(concurring in part and dissenting in part). I agree with Justice Levin that the scope of review to be used by the circuit court in reviewing the decision of the Civil Rights Commission is review de novo as expressly provided both in the constitution1 and in the statute.21 write separately to state that I cannot agree that, in this instance, the circuit court reviewed the Civil Rights Commission decision de novo. In that respect I agree with Justice Riley.
i
The full factual background is uniquely relevant to the claims of emotional distress, mental anguish, and humiliation. Mary Johnson began3 working as a waitress for the owners of the Silver Dollar Cafe, defendants Paul and Margaret Anson, in July 1983. Almost from the beginning, Ms. Johnson became the object of crude, explicit sexual remarks made by Mr. Anson.4 On one occasion, *119Mr. Anson "grabbed [Ms. Johnson’s] buttocks when she bend [sic] over to pick up her bill pad.” Ms. Johnson became upset and, in response, grabbed "a knife out of the drawer and threatened to use it if Paul Anson touched her again.”
Mr. Anson’s conduct "often reduced [Ms.] Johnson to tears.”5 She would become visibly upset and extremely nervous when Mr. Anson was present.6 Mr. Anson would on occasion wait in the cafe’s parking lot at closing time, and because of this, Ms. Johnson often asked patrons to stay with her until she closed the cafe.7
*120Mr. Anson made many of these remarks and performed many of these acts in his wife’s presence. While Mrs. Anson had assured Ms. Johnson that she would talk to her husband and promised that his conduct would stop, it did not. In fact, Mrs. Anson eventually told Ms. Johnson that "she had to wait on Paul Anson and be cordial toward him.” However, when Ms. Johnson threatened to quit, Mrs. Anson "begged” her "not to quit because she was needed.” In October 1984, Ms. Johnson was fired ostensibly for underreporting her tips. The Civil Rights Commission found this to be pretextual.
ii
"No area of administrative law has been subjected to more attention yielding less clarity than the scope of review of administrative action.”8 Scope of review refers to "how much a court will undertake to substitute its judgment for that of an agency which has made a decision.”9 To determine "how much” of its judgment a court may substitute for that of an agency, this Court must look to the underlying statute that created the agency.10
The Civil Rights Commission was created by Michigan voters when they approved the revised constitution in April 1963.11 Article 5, § 29 pro*121vides: "There is hereby established a civil rights commission . . . The constitution also specifies "how much” review decisions of the commission will receive from circuit courts: "Appeals from final orders of the commission . . . shall be tried de novo before the circuit court having jurisdiction provided by law.”12 Review of Civil Rights Commission decisions is also provided by statute. Under § 606 of the Civil Rights Act, "An appeal before the circuit court shall be reviewed de novo.”13 Thus, both the statute and the constitution provide for de novo review of Civil Rights Commission decisions.
This Court, in Walker v Wolverine Fabricating & Mfg Co, Inc, 425 Mich 586; 391 NW2d 296 (1986), interpreted the phrase "[ajppeals . . . tried de novo” in the constitution to require a review of the record before the Civil Rights Commission. Litigants were not entitled to an entirely new evidentiary hearing on appeal in the circuit court. We held that
the underlying intent of the framers would best be effected by requiring circuit courts to review final decisions of the crc de novo by taking a fresh look at the evidence and testimony in the record produced before the agency, and by determining whether the crc’s factual findings and legal conclusions are supported by competent, material, and substantial evidence. [Id. at 590.]
The parties in the instant case do not address *122the tension that exists between the statutory "review de novo” and the constitutional "tried de novo.” The portion of our holding in Walker that requires the court to review only the agency record without taking new evidence is therefore not called into question.
I am persuaded, however, that, despite my agreement with the Walker majority, our statement that the court must determine whether the crc’s factual findings are supported by competent, material, and substantial evidence was not necessary to our decision,14 and thus was obiter dictum.15 The exact scope of review, or just "how much” of the reviewing court’s judgment could be substituted for that of the agency, was not at issue in Walker. It is, however, the question that we must answer today.
