EQUAL EMPLOYMENT OPPORTUNITY COMMISSION, Appellee, v. RADIATOR SPECIALTY COMPANY, Appellant

JAMES DICKSON PHILLIPS, Circuit Judge: Defendant Radiator Specialty Company appeals from a judgment entered by the district court in this action brought by the Equal Employment Opportunity Commission 1 alleging violations of Title VII of the Civil Rights Act of 1964, 42 U.S.C. § 2000e et seq. The district court found racially discriminatory certain policies and practices used by Radiator Specialty Company (RSC) in the recruiting, hiring, transferring and promoting of persons into its clerical, sales, professional, and managerial job classifications. On this appeal RSC contends that the Commissioner did not comply with the procedural prerequisites to the institution of this suit and that the statistical evidence presented to the district court is insufficient to establish employment discrimination. Concluding that the evidence must be reexamined in light of critical intervening decisions of the Supreme Court and of this court, principally Hazelwood School District v. United States, 433 U.S. 299, 97 S.Ct. 2736, 53 L.Ed.2d 768 (1977), we reverse in part, vacate in part and remand to the district court for further proceedings. *181 I This suit arose out of a charge filed with the Commission by Willie Westbrook, a former RSC employee. Westbrook initially charged that he had been discharged because of his race and later amended his complaint to include an allegation that RSC maintained racially segregated departments. In January 1975 the Commission brought suit alleging that RSC discriminated against blacks in hiring and promoting into certain job classifications, in discharging employees, and in maintaining segregated departments. The Commission claimed that RSC had restricted blacks to less desirable and lower paying jobs. The Commission’s evidence showed that while, since the effective date of Title VII, the overall percentage of blacks employed by RSC (a Charlotte, North Carolina manufacturing corporation) has exceeded the percentage of blacks in the general population of the Charlotte area, the percentages of blacks employed in RSC’s professional, managerial, clerical,’ and sales positions were far lower than the percentages of blacks in the general population. 2 The evidence further showed that RSC had filled, by outside hiring or by promotion from within, over 100 vacancies in its sales, managerial, professional, and clerical classifications since 1971, but that the representation of blacks among the “new hires” in these classifications had been substantially less than their representation in the general population, and that relatively few of RSC’s black employees had been promoted into the jobs in question. In rebuttal, RSC presented statistical evidence showing the percentages of blacks employed in various job classifications in Mecklenburg County. These figures, based on statistics drawn from the Standard Metropolitan Statistical Area (SMSA) for the year 1970, were then compared with the percentages of blacks in comparable classifications at RSC from 1971 to 1977. In the professional and managerial categories RSC’s percentages of black employees were higher than the SMSA percentages; in the sales category RSC’s percentages were lower; in the clerical category RSC’s percentages were lower than the 1970 SMSA percentage until 1975 and were higher thereafter. Looking to evidence other than the statistical data, the district court found that RSC utilized unpublicized, unwritten, and highly subjective standards and procedures for recruiting, hiring, transferring and promoting persons into its upper level positions. Specifically, RSC was found to rely on “word of mouth” recruiting by employees among their friends and to have no policy of posting notice of vacancies or written descriptions of available job openings. Transfers and promotions to upper level positions were instigated either at RSC’s initiative by selecting a few persons as candidates or by permitting the employee to make an oral request to the employee’s immediate supervisor. Predominantly white managers and Personnel Department employees were given unlimited discretion in the appointment of persons to clerical, sales, professional, and managerial positions. The criteria for appointment to these upper level positions, used since RSC was first organized, were *182 both general and subjective. Various written tests used in certain jobs were found to be unstandardized, unvalidated, and haphazardly administered. Considering both the statistical data and the evidence showing these general “badges of discrimination,” see Brown v. Gaston County Dyeing Machine Co., 457 F.2d 1377, 1382-83 (4th Cir. 1972), the trial court concluded that plaintiff, by its statistical evidence showing gross disparities between the racial composition of RSC’s upper level positions and the racial composition of both the general population and RSC’s own work force, had established a prima facie case of racial discrimination in the recruiting, hiring, transferring and promoting of persons to clerical, sales, professional, and managerial positions. Defendant’s SMSA statistical evidence, offered as a more reliable statistical basis for assessing the impact of defendant’s practices, was rejected out of hand because the “skills and abilities of black persons in