Overview
Not long after the passage of the 1964 US Civil Rights Act (now commonly known as Title VII), the legislative and judicial fields began hammering out the concept of adverse impact. Nearly 50 years later, after thousands of cases and arbitrations and well over one billion dollars spent by employers, government enforcement agencies, special interest plaintiff groups, and law firms, the concept has been highly refined. It has also expanded to apply to settings other than that for which it was first designed (e.g., some US circuit courts have recently approved of using adverse impact calculations for age discrimination cases).
While the courts still struggle for a definitive explanation of what constitutes a “finding of adverse impact,” the term as used today essentially means the same as when it was first written: a substantially different rate of selection in hiring, promotion or other employment decision which works to the disadvantage of members of a race, sex or ethnic group (Uniform Guidelines Questions & Answers #10). The three most common methods for determining adverse impact are the 80% Rule, statistical significance tests, and practical significance tests. Each will be discussed in this Chapter.
Why does adverse impact occur?
Virtually every public sector employer has one or more entry-level positions with adverse impact in the testing process. Written tests typically have the highest degree of adverse impact, with the highest level of impact against blacks, then Hispanics, and sometimes Asians (Sackett, 2001; Neisser, 1996). Physical ability tests typically have adverse impact against women, especially when they measure upper body strength. Robert Guion (1998, p. 445) cites several reasons why adverse impact can occur:

1. Chance;
2. Measurement problems inherent to the test (e.g., poor reliability);
3. The nature of test use (e.g., ranking versus pass/fail);
4. Differences in distribution sizes (e.g., a selection process with 100 men and only 10 women);
5. Reliable subgroup differences in general approaches to test taking; and,
6. True population differences in distributions of the trait being measured.

Test bias is another possible reason why adverse impact can occur, but this can only be a valid reason if one or more of the first five reasons above exist and the sixth is rejected. Genuine discriminatory intent and actions can be yet another reason why adverse impact can occur in a selection process.
Adverse impact has become a loaded term, fraught with suggestions of ill intent on the part of the employer. It should be noted, however, that adverse impact simply describes differences between groups on a testing process. It is not a legal term that implies guilt, nor is it a psychometric term that implies unfairness or test bias. Virtually every employer that tests for relevant job skills will generate adverse impact in a testing process in one way or another, and most studies show that adverse impact is not normally due to forms of bias inherent to the tests (Sackett, 2001; Neisser, 1996).
History and development
The US government treatise that first mentions the concept of adverse impact was the Equal Employment Opportunity Commission (EEOC) Guidelines on Employment Testing Procedures (issued on August 24, 1966). This document, however, offered no indication as to how to determine (by calculation or otherwise) whether adverse impact existed in an employer’s selection practices. The government had defined some of the “what” but not the “how” of adverse impact. This left the higher circuit or supreme courts with the burden of using key legal cases to make judicial findings that provided insight into what is really meant by adverse impact.
The first major adverse impact case was the infamous Griggs v. Duke Power Company (1971). Duke Power was using a high-school diploma requirement and an off-the-shelf intelligence test as a screening device, both of which had adverse impact against blacks. Since the jobs being tested did not appear to really require a high school diploma to be performed successfully, the court held that the employer had to show a “business necessity” for these two requirements; otherwise, Duke Power would be in violation of Title VII. Responding to the adverse impact that these two requirements had on blacks, the court only stated a few words: “. . . they operated to disqualify blacks at a substantially higher rate than white applicants.” Once again, the EEO field had only a “what” but not a “how” regarding adverse impact.
Exactly what is a “substantially higher rate?” Wanting to answer this question (and many others pertaining to personnel testing) as well as build a set of criteria that employers could use for determining exactly how “adverse” a testing process needed to be to represent “adverse impact,” an advisory committee called the Technical Advisory Committee on Testing (TACT) was assembled by the State of California Fair Employment Practice Commission (FEPC) in 1971. TACT was charged with compiling the State of California Guidelines on Employee Selection Procedures (which were published in final form in October, 1972). These California Guidelines were designed to “supersede and enlarge upon” (Section 3 preamble) the earlier set of Guidelines on Employment Testing Procedures, issued by the U.S. EEOC on August 24, 1966. These California Guidelines were later incorporated into the Federal Uniform Guidelines on Employee Selection Procedures (1978), a document still in force at the time of this writing.
TACT included 32 specialists from various labor, employment, and technical fields, who deliberated the specific techniques and steps that would be taken to evaluate adverse impact. It was out of these deliberations that the (now infamous) 80% Rule was born (in short, the 80% Rule is calculated by dividing the focal group’s (the focal group is typically minorities or women) passing rate on a selection procedure by the reference group’s (typically whites or men) passing rate, and any value less than 80% is said to violate a “threshold test” for evaluating adverse impact).