Statistical Analysis of Adverse Impact

1. Introduction

The idea of disparate impact in the employment context was first mentioned in 1966, shortly after the passage of the U.S. Civil Rights Act. Despite the fact that the concept of disparate impact has been highly refined and is now widely understood, employers are still faced with hundreds of disparate impact lawsuits every year. The reason for this is, I believe, twofold.

First, employers have traditionally approached discrimination prevention through training. Recent estimates indicate that companies in the United States spend nearly $300 million annually on diversity training programs. Employees, managers and supervisors spend countless hours in seminars discussing conflict resolution and communications skills, the importance of having a diverse and inclusive workplace, and so forth.

Disparate impact is not a problem that can be solved completely through diversity training or anti-discrimination seminars. Disparate impact is not intentional discrimination and does not arise out of personal interactions. Because disparate impact resides in the policies and practices of the organization, conventional training for supervisors and managers will not effectively prevent disparate impact.

This is not to say that diversity and anti-discrimination training is not necessary – it is very necessary, and can prevent occurrences of intentional discrimination. But anti-discrimination training on its own is not an effective prophylactic for disparate impact.

Second, employers may be unaware that one of their policies or practices is creating disparate impact until they are confronted with a disparate impact lawsuit. The presence of disparate impact is less obvious than other forms of discrimination. Disparate impact doesn’t manifest in the obvious ways that disparate treatment manifests. If employers are not examining their policies and practices with an eye toward proactive identification of potential problems, they may not learn of the disparate impact until it is too late.

Managing the risk of disparate impact litigation cannot effectively be done through traditional means. It requires a new approach built on proactive quantitative analysis.

This book was written with human resources professionals, in-house counsel, and employment lawyers in mind. The goal is to introduce the reader to statistical analysis of disparate impact through the use of concrete examples and simple calculations. In most cases, members of the intended audience will not be performing the statistical analysis of disparate impact themselves; rather, the calculations will be performed by a statistical consultant. Having a basic understanding of the statistical tools commonly used in disparate impact analysis will give human resources professionals and legal counsel a better understanding of the information their statistical consultants are providing and how to correct – and prevent – problem areas that may exist within the organization’s employment policies and practices.

The book is organized into three sections. Chapters Two through Five provide a discussion of the disparate impact framework and the various tools that are used to examine an employment policy or practice for disparate impact. Chapters Six through Ten present how those tools can be applied to various questions arising within the hiring process, workforce analysis, promotion process, and reduction in force planning. Chapters Eleven through Thirteen present a discussion of some current issues regarding disparate impact, and how to minimize the risk of employment-related litigation with proactive statistical analyses.