Theory of justice

Theory of justice - is a systematic account of the principles that should govern how a society distributes rights, liberties, opportunities, income and wealth among its members. In economics, management and political philosophy the phrase usually points to one specific work: John Rawls's A Theory of Justice (1971, revised edition 1999), which opens with the sentence "Justice is the first virtue of social institutions, as truth is of systems of thought"^[1]. Rawls called his conception justice as fairness.

The book did something unusual. It took a question philosophers had argued about since Plato and turned it into a choice problem under uncertainty: pick the rules of society without knowing where you'll land in it. Rawls's answer (two principles of justice, ranked in strict order) became the reference point for half a century of debate on taxation, welfare policy and, more loosely, fair pay inside organisations. Critics came from every direction.

Background and history

Rawls (1921-2002) taught at Harvard for most of his career. The core idea appeared well before the book, in the 1958 paper "Justice as Fairness", where he already treated principles of justice as those that free and rational persons would accept from a position of equality^[2]. A Theory of Justice came out in 1971 and ran to some 600 pages. A revised text, which Rawls had prepared earlier for foreign translations of the book, was published in English in 1999, and later statements (Political Liberalism in 1993, Justice as Fairness: A Restatement in 2001) reworked parts of the argument^[3].

The book was written against a background dominated by utilitarianism, the view that the right institutions are those that produce the greatest total (or average) welfare. Rawls's complaint was blunt. Utilitarianism "does not take seriously the distinction between persons": it would accept serious losses for a few if the gains to the many were larger. For Rawls each person holds an inviolability founded on justice which even the welfare of society as a whole cannot override^[1].

The subject of the theory is the basic structure of society, meaning the main political, legal, economic and social institutions (the constitution, property rules, the organisation of the economy, the family) taken together. That point matters for business readers. Rawls did not offer rules for how a single firm should set wages; he offered principles for the framework within which firms, markets and households operate.

The original position and the veil of ignorance

How do you find fair principles when everyone has a stake in the result? Rawls's device is the original position, a hypothetical meeting in which representatives choose the principles that will govern the basic structure. They choose behind a veil of ignorance: they don't know their class, race, gender, natural talents, strength, intelligence, or even their own idea of a good life. They don't know which generation they belong to^[1]^[3].

The aim is simple. If nobody knows whether they'll be rich or poor, talented or not, nobody can tailor the rules to their own advantage. Whatever comes out of the procedure is fair because the procedure itself is fair. That is where the label "justice as fairness" comes from (pure procedural justice, in Rawls's vocabulary).

The parties are assumed to be rational and mutually disinterested. They aren't envious, and they aren't altruistic either. They simply want as large a share as possible of what Rawls calls primary goods: things any rational person wants whatever else they want. The list includes:

basic rights and liberties,
freedom of movement and free choice of occupation,
powers and prerogatives of offices and positions of responsibility,
income and wealth,
the social bases of self-respect.

Primary goods give the theory a common yardstick: holdings of goods rather than unobservable happiness.

The two principles of justice

Rawls argued that the parties would agree on two principles. In his final formulation they read roughly as follows^[3]:

First principle (equal basic liberties) - each person has the same claim to a fully adequate scheme of equal basic liberties, compatible with the same scheme for all. These include freedom of thought and conscience, political liberties, freedom of association, the rights that protect the integrity of the person, and the rule of law.
Second principle - social and economic inequalities must satisfy two conditions:
- (a) fair equality of opportunity: they are attached to offices and positions open to all under conditions of fair equality of opportunity, which means that people with similar talents and willingness to use them should have similar life chances regardless of the class they were born into;
- (b) the difference principle: they are to the greatest benefit of the least advantaged members of society.

Fair equality of opportunity is stronger than the formal rule that careers are open to talents. Banning discrimination in hiring is necessary but not enough; a society also has to correct for unequal access to education and training.

The difference principle is the most discussed part of the book, and the most misread. It doesn't demand equality. Inequality is permitted, even welcomed, when it works for the worst-off. If higher pay for surgeons or entrepreneurs draws out effort and skill that raise the incomes of the poorest group, those differences are just. If not, they aren't.

Lexical priority

The principles are ranked in lexical (or lexicographic) order. The first principle comes before the second, and fair equality of opportunity comes before the difference principle. A basic liberty can't be traded away for economic gain, however large. Think of a dictionary: the second letter matters only when first letters tie. Rawls admitted the priority holds only once a society has reached a reasonable level of material development.

Maximin reasoning

Why would people behind the veil choose these principles over, say, average utilitarianism? Rawls's argument draws on the maximin criterion from decision theory: rank alternatives by their worst outcome and choose the one whose worst outcome is best. He claimed that maximin is a sensible rule when three conditions hold together: the chooser has no reliable basis for estimating probabilities; the chooser cares relatively little about gains above a guaranteed minimum; and the rejected options carry outcomes one could hardly accept^[1]. The original position, he argued, has all three features. A choice of principles is made once, for a whole life, and for one's descendants. Gambling seems reckless.

Maximin is the pessimist's rule for decisions under uncertainty and risk, set against the expected-value logic of expected utility theory.

