Recently the administration called on corporations to pledge not to use unemployment status as a selection criterion in their hiring decisions. According to various media reports at least 300 corporations have taken this pledge including a number of major federal contractors. The purpose of this article is to discuss whether and to what extent the laws currently enforced by OFCCP can be used to address this issue.
OFCCP enforces Executive Order 11246, which prohibits discrimination on the basis of race, sex, national origin, color, and religion. It also enforces the Vietnam Era Veteran’s Readjustment and Assistance Act, commonly referred to as VEVRAA or Section 4212 and Section 503 of the Rehabilitation Act which prohibits discrimination on the basis of disability in federal contracting. OFCCP does not enforce any laws that directly prohibit discrimination on the basis of unemployment status. It is possible, on the right set of facts for unemployment status to fall within the enforcement power of the OFCCP, however, this outcome is highly unlikely.
Each of the laws enforced by OFCCP relies on two basic theories of discrimination — disparate treatment and disparate impact. Disparate treatment refers to treating individuals differently based on a protected status and requires an intent to discriminate. Intent can be proven by direct evidence or inferred from comparative evidence of differences in treatment from one group to another or from statistics. Disparate impact refers to using facially neutral criteria that disproportionately screen out protected groups. Disparate impact does not require any intent to discriminate only a discriminatory effect stemming from the exclusionary practice.
As enforced by OFCCP, disparate impact usually relies on a disparity in selection rates caused by a facially neutral criterion that is statistically significant at a level of 2 or more standard deviations. In other words, the facially neutral criterion is found to have a statistically significant negative impact on selection rates for the disfavored group. The regulations allow for a finding of disparate impact if the selection rate of the disfavored group is less than 80 percent of the selection rate for the favored group (the 80% rule) but in practice OFCCP relies more heavily on the 2 standard deviation threshold favored by the courts. Disparate impact on the basis of disability and/or protected veterans status does not require showing a statistically significant disparity in selection rates but only that the criteria excluded either an individual or groups of individuals because of disability or protected veteran status.
Using unemployment status as a selection criteria can be reached under disparate treatment if the criterion is applied differently on a basis covered under the laws enforced by OFCCP. For example, if a federal contractor pledges not to use unemployment status as a selection criterion but only honors that pledge with respect to white applicants and disregards the pledge with respect to Latino applicants, it would be engaging in disparate treatment. Similarly, if a federal contractor requires as a selection criterion that applicants must be currently employed to be considered but only applies that criterion to Latino applicants and ignores this criterion for white applicants, it would be engaging in disparate treatment in the application of the criterion. In either instance, it is the discriminatory implementation of the policy that would get the contractor in trouble, not the content of the policy.
Disparate treatment could also be shown if the evidence proves that the federal contractor adopted unemployment status as a selection criterion for the purpose of excluding a particular protected class of applicants, such as individuals with disabilities.
Under disparate impact, unemployment status would be considered a facially neutral selection criterion. It is facially neutral because it does not single out any group on the basis of race, sex, color, national origin, religion, disability or protected veteran’s status.
Executive Order 11246
As a facially neutral criterion it would only become problematic under Executive Order 11246 if it had the effect of disproportionately screening out a protected group at a statistically significant rate. An employer can defend the use of a criterion having a disparate impact by showing that the criterion is job related for the position in question and consistent with business necessity and that there is no less discriminatory alternative that would allow the employer to fulfill its legitimate business needs. The Uniform Guidelines on Employee Selection Procedures (UGESP) (41 CFR 60-3) apply to selection procedures under the Executive Order. In order to show that a selection procedure having a disparate impact is job related and consistent with business necessity a contractor must show that the use of the selection criterion complies with UGESP.
Data published by the Bureau of Labor Statistics show that the rate of unemployment varies by, among other things, race, gender and ethnic group.1 For example, for the same time period, December 2013, the non-seasonally adjusted civilian unemployment rate for whites was 5.7, for blacks 11.6, for Hispanic/Latino 8.3, and for Asians 4.1.
The fact that there are variances in unemployment rates by race, national origin/ethnicity, and gender does not automatically prove that using unemployment status as a selection criteria violates the Executive Order. In order to establish disparate impact in a particular hiring event, you would have to show that the facially neutral criteria operated in that particular hiring event to screen out members of the disfavored group at a statistically significant higher rate relative to the favored group.
There are several factors that go into this analysis. The first factor is the determination of which unemployment statistics are relevant. This requires making sure that the figures apply to the time period when the hiring occurred. It also involves making sure that the unemployment figures are based on the reasonable recruitment area. For example, national unemployment statistics are not particularly relevant if the reasonable recruitment area is local. The unemployment figures may also need to be refined to ensure that they only include individuals with the relevant job skills. For example, general labor force unemployment statistics, even if limited to the reasonable recruitment area, may not be appropriate if the job requires specialized skills, certifications or licenses.
