Page 377 - Discrimination at Work The Psychological and Organizational Bases

344
PAETZOLD
TABLE 14.1
Disparate Impact With Net Benefits to the Protected Class
Word-of-Mouth Recruiting
Pass Don't Pass
Pass Black 49 Black 1
White 70 White 10
Tests
Don't Pass Black 0 Black 50
White 0 White 20
Adapted from Paetzold, R. L., & Willborn, S. L. (1996). Deconstructing
disparate impact: A view of the model through new lenses. North Carolina
Law Review, 74,325-399.
persons, thus leading to a selection ratio of .70 (which is less than .80).
Thus, an adverse impact exists on Black persons, and the one Black person
who didn't "pass" through the word-of-mouth recruiting would be enti­
tled to relief. This is true even though the net affect of the word-of-mouth
recruiting operates to favor Black people as a group. In other words, only
one of 100 Black persons is actually adversely affected by the selection
mechanism, compared to 10 of 100 White persons. In fact, the word-of-
mouth recruiting mechanism increases the proportion of Black persons
in the pool available for hire from .385 to .412 vis-a-vis White persons
(i.e., without the requirement, 50 Black and 80 White persons pass and are
available for hire, but with the requirement, 49 Black and 70 White per­
sons pass and are available for hire). Thus, as a group Black people should
prefer that the word-of-mouth recruiting be left in place, but the one Black
person who is disadvantaged would be motivated to bring (and could
win, if no business necessity exists) a disparate impact lawsuit (Paetzold
& Willborn, 1996, p. 362). The example demonstrates the individual nature
of the disparate impact model, which is counter to traditional I/O or HR
understanding of the model, e.g., "[D]isparate impact discrimination oc­
curs when an apparently neutral employment practice disproportionately
excludes a protected group from employment opportunities" (DeNisi & Grif­
fin, 2001, p. 74, emphasis added); "[A]dverse impact... occurs when the
same standard is applied to all applicants or employees, but that standard
affects a protected class more negatively (adversely)" (Gomez-Mejia, Balkin, &
Cardy 1998, p. 91, emphasis added). It is certain members of a group who
may suffer an adverse impact; the group may not suffer at all.
This line of argument suggests that employers should think carefully
about the potential redundancy of their employment screens—not just in
terms of their validity for predicting performance (the typical I/O or HR
concern), but also in terms of their actual impact. An in-basket exercise,