Mecklenburg County are not necessarily reflected by their present or most recent employment.” The court further held that defendant had failed to rebut plaintiff’s prima facie case by showing that its policies and practices were justified by any business necessity. Concluding that race was therefore the only identifiable factor explaining the disparity between black and white employees in the upper level positions, the court held that RSC had engaged in unlawful employment practices and ordered extensive injunctive relief. II We first address RSC’s claims that the Commission has failed to comply with the statutory requirements for instituting this suit. The district court found that the Commission had complied with the procedural requisites of Title VII prior to bringing suit. RSC claims that this finding is infirm because of the Commission’s delay and its alleged refusal to engage defendant in conciliation. We are not persuaded by this argument, and affirm the district court’s conclusion of compliance. Westbrook filed his original charge with the Commission on September 23, 1970. The Commission was, at that time, not statutorily required to notify an employer of a charge within any particular time period. 3 Amendments to § 706 of Title VII, effective March 24, 1972, required the Commission to serve an employer with notice of any charge alleging his engagement in unlawful practices within 10 days of the filing of the charge. 4 On April 3, 1972, 10 days after the effective date of the 1972 Amendment, the Commission mailed notice of Westbrook’s 1970 charge to RSC. On March 15, 1973, Westbrook filed an amended charge, and notice of the amended charge was given to RSC. One year later, on March 19, 1974, the Commission issued a determination letter finding reasonable cause. RSC claims that the 17 months between Westbrook’s initial charge and the April 3, 1972, notification was inordinate and unreasonable. RSC asserts that the initial notice incorrectly specified the date of West-brook’s discharge and this error further aggravated the unreasonableness of the delay. Thus, RSC claims it received no hint of the action under investigation until the notice of March 1973 and received no specific information concerning the alleged violation until the March 19, 1974, determination letter. Because of these delays, RSC says it has been denied basic procedural protections necessary for the gathering and preservation of evidence. *183 The Commission had no statutory duty to inform defendant of the initial charge until the 1972 Amendments took effect, Chromcraft Corp. v. EEOC, 465 F.2d 745 (5th Cir. 1972); International Brotherhood of Electrical Workers v. EEOC, 398 F.2d 248 (3d Cir. 1968), 5 and it satisfied this statutory duty by informing defendant of the initial charge 10 days after the amendment took effect. 6 Therefore, defendant’s claim must be one of laches. Mere delay, however, is not sufficient to constitute a defense of laches; a defendant must also show prejudice. See D. Dobbs, Remedies § 2.3, at 43 & n. 19 (1973). A generalized allegation of harm from the passage of time does not amount to a showing of prejudice. EEOC v. South Carolina National Bank, 562 F.2d 329, 332 (4th Cir. 1977). Defendant in this case has shown nothing more. The dismissal of Westbrook’s claim by the trial court and the trial court’s finding that RSC’s discharge policies were not discriminatory demonstrate a virtual absence of prejudice. Defendant further argues that the Commission failed to make a bona fide conciliation effort as required by 42 U.S.C. § 2000e-5(b) and is therefore precluded from bringing suit. The Commission’s statutory duty to attempt conciliation is one of its most essential functions. EEOC v. Raymond Metal Products Co., 530 F.2d 590, 596 (4th Cir. 1976). Attempted conciliation is a condition to the Commission’s power to sue. Patterson v. American Tobacco Co., 535 F.2d 257, 272 (4th Cir. 1976). The law requires, however, no more than a good faith attempt at conciliation. EEOC v. Zia Co., 582 F.2d 527 (10th Cir. 1978). The record discloses that the Commission made such an attempt. The determination letter of March 19, 1974, contained the initial invitation to conciliate and fully informed defendant that there had been a reasonable cause determination with regard to segregated departments and facilities. At RSC’s request, the Director of the EEOC district office toured defendant’s plant and discussed the charges. Three months later, in response to a June 14 inquiry by the Commission suggesting a meeting to discuss the feasibility of an agreement, RSC mailed a letter to the EEOC stating that “[w]e therefore see no worthwhile reason for such a meeting and are convinced that such would not prove beneficial to either party.” On reviewing this letter, the Commission notified defendant that conciliation would be deemed to have failed unless RSC indicated it wished to resume conciliation within five days. Defendant did not respond. This evidence does not support RSC’s claim that the Commission made no attempts to engage in conciliation. Defendant, aware of the charges, was given several opportunities to participate in conciliation discussions; RSC’s failure or refusal to do so cannot be attributed to the Commission. Ill We turn now to RSC’s contention that the finding of discrimination is not supported by