1

The Making of Justice

WHEN JOHN RAWLS published A Theory of Justice in 1971, it made him the most famous political philosopher in the English-speaking world. Enormously detailed and painstakingly executed, it was, Rawls wrote, “a long book, not only in pages.”¹ Across six hundred pages and three parts, he worked out his vision of a just and “well-ordered society” regulated by the conception of justice he called “justice as fairness.”

The book was understood as a revival of the liberal social contract tradition, in part because of its most famous and evocative idea, the “original position.” There, “persons” meet behind a “veil of ignorance” that blinds them to their social characteristics—the things they have acquired because of social contingency and natural accident. Once behind the veil, they agree how a just society would be structured without letting these contingent facts get in the way of their choices. Rawls said they would choose two principles of justice as a set of standards for judging the justice of a society. The first was a principle of liberty, which affirmed citizens’ basic rights and liberties. The second was a principle of equality. It included the “difference principle,” which arranged social and economic inequalities so that they worked to the benefit of the least-advantaged members of society and stipulated that offices and positions must be open to all under conditions of “fair equality of opportunity.” Society was conceived as a “cooperative venture for mutual advantage.” One benefited more from being in it than outside it. The principles were there to make sure the advantages of membership were divided up in a fair and just way. They ensured that the things people had were not theirs because of luck, and that rewards for the efforts of individuals did not get in the way of social stability.

The second of five sons, two of whom died in childhood of diseases they contracted from him, Rawls knew a thing or two about luck.² He spent a lifetime working away relentlessly at a single theory that made him the most celebrated political philosopher of his generation, so he knew something about effort and reward too. After A Theory of Justice was published, Rawls’s ideas would be pulled in different philosophical and political directions. But he began to work on them long before, in a moment just after the Second World War when the contours of liberalism were being reconfigured.

It is often said that in the dry, dusty scene of midcentury analytical ethics, philosophers cared more about whether to cross the lawn of an Oxford college than politics. Yet when a young Rawls returned home from three years of service in the Pacific Theater to begin his graduate studies at Princeton in the spring of 1946, politics was inescapable.³ Rawls soon set about constructing a “social philosophy” to make sense of the era of total war and the world it created. Over the course of the 1940s and 1950s, the young Rawls would search widely across the human sciences for the conceptual tools to do so. A philosopher by training and by temperament, Rawls was also an intellectual magpie. Writing in the mid-1950s, he conceived of his aims as threefold. First, social philosophy, like sociology, economics, history, political theory, and jurisprudence, aimed to provide a “conception of society and the human person involving acceptance of certain general facts as true about society.” Second, like the rest of philosophy, it described a system of evaluative principles, in this case ethical ones. Third, it offered a vision, a “total picture of man and society,” Rawls wrote, borrowing from Joseph Schumpeter, that tied the whole together and made it “come alive for us.”⁴ Rawls’s search for these three pillars began within the ideological context of postwar liberalism.

By the time Rawls elaborated his mature system, it was so complex and introduced so many novel concepts that it was hard to see its original motivations. He began his philosophical career at a moment in the late 1940s when liberals were increasingly concerned with a defense of the freedom of individuals and a critique of the institution that, paradoxically, would be submerged in Rawls’s mature theory: the administrative state. In the aftermath of the New Deal and the Second World War, Rawls initially took on a barebones liberalism that tried to limit state intervention and planning. At the level of both argument and metaphor, he borrowed from various strands of anti-statist and pluralist liberalism. During the decade after 1945, Rawls’s conception of society and the person initially owed more to Wittgenstein, Hume, Tocqueville, and Hegel than to Kant, and it was concerned with the limitation of concentrations of power accumulated in the administrative state as much as with redistribution. His attempt to find procedures for ethical evaluation sprang from a set of worries—about how to limit the effects of prejudice and ideology and yet preserve the sanctity of individual judgments—that were common among midcentury moral and political philosophers. A Theory of Justice must thus be understood as a book of the postwar, not as a response to the years of the Great Society. Its preoccupations originate in these earlier years.

Rawls’s first efforts generated a number of lasting ideas. His explorations of how to set up a society in ways that made state intervention unnecessary led him, at the turn of the 1950s, to the notion of a “property-owning democracy” that later underpinned his ideal vision of society. A number of the ideas associated with the writings of the later Rawls began here too: his concern with consensus and deliberation and his quasi-Hegelian emphasis, which many later saw as underpinning his turn from Kant to Hegel.⁵ But in the first instance, Rawls left many of these early concerns behind. Over the next years, he developed the contours of his mature theory. He moved left in line with the debates about equality and social justice that preoccupied the revisionist wing of the British Labour Party, whose thought Rawls encountered during a year at Oxford in 1952. It was social democratic Britain as much as Cold War America that provided the political theories and orientation that shaped Rawls’s own, and that showed him the political work his ideas could do. In these years, as he broadened his political vision and expanded his toolkit to include not only welfare economics and ethics but sociology and moral psychology, Rawls developed the conceptual framework and terms of art that would become so influential: the original position, the basic structure, and his principles of justice. He was searching for the right forms of philosophical abstraction to give shape to the vision of person and society that he found in liberal ideas and institutions. By the end of the 1950s, the architecture of his theory was in place.

The liberalism of postwar America had a dual character. In the aftermath of the Depression, there had been widespread acceptance of the need for state control and intervention in social and economic life. The warfare state forged by the Second World War made many citizens see government as more legitimate than ever.⁶ After the war, a corporate liberalism—characterized by an openness to the state, opposition to radical labor, and a commitment to a corporative economy and “noncoercive” solutions—thrived at the highest levels of politics.⁷ Yet this liberalism also marked a retreat from the acceptance of planning and state intervention in the economy that characterized the First New Deal. The Keynesian consensus that underpinned the “growthmanship” of the 1950s circumscribed the role of the state to the task of stabilization, a move that to many marked the “end of reform.”⁸

This retreat from state-planned redistribution was in part a function of the rise of anti-totalitarianism and the redefinition of liberalism in opposition to the “totalitarian threat.”⁹ Fears about the collapse of liberal society into militarism proliferated, with worries that America was becoming what Harold Lasswell called a “garrison state.”¹⁰ These fears were invoked not only by those who worried about totalitarianism but by opponents and critics of the New Deal as well. Anti-statists described government’s role as limited to that of an “umpire,” or guardian of the “free enterprise system.” Business leaders claimed society was standing on the edge of a fateful line between capitalist freedom and statist slavery.¹¹ “Critical liberals” rejected the politics of planning and sought to check the power of the administrative state’s expert agencies with legal oversight and to roll back New Deal reforms.¹² Having for a time accommodated the labor politics unleashed by a decade of depression, reform, and war, many liberals were increasingly wary of labor radicalism.¹³ After the war emergency subsided, they deployed ideals of liberty, law, and the Constitution, as well as the Bill of Rights, or “constitution of rights,” against mass politics and executive power.¹⁴ When Friedrich Hayek published The Road to Serfdom in 1944 to widespread acclaim, it tapped into this persistent and “powerful strain of Jeffersonian anti-statism.”¹⁵

