One plane crash is also n=1 for whether the maintenance interval of the element that failed should increase, not n=num_victims, since there's only one aircraft that failed after all. Here, it's one organisation

I'm not asking for data points to do a study into individual companies applying their policies and compiling anecdotal results to do statistical or qualitative tests on all by myself in spare time when there's people being paid to do this properly. I was wondering if someone knew if the evidence has already been compiled and we know the overall effect, rather than stories which are surely plenty on both sides (assuming we agree that racism and other biases are not legitimate selection reasons, thus there being a reason to push against them). The question is: where is the balance? What policy creates the greatest median happiness in the population, all else being equal? This isn't trivial to determine, I can't do that by myself in an afternoon based on a few news reports of individual cases