Most people interested in thinking about inequality, will have come across the dry and sarcastic saying from Anatole France that “In its majestic equality, the law forbids rich and poor alike to sleep under bridges, beg in the streets and steal loaves of bread.” France shows us how the reality of class, economic inequality, makes a mockery of equality before the law because of the different real-life effects of the same law.
Now why is sleeping under bridges bad? Well, presumably, it is bad because it makes you more vulnerable to certain harms: exposure to the cold, or being beaten up by gangs of strangers. We can imagine a kind of objection to Anatole France, albeit a very obviously silly one. It would go like this: the thing we should really be concerned about is the harms that people are exposed to. And when we investigate we find no difference in those harms between rich people who sleep under bridges and poor people who sleep under bridges. (We can assume that a few rich people, inebriated after a night at their club, end up under bridges too). According to the silly objection, what we should concentrate on is the group of people who sleep under bridges: there’s a perfect match between membership of this group and those who suffer the harms, whereas it turns out that lots of poor people, because they never sleep under bridges, are not at risk of such harms.
It is a silly objection, and obviously so. And yet we come across something very similar in form in many arguments about race and class. There are harms reliably associated with low socio-economic status and those harms fall on people regardless of their race. Kerching! – it is claimed – race doesn’t matter in the explanation of those harms! But obviously, if being black increases your relative propensity of being sorted into a poor working-class group that is exposed to such harms, and if being white reduces your relative propensity of being so sorted, then race is actually a big part of the picture. Showing that, of those who are in a category that is strongly pre-selected for by race, harms were not associated with race, does not lead to the valid conclusion that those harms are not associated with race.