Adverse Impact Analysis: Aggregation of Races
Adverse Impact Analysis: Aggregation of Races
Share
In August 2013, Adverse Impact (AI) analysis received one of the most important judicial clarifications rendered since its formal Supreme Court codification in 1971. Administrative Law Judge Kenneth A. Krantz ruled in favor of VF Jeanswear (VF) and denied OFCCP’s request for summary judgment against VF. In his decision, Judge Krantz formally clarified and articulated a judicial position on methods of grouping races for AI analyses. While it is common for White versus non-Whites (grouping all non-White individuals) in AI analyses, Judge Krantz concluded that other race groupings, such as “non-Asians” (in VF), is not a race group, both in common parlance and by law, and therefore, the OFCCP’s race discrimination claim against VF had no basis.

In October 2016, a little over three (3) years after the VF decision, LandCare USA, LLC was found to have discriminated against non-Hispanics at their Las Vegas, Nevada facility and entered into a financial settlement with the OFCCP.

Beyond the obvious significance of the LandCare outcome as it relates to social justice, the outcome appears to be inconsistent with the VF decision where use of a non-traditional race group (non-Asians) in an adverse impact analysis was not allowed. This inconsistency raises many important questions that we will analyze in this paper:

  1. Is the LandCare outcome a reversal of the ruling on the VF Jeanswear case?

  2. Does the LandCare outcome allow for aggregation of other race groups outside of the more familiar and traditional minority groups?

  3. How did the OFCCP successfully obtain a finding of AI against non-Hispanics in LandCare? On a related note, could the OFCCP have applied a similar strategy with equally successful outcomes in the VF matter?
Tale of Two OFCCP Cases – Details

VF Audit

In the VF matter, the OFCCP alleged that VF Jeanswear (VF) discriminated against qualified non-Asian applicants for the Operative job group positions in their Winston-Salem, North Carolina facility. The OFCCP’s theory of discrimination was simple: qualified non-Asian applicants experienced significantly lower hiring rates than Asian applicants.

VF Outcome

On August 5, 2013, Judge Krantz denied the OFCCP’s motion for summary judgment, citing that non-Asians (in this case, Whites, African-Americans, and Hispanics combined) is not a “race” as defined by the regulations or in common usage and, therefore, the OFCCP’s assertion of discrimination based on race could not be enforced. To date, Judge Krantz’s decision has not been challenged by the OFCCP and hence, the aggregation of all minority groups (African-American, Asian, Hispanic, and American Indian/Alaskan Native) remains the only acceptable race/ethnic group as defined in 41 CFR 60-3.4b. This group is commonly referred to as Total Minority. This decision seemed to have taken a different turn in the LandCare case.

LandCare Audit

In the LandCare matter, the OFCCP alleged that LandCare discriminated against qualified non-Hispanics for the Laborer, non-driver position in their Las Vegas, Nevada facility. OFCCP’s theory of discrimination was simple: qualified non-Hispanic applicants experienced significantly lower hiring rates than Hispanic applicants.

LandCare Outcome

In October 2016, LandCare entered into a conciliation agreement with the OFCCP and agreed to distribute $161,899 and extend 29 job offers to all eligible non-Hispanic class members (i.e., Whites, African-Americans, and Asians).
Discussion – Apples-to-Oranges Comparisons

On the surface, both LandCare and VF appear to be parallel but with seemingly different outcomes. In reality, the method and approach in establishing a case of AI discrimination between the two cases are very different.

VF – OFCCP’s AI Analysis Strategy

In the VF matter, the OFCCP was solely focused on AI against non-Asians and instructed their expert, Dr. David L. Crawford, to focus only on Asians versus non-Asians. In reading Judge Krantz’s opinion, it is clear that he wanted to know if Dr. Crawford investigated for AI in individual race group comparisons (e.g., Asians vs Whites, Asians vs Hispanics). Dr. Crawford did not investigate for AI in individual race group comparisons.

LandCare – OFCCP’s AI Analysis Strategy

In the LandCare matter, the OFCCP wanted to conclude that there was AI discrimination against non-Hispanics, but analytically, they adapted their AI analysis strategy to reflect the lessons learned from the VF matter. In addition to investigating for AI in a Hispanics versus non-Hispanics comparison, the OFCCP analyzed and found AI in individual race group comparisons (Hispanics vs Whites, Hispanics vs African Americans, Hispanics vs Asians). Rather than solely relying on one analysis (Hispanics vs non-Hispanics), the OFCCP established that Hispanics were hired at significantly higher rates than Whites, African Americans, and Asians in three additional separate analyses.

Apples-to-Oranges Comparisons

By analyzing individual race group comparisons in the LandCare matter, the OFCCP successfully addressed the major weaknesses that Judge Krantz had noted in the VF matter. This major shift in analytical strategy between VF and LandCare clearly shows that the two audits do not represent a true apples-to-apples comparison. Returning to the original questions proposed earlier:

Q1: Is the LandCare outcome a reversal of the ruling on the VF Jeanswear case?

A1: No. The outcomes of the two audits are based on two significantly different analytical strategies. In the VF matter, Judge Krantz identified deficiencies in OFCCP’s analytical strategy and body of evidence. In contrast, in the LandCare audit, the OFCCP changed their analytical strategy to address all the deficiencies Judge Krantz noted in VF.

Q2: Does the LandCare outcome allow for aggregation of other race groups outside of the more familiar and traditional minority groups?

A2: No. Individual race group analyses are needed to formally establish an AI theory that there is a pattern of favoring one race group versus all others.

Q3: How did the OFCCP successfully obtain a finding of AI against non-Hispanics in LandCare? On a related note, could the OFCCP have applied a similar strategy with equally successful outcomes in the VF matter?

A3a: The OFCCP was successful in the LandCare matter because they analyzed and established AI in individual race group comparisons (Hispanics vs Whites, Hispanics vs African Americans, Hispanics vs Asians). In doing so, the OFCCP successfully established a consistent pattern of AI where Hispanics were favored and non-Hispanics were negatively impacted.

A3b: Whether or not the OFCCP could have successfully applied the analytical strategy in the LandCare audit in the VF matter is largely speculative since the detailed counts were not supplied in the VF matter. Based on the numbers that were supplied in the VF decision, it is clear that both statistical power and magnitude of effect were in favor of supporting OFCCP’s claims; the OFCCP most likely would have prevailed in VF if they had provided individual race group analyses.
Conclusion

Comparing and contrasting the VF and LandCare audits provides invaluable insight into both the OFCCP and best practice strategies in race-based AI investigations. First, while many experts and attorneys make light of the OFCCP’s technical expertise, the adjustments between the two audits clearly show that the OFCCP is a very capable enforcement agency that learns and adapts, and wins. Second, federal contractors should be analyzing AI among individual race groups. If there is a pattern where one race group is consistently advantaged, then that becomes an added concern that AI exists among aggregated race groups.