"In the context of appellate review, the term [de novo] suggests a new look at a case, starting from scratch, without deference to what has gone before. ... As one might expect, the de novo standard of review has never applied this literally.”16 In Beason v Beason, 435 Mich 791, 799-800; *123460 NW2d 207 (1990), we noted that "[v]ery early, this Court recognized the superior position of the trial court in making factual determinations and, consequently, limited review of such determinations . . . ”17
4. "We review the findings of a court of equity de novo. We will
*124Indeed, in Kar v Hogan, 399 Mich 529, 553, n 8; 251 NW2d 77 (1976), Justice Levin wrote, in dissent:
Formerly in chancery cases, although it was commonly said that the issues of fact were tried de novo on appeal or that the Supreme Court must weigh the evidence and reach an independent conclusion on review of the facts, this did not mean that the findings of the trial judge were entitled to no consideration. Notwithstanding the right and duty of the Supreme Court to make its independent evaluation of the evidence, it would not set aside the findings of the trial judge unless it was convinced that a clear showing of error had been made. Running through the opinions are phrases such as "clear error,” "manifest error,” "palpably erroneous.”. . . Especially when there was a sharp conflict in the evidence, the reviewing court would not disturb the trial court’s determination of fact questions, unless it was clear that a wrong conclusion had been reached. The Supreme Court was most reluctant to disturb the findings of a trial judge based on credibility, since the trial judge, as the trier of facts, had the advantage of observing the witnesses. [Quoting 2 Honigman & Hawkins, Michigan Court Rules Annotated (2d ed), Authors’ Comments accompanying Rule 517, p 596.]
More recently, in Sparks v Sparks, 440 Mich *125141, 147; 485 NW2d 893 (1992), we again discussed, albeit also in a different context, the meaning of de novo review. Chief Justice Cavanagh noted that because the trial court is in a better position to judge the credibility of the witnesses, the trial judge’s findings of fact are entitled to considerable weight.
While there does not appear to be a consensus regarding the exact meaning of the term de novo, one thing is clear: This Court has consistently recognized the superior position of the trial court in making findings of fact. In fact, one early article discussing the newly created Civil Rights Commission recognized the superior position of the factfinder in cases brought before the Civil Rights Commission. In discussing the potential interpretation of the term "de novo” in the constitution, he noted that the reviewing court would be required to
exercise its independent judgment as to questions of fact, as well as law, on the record complied before the Civil Rights Commission. The difficulty with this view is that findings based on the credibility of witnesses (which are very important in the resolution of discrimination cases) are made by a trier who has neither seen nor heard the witnesses.[18]
Thus, while I agree with Justice Levin that "[t]o give meaning to the term 'de novo,’ we must hold that a circuit court, in reviewing a decision of the Civil Rights Commission, may substitute its assessment for the findings, conclusion, and decision of the Civil Rights Commission,” ante at 116,1 do not agree that the circuit court’s review is one of unfettered substitution of judgment. The reviewing *126court must recognize the superior position of the initial factfinder in resolving issues of credibility.
I also cannot agree that the circuit court reviewed this matter de novo, as Justice Levin and the Court of Appeals suggest. As Justice Riley notes in her opinion, the circuit court applied the substantial evidence test to the record.19 Indeed, the circuit court’s order supports this conclusion. In it the court expressly stated that the Civil Rights Commission’s findings of fact either were supported or not supported by the record. Moreover, the circuit court on review made no independent findings of fact. Thus, I would remand the case to the Court of Appeals for remand to the circuit court for review de novo of the findings and conclusions of the Civil Rights Commission, keeping in mind the superior position of the factfinder in matters of credibility.
in
The scope of the Court of Appeals review of the circuit court’s decision presents "[t]he question . . . whether each succeeding reviewing court should apply the same standard of review to the agency fact-finding or should instead limit their review to the decisions of the previous court.”20
Because the parties do not dispute that we must apply the clearly erroneous standard set forth in Dixon v Ford Motor Co, 402 Mich 315; 262 NW2d 666 (1978), this is not the appropriate case to reexamine the clear error standard of appellate review.21 Both Justice Levin and Justice Riley adopt that standard, and I agree that reexamination of the question should await another day.
