That was Keen's point about Pareto optimality, wasn't it? That this kind of efficiency doesn't favour free-market capitalism?

I think Chang was saying that pareto optimality is conservative or myopic compared to the more philosophical, evaluative utilitarianism, which brought to the fore questions about sacrificing some social classes for the sake of others.

Now, with Pareto optimality, the economist just has to think like a bureaucrat, bracketing questions of value, and taking the system as it is, merely tinkering the parameters so that no one is sacrificed directly for someone else's benefit. Instead, under capitalism, the sacrifice happens indirectly due to initial inequalities that make for different outcomes in the competition for resources and profits. It's just that economists blind themselves to that unfairness because considering it isn't scientific.