the evidence. The trial court’s finding of discrimination was based primarily upon plaintiff’s evidence of statistical discrepancies between the percentages of blacks appointed to defendant’s white collar positions compared with the percentages of blacks in both the area population and defendant’s entire work force. Relying on Teamsters v. United States, 431 U.S. 324, 97 S.Ct. 1843, 52 L.Ed.2d 396 (1977), and Griggs v. Duke Power Co., 401 U.S. 424, 91 S.Ct. 849, 28 L.Ed.2d 158 (1971), the court held this evidence sufficient to establish a prima facie case of discrimination, and ruled that RSC had failed to show that the statistical data adduced by plaintiff did *184 “not reflect the pool of qualified applicants for these white collar positions.” RSA contends that the court’s finding of liability is not supported by the evidence. Specifically it argues that its various clerical sales, professional, and managerial positions require special qualifications for their effective perforemance, and that under the Supreme Court’s decision in Hazelwood School District v. United States, 433 U.S. 299, 97 S.Ct. 2736, 53 L.Ed.2d 768 (1977), and this court’s decision in Roman v. ESB, Inc., 550 F.2d 1343 (4th Cir. 1976) (en banc), plaintiff’s statistical evidence was, as a matter of law, not sufficiently probative to establish the inference that RSC had engaged in unlawful discrimination under Title VIL RSC further contends that its own evidence, comparing the racial composition, position by position, of defendant’s work force with that of the employed work force in Mecklenburg County, is, under Hazel-wood and Roman, a more relevant indication of the level of RSC’s compliance with Title VII because these figures adequately represent the qualified labor market and available labor pool for RSC’s white collar and supervisory positions. In order to resolve this issue, it is necessary to take into account the specific teaching and implications of several recent decisions, most notably Hazelwood, that bear importantly upon it. Taken together, we conclude that these decisions require reversal in part and a remand in part for reconsideration of the evidence, possibly upon a reopened record. In Teamsters the Supreme Court reaffirmed the probative value of general population statistics where the jobs in question do not require special qualifications. The Hazelwood Court noted that truck driving, the skill in question in Teamsters, was one generally possessed or readily acquired, and, therefore, the use of general population statistics was appropriate. 433 U.S. at 308 n. 13, 97 S.Ct. 2736. If plaintiff demonstrates gross statistical disparities, defendant must then demonstrate that plaintiff’s proof is “inaccurate or insignificant.” United States v. Teamsters, 431 U.S. at 360, 97 S.Ct. 1843. Defendant may make this showing by demonstrating that its post-Act hiring has been nondiscriminatory. Id. Hazelwood involved alleged discriminatory hiring of teachers. When special skills are required statistical comparisons between defendant’s work force and the general population will not normally be sufficiently probative to make out a prima facie case. 7 433 U.S. at 308 n. 13, 97 S.Ct. 2736. In such a situation, Hazelwood requires a comparison between the racial composition of the relevant portion of defendant’s work force (after the effective date of Title VII) and the racial composition of qualified workers in the relevant labor market. See id. at 308, 97 S.Ct. 2736. In EEOC v. Chesapeake & Ohio Railway, 577 F.2d 229 (4th Cir. 1978), the Commission relied on a disparity between the racial composition of defendant’s upper and lower positions to establish discrimination in promotions. We held that, because most of the higher positions required special qualifications, a prima facie case was not made out since the Commission failed to show how many lower level employees were qualified for promotion. See id. at 233. Similarly in Hill v. Western Electric Co., 596 F.2d 99 (4th Cir. 1979), use of all hourly employees as the labor pool for promotions was held to be inappropriate when the special qualification of experience was clearly required for the job positions to which promotions were denied. These decisions show the critical importance of establishing the existence or nonexistence of special job qualifications when general population statistics are offered as the base data to prove a prima facie case *185 based on gross racial disparities in employment results. They do not, however, address in specific terms a further procedural question of equal importance in such cases: how, and by whom, this critical point is to be established in proof. Though not given in specific terms, we think the answer may be fairly deduced from the decisions just analyzed and from general principles. In some cases, the fact that no special qualifications exist will be manifest as a matter of law from the mere identification by the plaintiff of the job positions in question, e. g., Teamsters. In such cases, a court may properly look to general population statistics in assessing plaintiff’s prima facie proof. In other cases, the fact that special qualifications do exist will be equally manifest from the same source, e. g., Hazelwood, Hill, and Chesapeake & Ohio. In these cases, a court may not usually find a prima facie case established by use