If you’re examining your selection decisions using the Four-Fifths Rule (aka the 80% Rule), it is arbitrary.
But it’s not your fault – it’s the way the Four-Fifths Rule was designed back in the 1970s.
The Four-Fifths Rule was originally framed in 1971 by the Technical Advisory Committee on Testing (TACT). TACT was assembled by the State of California Fair Employment Practice Commission (FEPC) and consisted of 32 professionals from various labor, employment and technical fields. TACT was tasked with outlining the specific methodology for evaluating adverse impact. The Four-Fifths Rule was the methodology they proposed.
But to see why its design is inherently arbitrary, we have to look deeper at the origin of the Four-Fifths Rule. Here is one committee member’s description of how the rule came to be:
During the negotiations of the FEPC Guidelines (which went on for months), one session had a significant debate on an appropriate statistical tool for determining adverse impact. We wanted to put an operational definition to some words defining what constituted adverse impact. There were about twenty of the committee members in the room. The members agreed that a statistical test was appropriate, but not enough. They also agreed that those who would implement these guidelines (the FEPC consultants) would never have the appropriate training to implement statistical tests (prior to the common use of computers, calculating probability statistics was a difficult task only completed by the technically savvy). Therefore, we needed an administrative guideline as well as a technical one for cases. I recall a heated debate that went on for way too long (as usual) with two camps: a 70% camp and a 90% camp. The 80% Test was born out of two compromises: (1) a desire expressed by those writing and having input into the Guidelines to include a statistical test as the primary step but knowing from an administrative point of view a statistical test was not possible for the FEPC consultants who had to work the enforcement of the Guidelines, and (2) a way to split the middle between two camps, the 70% camp and the 90% camp. A way was found to use both. In the way the 80% Test was defined by TACT, if there was no violation of the 80% Test, then there would be no reason to apply statistical significance tests. This hopefully would eliminate many calculations and many situations where TACT would not be necessary and the decision could be made in the field. So from a practical point of view, the 80% Test became a first step. If there was no 80% Test violation, there was no need to go further and use a statistical test. If there was a violation of the 80% Test, statistical significance was needed and the 80% Test then became a practical significance test for adverse impact.
So there’s no scientific reason behind the selection of 80% as the cutoff point. It was simply a way of splitting the baby in two. If either the 70% camp or the 90% camp had argued longer and/or more strongly for their cutoff preference, we might have the “Seven-Tenths Rule” or the “90 Percent Rule”.
Next time you’re looking at whether your selection decisions have an adverse impact, think about the origins of the Four-Fifths Rule. It may be the “predominant method” for looking at adverse impact, but it’s not the preferred method. There are several statistical tests that are easy to calculate, generate scientifically-based results, and take the arbitrariness out of determining whether adverse impact exists.