MEMORANDUM
Plaintiff Buck Green commenced this class action upon the claim that the use by defendant Missouri Pacific Railroad Company (“Mo Pac”) of arrest and conviction records as barriers to employment is racially discriminatory in violation of the Civil Rights Acts of 18701 and 1964 2.
This action was tried to the Court sitting without a jury.
Defendant’s prior motion to dismiss for lack of subject matter jurisdiction was overruled upon the ground that such jurisdiction existed as granted by 28 U.S.C. § 1343. Brady v. Bristol-Meyers, Inc., 459 F.2d 621, 621-622 (8th Cir. 1972). The Court now concludes, supported by evidentiary findings described below, that subject matter jurisdiction also exists as granted by 42 U.S. C. § 2000e-5(f)(l).
On April 13, 1973, the Court ordered this action maintained as a class suit pursuant to Rule 23(a) and (b)(2), Federal Rules of Civil Procedure. 7 FEP Cases 140. The class was then defined as those Negroes who have been denied employment at defendant’s General Office in St. Louis, Missouri, because of defendant’s policy of refusing to hire persons arrested and convicted of a criminal offense other than a minor traffic violation. At trial plaintiff moved to broaden the class which motion was taken with the case. I find no reason to alter the class definition and will not disturb the present class definition.
In Equal Employment Opportunity Commission v. Missouri Pacific Railroad Company, 493 F.2d 71, 75 (8th Cir. 1974), the Court of Appeals encouraged this Court to permit the Equal Employment Opportunity Commission (“EEOC”) to intervene in this action. The EEOC, however, took no step to participate in this action.
On September 29, 1970, plaintiff Buck Green, a black man, then twenty-nine years of age, applied for employment as a clerk3 at defendant’s Personnel Office, located in its General Office in the City of St. Louis, Missouri. Green had been referred there by the Missouri State Employment Service. As do all job applicants at Mo Pac’s General Office, Green filled out an application form. On it he stated that his last employment terminated on November 15, 1968, when he began a term of imprisonment for refusing military induction4. He stated he was paroled on January 24, 1970, after twenty-one months imprisonment. In a personal interview conducted by a Mo Pac personnel office employee, plaintiff was informed that his employment application could not be considered because of his conviction and prison record.
After the rejection of his employment application plaintiff corresponded with certain officials of defendant concerning its policy of not hiring ex-offenders. Thereafter, by letter of December 13, 1974, plaintiff wrote to the Kansas City, Missouri, office of the EEOC relating his employment application rejection by Mo Pac and requesting information concerning Mo Pae’s hiring policy. In response, the EEOC office sent plaintiff a discrimination charge form. On January 17, 1971, plaintiff executed a formal EEOC charge alleging Mo Pac’s failure to hire him because of his conviction record. On August 9, 1971, the charge *994was amended. On August 11, 1972, the EEOC issued Green a ninety day notice of his right to sue. This action was commenced on November 7, 1972.
Defendant is a railroad corporation that conducts business in several states, including Missouri. It employs 22,GOO-23,000 persons company-wide. In the St. Louis metropolitan area it employs approximately 3,500 persons. Approximately 2,000 persons are employed in the General Office.
Mo Pae’s Central Personnel Office was established in the General Office in 1964. The Central Personnel Office administers defendant’s company-wide employment policies but processes employment applications for positions in the St. Louis metropolitan, area only, including the General Office. When there are vacancies in entry level positions in this area, this personnel office recommends to the subject department head the applicant best qualified for the position.
Mo Pac has had a general policy against hiring persons with criminal convictions other than minor traffic offenses since 1948. Ray Breedlove, Mo Pac’s Personnel Manager, testified without factual specification, that this policy was initiated in reaction to a personnel problem that then existed.
Prior to December, 1972, when the United States Court of Appeals for the Ninth Circuit ruled in Gregory v. Litton Systems, Inc., 472 F.2d 631, that excluding from employment persons because of their arrest records was unlawful discrimination, defendant investigated applicants’ arrest records and disqualified those who had been arrested a substantial number of times or for a serious crime. After Gregory, defendant sought information regarding arrest records only of the applicants themselves. This has been done by written question on the application form and orally in personal interviews. In early 1973, on advise of legal counsel, defendant eliminated the question regarding arrests from the application form, and ceased using arrest records as an absolute bar to employment. However, as late as April, 1973, defendant inquired orally of applicants in the personal interviews about arrests as a prelude to the inquiry about convictions. Currently, if an applicant has been arrested and a charge is still pending, the applicant is not automatically disqualified. However, all circumstances of the applicant, including the arrest and the pending charge are considered in evaluating the applicant. Information concerning the charge is sought from local police departments. Since September, 1971, no employment applicant has been rejected solely because of his arrest record.
