How to statistically prove or disprove pattern of employment discrimination?

Suppose a US company employs 100 people, of whom 20 are officers, and the company is located in a major US city with a large racial minority population. Suppose only 1 of the 20 officers is a racial minority. How would the EEOC statistically prove a pattern of employment discrimination? How would the company statistically disprove it? Is there a good online citation for statistical calculations in employment discrimination?

How to statistically prove or disprove pattern of employment discrimination?
It seems you would have to first prove, statistically speaking, that qualified minorities were not considered for officer positions.





If the position required a business degree or some other such speciality and only 1 out of 20 minority applicants for the job had a business degree, that complicates the figures somewhat.





Of course it could be argued that the employer was writing the job requirements in order to exclude certain minorities, too. Tough question
Reply:The statistics that the EEOC looks at and that the accused employer submits, are the statistics of that particular company. In your example of the 1 out of 20 company officers being a certain racial minority, the EEOC would only get this when someone file a complaint. The complaint could be that a person of that race was not promoted to the officer position or that this person was an officer but was demoted/fired due to his/her race.





The EEOC then asks the company to produce personnel records of those who have had, who have applied for, and who have been fired from that officer position --- with a listing of the race of these people. The EEOC investigator also speaks with the company to find out certain things about becoming an officer --- is there a normal career path? is this always an outside hire? and if so is the recruitment done within the guildlines of equality? what are the criteria used for the hiring? The EEOC may ask to speak with former officers of this race and those who select people into the officer position.





The proof of a pattern of employment discrimination is done by looking at all those of this race who have ever worked for this company, their position, and if they were terminated. It takes the EEOC investigator months to wade through this type of evidence.





Many times the EEOC looks at its own history of complaints. For example a factory in Illinois routinely refused to let the women on its assembly line leave the line for bathroom breaks, except for the breaks already assigned. The women were told to use adult diapers and many did. Over a 10+ year time period women filed EEOC complaints on this company. In every case the EEOC did what they could, but in the end told the complainant that the EEOC would not take the case and that they should get their own attorney. After the 10+ years of many complaints the EEOC finally decided to investigate the company and found many more violations and evidence of sexual discrimination -- they found a pattern of sex discrimination. The EEOC then contacted all those women who had filed complaints, but were blown off and included them in a class action. The class action was settled for millions of dollars.





On the company side the company wants to show that both hiring and firing in this position has nothing to do with race. If the initial complaint that started this was due to a termination, the company will bring in supervisors to testify that this employee was not terminated due to race. Many times the company will provide a list of all former employees and their race, who were fired for this same violation.





I think what you are getting at is the company saying that in this town people of this race who are qualified for this position only make up a certain percent of the population. I have never seen the census type of statistics used successfully in an EEOC case.