Among certain liberal thinkers in this moment, there was a growing effort to develop moral theories to judge the limits of state institutions.¹⁶ This marked a departure from the interwar years and the triumph of Progressivist theories. Then, many legal realists, sociologists, economists, and theorists of public administration who supported the New Deal state and economic planning had appealed to social and economic “facts” instead of moral rules and principles, separating value from fact, and ethics from science. Logical positivism and its offspring emotivism had rendered the study of substantive ethical questions nonsensical.¹⁷ But since then, critics of the administrative state had challenged “relativistic,” pragmatist, and “value-free” strains of American democratic theory for enabling totalitarianism.¹⁸ In the early 1940s, with anti-totalitarianism at fever pitch and even John Dewey attacked as a “threat to democracy,” many lawyers, philosophers, and political scientists claimed democratic morality as a bulwark against totalitarianism.¹⁹ Natural law theory underwent a revival.²⁰ A wariness of concentrated political and economic power—of the state, of corporations, of labor unions—returned to certain quarters of liberalism, which now elevated the individual and associational life. Some argued that a new constitutionalist order had to be built, by restoring what the Harvard liberal theorist Carl Friedrich called “faith” in the “rationality of the common man.”²¹ The philosopher and former colonial official Walter Stace, who was Rawls’s adviser at Princeton, called for a moral defense of democracy from totalitarianism’s attack on the “belief in the individual’s infinite value.”²² Anti-totalitarian democracy required a new kind of “objective ethics”—a universal but non-absolutist moral theory.

Rawls grew up on these ideas. He was born in 1921 in Baltimore to an affluent Episcopalian family. As an undergraduate at Princeton in the early 1940s, he was already concerned with ethics in non-state communities and with individuals as the source of value. At first, this took a theological form. Rawls developed a vision of community in which morality was not defined by the state or the pursuit of a highest good, but located in interpersonal human relations.²³ Salvation was not earned through work and action, but through the proper “recognition” of human persons “as persons”—as members of a universal moral community.²⁴ Against the social contract tradition, with its “egotism” and view of society as the result of “bargaining” between atomized individuals, and against Pelagian moralities based on merit, which rewarded individual actions, Rawls took “persons” as the basic unit of his ethics.²⁵ He carved a space between collectivist theories, which gave little space to individuals, and individualist ones, which abstracted individuals from their social contexts.²⁶ Rawls’s theological interests were dulled by the war, though his philosophy retained traces of theodicy and in many respects took the form of a secularized liberal Protestantism.²⁷ When he began his doctoral work on the GI Bill, he embarked on a project that shared much with those who saw in totalitarianism a crisis for social science and its capacity to explain social developments: Rawls sought to construct a system of objective standards for judgment that would stand without a God, or a state, to ground it.

Rawls entered the Cold War university at a time when “Western Civilization” courses were reinventing liberalism as an anticommunist and anti-totalitarian “fighting faith.”²⁸ Social scientists had grand ambitions, both intellectual and to serve the interests of the state. The behavioral, cybernetic, and systems sciences were flourishing, while the tools of modernization and rational choice theories were taken up across disciplines.²⁹ Sharing these system-building ambitions, Rawls benefited from the “golden age” expansion of research universities.³⁰ He was able to capitalize on the “tool-trading” of the “depoliticized” postwar human sciences without being constrained by the practical demands that faced policy and defense intellectuals.³¹ Rawls began a lifelong search for a general theory to evaluate political institutions. In the dissertation he completed at Princeton in 1949, he put in place a framework that he retained, in outline, for decades.

At the core of social life in democracies, Rawls suggested, there was a “consensus.”³² The aim of philosophy was to find a “reliable method” to appeal to that consensus, a “heuristic device” to yield justifiable principles for judgment that would allow room for change. “In the face of numerous ideological warfares, waged by means of institutionally-supported propaganda machines,” Rawls wrote, “men are likely to doubt not only the efficacy of reasonable principles, but their existence.” This had happened to many philosophers, who now either thought ethical principles were impossible to find or accepted their imposition by “authoritarian” means. Rawls was hostile to these positions, which he saw as common to “intuitionism” and utilitarianism.³³ He set himself a task: to find open and noncoercive procedures that specified under what conditions the reform of common-sense judgments could be justified.³⁴ He examined the justification of ethical beliefs, the norms of decision and action, by exploring what principles were “implicit” in “intuitive judgments.”³⁵ To do so, he rejected appeals to “certainty,” “by fiat and proclamation,” and to “exalted entities” like “the State, or the Party, or the Tradition, or the Church, or any one of a number of agencies.”³⁶ He argued that “rational discussion” was fundamental to discouraging those “social elements which, in democratic countries, we have tried to get rid of: the authoritarian, the arbitrary and the irrational.” It was through discussion that social consensus could be revealed.³⁷

In democracies, Rawls argued, “the law,” not the state, was the primary source of authority. He gave a radically minimalist, eliminativist account of the state as “the collection of men—senators, administrators, judges, police etc.—carrying out certain tasks according to laws under a constitution of sorts.”³⁸ The men and the laws were the state; the state was simply “their legal mask in a court of law.”³⁹ Laws were conceived as the rules that “public discussions” had shown to be “right and reasonable,” and that “citizens, as a group of intelligent men,” had “voluntarily consented to” as binding. Rational discussion, Rawls wrote, was the “essential precondition of reasonable law.” This process of reasoning was crucial to democratic practice, but also to theory: it could provide a foundation for ethical principles, locating authority “only in the collective sense of right of free and intelligent men and women.”⁴⁰

Rawls thus proposed as his heuristic device a “discussion” between “reasonable men”—imagined as “average, rational, and right-thinking and fair men, irrespective of wealth, social stratification, nationality, race, creed or religion,” with average emotional and intellectual intelligence and sufficient knowledge and education, defined by a capacity to judge evidence in a court of law.⁴¹ He borrowed the idea of the reasonable man from tort law cases involving negligence, where the standard of a reasonable man who could have “foreseen the risk” functioned as a “criterion” for establishing responsibility.⁴² He borrowed the idea of the average man—who, just as the scientist was an expert in science, was expert in “moral matters”—from democratic theory.⁴³ Democratic men, not technocratic or scientific experts, were judges of morals. Rawls implicitly took Dewey’s side in his famous debate with Walter Lippmann about the role of citizens and experts in democracies.⁴⁴ Yet science was not “the model of rational inquiry,” and Rawls thought Dewey went too far in making it so.⁴⁵ Rawls also injected an Enlightenment idea into the criterion of reasonableness, by including a “capacity for understanding suffering.” In the “free and uncoerced opinion” of those men, cleansed of the influences of propaganda, interest, and ideologies—of what Rawls called “the race soul,” “the dialectics of history,” and “the distant revelation”—an ethical consensus could be located.⁴⁶