Mo Pac’s policies of rejecting employment applications of persons who have conviction and arrest records have not been applied with intentional racial bias. While exceptions have been made for persons convicted of minor crimes5 where no imprisonment was served, such exceptions have been rare. In the St. Louis metropolitan area only fourteen such exceptions were found among persons employed since 1963 — four of them in the General Office.
Plaintiff seeks to prove a prima facie case of racial discrimination with statistical evidence. See e. g., Parham v. Southwestern Bell Telephone Company, 433 F.2d 421, 426 (8th Cir. 1970); Carter v. Gallagher, 452 F.2d 315, 323 (8th Cir. 1971), mod. on other ground, 452 F.2d 327, cert. den. 406 U.S. 950, 92 S.Ct. 2045, 32 L.Ed.2d 338 (1972). The determinative issue in this regard is whether the subject employment policies bar from employment with defendant a percentage of black persons among all applicants that is disproportionately large when compared with (a) the percentage of black persons in the relevant population area (that area from which applicants are more likely to come than not, Sagers v. Yellow Freight System, Inc., 58 F.R.D. 54 (N.D.Ga.1973)) and *995(b) the percentage of white persons among the applicants also barred from employment by the same policies. The Court believes the relevant geographical area is the St. Louis metropolitan area, rather than the City of St. Louis alone.
Statistical evidence6 indicates that from 1960 to 1970, inclusive, blacks comprised approximately from 10.5% to 11.-1% of the population of the United States; from 9% to 10.3% of the Missouri population; from 10.3% to 12.8% of the Illinois population; from 28.6% to 40.9 % of the population of the City of St. Louis; and from 14.3% to 16% of the population of the St. Louis metropolitan area.
Certain statistical data and treatises offered into evidence by plaintiff indicate that the percentage of black or non-white persons convicted of criminal offenses is two to three times the percentage of black or non-white persons in the populations of certain geographical areas. The same conclusion was indicated by a comparison of the percentage of black persons arrested with the percentage of black persons in the populations of certain geographical areas.
Dr. Ronald Christensen concluded at trial that it is between 2.2 and 6.7 times as likely that a black person will have a criminal conviction record as it is a white person will have such a record. He further concluded that in urban areas from 36.9% to 78.1% of all black persons would have a conviction during their lifetime, but that from 11.6% to 16.8% of all white persons would acquire a conviction record.
This general statistical information, however, must yield to the extent records of employment applications at defendant’s General Office. These records show, and the Court finds, that the subject policies did not have a discriminatory effect upon black employment applicants. From September, 1971, through November, 1973 7, at defendant’s General Office 3282 black persons and 5206 white persons applied for employment. Of these 174 blacks (5.3% of the black applicants) were rejected because of their conviction records and 118 whites (2.23% of the white applicants) were likewise rejected. Of the same applicants 73 blacks and 99 whites had arrest records.
Plaintiff places great weight upon a comparison between the 5.3% and 2.-23% figures to show discriminatory effect. However, such a comparison is, alone, inconclusive on the issue of discriminatory impact because it fails to allow for a consideration of the percentage of blacks in the total number of applicants and for a consideration of the percentage of blacks in the St. Louis metropolitan area. The proper computation is as follows:
118 (whites rejected due to conviction) 8488 (total number of applicants) x 100 = 1.4%
174 (blacks rejected due to conviction) 8488 (total number of applicants) x 100 = 2.05%
99 (whites with arrest records) 8488 (total number of applicants) x 100 = 1.2%
73 (blacks with arrest records) 8488 (total number of applicants) x 100 = .9%
*996A comparison of the 2.05% and .9% figures, next above, with the 16%8 (percentage of blacks in' the St. Louis metropolitan .area in 1970) shows that the percentage of blacks adversely affected by the subject hiring policies is not disproportionately large when compared with the percentage of blacks in the subject population. The 2.05% figure shows a de minimis discriminatory effect against blacks with conviction ' records, and the 1.2% figure-shows such an effect against whites with arrest records. Plaintiff discounts this analysis by arguing without evidentiary support, that the discriminatory reputation of defendant in the black community dissuaded other' blacks with arrest or conviction records from applying for employment. This argument is based solely upon speculation and must be rejected.
