Although neither guidance addresses the issue, and I am not aware of any other guidance by either agency that does, it would appear that the principles set out for the evaluation of the discriminatory use of arrest and conviction records would be equally applicable to challenges to gender discrimination arising out of the use of such records. According to the same Department of Justice report cited by the EEOC for racial and national origin disparities in imprisonment rates, there were a total of 1,445,628 sentenced male prisoners under the jurisdiction of Federal and State correctional institutions nationwide in 2010 compared to 104,629 sentenced female prisoners under the jurisdiction of such institutions during the same period. These disparities persist even among males and females of the same race or national origin group. In other words, African American males are significantly more likely to be incarcerated than African American females, White males more than White females, Hispanic males more than Hispanic females. The numbers of sentenced males of all races combined is more than 13 times as high as the number of sentenced females. You will note that the statistics cited by the EEOC above, discussing the incarceration disparities by race and national origin, used male rates of incarceration. Female incarceration rates also demonstrate racial and national origin disparities in incarceration rates but the numbers are not nearly as dramatic because of the lower female incarceration rate generally. In other words, the gender disparity is even wider than the racial and national origin disparities that are the subject of the guidance.
Since the disparate treatment and disparate impact theories apply to gender as well as race and national origin, it is safe to say that the guidance of both agencies can be used in evaluating the gender impact as well as the racial and national origin impact of arrest and conviction records. The discussion that follows focuses on race and national origin bias because this article is intended to be a review of the EEOC and OFCCP guidance. However, you should keep in mind that since the principles set out in the guidance apply equally to gender based claims, males, regardless of race or national origin, potentially have a cause of action for gender-based discrimination arising out of the use of arrest and conviction records to limit their employment opportunities.
The principles articulated in the EEOC and OFCCP guidance apply to both employees and applicants. The issue of arrest and conviction records comes up for most employers when they conduct criminal background checks at or near the end of the hiring process. According to a press release published by the Society for Human Resource Management (SHRM), "SHRM found an increase in the percentage of employers that don’t conduct criminal background checks, from 7 percent in 2010 to 14 percent in 2012." This suggests that potentially 86% of SHRM's survey respondents still conduct criminal background checks, a decrease from the 93% of respondents who conducted background checks in 2010 (assuming all responses were either yes or no on the question of criminal background checks). The downward trend in criminal background checks may continue in light of the issues raised in the OFCCP and EEOC policy guidance referenced above. However, despite this decline, it is clear that a significant number of employers continue to conduct such investigations and need to be careful in how they use this information in making employment decisions.
In general, disparate treatment occurs when an employer treats an employee or applicant differently (and usually less favorably) on a protected basis such as race or national origin. As noted in the OFCCP and EEOC guidance, having a criminal history is not, by itself, a protected basis under the laws each agency enforces. In the context of disparate treatment, the agencies are looking to see if applicants and employees are treated differently with respect to arrest and conviction because of their membership in certain racial or national origin groups.
It would be disparate treatment to apply a different standard of acceptable criminal history to one race as compared with another. For example, if a felony conviction is treated as disqualifying for an African American applicant but overlooked for a similarly situated White applicant, you have a case of unlawful disparate treatment. Similarly, terminating an Hispanic employee because it is discovered that he served time for credit card fraud, while retaining a similarly situated White employee who is also discovered to have served time for credit card fraud, is also disparate treatment.
In investigating this termination, it is not enough to simply verify that the terminated employee actually had the criminal history that led to his termination, rather the compliance officer should examine how the employer has dealt with other employees when it became aware of prior criminal convictions. This investigative step should turn up the similarly situated White employee who had a comparable criminal history but was not terminated. In such an investigation, the actual commission of the offense is not the issue since both comparators committed the offense. The issue is whether the consequences with respect to employment were harsher on the applicant or employee because of his or her race or national origin.
As noted in the EEOC Guidance, disparate treatment can also occur if applicants and/or employees are treated differently with respect to the background investigation process. This can occur when background checks are only conducted on some races and national origins but not on others. It can also occur if the type or extent of the criminal background check varies by race or national origin.
Assuming there was enough meaningful statistical evidence to support an inference of intentional race or national origin discrimination based on the employer's history of exclusions related to arrest and conviction records, the agencies could also potentially establish a pattern and practice of intentional race or national origin discrimination. However, it is more likely that any statistical arguments would be brought under a disparate impact theory.
In disparate impact, a facially neutral policy or practice is shown to have a disproportionately negative effect on a protected group and the policy or practice either cannot be shown to be job related and consistent with business necessity, or there is a less discriminatory means of achieving the employer's legitimate policy objectives. For example, a blanket policy of not hiring anyone with an arrest record is facially neutral because it does not single out a particular racial or national origin group for different treatment. Everyone, regardless of race or natural origin, is subject to the policy. In this scenario, disparate impact is an issue because certain racial and national origin groups have historically had significantly higher arrest rates.