Though Rawls left much unsaid, about certain things he left no doubt. He conceived this exercise in reasoning as his contribution to the project of shoring up democratic thought against authoritarian threats. Philosophy and the Constitution were invoked against alternative ideologies and interests. The resonances of this move were not only anti-totalitarian. The law, the Supreme Court, and the “court” of interest-free public opinion were part of a rejection of a vision of politics as state control and as collective bargaining between labor and capital.⁴⁷ By the mid-1940s, the ideas of consensus and the common man were invoked by liberals against the threat of the “authoritarian personality,” but they were also part of a defense of a “rationalist” alternative to class politics.⁴⁸ Rawls likewise wanted to avoid interests and ideology and to find an appropriate definition of the “we,” without “falling into the error of relying on a limited or biased group morality.”⁴⁹ It was not easy for citizens to step outside their prejudices—the “public” prejudices “against the Negro in the South (and elsewhere), on the West Coast against the Oriental.” But these prejudices did not “accord with the democratic ideal” expressed in the Constitution; that was why such beliefs were called “prejudices.”⁵⁰

Rawls’s racial liberalism was twinned with a worry about appeals to class and group interests. He was committed to the idea that the right “device” could help get beyond both.⁵¹ Rawls also thought such a device could get beyond something more universal: the partiality of persons. His vision of ethical and social life was associational, and he saw actions in the family and the “fellowship of close associations” as necessarily and appropriately motivated by love and “direct, free and spontaneous” affections. That partiality had to be socially constrained by “rational” procedures. Rawls did not want ethical judgments at a social or political level determined by love, prejudice, interest, or ideology.⁵²

When it came to explaining why, he was just getting started. These concerns with persons and procedures were quasi-Kantian, but Rawls did not yet invoke Kant. Over the next decade, he ranged widely in his exploration of potential procedures. Economics, political science, psychoanalysis—Rawls experimented with ideas drawn from each but faulted all for lacking an objective standard for common judgment. He would also develop the vision of person and society that was here embryonic—one that was hostile to authoritarianism, coercion, and social control, and saw society as ethical, associational, and underpinned by a consensus that could be accessed through rational discussion.

In the postwar years, a new generation of philosophers developed a view of morality as something that existed in the world that could be studied using rules of induction and observation. They devised an approach to ethics that focused on the “natural” facts of ordinary human relations—above all, the “facts” of moral psychology—and took as its cornerstone Ludwig Wittgenstein’s later work, particularly the posthumously published Philosophical Investigations (1953).⁵³ Under Wittgenstein’s influence and that of J. L. Austin’s ordinary language analysis, the study of ethical terms was replaced by the study of their meanings understood through use—the study of what words do and how they are embedded in practical life. This enabled a different kind of ethical inquiry: the study of action and institutions. Wittgenstein’s shadow loomed large over analytical philosophy, particularly at Cornell and Oxford, where Rawls studied in 1947 and 1952, respectively, before returning to Cornell to teach in 1953.⁵⁴ Rawls became immersed in Wittgensteinian ideas, initially through his teacher at Princeton, Norman Malcolm. He joined a transatlantic community of ethical theorists, which included Richard Brandt, Roderick Firth, William Frankena, Stephen Toulmin, and Kurt Baier, in their search for a naturalistic, objective foundation for ethics that took Wittgenstein as inspiration.⁵⁵

Wittgenstein meant many things to many people. Rawls’s Wittgenstein initially came mediated through Malcolm and Max Black, whose ideas—particularly their defense of the value of inductive reasoning as a way of explaining ethical facts—shaped how Rawls developed and justified his method in his dissertation.⁵⁶ Next, Rawls read the philosophical psychology and grammatical analysis of the likes of Malcolm, Elizabeth Anscombe, and Philippa Foot, who saw in Wittgenstein’s naturalistic understanding of the person crucial insights into the conventions of human life.⁵⁷ Rawls already thought that morality was interpersonal, existing in the recognition of other persons as persons in communities. Soon he used Wittgenstein to explain that morality was social, defined by its use—there in the world to be discovered, not chosen.⁵⁸ Having a morality was like having a sense of humor. It was part of what it meant to be human, part of a “form of life.” The phrase was Wittgenstein’s, but Rawls recognized that it was a “Hegelian notion.”⁵⁹ He used it to describe morality as a natural phenomenon, continuous with other aspects of life—with “natural feelings” like pity, sympathy, compassion, and the “fellow-feelings.” It developed in childhood through the love that flourishes and the guilt that is experienced in the family, in friendships, and then in larger associations—a process of development Rawls studied via Freud, Lévi-Strauss, and Melanie Klein and his preferred accounts, Rousseau’s Émile and Jean Piaget.⁶⁰ Moral principles had to be understood in terms of the moral feelings, like shame, remorse, or guilt, which were responses to breaches in natural interpersonal relations of recognition and “part of the way we show our recognition of persons as persons.”⁶¹ To fail to differentiate between persons and things was thus a failure of morality. A person who failed to do so would be, to use a phrase of Wittgenstein’s, a “person who pitied only dolls.”⁶²

This was a vision of ethical life built from the bottom up, where morality was universal, natural, and constitutive of personhood, yet was developed and earned in communities.⁶³ Rawls’s youthful communitarian vision of ethical life persisted. He used Wittgenstein to think about how philosophy might understand it. Wittgenstein suggested that philosophical problems could not be resolved through a process of penetrating or getting behind phenomena. The search to uncover hidden truth was mistaken. “What we want to know lies already in sight,” Rawls wrote, glossing Wittgenstein.⁶⁴ In the “Wittgenstein Lexicon,” which Rawls constructed to grapple with these ideas, he wrote that “philosophy is an attempt to bring these things which we know in view.”⁶⁵ Social philosophy, he later said, “can be viewed as self-analysis, an examination of one’s own moral opinions.”⁶⁶ Rawls’s major departure was to say that additional devices—models of discussions and procedures for self-inquiry—were required to help us see what was already there and how we might change it.

What did Rawls think was already there? His portrayal of interpersonal, natural morality was universal, but his vision of society was historically specific, adapted from the pluralist theories that accompanied the midcentury attacks on concentrated state power. Political pluralism, along with its industrial variant, collective bargaining, was meant to “institutionalize” industrial conflict.⁶⁷ In the postwar years, it provided a depoliticized vision in which interests and institutions replaced passions, ideologies, and solidarity, and voluntary associations and “civil society” supplanted and devalued class-based political action.⁶⁸ Enthusiasm for Rousseau among Anglo-American pluralists was abandoned for the warnings about the perils of excessive centralization and federal bureaucracy contained in Alexis de Tocqueville’s long-ignored Democracy in America (1835–1840), reprinted in 1945.⁶⁹ With its appreciation of American localism, it became the lynchpin of newer theories of the “civic culture” and of those that celebrated associational life as the barrier against the administrative state at home and the totalitarian state abroad.⁷⁰

These ideas were hard to miss. Rawls encountered a range of critiques of concentrated power in a seminar on American democratic thought in 1949–1950 taught by the Princeton political scientist Alpheus Mason. There Rawls read Tocqueville, Jefferson, and nineteenth-century critics of capitalism like Brooks Adams and Richard T. Ely, both of whom Rawls saw as sketching complex dynamics between the capitalist class, the administrative state, and centralized bureaucracy. He largely agreed with the sociologist Robert Lynd’s diagnosis, in the context of anticommunist repression, of the threat to academic freedom posed by “big business” and its “scapegoating” of African Americans, Jews, and organized labor. The social gospel critiques of the capitalist had concerned Rawls since his undergraduate days.⁷¹ Now he also read constitutionalists like Edward Corwin and explored how the courts, particularly judicial review, functioned to protect rights from the executive and legislature.⁷²

