> > Standard wage equations, in which pay is dependent on an array of
>> attributes, like age, experience, occupation, education, etc.
>
>... but not productivity?
Try measuring it. Maybe you & Kelley can collaborate and split a Nobel.
Economists who deny discrimination usually explain statistical evidence of it away by citing "unobservable characteristics" - like productivity (and never the boss's own prejudices, of course). If you take a large pool of workers with roughly equal education and experience, it's likely that their average productivity is very similar.
Talk to black or women workers and they'll tell often you about crummier white male workers getting better pay and promotions. Of course, that's anecdotal evidence.
Doug