To be lawful, policies that have a disparate impact must be job related and consistent with business necessity, and there must be no less discriminatory alternative for achieving the employer's legitimate business needs. A blanket policy that excludes everyone with an arrest record cannot satisfy that requirement. Since arrest does not prove that any crime was committed and there is a presumption of innocence, it will not be possible for the employer to show that the blanket arrest record exclusion is job related and consistent with business necessity. It should be noted that while a blanket exclusion based on arrest records cannot satisfy the disparate impact standard, employers are still permitted to make employment decisions based on the actual conduct that led to the arrest. However, such a decision will need to be an individualized inquiry that includes evidence that the underlying conduct goes to the ability to perform the job in question.
There is nothing new in the proposition that arrest records cannot be used to disqualify applicants and employees for employment opportunities. The more complex question is how disparate impact applies with respect to exclusions based on conviction records. The easiest case is with blanket policies that erect an absolute bar to hiring anyone with a conviction record. It would be difficult, if not impossible, to show that such a broad exclusion was either job related and consistent with business necessity, or that there was no less discriminatory alternative. Virtually the only situation where this might be justified is if there was a Federal requirement that the job exclude anyone with a conviction record. Since Federal equal opportunity requirements trump state and local laws, those laws would not be sufficient to shield an employer who instituted such a policy.
Both guidance documents acknowledge the possibility that a more tailored exclusion for certain types of convictions may be defensible. They offer two ways to establish that an exclusion for convictions is job related and consistent with business necessity. First, the exclusion can be validated pursuant to the Uniform Guidelines on Employee Selection Procedures (UGESP). These are the same guidelines used to determine if a test is valid. Essentially an exclusion for convictions operates as a test of whether the individual should be hired or otherwise afforded an employment opportunity. The important issue for validation studies in the context of testing is the examination of the particular job in question to determine the predictive value of the applied criteria. While validation under UGESP is theoretically possible for exclusionary conviction criteria, it is not likely to be very practical or cost effective. Even convictions for the same general offense may arise out of very different circumstances that would have a bearing on the likelihood of a successful employment outcome in a particular job.
Where traditional validation is not feasible, such as where there is not sufficient pertinent data about the connection between the exclusionary criteria and the requirements of the job, both agencies point to three factors that should be considered in determining whether the exclusion is lawful. These factors are:
- the nature and gravity of the offense or conduct;
- the time that has passed since the offense, conduct or completion of the sentence; and
- the nature of the job held or sought.
In my experience, two scenarios usually emerge when the agencies are asked to evaluate matters about which they know fairly little. Either they resort to gut feelings about what to them seem "common sense" requirements or they bring in outside experts and rely on professional opinions. The outside expert route is usually reserved either for litigation or for large cases. What this means for you as a contractor is that you should avoid exclusions where possible, and where you feel compelled to implement exclusionary policies, focus on tailoring any exclusionary requirements to the job in a way that a reasonably intelligent person could feel comfortable that the connection makes sense.
For example, if an exclusion is going to be implemented, see if there are any Federal requirements that the holder of such a position not have the particular criminal background that causes you concern. An external, Federal obligation will usually end the inquiry. Absent any external requirement, draw a clear connection between the requirements of the job and the nature of the offenses that cause concern. Basically, this is a combination of factors 1 and 3. Does the nature of the offense relate to the nature of the job held or sought? Often cited examples of such a relationship are where money handlers are required not to have a history of offenses related to the mishandling or misappropriation of funds, or child care providers are required not to have convictions relating to the abuse of children. The clearer the nexus between the concerning offense and the essential functions of the job, the more likely the agency will be comfortable with the exclusion. Challenge yourself to clearly articulate what the nexus is that requires exclusion. Also consider if a probationary period or other strategy might constitute a less discriminatory alternative and be able to articulate exactly why probation poses an unacceptable risk if you reject it as a possible alternative.
How much time has elapsed since the conviction should also be a part of the calculus in determining whether a conviction record exclusion is discriminatory. Again, you should be able to articulate a rationale for your timeframe that a reasonably intelligent, but not particularly knowledgeable, individual can understand. The agencies like studies that support your time frame but I doubt relevant ones will be easy to come by. What you may be able to point to are norms observed by others in the same or similar industries. At least then you will be able to show that you gave some thought to what constitutes an appropriate interval between the offense and employment in the position held or desired. OFCCP reviews a large number of companies and may be able to compare your timeframes with other similar companies. If you do the comparison up front, at least you will not draw attention to yourself by being significantly out of the mainstream by having a much longer interval between the offense and employment or a much longer timeframe over which you review criminal convictions.
The following recommendations may be helpful in avoiding discrimination violations related to arrest and conviction records:
- Avoid blanket exclusions
- Use individualized assessments tailored to the applicant or employee and the particular job held or desired
- Monitor the implementation of any exclusions for disparate treatment
- Re-evaluate your use of background checks to limit them to situations where they are absolutely necessary
- Be able to explain any time frames, such as how many years are covered by your criminal background check or how much time must elapse between conviction and eligibility for the position at issue
- Be able to clearly articulate the nexus between the conviction and the job
- Be able to identify any alternatives that were considered and provide explanations as to why they were not acceptable