of general population statistics, and a plaintiff may safely rely only upon specially qualified market statistics in attempting to show the requisite disparities. In still other cases, it will not be manifest as a matter of law simply from plaintiff’s identification of job positions whether or not special qualifications exist for them. In such cases, the burden should be upon defendant to establish that the positions in fact do require special qualifications not possessed or readily acquired by the general population, at peril of having the general population statistics presumed appropriate in assessing plaintiff’s prima facie proof. If the defendant succeeds in this the plaintiff should have an opportunity to adjust his statistical proof to reflect a labor pool base with the special qualifications found required. 8 Given the centrality that these principles have now assumed in statistical proof cases, it is apparent that for orderly trials and effective review, trial courts must now specifically address in all cases where general population statistics are sought to be used the threshold issue of special qualifications. Where this is determined as a matter of law from the mere identification of the job positions in issue, a simple conclusion of law to this effect is appropriate. When it is not so manifest, the conclusion that it is not so should be followed by findings of fact determining the matter. When these newly refined principles are applied to the proof and its assessment in the district court, the necessity for reversal in part and remand for reconsideration in part is apparent. In the first place, RSC’s professional positions as identified and described in plaintiff’s proof do so manifestly require qualifications neither corn- *186 monly possessed nor readily acquired by the general population that plaintiff’s general population statistics were not appropriately used to find plaintiff’s prima facie case established as to these. Hill v. Western Electric Co., 596 F.2d at 105-06. The district court’s conclusion of violation as to these positions is accordingly reversed, and any injunctive decree finally entered must be modified accordingly. Defendant asserts that the remaining positions, clerical, managerial, and sales, also require special qualifications. The trial court made no findings with respect to this threshold issue. Unlike the professional positions, we conclude that the questions whether special qualifications exist for these other positions cannot be resolved as a matter of law on the present record and must be addressed upon remand as issues of fact. The court may find it necessary to reopen the record and receive further evidence from both parties regarding the actual character of the jobs in question and the extent of any special qualifications genuinely required. See Pennington v. Lexington School District 2, 578 F.2d 546 (4th Cir. 1978). If the court finds that none of the positions in question require qualifications beyond those commonly possessed or readily acquired by the general population, then the finding of violations in RSC’s hiring practices may be reinstated. If, however, special skills are required, the Commission must show, in order to make out a prima facie case, that a disparity exists between the percentage of qualified blacks in the population and the percentage of blacks holding those positions at RSC since the effective date of Title VII. With respect to the qualified labor market, the trial court apparently assigned no probative value to the Company’s SMSA evidence comparing the percentages of blacks actually employed by RSC in certain job classifications with comparable SMSA percentages. It is true, as the trial court stated, that the SMSA figures do not necessarily indicate the percentage of blacks in the Charlotte-Mecklenburg area who are qualified to perform in RSC’s blue collar supervisory and white collar positions. The SMSA figures do not purport to measure the percentage of qualified persons in the population and may, because they were prepared in 1970, present an outdated picture of actual employment in the area. If special skills are required for any of RSC’s positions, the Commission may not rely upon general population statistics as the basis for showing disparities as to those, and in the absence of more relevant evidence, 9 RSC’s statistics should be considered as at least one indicator of the percentage of qualified blacks in the labor pool. 10 A similar analysis should be used in assessing defendant’s liability for discrimination in promotions. If no special qualifications are required for the upper level positions, defendant’s lower level work force is the appropriate labor pool. If special skills are required for the promotion jobs, the Commission must produce evidence of the number of qualified blacks in lower level positions at RSC. Defendant’s SMSA qualified market data does not indicate an available labor pool in promotion and transfer cases, but such evidence may nonetheless be relevant, as we indicated in Patterson v. American Tobacco Co., 535 F.2d 257 (4th Cir. 1976), in assessing defendant’s compliance with the requirements of Title VII. Accordingly, proper regard must be given to RSC’s qualified labor market statistics in *187 the absence of evidence indicating the actual percentages of blacks in lower positions