There's a liberal story that limited opportunities, and barriers, lead to employment problems and criminal records, but then there's another story that has to do with norms, behaviors, and oppositional culture. You can't prove the latter statistically, but it still might be true.' Holder thinks that both arguments contain truth and that one doesn't preclude the other. Fair enough. Suffice it to say, though, that the evidence supporting structural inequality is compelling. In 2001, a researcher sent out black and white job applicants in Milwaukee, randomly assigning them a criminal record. The researcher concluded that a white man with a criminal record had about the same chance of getting a job as a black man without one. Three years later, researchers produced the same results in New York under more rigorous conditions.