# Disparate Impact **URL:** \ **Category:** Great Debates\ **Created:** [September 1, 2009, 3:03pm UTC](https://boards.straightdope.com/t/disparate-impact/508485 "2009-09-01T15:03:38Z")\ **Posts on this page:** 1\ **Page:** 6
**Author:** ![Chief\_Pedant](https://avatars.discourse-cdn.com/v4/letter/c/a6a055/32.png) [@Chief\_Pedant](https://boards.straightdope.com/u/Chief_Pedant)\ **Post date:** [September 3, 2009, 7:22pm UTC](https://boards.straightdope.com/t/disparate-impact/508485/101 "2009-09-03T19:22:44Z")
> [@brickbacon](#): > > First, you don’t understand the law in these cases if you are making this argument. It’s not a problem if there is disparate impact if the tests, etc. are required for the job. There would be no problem with prevent x group from doing y if y required something they are not capable of doing. The issue arises mostly when people create tests as a means of rationing rather than a way to evaluate competency. The New Haven test was thrown out in large part because experts testified to the bias in the test and the inclusion of questions that were not germane to the job. Others also stated that there were other tests that could have done a better job in both minimizing disparate impact, and evaluating candidates. After that testimony, they threw out the test. Just because Black firefighters did poorly in and of itself would not have been as problematic. The test has been posted online. I’d be curious about your opinion, and an example, of any question which could be construed as being favorable toward one racial background over another. Every test has criticizable features but surely the main consideration should be whether or not it favors any group somehow. While anyone can state that other tests would do a better job, it’s a hollow statement without concrete alternatives. In general what many testing companies have done is to substitute soft evalulations such as oral questions, or other non-quanitified metrics because of the persistent disparity on quantified, standardized tests. And while it may well be the case that such tests are inadequate, there remains the issue of explaining away why an inadequate test would produce disparate impact if no race-based biasing is present. An inadequate test is equally inadequate for all. Simply criticizing the test in general does not address the key issue. Moreover, the key sticking point of current disparate impact law is that the test for whether or not a test is biased is disparate outcome. Disparate outcome alone is sufficient to question the test itself. _“…an adverse impact if members of a protected class are selected at a rates less than four fifths (80 percent) of that of another group. For example, if 50 percent of white applicants receive a passing score on a test, but only 30 percent of African-Americans pass, the relevant ratio would be 30/50, or 60 percent, which would violate the 80 percent rule.”_[EEO Information and resources for HR Professionals and Students](http://www.hr-guide.com/data/G702.htm) Such a rule only makes sense if there is an a priori assumption that the cohorts being compared are equal in potential to pass the test. [Previous page](https://boards.straightdope.com/t/disparate-impact/508485.md?page=5)