*127It is worth observing, however, that concerns regarding deference to agency expertise are raised not only by the concept of de novo review of the commission by the circuit court, but also by application of the clearly erroneous standard for appellate review articulated in Dixon.
Professor LeDuc differentiates between the substantial evidence standard of review and the clearly erroneous standard and concludes that
the clearly erroneous test gives a reviewing court greater flexibility and power than does the substantial evidence test. A decision could meet the substantial evidence test and be reasonable and still be overturned on review under the clearly erroneous test. ... If the appropriate standard is the substantial evidence test, a reviewing court cannot set aside a reasonable, but mistaken finding of fact. If the appropriate standard is the clearly erroneous standard, it can.[22]
Dixon assumed, without expressly deciding, that the Court of Appeals would review the circuit court and not the Civil Rights Commission, the method of review of the decision of the inferior tribunal that comports with the roles of the reviewing courts in the federal system.23 In Universal Camera Corp v Nat’l Labor Relations Bd, 340 US 474; 71 S Ct 456; 95 L Ed 456 (1951), the Court characterized its own role in the review of agency findings of fact as follows:
Our power to review the correctness of application of the present standard ought seldom to be called into action. . . . This Court will intervene only in what ought to be the rare instance when *128the standard appears to have been misapprehended or grossly misapplied. [Id. at 490-491.]
The effect of the Court’s decision in Universal Camera is that
the initial court of review operates as the primary locus of judicial review of agency fact-finding and bears the responsibility of applying the appropriate scope of review to such fact-finding. Each subsequent reviewing court reviews the findings of the earlier reviewing court, rather than the findings of the agency. This defines the function of the first reviewing court as the primary fact-finding check on agencies and represents a much better use of judicial resources.[24]
IV
In my view, the circuit court applied an incorrect legal standard when it reduced the compensatory damages award of the Civil Rights Commission.
The purpose of the Civil Rights Act is "to eradicate discrimination and to make persons whole for injuries suffered as a result of discrimination.”25 As one treatise notes, "one can easily imagine that some of the most damaging aspects of discrimination affect the person suffering the discrimination rather than his or her pocketbook. . . . [M]any victims of such discrimination describe the occurrence as the singularly most humiliating or enrag*129ing event in their life.”26 Damages for humiliation, outrage, and embarrassment are allowed under the Civil Rights Act to attempt to make persons whole for the injuries suffered.27
We have held when reviewing a damages award *130for pain and suffering "that there is no absolute standard by which to measure” awards for pain and suffering,28 and that "awards for personal injury should rest within the sound judgment of the trier of fact, particularly awards for pain and suffering . . . .”29 We have also recognized that "since no trier of fact can value pain and suffering with mathematical certainty, a reviewing court must offer 'something more tangible than a difference of opinion as to amount’ before it sets aside a non-jury award as clearly erroneous.”30
In Precopio v Detroit, 415 Mich 457; 330 NW2d 802 (1982), we cautioned against using past awards for analogous injuries as the guidepost for determining whether damages were excessive. Analogous cases are only one factor in this decision. Other factors include the evidence presented at the hearing on the plaintiff’s suffering. However, we stated that "if research uncovers a sufficient sample of reviewed awards, comparisons with analogous cases may prove of some value.”31 Thus, if awards are analogous, and if research uncovers a sufficient sample, a reviewing court may rely on those amounts._
*131In the instant case, however, the circuit court compared damages for the mental anguish caused by sexual harassment to damages for wrongful death and injury resulting from automobile accidents. In my view, three inapposite cases were not a sufficient sample from which to compare awards.32
CONCLUSION
I would remand the case to the Court of Appeals for remand to the circuit court for further review consistent with our decision today.
