Cir. 1985) (failure to promote); Robinson v. City of Fairfield, 750 F.2d 1507, 1512 (11th Cir. 1985) (failure to promote). Here the commission decided that had Blood been treated as well as the man who was hired four months before she was, she would have achieved full-time status in 1977. The male employed achieved full-time status four months after he was hired. The commission calculated Blood's back pay accordingly.

As we said, Iowa Code section 601A.15(8) gives the commission considerable discretion in fashioning an appropriate remedy that will accomplish the purposes of chapter 601A. So the district court saw the award of back pay with full-time status to 1977 as a judgment call within the province of the commission and not the court. It was, as our cases recognize. See, e.g., Mercy Health Center v. State Health Facilities Council, 360 N.W.2d 808, 809 (Iowa 1985).

B. Additional hours. The commission ordered Hy-Vee to offer Blood the first aisle coordinator job available in the greater Cedar Rapids area. In addition, the commission ordered Hy-Vee to give Blood additional hours needed to make her a full-time employee.

The district court believed there was substantial evidence to support the award as to the promotion but not as to additional hours. The court then affirmed the award as to the promotion but vacated the award as to additional hours.

As we said, on remand the district court concluded there was substantial evidence to support an award of back pay with full-time status. The court's conclusion as to the full-time status award under the back pay issue is inconsistent with its earlier conclusion that there was no substantial evidence to support an award of additional hours. We think the same evidence the district court relied on to affirm the award of back pay with full-time status supports the award of additional hours. So we reverse that portion of the district court judgment that vacates the award of additional hours.

C. Front pay, and back pay to June 1983 at the increased rate. The commission ordered Hy-Vee to pay Blood at the same rate of pay an aisle coordinator receives from the date of the commission's decision until she receives the first available aisle coordinator position. Such an award is commonly referred to in job discrimination cases as "front pay." Schlei Grossman, at 1434-36.

In addition, the commission ordered the increased rate of pay to start June 1, 1983 — about the time Zahari was promoted to aisle coordinator. This latter award is a form of back pay at an increased rate.

We think the same principles that we applied to the back pay award to 1977 apply here. Cf., e.g., Stallworth, 777 F.2d at 1431 (failure to promote — defendant ordered to pay claimant at highest paid administrative position); Grimes v. Athens Newspaper, Inc., 604 F. Supp. 1166, 1167 (M.D.Ga. 1985) (female newspaper copy editors who performed equal work but were paid less than men sportswriters were entitled to back pay measured by highest paid male sportswriter). Here, Wickman — the individual who was hired four months before Blood — was earning more as an aisle coordinator than Blood as a checker. We think the evidence supports the two awards — front pay to June 1983, and back pay to the same date at the increased rate. We reverse the order of the district court that vacated these awards.

IV. Disposition.

We agree with the district court that there was substantial evidence to support the commission's findings of discrimination based on sex and national origin. We also agree there was substantial evidence to support the commission's findings that Blood was entitled to damages for emotional distress, that she filed her complaint in time, and that she was entitled to back pay with full-time status to 1977. So we affirm the district court's judgment on these issues.

The district court, however, incorrectly determined that there was no substantial evidence to support the commission's award of additional hours, front pay, and back pay at an increased rate. So we reverse the district court's judgment vacating these awards.

AFFIRMED ON THE APPEAL; AFFIRMED IN PART AND REVERSED IN PART ON THE CROSS-APPEAL.