Rawls was especially taken with Tocqueville, both his idea of the New England townships as “great schools of democracy” and his suggestion that associations might play a role, like law, in restricting the state and distributing power. Tocqueville’s analysis seemed acute to Rawls, particularly his argument—Rawls thought it sounded Marxist—that “the constellation of social conditions binds the capitalist as well as the worker.” Tocqueville was afraid that industrial capitalism would eliminate social bonds and create a new aristocracy. That aristocracy, Rawls wrote, could only then be checked by a “‘welfare state’ to use a phrase with its present coin.” Yet that welfare state was itself a “gigantic engine of encroachment on all areas of life.”⁷³ In the face of it, associations should provide order and stability. They served as bulwarks against tyranny and instruments of power against the government and the “manufacturing class.” Without them, the individual would be “left helpless before the State, and against other concentrations of power.” “The totalitarian state,” Rawls wrote, “is the state of the ‘unassociated man.’”⁷⁴

Rawls borrowed from these ideas in a pair of lectures probably delivered in 1951, where he developed for the first time his own substantial social and political vision.⁷⁵ He rejected theories that compared society to an organism or a mechanism. “Society,” Rawls declared, “is like a game,” conducted according to man-made, agreed-upon rules that allowed players to have reasonable expectations and make reasonable decisions. It was a game of multiple players acting in their own interests. “The motor of society,” Rawls wrote, “lies in many small groups.” Self-interest meant looking out for your association. Government was a regulatory authority, a system of rules that stabilized associational life. The game was not centrally directed. If it were, society would be more like “an army” than a game. But it did need an “umpire” to enforce the rules and secure the “general conditions of social order.” Laws and government were there to ensure the game did not break down.⁷⁶

Individual persons were the primary unit of Rawls’s ethical theory, but his social vision looked to the “social units” that connected them: “colleges, counties, cities, churches, corporations, trade unions.”⁷⁷ What mattered most was the family, church, and firm. These Rawls would describe in naturalized terms. He did not use the legal language of personality, deployed to describe the artificial groups of union, corporation, and state and associated with the partisan politics Rawls tried to diffuse.⁷⁸ The family was the primary association and, conceived as the household, the basic economic unit. In this, Rawls followed contemporary social theory and policy, which institutionalized the family in a variety of ways, from the suburban home to the family wage.⁷⁹ He saw families and small associations as sites of ethical development and meaning. They had a structural role, enabling stability and preventing the centralization of power.⁸⁰ Rawls framed this stabilizing function in terms of the game: the morality of players, those “motors” of society, were irrelevant to the smooth functioning of the game. While relations within families would be altruistic, it was “safe” to assume that relations between “teams”—between households, firms, and associations—would be the opposite.⁸¹ Stability relied not on altruism but on associations playing the game, which meant, by definition, acting in their own interest.

Game analogies were widespread in postwar thought. Rawls later listed prominent examples: Wittgenstein, Karl Popper’s politics, the psychology of Piaget, the legal thought of the Oxford philosopher H.L.A. Hart, and game theory.⁸² The idea of the game, Clifford Geertz observed retrospectively, seemed “to explain a great deal about a great many aspects of modern life.”⁸³ When Rawls first described society as a game in 1950–1951, he was at Princeton, the home of game theory, and he cited John von Neumann and Oskar Morgenstern’s already influential Theory of Games and Economic Behavior (1944).⁸⁴ But it was less their account of rational behavior than the tradition of analogizing games with institutions Rawls had in mind. Rawls was studying economics with William Baumol and had previously taken a class with Jacob Viner.⁸⁵ In these years, political problems on both sides of the Atlantic were parceled off to economists. Much of Rawls’s political education came from following debates in the new welfare economics, theories of utility, consumer behavior, and general equilibrium, and from his encounter with the utilitarian tradition of “institutional design” (in which Rawls placed Hume, Adam Smith, and Bentham).⁸⁶ It was in the game analogy that Rawls found the “most important proposition of social philosophy . . . that it is not necessary to have central direction to achieve a rational social order.”⁸⁷ The rules of the game can be structured so the players achieve social ends simply by taking part; the “ideal entrepreneur,” he later wrote, can do good while playing the “capitalist game of maximizing profits.”⁸⁸ Rawls here adapted from visions of commercial society characterized by virtues of fellowship and sympathy, where interpersonal relations were shaped from the bottom up by rules and conventions, not a strong sovereign authority. Adding a Humean twist to his Wittgensteinian account of natural sociability, Rawls welcomed this society in which rules were man-made but created out of customary institutions like promises, contracts, and agreements.⁸⁹ He would write in 1959 that “the conception of justice I set out is perhaps closer to Hume’s view than any other.”⁹⁰

There was one use of the game analogy that stood out: the description of business as a competitive game offered by the Chicago economist and teacher of the libertarian right Frank Knight. A staunch defender of the price system as a tool for freedom, Knight was also a critic of the ethical consequences of capitalism for moral character.⁹¹ Despite Knight’s view that any search for an absolute standard in ethics was a “fetish” of scientific method, Rawls took great interest in his social philosophy.⁹² The task Knight defined as the key problem of political life—“to find the right proportions between individualism and socialism”—Rawls underlined in his copy of Knight’s The Ethics of Competition (1935) in three different pens.⁹³ Like Rawls, Knight saw a politics of “discussion” as democracy’s bulwark against bureaucratic overreach. In the tense climate of McCarthyism, Knight’s anti-authoritarianism was attractive.⁹⁴ Yet it was in a critical essay of Knight’s on the moral limits of markets, which Rawls annotated heavily, that he seemed to find a persuasive understanding of society. Knight described business as having the characteristic ingredients of a game—luck, skill, effort. In the business game, income accrues to owners of productive capacities (where ownership itself is based on inheritance, luck, and effort). Effort mattered morally, but Knight thought it ought not to be elevated to the status of a moral principle. Inheritance was a matter of luck. Yet luck tended to be rewarded. Rewards tracked marginal productivity, which was influenced by luck all the way down—the luck of natural talent, inheritance, and circumstance. That luck accumulated through success, while the participation of the losers in the game was reduced to “mechanical drudgery.” Inequality of winnings led to an inequality of starting places and stakes in the game. The game of business was not, therefore, a good game: it did not cultivate the ideal of good sportsmanship or the other aspects of character necessary to the improvement of society.⁹⁵