Mallett, J., concurred with Boyle, J.Const 1963, art 5, § 29.
MCL 37.2606; MSA 3.548(606).
The Ansons had once before asked Ms. Johnson to work for them in their restaurant. Because she had previously been the object of Mr. Anson’s sexual advances, she declined their offer. The Ansons were "anxious for [Ms. Johnson] to work for them because she was a jolly person, had a good personality, smiled all the time, got along with people and was honest.”
Although the conduct that Ms. Johnson was subjected to is vulgar, I find it necessary to set out the findings of the Civil Rights Commission in that regard. The Civil Rights Commission found:
15. Respondent Paul Anson told Claimant he would like to suck her breasts until her nipples stuck out.
*11916. Respondent Paul Anson told Claimant he couldn’t understand why she had such big breasts and skinny legs.
17. Respondent Paul Anson told the Claimant: "You have a love button. I would love to eat that.”
18. Respondent Paul Anson told the Claimant he would like to "eat” her until she came all over his face.
19. Respondent Paul Anson told Claimant that her husband, a Vietnam veteran who had been wounded five times and had been in a serious truck accident, was not a man and that if the Respondent "had” her, she would leave her husband.
20. Respondent Paul Anson told Claimant she did not have to do her job; she could do "another” job if she did him sexual favors.
21. Respondent Paul Anson asked the Claimant to go to bed with him.
22. On different occasions, Respondent Paul Anson offered Claimant one hundred dollars for ten minutes of her time for sexual favors.
23. Respondent Paul Anson would throw Claimant’s bill pad on the floor and, when she bent to pick it up, tell her how "good looking” her "ass” was.
24. Respondent Paul Anson would stare at the Claimant as though he was undressing her.
25. Respondent Paul Anson would make motions with his mouth to the Claimant such as licking his lips and puckering up.
26. Respondent Paul Anson would approach the Claimant in a narrow aisle where she washed dishes and rub the front of his body against her.
188 Mich App 147, 151; 469 NW2d 42 (1991).
Id.
Id.
LeDuc, Michigan Administrative Law, § 9:01, p 3).
Id., p 2
It is not only the statute that created the agency that defines the scope of review, but also the constitution and related statutes that may define its scope. LeDuc, n 8 supra, p 3. In fact, in a report to the Michigan Legislature, the Michigan Law Revision Commission identified nearly two hundred statutes dealing with judicial review of administrative decisions. See Michigan Law Revision Commission, 25th Annual Report 1990, p 21 and attachments 1-5, pp 47-140.
Cramton, The powers of the Michigan Civil Rights Commission, 63 Mich LR 5 (1964).
Another section of the Michigan Constitution "provides for extensive judicial review of decisions of 'any administrative officer or agency existing under the constitution or by law.’ ” Cramton, n 11 supra at 20, quoting Const 1963, art 6, § 28. Professor Cramton, however, notes that "this provision is largely displaced insofar as the Civil Rights Commission is concerned by the more speciñc judicial review provisions of the third paragraph of article V, section 29 . . . .” Id. (Emphasis added.)
MCL 37.2606(1); MSA 3.548(606)(1).
As Professor LeDuc points out, "The holding also exceeds the issue presented, since the challenged action was not one based on a hearing and is therefor one to which the substantial evidence test has no application.” LeDuc, n 8 supra, § 9:43, p 56. Professor LeDuc is correct. Article VI, § 28 of the constitution expressly states that the "competent, material and substantial evidence on the whole record” standard applies "in cases in which a hearing is required[.]” Other agency actions — final decisions, findings, rulings, and orders— are reviewed to determine whether they were authorized by law.