who possess the required qualifications. REVERSED IN PART; VACATED AND REMANDED FOR FURTHER PROCEEDINGS IN PART. Footnotes [1] Authority to bring this suit arises under § 706(f) of Title VII, 42 U.S.C. § 2000e-5(f)(1). [2] For example, in April 1971 RSC employed 207 white persons and 273 black persons. About 82% of the employed whites were either “white collar” employees or were supervisors or foremen in “blue collar” units. Less than 5% of the blacks employed were white collar, or were supervisors or foremen in blue collar units. For the years following 1971 the approximate figures are: White Employees March 1972 201 March 1973 239 March 1974 229 March 1975 254 Percentage in white collar, or supervisory positions: Black Employees White Black 287 80% less than 5% 298 71% less than 5% 310 75% less than 5% 247 67% 7% [3] Section 706(a), 78 Stat. 259 (relettered § 706(b) by 1972 Amendment), then read in part: Whenever it is charged . . . that an employer . . . has engaged in an unlawful employment practice, the Commission shall furnish such employer . , with a copy of such charge and shall make an investigation of such charge. [4] Section 706(b), 42 U.S.C. § 2000e-5(b), now reads in part: Whenever a charge is filed . alleging that an employer . . . has engaged in an unlawful employment practice, the Commission shall serve a notice of the charge (including the date, place and circumstances of the alleged unlawful employment practice) on such employer . . within ten days . [5] The Commission’s former practice of delaying service of a charge until investigation began was designed to protect a charging party from retaliatory action and to prevent deliberate destruction of records during the intervening period. See Chromcraft Corp. v. EEOC, 465 F.2d 745 (5th Cir. 1972); B. Schlei & P. Grossman, Employment Discrimination Law 774 (1976). [6] Defendant’s claim that the April 3, 1972, notice incorrectly stated the date of the alleged violation is not an error that can be considered jurisdictional. See EEOC v. Airguide Corp., 539 F.2d 1038 (5th Cir. 1976). [7] The Hazelwood court did not entirely rule out the use of general population comparisons in special qualification cases: “When special qualifications are required to fill particular jobs, comparisons to the general population . may have little probative value.” 433 U.S. at 308 n. 13, 97 S.Ct. at 2742 (emphasis added). When the statistical disparities are “gross,” it may be unnecessary to “fine tune” the statistics. See United States v. Teamsters, 431 U.S. at 342 n. 23, 97 S.Ct. 1843. [8] Requiring the defendant to show the inappropriateness of general population statistics in such situations follows the general principle of allocation of proof to the party with the most ready access to the relevant information. A practical awkwardness at this stage of the increasingly elaborate proof scheme being worked out for Title VII litigation should be frankly noted. As has generally been recognized, these proof schemes are aimed more at defining the appropriate mode of judicial analysis of the proof after all the evidence has been adduced in the Title VII bench trial, than at prescribing the sequential order of proof by litigants. See B. Schlei & P. Grossman, Employment Discrimination Law 1159 — 61 (1976). A plaintiff attempting to establish his prima facie disparate impact case by statistical proof using general population statistics has no means under general procedural rules to force during the bench trial itself a judicial ruling on the appropriateness of the statistics. If not forced earlier, the ruling may appear only after submission of the case. A defendant’s motion to dismiss under F.R.Civ.P. 41(b) is not likely to force such a ruling, see generally 9 Wright & Miller, Federal Practice & Procedure: Civil § 2371, and in any event cannot be prompted by plaintiff. As a result, a plaintiff may learn only upon the filing of findings of fact and conclusions of law in the trial court, or even only after appeal, see, e. g. Hill v. Western Elec. Co., 596 F.2d at 106, that his prima facie case has failed because general population statistics were not appropriate. While an opportunity to reopen proof using appropriate statistics might, of course, be accorded even at these late stages, it would be by an exercise of grace. The obvious practical answer to this awkwardness is the judicious use of discovery and other pre-trial procedures to identify this as an issue of fact or law requiring timely resolution in the trial court. [9] The Supreme Court noted in Hazelwood that applicant flow data would be “very relevant”. proof. 433 U.S. at 308, 97 S.Ct. 2736 n. 13. In Dothard v. Rawlinson, 433 U.S. 321, 97 S.Ct. 2720, 53 L.Ed.2d 786 (1977), the Court recognized that applicant data might not adequately reflect the pool of potential applicants if the effect of the discriminatory standard has been to discourage otherwise qualified persons from applying. Id. at 330, 97 S.Ct. 2720. [10] In Patterson v. American Tobacco Co., 535 F.2d 257, 275 (4th Cir. 1976), we held, in assessing the appropriateness of a remedial order establishing quotas for the appointment of minorities to upper level positions, that SMSA evidence furnished a more realistic measure of a company’s conduct than statistics indicating the gross percentages of minorities in the general population.