Rawls applied Knight’s model of the good game to his vision of society. A good game, Rawls wrote, involves a certain amount of unpredictability, luck, and chance. For it to be worthwhile, it cannot depend on “pure luck” but requires effort and skill.⁹⁶ Players need to be able to win if they play by the rules: every game must have the feature that “we think we have control over our fate.” One of the tricky things about games is that success accumulates: if someone wins all the time, the outcome can be foreseen, and “other players lose the zest for playing.” For a game to be worth playing, inequalities need to be broken down. The outcome “can only be unforeseeable, and effort can only be efficacious, if the players enter the game with roughly equal resources,” both material and spiritual. What is needed “is some kind of control over the game,” to redistribute “some of the rewards of winning.”⁹⁷

But how much control? Opposition to concentrated power here translated into a minimalist politics. Rawls departed from his early eliminativist image of the state as a mere collection of individuals, which denied the possibility of a group agent. Now he framed government as capable of enacting change, but highly limited. Governments, Rawls argued, should enforce only the kinds of rules that everyone knows they want: rules of the highway that maintain the direction of traffic, rules for public health and safety, and rules for defining crimes, contracts, weights and measures, as well as rules to solve basic disagreements and answer to “certain general conditions of social order regardless where we are going.” “We don’t need agreement all down the line from theology to tastes in tea,” Rawls wrote, crossing out his earlier, even more minimalist example, “the day of the week.” Rules set the basic terms for common life, but left people alone. People want “settled rules,” enforced by “local governments,” about “how we shall drive on the highways,” but they “all want to go in different directions.” Governments cannot tell them where to go, Rawls insisted. To do so would be to “tyrannize” over them.⁹⁸ It is perhaps not surprising that Hayek could write twenty years later that he had “no basic quarrel” with Rawls.⁹⁹

At the turn of the 1950s, they had a lot in common. Indeed, in a political landscape where many attacked planning but nonetheless accepted the need for substantial government intervention in the economy, the constraints that Rawls placed on the scope of legitimate state action and the metaphors he deployed put him in the company of those who used the language and rhetoric of radical anti-statism. His use of the image of the highway was telling: borrowed from Walter Lippmann’s The Good Society (1937), it was commonly used among early ordo- and neoliberals, including Hayek and Lionel Robbins, to differentiate the liberal from the planned state.¹⁰⁰ They envisaged the state and the rule of law as responsible only for establishing and enforcing the right kind of economic life, and they likewise used the metaphor of the game.¹⁰¹ Moreover, Rawls’s examples of appropriate government policies for concerns that did not have their interests represented in the “normal run” of the competitive game, like “future generations” and the “conservation of resources,” also included the classical liberal tropes of national defense and lighthouses.¹⁰² The functions of government, the umpire of the game, hung on a single idea—the common rules on which people could agree. Rules of government interference changed over time, as beliefs changed: Rawls’s example of a new sort of government action was fiscal policy to limit the “hazard” of unemployment and maintain the “fullest employment possible” insofar as it was consistent with the free price system.¹⁰³ This was a nod to the Keynesian consensus and the commitment to maximum, not full, employment in the Employment Act of 1946, which won out over alternative visions of state economic planning and the robust Keynesianism of the Full Employment Bill of 1945.¹⁰⁴

Rawls was less Keynesian than the consensus. He wanted changes to be slow and rule-bound. Worried about the discretionary powers of administrative agencies, he looked for stable rules to limit discretion and “avoid arbitrariness” by providing a “steady background” against which choices could be made.¹⁰⁵ He argued that an ideal legislator ought not to take rules on or off the books unless doing so affects “the general system of rules enforced by government” in that “there is a reasonable expectation that any person, taken at random, would be benefitted by it when the system is thought of as being held to over the long pull.” “Redistributive changes” could be worked out within this general framework, but government ought not to be a “purposeful factor in their occurrence.”¹⁰⁶ Rawls described his vision not as one of a “laissez-faire” society, but a “prenez-garde” one. He agreed with many of his contemporaries that it remained impossible to return to the laissez-faire ideals that a half century of state-building and war had eclipsed. Yet the functions of government were still to be constrained by the idea that we should “take care” and watch out for those who might want to define our purposes, values, and ends, both in times of “peace” and in times of “emergency.”¹⁰⁷

Here Rawls invoked a wartime anti-totalitarianism in service of a peacetime anti-statism: the attempt to define the ends of citizens, he wrote, may mean “the use of assassins, informers, gas-chambers.” The United States had managed to escape this fate thanks to its “political maturity” and its lack of a “dynamic political creed,” like fascism or communism, geared to a single aim.¹⁰⁸ Nonetheless, “we should beware” of centralized power. The proper function of government should be limited to the enforcement of basic rules that allow people to live together, leaving persons and associations to pursue their own ends.¹⁰⁹

This entailed a series of demands typical of postwar liberalism: government had to allow consumer sovereignty, the “entrepreneurial control of resources,” a constitutional Bill of Rights, and equality of opportunity. But “political processes could not be relied upon . . . to run the economy.”¹¹⁰ Elsewhere, in an early 1950s essay concerned with the “civil and economic rights” of individuals and their violation by associations, particularly religious and economic associations, Rawls wrote that the state’s role was not to intervene in such groups, but to allow people to disassociate from them. Where an entrepreneur hires only individuals of his own religion, Rawls saw this as permissible, so long as the state “affords sanctuary.”¹¹¹ Yet the state would have to intervene sometimes—to halt the buildup of economic and political power, or to stop the game stalling. Otherwise society might become, in Rawls’s metaphor, more like an army—this time a private army of economic actors.¹¹² Pluralism must protect against the garrison state, but it also ought to prevent private associations from becoming militias.

It was in his discussion of state power that Rawls introduced for the first time an idea that he would later use to designate the ideal regime for his just society. For the young Rawls, the challenge was to sustain non-interference while ensuring the stability of the game over time—to balance “freedom” and “order.”¹¹³ Stability requires guaranteeing that people have enough of an equal start that they want to play. To do that, the prenez-garde state, Rawls wrote, would aim at a “property-owning democracy”:

Government should follow a program designed both to protect the necessary rewards of playing, as well as to guarantee, if not an equal start, for this is impossible, at least that sort of situation in which most persons we want to play will think that they have an equal enough start to keep the game moving. More definitely, this means assuring a fairly wide distribution of property; we want a property owning democracy where everybody has a stake. But on the other hand we must allow numerous and substantial rewards above the average to draw forth the efforts of the players. This obviously is difficult to do: it requires the most difficult sort of balancing . . . and it requires a constant policing, not only in the shape of enforcing those rules of the game which specify the game, but in the shape of keeping things even.¹¹⁴

This puzzle came to preoccupy Rawls: the “difficult sort of balancing”—how to achieve the “necessary wide distribution of resources and power and prevent the concentration of resources and power while still rewarding success so as not to stall the game.”¹¹⁵ The players of the game should begin with a roughly equal start. Rawls thought people had different abilities and were willing to take different risks. Inequalities that resulted from these differences need not be eliminated, provided that the general system of rules functioned to keep things “even” enough for the game to carry on. The challenge was how to balance the need to reward effort against the risk that winners would accumulate too much and dissolve what Rawls, following Knight, called the “luck element of the game.”¹¹⁶

