Chasing a single perfectly just institution is the wrong goal

Sen's central target is transcendental institutionalism, the approach — most associated with Rawls — of trying to identify the ideal set of just institutions a society should have, as if justice were a single correct answer waiting to be derived through careful reasoning. He argues this project, even when brilliantly executed, doesn't tell us what to do when comparing two real, imperfect options in front of us right now.

Knowing what perfect justice would look like, Sen points out, doesn't tell you which of two flawed, actually available policies is more just, any more than knowing the best possible painting tells you how to rank two mediocre ones. Ranking real alternatives is a distinct exercise from defining an ideal, and Sen thinks philosophy has neglected the former while over-investing in the latter.

He doesn't dismiss ideal theory as worthless, but insists it's insufficient for the practical work of actually reducing injustice in the world as it exists.

Takeaway: knowing what perfect justice looks like doesn't help you choose between two imperfect real-world options — and that choice is the one that actually matters.