Professor LeDuc also criticizes the Walker decision as failing "to recognize the basic difference between . . . de novo judicial review and the more limited nature of the substantial evidence scope of judicial review . . . .” LeDuc, n 8 supra, § 9:43, p 56. Moreover, "the entire opinion is devoid of any basis whatsoever for determining that the use of the words de novo meant the same as the provision for substantial evidence in Article VI, Section 28.” Id., p 56. Professor LeDuc continues: "The Walker decision is the Constitution as the majority would have written it, not the one which was written.” Id., p 56.
Powell & McAlpine, Standards of review in Michigan, 70 Mich B J 28, 29-30 (1991).
*123Powell & McAlpine recognize several formulations of the de novo standard. For instance:
1. "In equity cases, this Court’s review of the record is de novo with due deference being given to the findings of the trial court. This Court will sustain those findings unless its ruling would have been contrary to that of the trial court.” [Marconeri v Village of Mancelona, 124 Mich App 286, 287-288; 335 NW2d 21 (1983).]
2. "A suit to quiet title is equitable in nature and subject to de novo review. However, this Court will give great weight to the findings of fact made by the trial court and will not disturb those findings unless convinced that it would have reached a different result had it been in the position of the lower court.” [Connelly v Buckingham, 136 Mich App 462, 467; 357 NW2d 70 (1984).]
3. "Preliminarily, we note that our review of an equitable accounting is de novo, although the lower court’s findings will not be disturbed unless the evidence clearly preponderates in the other direction.” [Nogueras v Maisel & Associates of Michigan, 142 Mich App 71, 81; 369 NW2d 492 (1985).]
sustain the findings unless convinced that had we heard the evidence in the first instance we would have been compelled to reach a contrary result.” [Orion Charter Twp v Burnac Corp, 171 Mich App 450, 459; 431 NW2d 225 (1988).]
We also recognized this anomaly in Beason v Beason, 435 Mich 791, 800; 460 NW2d 207 (1990).
It has been variously stated that the appellate court "generally does not reverse or modify unless convinced that it would have had to reach another result had it occupied the position of the trial court,” or "unless convinced that we would have reached a different conclusion had we occupied the situation of the trial court,” "that we should have reached a different conclusion had we occupied the position of the trial court,” or "unless it is clear the reviewing court could have reached a different result had it occupied the position of the trial judge.” [Citations omitted.]
We also noted:
*124"There is much to believe and much to disbelieve, and the case is one of those where the decision of the judge exercising primary jurisdiction on matters of fact ought not to be overruled by a court of appeal except upon clear and satisfactory grounds. The appellate tribunal ought to be fully persuaded that it must have reached a different conclusion had it occupied the position of the court appealed from and been favored with all the advantages of that court for judging rightly.” [Beason, supra at 800, quoting Nicholas v Nicholas, 50 Mich 162; 15 NW 64 (1883).]
Moreover, "It is also said that in a chancery case the decree of the trial court will not be reversed unless the record fails to sustain the decree.” Beason at 801.
Cramton, n 11 supra at 24.
Post at 135-138.
LeDuc, n 8 supra, § 9:49, p 68.
See id., § 9:48, pp 64-67.
Id., pp 66-67.
See id., § 9:49, p 68. LeDuc suggests that "[t]he setting in Michigan is ripe for application of the clearly erroneous test as applied [by] the United States Supreme Court . . . .”
Id.
Dep’t of Civil Rights ex rel Cornell v Sparrow Hosp Ass’n, 423 Mich 548, 564; 377 NW2d 755 (1985).
1 Michigan Law of Damages (2d ed), Damages Recoverable in Civil Rights Cases, § 9.9, pp 9-7 to 9-8. (Emphasis added.)
Some professionals believe that sexual harassment can be worse than rape for some victims, since the dynamics include repetitive reinforcement of degradation and a sense of powerlessness, sometimes on a daily basis for many years. "It’s gradual,” comments one professional, "if you’re raped, it’s an immediate and intense intrusion on your being . . . [sexual harassment] erodes away at your basic being as a woman.” [Sepler, Sexual harassment: From protective response to proactive prevention, 11 Hamline J of Pub L & Policy 61, 67 (1990). Citation omitted.]