Since 1968 Mo Pac has put into effect an affirmative action program that increased minority employment opportunities9. Also, since 1968 Mo Pac has cooperated in a program sponsored by The National Association of Businessmen by hiring 412 disadvantaged blacks, including ex-offenders.
Upon the evidence before me I find that plaintiff has not proven a prima facie case that the subject policies had a discriminatory impact upon those blacks who applied for employment at the defendant’s General Office.
Assuming arguendo that plaintiff has proven a prima facie case, I find that defendant has proven that the subject policies are founded upon “business necessity”. Griggs v. Duke Power Co., 401 U.S. 424, 431, 91 S.Ct. 849, 28 L.Ed.2d 158 (1971). The burden of proving business necessity has been described as requiring proof at least that defendant’s policies foster employee productivity, see, Griggs, supra; and see, Wallace v. Debron Corporation, 494 F.2d 674, 677 (8th Cir. 1974), citing Note, 85 Harv.L.Rev. 1482, at 1487 (1972)10, and as requiring proof that no acceptable alternative exists that will accomplish the same employee productivity with a “lesser differential racial impact”. United States v. St. Louis-San Francisco Railway Co., 464 F.2d 301, 308 (8th Cir. en banc, 1972), cert. den., 409 U.S. 1116, 93 S.Ct. 913, 34 L.Ed.2d 700 (1973); Williams v. Matthews Company, 499 F.2d 819, at p. 828 (8th Cir. filed June 20, 1974).
In this regard, defendant admits that it has not empirically validated the conviction and arrest record policies as required by the EEOC Guidelines on Employee Selection Procedures, 29 C.F. R. Part 1607. While these Guidelines are recommended in appropriate cases, Griggs, supra, 401 U.S., at 434, 91 S.Ct. 849, they are not binding per se upon *997this Court. I do not believe the Guidelines are appropriately applied here. The employment policies of which plaintiff complains herein are radically different from the testing criteria with which Griggs was involved.
The accumulation of empirical data regarding the job-success of ex-convicts has been the difficult task of the directors of many programs. Data that has been generated is often drawn from a too-narrow selection of subjects to be of much predictive value when used in a specific employment context, such as in the instant case.
Robert N. McMurray, an industrial psychologist who participated in devising Mo Pac’s employment policies, considers a criminal record to be an objective indication of an applicant’s personality. He is familiar with statistics that show that at least 50% of those convicted will be convicted again. However, there is no sure method for determining who will be recidivists. The state of the art of psychology has not developed to a degree that permits the use of definite rules.
The Court has carefully considered the arguments of plaintiff and defendant regarding those factors defendant believes support its position on the business necessity of the subject policies. The prevention of cargo theft, theft of company property, refusal to comply with employment directives, and the employment disruption caused by recidivism are among the factors that concern defendant. These factors are relevant to the safety and efficiency of defendant’s business and establish that the subject policies are based upon sound business necessity. See footnote 9, supra.
Plaintiff indicates that defendant might have attempted to validate an alternative policy of correlating specific crimes with specific jobs and disqualifying for employment persons convicted of certain crimes for certain jobs. To judicially order or review the implementation or validation of such a policy is to open a Pandora’s box of difficulties in judicial administration. Such factors as nature of the offense, length and nature of the punishment, and remoteness in time of either the conviction or the punishment do not reasonably,indicate with certainty whether ex-offenders will be successful employees. By what standard should the Court proceed, when the science of psychology has not established certain rules on this subject?
The Court certainly does not believe that all ex-offenders have been or will be bad employees and the employment of ex-offenders can only be encouraged in appropriate cases., However, the Court concludes, under the facts of this action, that defendant’s subject policies are not violative of the Civil Rights Acts of 1870 or of 1964.
For the reasons given above, I conclude that plaintiff’s claim is without merit and I will dismiss the action.
This memorandum contains the findings of fact and conclusions of law required by Rule 52(a), Federal Rules of Civil Procedure.
APPENDIX I
APPENDIX TI
APPENDIX III
*1000APPENDIX IV
APPENDIX V
*1001APPENDIX VI
GENERAL OFFICE