This was the first time Rawls described the solution to this puzzle as a “property-owning democracy”—a vision of an “open” society where capital is widely dispersed and the rules and stability over time are secured without government interference. He did not yet use it as a term of art, and provided no source: the later source he cited, the economist James Meade’s Liberty, Equality, and Efficiency (1964), would not be published for over a decade (though Rawls was already aware of Meade’s work and was reading similar sources). If it had a specific provenance, it was likely drawn from the republican tradition and its commitment to property dispersion, and in particular from Tocqueville (especially his view of inheritance laws as a means of securing a wide distribution of property) and Jefferson (whose denouncement of the “concentration of property . . . and the uncertainty of tenure” and belief that a “wise and fair distribution of land” was necessary to good government Rawls noted with approval).¹¹⁷ In these thinkers, Rawls found both a defense of inheritance laws rather than active redistribution and a set of recognizable anxieties about the demands of labor and the power of the state.¹¹⁸ Rawls did not want a politics of income and pay disputes, but one of property. He insisted that keeping the game going required wide distribution of powers and resources, as well as “considerable social mobility between classes; of seeing to it that one holds position in society based on one’s efforts and skill. It involves, I think, the sort of thing Jefferson had in mind when he spoke of an aristocracy of talent.”¹¹⁹ Rawls did not yet speak of “natural talents” as “arbitrary.”

This linking of decentralization with rewards for effort and entrepreneurial incentives within a free price system implied that private ownership was the bedrock of an open society. What was crucial was less redistribution or limiting inequality, and more the dispersal of power away from centers and toward peripheries. Rawls’s account of what it took to stabilize the game had an ambiguity that persisted in his later work: what mattered was that the players think they have an equal enough start. Was it their experience of the game that mattered—that the players think they have an equal enough start, not that they do—or did the extent of equality matter from another point of view? As Rawls explored what the protection of stability and moral community in an “open society” required, he carried on exploring this question. His concern with how much and what kind of redistribution would be necessary became central. Soon Rawls would focus on the extent to which upholding this vision required the elimination of the effects of contingency and luck by the rules of the political game themselves.

Rawls would leave behind this barebones liberalism, with its anti-interventionist commitment to small government. He swapped the analogy of society and the price system for different examples of endurance over time, such as constitutions. Though he never dropped the metaphor of the game, he supplemented it.¹²⁰ The game’s regulatory rules and players did not capture the complex ways in which rules shaped character, nor the fact that persons pursued their own ends not as rational egoists or strategists but because of their deep partiality and love for their families and associations. Rawls wanted to protect that partiality. But that partiality was what made impartial rules and institutions necessary. In notes from 1952–1953, Rawls gave these ideas a Wittgensteinian inflection. He described how the rules of society existed to protect forms of life:

What we can do is to make our society a free society in which various forms of life are tolerated by the state and in which mutual toleration between forms of life is encouraged. It is true—trivially so—that men, to live together, must agree on something; but they must not agree on fundamentals, or things they think in some religious sense most important, if by this one means the one proper form of life for a reasonable man to adopt. All they need agree upon . . . is in taking common action, via the state and other general rule mechanisms, to foster the necessary and fostering conditions for any form of life at all. . . . One avoids the forced choice, the either/or, of this or that form of life here and now, on this spot at this time, for everybody, by making a society which allows these differences to develop within a commonly accepted general structure.¹²¹

Here was a vision of an institutional structure to secure the existence of the associations in which persons could be treated as the equal persons Rawls thought they were, if only we could see them clearly. It was a model built on the potentiality of consensus and agreement, on rules, uncovered and specified through procedures, designed to dull the force of class and group conflict, ideology, prejudice, and passion. This was a long way from the egalitarian theory for which Rawls would be known, yet its contours were coming into view.

Rawls’s mature theory of justice was born over the course of the decade that followed his trip to Oxford in 1952–1953. Cold War liberalism was deepening its hold on intellectual life, and McCarthyism was stifling debate.¹²² As such, it was unsurprising that, like American liberals in the aftermath of the last world war, Rawls found a number of the resources for his theory not in the United States but in Britain.¹²³ Oxford was an exciting place to be a philosopher. J. L. Austin held his famous “Saturday Mornings” group on language and lectured on “Words and Deeds.” Anscombe gave lectures on Wittgenstein’s Philosophical Investigations, published the same year. P. F. Strawson presented the material that became his Individuals, and H.L.A. Hart lectured on The Concept of Law.¹²⁴ The linguistic analysis of Richard Hare, Gilbert Ryle, and Geoffrey Warnock reigned. Yet Isaiah Berlin and G.D.H. Cole taught utopian socialism, and Berlin, Hart, and Stuart Hampshire ran a seminar on political philosophy, which Rawls took, where they read Rousseau, Kant, and Keynes.¹²⁵ Thanks to ties between philosophers and the British Labour Party, Oxford was not only the crucible of language philosophy but also aflame with debates about inequality.

In Britain, where the aftermath of war saw the establishment of major institutions of welfare provision—the National Health Service, universal education, and national insurance—the debate about state planning, investment, and public ownership loomed large. For the Labour Party and its fellow-travelers, the question was what the relative success of these institutions meant not only for the future of welfare provision and social services, but for the left.¹²⁶ When Rawls arrived in Oxford, Labour had lost an election. The revisionist right wing of the party, under the direction of Anthony Crosland and Hugh Gaitskell, was leading a “modernizing” push to drop the commitment to nationalization and public ownership—Clause IV of the Labour Party’s constitution—and foreground a moral concern for “social equality” and “social justice.”¹²⁷ Arthur Lewis’s Fabian Society study, Principles of Economic Planning (1949), which had recently been published, rejected the importance of public ownership and advocated redistributive policies by progressive taxation in a market economy. James Meade’s Planning and the Price Mechanism (1948) envisaged the role of the state as securing the preconditions by which the market could efficiently allocate and fairly distribute resources and achieve the Keynesian objectives of full employment, stability, and equity.¹²⁸

Over the course of the subsequent decade, many would become gripped by a transatlantic debate about sociological changes in modern capitalism. Social theorists in the 1940s had tied critiques of concentrated power to deeper transformations in the nature of capitalism and the state. Pessimistic accounts of long-term decline proliferated, from Karl Polanyi to Joseph Schumpeter. In his bestseller The Managerial Revolution (1941), James Burnham adapted an argument made a decade earlier by Adolf Berle and Gardiner Means, and separately by Keynes, to popularize the idea that the historical transition between proprietor and corporate capitalism had generated a new form of economic life: ownership of corporations was now divided from their control and management, and excessive power lay in the hands of a new managerial class (rather than property owners).¹²⁹ In mid-1950s America, as the pessimistic mood lifted and the concern with stability gave way to a hunger for growth, social liberals drew on this analysis to make sense of the new age of affluence.¹³⁰

Amid the relative political calm, many claimed that transformative ideas were exhausted. The ideology of the “end of ideology” arrived.¹³¹ For some this was a cause for concern and pessimism. They worried about what David Riesman called the new “social character,” the cultural consequences of the consumer republic.¹³² The political theorist Judith Shklar observed a “cultural fatalism” that led to “not only a decline of social optimism and radicalism but also the passing of political philosophy.”¹³³ For the political theorist Sheldon Wolin, “the political” was submerged beneath the economic and the social. Traditional political problems were confronted, but only in the study of “nonpolitical groupings,” like “trade unions, corporations, bureaucracies and neighborhood gangs.”¹³⁴ Managerialism meant that both politics and political theory itself were waning. Others were more optimistic. Daniel Bell saw in the logic of managerialism the seeds of a “post-industrial society” beyond capitalism—a mixed economy in which the distinction between private and public ownership no longer mattered. If it was management and control rather than ownership that was tied to power, then a concern with public ownership was irrelevant.