In addition to the emotional and psychological effect, victims of harassment suffer economic consequences. A woman may leave a job in which she was sexually harassed. "[T]his often means moving from one entry-level position to another, failing to achieve sufficient tenure at a single company for merit increases and promotions . . . .” Id.
Sexual harassment in the workplace affects not only the victim of the harassment but also affects the organization in which the harassment occurred.
[T]he problem of sexual harassment costs a typical Fortune 500 company $6.7 million dollars each year in absenteeism, employee turnover, low morale and low productivity. Add to this the increasing costs of litigation and the growing number of financial awards to victims of such harassment, and sexual harassment becomes a problem of substantial economic proportions. [Id. at 68, citing Sandroff, Sexual harassment in the Fortune 500, Working Women Magazine, December 1988 at 69.]
See, generally, Final Report of the Michigan Supreme Court Task Force on Gender Issues in the Courts; Final Report of the Michigan Supreme Court Task Force on Racial/Ethnic Issues in the Courts.
See Eide v Kelsey-Hayes Co, 431 Mich 26; 427 NW2d 488 (1988). See also Reithmiller v Blue Cross & Blue Shield of Michigan, 151 Mich App 188, 202; 390 NW2d 227 (1986) ("damages for mental and emotional distress are recoverable under the Civil Rights Act”); Slayton v Michigan Host, Inc, 122 Mich App 411, 417; 332 NW2d 498 (1983) ("These types of injuries are the kind that the Elliott-Larsen Civil Rights Act was designed to protect against, and to hold otherwise would undercut the legislative scheme to remedy discriminatory wrongs”).
Precopio v Detroit, 415 Mich 457, 464-465; 330 NW2d 802 (1982). But see Howard v Canteen Corp, 192 Mich App 427, 436; 481 NW2d 718 (1991) (Evidence to support a $200,000 award for mental anguish, humiliation, and emotional distress was "found in the harassment and discrimination inflicted upon [the plaintiff! for a lengthy period of time”); Brunson v E & L Transport Co, 177 Mich App 95, 106-107; 441 NW2d 48 (1989) (Evidence to support an $80,000 award for emotional distress was "found in the long history and obvious pattern of gender discrimination practiced upon plaintiff by defendant”). However, the exact length of time involved in each case is unclear.
Precopio, n 28 supra at 464.
Id. at 470-471. While it is true that the Court was reviewing the damages award under a clearly erroneous standard, the de novo standard of review also contemplates some deference to the trier of fact. Again the trier of fact is in the best position to view the witness’ demeanor when describing the mental anguish suffered as a result of the sexual harassment.
Id. at 471.
The circuit court based its decision regarding compensatory damages in part on its experience in Hillsdale County. The court stated:
I would remind both of them [the attorneys representing the parties] that this county is not Wayne County, it’s not Ingham County, it’s not even Jackson County, our neighbor to the north. We’re a small rural county. When we speak of damages in this county we speak of them in much lesser terms of amount than we do in some of the more metropolitan and populace counties.
However, a recent survey of Michigan jury verdicts shows that Hillsdale county jury awards, on average, rank two percent above national jury awards, as do awards in Ingham and Jackson Counties. Wayne County ranks ten percent above the national average. Jury Verdict Research Series (Michigan ed, 1992-93), Personal Injury Verdict Survey, pp 1-2. Thus, the circuit court’s stated frame of reference was incorrect as far as damages "in much lesser terms of amount” than those awarded in Ingham and Jackson Counties.
Moreover, according to the survey, the midpoint verdict, that is the "middle (median) verdict value among awards listed in ascending order,” for emotional distress damages in Michigan was $113,000. Id. at 7-8. The verdict average for emotional distress damages was $323,313. Id. at 9. Both of these amounts are considerably higher than that awarded by the Civil Rights Commission.