Sen vs Rawls: comparing real options or designing the just society?

Amartya Sen dedicated The Idea of Justice to the memory of John Rawls, and then spent much of the book arguing with him. The argument between the two is the most useful debate in modern political philosophy for anyone who wants to govern a divided European democracy, because each of them answers a question the other leaves open.

What Rawls built

Rawls's A Theory of Justice (1971) asked what principles free and equal people would choose if they did not know their own place in society. Behind this veil of ignorance, he argued, they would pick two principles. The first guarantees equal basic liberties, and it takes priority. The second allows social and economic inequalities only if positions are open to all under fair equality of opportunity and if the inequalities benefit the least advantaged, the famous difference principle. Justice is measured by the distribution of primary goods: rights and liberties, freedom of movement and occupation, income and wealth, and the social bases of self-respect.

Political Liberalism, published by Columbia University Press in 1993 and expanded in 2005, takes on a harder problem. In a free society people hold conflicting religious, moral and philosophical worldviews, and they will keep doing so. Rawls calls this reasonable pluralism. Yet they all live under one set of laws. How can that be legitimate and stable? His answer has three moves, summarised in the Stanford Encyclopedia of Philosophy. Political power is legitimate when it is exercised under a constitution that all citizens, as free and equal, could reasonably be expected to endorse. The political conception of justice is freestanding, built from ideas already present in a democracy's public culture, so it needs no single worldview behind it. And citizens can reach an overlapping consensus, each endorsing the same principles for reasons drawn from their own beliefs. Officials and voters then owe each other a duty of civility: to justify decisions on fundamental questions with reasons others can share, which Rawls calls public reason.

What Sen objects to

Sen calls Rawls's project transcendental institutionalism. It looks for the principles and institutions of a perfectly just society. Sen raises two objections, as Charles Barclay Roger lays out in openDemocracy.

The first is about feasibility. Sen doubts that the original position would yield Rawls's two principles and no others. His example is three children and a flute: one plays it best, one is the poorest, one made it. Utility, equity and entitlement each support a different child, and each case can be made impartially. Roger points out that by the same logic a utilitarian principle of the kind John Harsanyi defended could be chosen over Rawls's maximin for distributive questions. If impartial reasoning can produce several defensible answers, the step-by-step route from the original position to perfect institutions breaks.

The second objection is about redundancy. A description of the perfect society does not tell us how to rank two imperfect ones, and we can rank them without it. Sen also faults Rawls for relying too heavily on institutions and underestimating how hard it is to get people to follow just norms. He wants to judge the lives people actually lead, measured by their capabilities, the real freedoms they have to be and do what they value. Rawls's primary goods look at the means; Sen looks at what people can do with them, which varies enormously with health, disability and surroundings.

Rawls answers back, through his readers

Laura Valentini's 2011 reply in Economics and Philosophy is the best-known defence of Rawls. She argues that Sen's call for a paradigm shift is unwarranted: his criticisms either misread Rawls or are accurate but of little consequence, and the Rawlsian approach already delivers much of what Sen wants. Rawls himself distinguished ideal theory, which assumes compliance and favourable conditions, from non-ideal theory, which deals with injustice and hardship by reference to the ideal. An ideal, on this reading, tells reformers which direction counts as improvement.

There is a quieter point of convergence too. Sen's plural grounding, the idea that people with different values can agree on comparative judgements without agreeing on all their reasons, looks very like Rawls's overlapping consensus. The two men share more than their debate suggests.

How a radical centre uses both

Each philosopher is strongest on a different question. Rawls is the better guide to legitimacy: how a Spain or a European Union of people who disagree deeply about religion, nation and the good life can share rules that all of them can accept. Sen is the better guide to policy: given the rules, which reform should we pick from the options on the table this year?

That division of labour gives a position with a piece from each tradition. From the left, Rawls's difference principle sets the test for any reform that widens inequality: it must leave the least advantaged better off. From the liberal tradition, political liberalism asks the state to stay neutral between ways of life and let each citizen pursue their own good, which rules out a government that picks one worldview and legislates it. And from neither camp comes the shared insight of plural grounding and overlapping consensus: citizens can agree that one option is better than another without agreeing on why.

In practice this changes how a coalition argues. Partners from different traditions stop trying to win the argument about first principles, which they never will. They agree on the constitutional rules of the game, Rawls's territory, and then compare concrete alternatives on evidence, Sen's territory, checking each against its effect on those with least. A Christian Democrat, a social democrat and a liberal can all back a reform for their own reasons. That is enough. The radical centre asks for agreement on the comparison and leaves each person their own reasons.