Example of utilitarian and Rawlsian choice

Consider four hypothetical societies, each made up of five equally sized income groups. Figures are annual incomes in thousands of euros for the representative member of each group. The numbers are invented for illustration.

Society	Group incomes (k€)	Total	Average	Minimum (worst-off group)	Gini coefficient
A	10, 20, 30, 40, 150	250	50.0	10	0.480
B	25, 30, 35, 40, 60	190	38.0	25	0.168
C	22, 28, 40, 55, 100	245	49.0	22	0.299
D	28, 32, 38, 50, 90	238	47.6	28	0.239

Now apply three criteria.

Classical or average utilitarianism (treating income as a stand-in for welfare) ranks the societies by total or average: A (250) > C (245) > D (238) > B (190). Society A wins, even though its poorest group earns only 10 thousand, a fifteenth of the top group.

Rawlsian maximin looks only at the worst-off group: D (28) > B (25) > C (22) > A (10). Society D wins.

Strict equality, judged here by the lowest Gini coefficient, would pick B (0.168).

Look at D against B. D is clearly more unequal (Gini 0.239 against 0.168; the top group earns 90 rather than 60). Yet the poorest are better off in D, so the difference principle prefers it. That's exactly the point Rawls made: his theory isn't egalitarian in the simple sense. Inequality that raises the floor passes the test.

If welfare grows with the logarithm of income (a crude way to model the law of diminishing marginal utility), the sums of log-incomes come out at about 17.40 for A, 17.96 for B, 18.72 for C and 18.85 for D. D still wins, C comes second, and A falls to last. Diminishing marginal utility pushes a utilitarian toward Rawls's ranking without fully reaching it.

A caveat: Rawls measured the position of the least advantaged in primary goods across a whole life, not in a single year's income, and his "groups" are social positions, not individuals. The table is a teaching device, not a model of his full theory.

Critics and rival theories

Nozick and the entitlement theory

Robert Nozick, Rawls's Harvard colleague, answered with Anarchy, State, and Utopia in 1974^[4]. His entitlement theory holds that a distribution is just if it arose from just acquisition and just transfer of holdings, with rectification of past injustice. Justice is historical, not a matter of the shape of the outcome. Nozick attacked "patterned" principles such as the difference principle with his famous basketball example: if a million fans each freely pay to watch Wilt Chamberlain play, the resulting inequality came about through voluntary exchange, so on what grounds is it unjust? Keeping any pattern in place, he argued, requires continual interference with people's free choices. Taxation of earnings, on this view, is morally close to forced labour. His minimal state sits close to the ideal of laissez-faire government.

Harsanyi's utilitarian critique

John C. Harsanyi, later a Nobel laureate for his work in game theory, had already argued in the 1950s that a rational person choosing behind a veil of ignorance, with equal chances of being anyone, would maximise expected utility and so pick average utilitarianism. In a 1975 paper in the American Political Science Review he attacked the maximin rule directly^[5]. Applied consistently, maximin gives absurd advice: it tells people to refuse any action that carries even a tiny probability of disaster, regardless of the benefits. It also gives absolute weight to the worst-off group even when a very small gain for them costs everyone else a great deal. The debate between maximin and expected utility never really closed.

Sen and the capability approach

Amartya Sen accepted much of Rawls's spirit but questioned the metric. Primary goods, he argued, measure means, not what people can actually do and be. A person with a disability, a pregnant woman and a healthy young man may need very different amounts of income to reach the same level of mobility or nutrition. Sen's capability approach shifts attention to the real freedoms (capabilities) people have to achieve valuable functionings. In The Idea of Justice (2009) he went further, criticising what he called "transcendental institutionalism", the search for perfectly just institutions, and arguing for comparisons that help remove manifest injustices in the world as it is^[6]. The capability idea fed directly into the Human Development Index.

Communitarian critique: Sandel

Michael Sandel's Liberalism and the Limits of Justice (1982) targeted the person behind the veil^[7]. Rawls's parties are "unencumbered selves", stripped of the communities, loyalties and shared purposes that actually make people who they are. Sandel argued that such a self can't ground the solidarity the difference principle demands: why would I share my talents' rewards with strangers unless we already belong to some common enterprise?

Later Rawls: Political Liberalism

Rawls's own rethinking was driven less by these critics than by a problem he found in his book's account of stability. Free institutions inevitably produce reasonable pluralism: citizens hold different religious, moral and philosophical doctrines, all reasonable. A theory of justice that depends on one all-embracing moral doctrine can't win the support of all of them. In Political Liberalism (1993) Rawls recast justice as fairness as a "political" conception, one that citizens with different doctrines can endorse from their own perspectives in an overlapping consensus, and introduced the idea of public reason^[8]. The two principles survived; the philosophical justification changed.

Applications in economics and management

Distributive policy and welfare economics

In welfare economics, the "Rawlsian social welfare function" is shorthand for a function equal to the welfare of the worst-off person. It sits at one end of a spectrum whose other end is the utilitarian sum. Policy analysts use it as a benchmark: how would this reform look if we cared only about the bottom? Rawls himself argued that neither laissez-faire capitalism nor the welfare-state version of capitalism met his principles. He favoured a "property-owning democracy" (wide dispersal of capital and education) or liberal socialism^[3].

