As expected, it looks like the Republicans are going to try to use Sonia Sotomayor's ruling in the case of the white firefighter in New Haven to throw whatever muck they can at her in her confirmation hearings next week:
Republicans will call two New Haven firefighters to testify in the confirmation hearings of Sonia Sotomayor next week, making clear the GOP's intent to place affirmative action at the center of Senate battle over Sotomayor's nomination.
A Judiciary Committee press release lists Frank Ricci and Ben Vargas as expected Republican witnesses. Ricci was the lead plaintiff in Ricci v. New Haven, the controversial case in which Sotomayor ruled the New Haven fire department acted constitutionally when it promoted black firefighters who scored lower than their white counterparts on a qualifying test. Vargas, who is Hispanic, was the only non-white co-plaintiff in the racially charged case.
The Supreme Court later overturned that decision.
Actually, I believe The Hill has that wrong. I believe New Haven simply threw out the results and didn't promote at all on the basis of the test. But no matter. I'm sure it's not the only inaccuracy that will be dragged into the fray.)
There's an angle on this story that I haven't seen followed up at all that I think is quite important. That's the question of the test itself that New Haven used. The law provides for accepting a discriminatory result if it can be proven that the tested requirements are actually essential to doing the job.
According to the Hartford Business Journal, the tests were designed by a company that specializes in testing for public safety professions, I/O Solutions.
The only way to decide whether the test really does test for requirements of the job or whether it (deliberately or accidentally) is selecting on a racial/cultural basis instead, one would have to examine the test itself and consult with a wide variety of people with expertise in that field to come to anything like an informed answer. Obviously, that's beyond the scope of what I can look at here. But a couple of things do jump out at me.
One, I/O Solutions may or may not base its tests on the core competencies for firefighting developed by the National Fire Protection Agency and the International Association of Fire Chiefs. (Here it is in Google Books.) If I knew that they were, that would give me more confidence that they were basing their testing on valid standards rather than on something else.
What they say instead is:
I/O Solutions works diligently to help modern departments and agencies meet the increasingly stringent demands of public safety personnel assessments. We are dedicated to developing hiring and promotional assessments of the highest possible quality in order to reduce adverse impact against minority groups and females, and increase levels of test validity and selection utility to previously unattainable levels.
They apparently have a series of promotional exams for firefighters, listed here.
Now the complaint from minority firefighters is that there's widespread discrimination against them in hiring and promotion. If so, they're being systematically prevented from reaching positions where they can set the standards by which hiring and promotion decisions are made. Which begs the question - are tests being designed to test for real needs of the job or to reflect the innate cultural preconceptions and prejudices of the current elite making the rules?
As Justice Ginsburg points out about the Supreme Court, cases turn out differently when the mix of judges is different, because "people relate to their own experiences."
According to at least some firefighters who've taken I/O Solutions tests, important portions of the tests are very subjective:
Has anybody here had any issues with the National Firefighter Selection Inventory written test? I recently took the test and was very disappointed with my results. The main issue I have is, how does one possibly calculate a percentage score for personality based questions? I'm even further confused by the NFSI test because their study guide specifically states that the personality questions "have no correct answer". If there is no correct answer, how can they score it?
The last version of this test I took had 105 general knowledge questions and 50 personality questions. I can only recall about 5 questions that were even remotely difficult to answer. Even if I assume I missed 10 general knowledge questions (which I know is not the case), that would still give me a 90%. My final score was 74%! That means my score on the personality portion was under 50%.
For the record, I took the practice exam and scored a 96%. Also, I have taken and passed the MMPI 2 personality test without any difficulty.
...
I have taken the io solutions test 4 times and have done a lot of research in to their testing methods. Here is where I disagree with their testing methods. Although they tell you there are no right or wrong answers on the personality portion, points are deducted from your final score based on their formula for scoring the questions. You could ace the general knowledge section and end up failing the test based on your answers to the personality questions. I was able to get this information from a city who allowed candidates the chance to review the test and dispute any questions. I only missed 2 questions on the general knowledge portion but my overall test score dropped to a 79% after the personality portion was figured in and you can't dispute those questions. This particular city told me to call Io Solutions to dispute the personality based questions. IO Solutions was very tight lipped and basically told me they could not give me that type of information so they could protect the integrity and validity of the test.
Granted, this is coming from anonymous blog commenters. But it does raise questions in my mind about the supposed strict professional objectivity claimed for the testing procedure at the heart of the discrimination case. If the above accusations about the tests are true, they throw a very different light on what exactly the New Haven fire department was really testing, and it suggests it was a correct decision to throw out promotional decisions based on them.