This analysis was used to justify the anti-Marxist focus of Cold War social thought: capitalism had eradicated the need for socialism, or would eventually provide for it from within, without the need for class struggle. Soon, a new wave of social democratic liberalism came to occupy the space cleared by this analysis.¹³⁵ This was particularly true in Britain, where, earlier than in the United States, these arguments justified prioritizing equality and social democracy above socialism. Labour thinkers, from liberals like Crosland to Richard Crossman and the Bevanite left, read Burnham closely, particularly his argument that state ownership and central planning entrenched privilege and empowered the managerial elite.¹³⁶ Many feared managerialism was damaging both nationalized industries and corporations. American interpreters used Burnham to bolster an anti-totalitarian anti-statism, but in Britain, Crosland wrote, no one any longer believed Hayek’s “slippery slope” argument that interference in the market led to totalitarianism.¹³⁷ Instead, British Labour thinkers borrowed these ideas to justify pluralism and decentralization and to redefine socialism’s aim as equality, not common ownership. Crosland argued that “the rights of property, private initiative, competition and the profit motive” were no longer dominant.¹³⁸ There was no need for the state to seize property to secure socialism. “Post-capitalism” was on its way. In fact, its theorists thought it was already here, if only they could see it clearly.

These ideas extended deep into Oxford, where the lines between the Labour Party and academic liberals were porous. Hart and Berlin had close ties to the revisionist and Fabian wing of Labour, whose MPs were frequent visitors to Oxford high table. Hart had been an undergraduate with Gaitskell and during the war lived with Douglas Jay, who had authored an influential attack on state ownership that defended an egalitarian idea of social justice.¹³⁹ As Labour thinkers swapped public ownership for a new priority of limiting inequality, philosophers followed. The call for a reinvigorated “distributive justice” merged with the effort to inject ethics into economics. In 1950, Crosland’s close friend Ian Little’s critique of welfare economics attempted to reverse the split of ethics and economics advocated by Lionel Robbins in his defense of a value-free “economic science” twenty years earlier, joining economic theory with a liberal socialism that prioritized justice and equality over efficiency and class-based redistribution.¹⁴⁰ In 1952, a new edition of R. H. Tawney’s Equality was published, renewing attacks on the idea of equality of opportunity. Philosophers like Ayer, Iris Murdoch, Hart, Berlin, and Richard Wollheim debated the relative value of equal rights to property and liberty and the place of equality of outcome and opportunity in managerial societies.¹⁴¹

In response, the philosophical wing of the British New Left was born. A young Alasdair MacIntyre combined critiques of analytical ethics with Marxism. Charles Taylor, who studied with Berlin and Anscombe, became an editor of Universities and Left Review, one of the predecessors of New Left Review.¹⁴² For this first New Left, the “managerial revolution” was not shepherding socialism into existence. Its “power elite” were the shock troops of corporate capitalism.¹⁴³ When Taylor returned from studying in Paris in 1957, enthused after reading the French edition of Marx’s Economic and Philosophical Manuscripts, he helped ignite the philosophical rediscovery of the early Marx among the New Left.¹⁴⁴ Many joined studies of alienation to critiques of the bureaucracy that the Fabian vision of equality required for its implementation. On this view, the affluent society led not to socialism but to alienation.¹⁴⁵ For Taylor, the dream of a “reform of capitalism ‘from within’” was an illusion. What was needed was “common ownership (not state monopoly).”¹⁴⁶ Revisionism put procedural stability above worker democracy. It was an elite project to secure the workings of private enterprise.¹⁴⁷

When Rawls encountered these debates about equality and ownership, he had seen some of the arguments before. But it was their interpolation in the hands of the British Left that mattered for his trajectory. Though Rawls arrived in Oxford before these debates peaked, they nonetheless helped him see what political work his ideas might do. The revisionists’ teleological optimism about the tendencies of post-industrial society complemented Rawls’s optimism, particularly his earlier claim that a “core” consensus existed within democracies. In Britain, this optimism was not combined with sociology, as in its American iteration, but with ethics and economics. Rawls’s familiarity with both, as well as his emphasis on decentralized ownership, oriented him toward the theorists who defended equality over common ownership. As he explored their ideas, he encountered socialism too. After he left Oxford to return to Cornell in 1953, he kept abreast of the British debates—about equality and opportunity, effort and reward, and moral psychology—and these helped him rethink his earlier support of simple equality of opportunity and consumer sovereignty. He took notes on essays by Murdoch and others on equality of opportunity as they appeared, and he kept up his personal connections with Oxford philosophers, many of which lasted his life. During the year Hart spent at Harvard in 1956, he visited Rawls at Cornell. That was also the year Crosland’s The Future of Socialism was published, which Hart had sent to him in the United States.¹⁴⁸ Rawls cited it when teaching equality thereafter—alongside Tocqueville and others whose thought he encountered at Oxford, including Tawney, Wollheim, Berlin, John Plamenatz, and D. D. Raphael, as well as post-Keynesian economists like Joan Robinson.¹⁴⁹

While this debate about equality and the Labour Party thrived, Rawls began his work on justice at Oxford, proceeding first by ordinary language analysis. A page of notes began: “Justice. Rules for a Game. Game of Society. Social Justice.”¹⁵⁰ Rawls now defined the “problems of justice” that would determine the trajectory of philosophy: “what is the relationship of justice and fairness”; “extent of justice: what sort of actions does it apply to”; “ranking principles of justice.” He extended his interest in games and his device for judgment to social institutions, setting up a “discussion” or “reasoning game” to explore the “maxims of justice” necessary to help people choose what Berlin called a “pattern” or form of life—or, as Knight had written, a “kind of life.”¹⁵¹ These would create the kind of habits needed to standardize choice and (continuing his critique of intuitionism) protect people from the “tyranny” of having too much. The point of analyzing justice was to find the best maxims of organization, distribution, and freedom in situations where persons make claims on a certain “stock of goods.”¹⁵²

Surveying notions of justice, Rawls wrote that justice could be understood in terms of taxes, wages, social structures, forms of penal justice, the distributions of political office, and the structure of society itself. Maxims of justice could help decide what kind of social organization was just: was it just to organize society like a “joint stock company,” or a corporation, where everyone has a stake? Maxims would also guide decisions between forms of freedom (he listed freedom of consumer sovereignty, choice, occupation and work, leisure, freedom from want), forms of equality (of income, opportunity, or stake in society), and different economic goods (economic progress, efficient allocation, full employment, and price stability). They would also discriminate between values that determined distributions. The fairness of wages, for instance, could either be calculated according to labor time, in accordance with effort, need, or equality, or defined by the value of the marginal product of labor. Taxes in open societies could be tied to income, fixed or flat. Each of these choices was supported by different reasons: that man “ought to have the fruit of his labor”; that a man’s worth “depends on his intention and effort”; that “distribution, or payment, should be in accord with needs and wants”; or that the “rule of equality” was primary.¹⁵³