Taxation

Tax debates borrow the vocabulary freely. Traditional taxation principles and horizontal equity say that people in similar situations should pay similar amounts. Rawls adds a vertical question: is the tax-and-transfer system as a whole arranged so that the least advantaged do as well as possible? In optimal tax models a maximin objective implies setting top rates where they raise the most revenue for transfers, rather than where they maximise total output. Nozick's view points the other way, treating progressive taxes on earnings as an encroachment on entitlements.

Pay structures

Inside firms the difference principle works as an analogy, not a rule; Rawls never claimed it should set individual wages. Still, managers and compensation committees use Rawlsian language when asking whether pay differentials can be justified to the lowest-paid staff. Do higher rewards for executives or star performers create value that raises everyone's position (job security, wages, training), or do they simply capture it? Pay for performance and merit pay schemes can be defended this way, if the incentives really do enlarge the pool. Fair equality of opportunity translates more naturally: transparent promotion rules, access to development programmes, and attention to barriers that keep capable people from advancing.

Organizational justice

The research field of organizational justice studies how employees perceive fairness at work: distributive (outcomes), procedural, interpersonal and informational. It's an empirical, psychological literature and should not be confused with Rawls's normative theory, though the two share terms. Rawls's emphasis on fair procedure has an echo in work on procedural fairness, and the veil of ignorance is sometimes used as a practical heuristic in ethical decision making: design the policy as if you didn't know which employee you would be. The same thinking appears in debates on corporate social responsibility and stakeholder treatment.

Advantages of the theory of justice

Gives a clear criterion (look at the worst-off position) that can be applied to policies.
Protects basic liberties against trade-offs with economic gain through lexical priority.
Allows inequality where it pays its way, which fits real economies better than strict equality.
The veil of ignorance is a simple, portable tool for impartial reasoning in public policy and in organisations.
Takes the separateness of persons seriously.
Primary goods offer a measurable, public metric instead of unobservable happiness.

Limitations of the theory of justice

Maximin is extremely risk-averse; critics such as Harsanyi argue that rational choosers would maximise expected utility instead.
Focusing on the single worst-off group can ignore large losses for everyone else.
Primary goods ignore differences in people's ability to convert resources into well-being (Sen's objection).
The entitlement critique: a pattern-based principle may require continual interference with voluntary exchange.
Identifying "the least advantaged" in practice is contested: by income, wealth, health, or a mix?
The theory applies to the basic structure of a society; transferring it to firms, pay systems or individual decisions involves interpretation Rawls didn't provide.
Ideal theory: it assumes broad compliance and gives limited guidance for deeply unjust, real-world conditions.

Related approaches

Prioritarianism - extra weight for gains to the worse-off, without the absolute priority of maximin.
Luck egalitarianism - neutralise inequalities due to brute luck but not those due to choice.
Social exchange theory - an empirical account of reciprocity used in workplace fairness research.

References

Harsanyi J.C. (1975), Can the Maximin Principle Serve as a Basis for Morality? A Critique of John Rawls's Theory, American Political Science Review, Vol. 69, No. 2, pp. 594-606
Nozick R. (1974), Anarchy, State, and Utopia, Basic Books, New York
Rawls J. (1958), Justice as Fairness, The Philosophical Review, Vol. 67, No. 2, pp. 164-194
Rawls J. (1999), A Theory of Justice, revised edition, Belknap Press of Harvard University Press, Cambridge MA (first edition 1971)
Rawls J. (2005), Political Liberalism, expanded edition, Columbia University Press, New York (first edition 1993)
Sandel M.J. (1998), Liberalism and the Limits of Justice, 2nd edition, Cambridge University Press, Cambridge (first edition 1982)
Sen A. (2009), The Idea of Justice, Belknap Press of Harvard University Press, Cambridge MA
Wenar L. (2025), John Rawls, Stanford Encyclopedia of Philosophy, Stanford University

Footnotes

↑ Rawls J., A Theory of Justice, revised edition, Harvard University Press, 1999, p.3
↑ Rawls J., Justice as Fairness, The Philosophical Review, 1958
↑ Wenar L., John Rawls, Stanford Encyclopedia of Philosophy, revised 2025
↑ Nozick R., Anarchy, State, and Utopia, Basic Books, 1974
↑ Harsanyi J.C., Can the Maximin Principle Serve as a Basis for Morality?, American Political Science Review, 1975, pp. 594-606
↑ Sen A., The Idea of Justice, Harvard University Press, 2009
↑ Sandel M.J., Liberalism and the Limits of Justice, Cambridge University Press, 1982 (2nd edition 1998)
↑ Rawls J., Political Liberalism, Columbia University Press, 1993

Theory of justice — recommended articles

Organizational justice — Maximin criterion — Amartya Sen — Expected utility theory — Horizontal equity — Procedural fairness — Ethical decision making