It's kind of a sleight of hand to piggyback on the panic about income inequality but target your attack based on consumption. If consumption is what matters, we

It's kind of a sleight of hand to piggyback on the panic about income inequality but target your attack based on consumption. If consumption is what matters, we should also use it to measure the level of inequality. And when we do that, we find (as in your parent comment) that it's not nearly so dire. How much variation in outcome should we allow? Beating the median by a factor of 2 doesn't seem particularly crazy.