Once the appropriate unemployment statistics are identified, they have to be applied to the appropriate pool for comparison. If the contractor advertised that only the currently employed need apply this may have a chilling effect on the size of the applicant pool. In this situation, it may be more appropriate to use relevant labor market statistics rather than applicant flow statistics. Applicant flow statistics are generally more appropriate when the criterion is applied later in the application process.2
Once the appropriate statistics are isolated and the pool for comparison identified, the agency would still have to show that the exclusion rates meet the thresholds for establishing disparate impact. Even where the rates of exclusion match the unemployment rates, it may be quite difficult to get to a number that is statistically significant depending on the size of the pool used for comparison.
- Consider the following hypothetical example of a technology developer looking to fill 11 open positions. Let’s assume that the candidate pool consists of 100 individuals, 70 of whom are white and 30 of whom are Hispanic. Applying the current unemployment rates, let’s assume that 8 percent of the white applicants (a total of five) were unemployed and stricken from consideration for that reason. Let’s also assume that roughly 12 percent (a total of four) of the 30 potential Hispanic candidates were unemployed and stricken from consideration for that reason. The adverse impact ratio is 93 percent. Applying the four-fifths rule, there is no statistical evidence of disparate impact. Also, if you look at standard deviations, there is no statistical evidence of adverse impact because the proportion of Hispanics who received further consideration would be only 0.991 standard deviations below the proportion of whites who received further consideration.3
Thus, while a disparate impact case can theoretically be made based on the impact of unemployment status as a selection criteria, it is not an easy case to make.
On the other hand, if disparate impact is shown, it would be even more difficult to prove that an unemployment status criterion is job related and consistent with business necessity. Unemployment status is basically a proxy or shortcut to weed out individuals because of concerns about their skills, the currency of their knowledge of developments in the field and the like. Direct testing of possession of the requisite knowledge, skills and abilities is a readily available less discriminatory alternative for such a proxy.
Section 503 (disability) and Section 4212 (protected veteran’s status)
OFCCP disability regulations provide:
- (g) Qualification standards, tests and other selection criteria—(1) In general. It is unlawful for the contractor to use qualification standards, employment tests or other selection criteria that screen out or tend to screen out an individual with a disability or a class of individuals with disabilities, on the basis of disability, unless the standard, test or other selection criterion, as used by the contractor, is shown to be job-related for the position in question and is consistent with business necessity. … §60-741.21
An individual with a disability would be able to state a disparate impact claim by showing that an unemployment status selection criterion operated as a barrier to employment and that there is a nexus between disability status and unemployment status.
The first hurdle is finding proof that unemployment status actually operated as a barrier. The fact that the disability claimant does not need statistics also means that statistics are not available to serve as proof that unemployment status was the reason for non-selection. The mere fact that the applicant with a disability is currently unemployed is not sufficient to show that the unemployment status was the barrier to hire. Other evidence will have to be produced.
The second hurdle is establishing a nexus between disability status and unemployment status. Data showing the relative unemployment rates between individuals with disabilities and non-disabled individuals would most likely be relied on to establish this nexus. BLS data on unemployment status of the civilian population by disability status shows a December 2013 rate of 11.9 for persons with a disability compared to 6.3 for persons with no disability.4
If unemployment rates of individuals with disabilities and/or protected veterans are to be used to establish this nexus, as with the Executive Order, you would have to ensure that the statistics used are relevant with respect to the timing of the hire and related to individuals with the requisite skills in the reasonable recruitment area.
The principles applicable in the disability context apply equally to protected veterans. The regulatory language is virtually identical to the Section 503 regulations quoted above.5
Except where employers overtly identify current employment status as a selection criterion, it is extremely difficult to prove such a criterion was applied. Even where such a criterion can be shown to have been applied, it may be difficult to demonstrate unlawful intent or unlawful impact in the manner of its application. Thus, while it is theoretically possible for the laws enforced by OFCCP to touch upon the problem of discrimination on the basis of unemployment status, they are clearly not an effective mechanism for addressing this issue.
1. See BLS Tables A-2, Employment status of the civilian population by race, sex and age and A-3, Employment status of the Hispanic or Latino population by sex and age. Last updated January 10, 2014. ↩
2. See Written Testimony of Helen Norton, Professor at University of Colorado School of Law, Meeting of February 16, 2011 – EEOC to Examine Treatment of Unemployed Job Seekers (www.eeoc.gov) ↩
3. See Written Testimony of James S. Urban, Partner, Jones, Day, Meeting of February 16, 2011 – EEOC to Examine Treatment of Unemployed Job Seekers (www.eeoc.gov) ↩
4. Table A-6. Employment status of the civilian populations by sex, age, and disability status not seasonally adjusted. ↩
5. (g) Qualification standards, tests and other selection criteria—(1) In general. It is unlawful for the contractor to use qualification standards, employment tests or other selection criteria that screen out or tend to screen out individuals on the basis of their status as disabled veterans, recently separated veterans, other protected veterans, or Armed Forces service medal veterans unless the standard, test or other selection criterion, as used by the contractor, is shown to be job-related for the position in question and is consistent with business necessity. … §60-300.21 ↩