These reasons reflected the contours of the postwar British debate about the distribution of goods in welfarist or social democratic societies.¹⁵⁴ The claim that rewards should be proportional to an individual’s or collective’s productive contributions was common both to Ricardian and socialist labor theories of value and to neoclassical versions of marginal productivity theory, which saw contribution as a measure of desert but calculated income in terms of marginal product. Defenses of state welfare provision relied on a different ethic—that of meeting needs and wants. Universal social insurance schemes, designed to meet needs in market societies, provided income, services, and non-means- or contribution-tested “social assistance” to those excluded from the workforce. In postwar Britain, many debated whether welfare provision should aim at the “relief of poverty or the maintenance of a national minimum” or also at equality and, in T. H. Marshall’s framing, the reinforcement of social citizenship and altruism.¹⁵⁵ Rawls’s rule of equality corresponded to the position associated with Labour revisionists, who went beyond meeting needs and wants but also rejected the labor theory of value and common ownership. Instead, they focused on the size of income to be distributed, not its sources. What they debated was equality of opportunity.¹⁵⁶ Only a “pure form of laissez-faire society,” Isaiah Berlin wrote, would permit all inequalities that arose from equality of opportunity.¹⁵⁷ Revisionists like Crosland, Jay, and Michael Young took up Tawney’s condemnation of equality of opportunity. But they also tried to rehabilitate the idea by conceiving it as including not just equal access to play the game but the fair starting places of players too.¹⁵⁸

What these theorists discussed less was what criteria should determine starting places or limit inequalities, particularly those arising from returns for talent and effort. In the early 1950s, Rawls began to formulate answers to this “balancing” problem. His early efforts were not always clear. He wrote on a scrap of paper during his time at Oxford that “to provide every man with such an opportunity that if he has but the initiative to take he may acquire such a stake in society as to make him conservative is the prime objective of radical politics.”¹⁵⁹ Equality of opportunity, Rawls implied, was the aim of a radical politics. Opportunity twinned with initiative allowed for the acquisition of property, which was the basis for a property-owning democracy. Yet too much property tended to conservatism. Rawls’s task was to pinpoint how much was enough.

Already in his youthful theological writings, Rawls had indicated his skepticism that persons could earn the favor of God or “merit” election; he had emphasized that the performance of righteous acts was less important to Christian ethics than adopting ethical attitudes.¹⁶⁰ He did not yet fully theorize how this level of judgment about the irrelevance of merit coexisted with considerations about desert in social and economic life. But he now paid close attention to how rewards for efforts and talents could be distributed at a social and economic level. He tacked between Knight and the Labour revisionists. Knight had rejected socialist arguments that tied desert to labor, as well as the marginal theory of productivity, on grounds that contribution as measured by market prices was a matter of luck over which individuals had little control. He saw an inconsistency in the position of socialists who objected to inheritance of wealth while defending unequal reward based on differential productive capacity. One was a product of social circumstance, the other of natural talent, but both were caused ultimately by luck and contingency.¹⁶¹ Inheritance of wealth was no more arbitrary than inheritance of talent. Crosland also rejected “vague” criteria of “worth” on the grounds that it seemed “unjust and unwise to reward or penalize people . . . for inherited characteristics.” He retained the idea of the “rent of ability”—the “additional reward which exceptional ability can in practice command from the community”—which raised the “question of incentives” and the importance of balancing equality with “the supply of ability (and also of effort, risk-taking, and so on)” and “economic growth.”¹⁶² Rawls noted approvingly Knight’s claim that “productive contribution can have little or no ethical significance from the standpoint of absolute ethics,” or as an “ethical measure of desert.” He likewise made clear his aversion to theories that tied distribution to the “moral worth” and merit of recipients, though he agreed that talent and effort could be rewarded.¹⁶³ Gradually, Rawls formulated a theory that could account for these ideas.

In 1957, Rawls presented the first version of the theory he now named “Justice as Fairness” at the American Philosophical Association annual meeting. His solution to the balancing problem was his two principles of justice. The liberty principle stated that each person “has an equal right to the most extensive liberty compatible with a like liberty for all.” The equality principle stated that inequalities were “arbitrary” unless “they will work out for everyone’s advantage and unless the offices to which they attach, or from which they may be gained, are open to all.” Together they expressed “justice as a complex of three ideas: liberty, equality, and reward for contributions to the common advantage.”¹⁶⁴ The next year, Rawls clarified this last idea as “reward for services contributing to the common good.”¹⁶⁵ Inequalities, he argued, had to work for the social advantage. They were only permissible if “every party” gained from them. It was “reasonable” to “acknowledge equality as an initial principle,” but this should not be regarded as “final”: “for if there are inequalities which satisfy the second principle, the immediate gain which equality would allow can be considered as intelligently invested in view of its future return. If, as is quite likely, these inequalities work as incentives to draw out better efforts, the members of this society may look upon them as concessions to human nature.”¹⁶⁶ Talents were rewarded not because they were valuable, but only if everyone in a society gained from them. If inequalities were “won in return for a contribution to the common advantage,” then there was no reason for complaint.¹⁶⁷

This account of distributive justice soon defined a new tradition of liberal philosophy. With these principles, Rawls brought philosophical order to the ideas of the Labour revisionists. Redistribution and inequality had become far more central to Rawls’s thought, and his earlier concerns with moral worth and reward found new form. Yet Rawls would continue to develop his ideas, arguing that the equivalence between “social contingencies” and “natural chance” required going beyond equality of opportunity.¹⁶⁸ Natural talents and the contributions or deservingness of players could not determine the starting places of players in the game of society. If society were to be just, it would have to treat inequalities of wealth and income the same way as inequalities produced by rewards for differentials in talent, ability, and the capacity for effort. Rawls would make openness to all a condition on the justification of inequality of reward. When Michael Young’s long-awaited critique of meritocracy was published in 1958, Rawls agreed not only with Young’s claim that public ownership mattered less than equal chances in the social game, but also with his argument that “meritocracy” or a system of “natural liberty” where natural talents themselves determined life chances simply amounted to “an equal chance to leave the less fortunate behind.”¹⁶⁹ Rawls later argued that if “formal equality of opportunity” existed without other measures to keep the game even, success would accumulate too much. “Fair equality of opportunity,” where “free market arrangements” were set within an institutional framework to regulate the “overall trends of economic events” and “preserve” social conditions, was better at “eliminating the influence of social contingencies.” But it still permitted abilities and talents to affect the distribution of wealth and income and would be imperfect “so long as the institution of the family exists.”¹⁷⁰ Already, Rawls was clear that equality of opportunity alone was not enough. A just society required further limits to inequality.

By the late 1950s, a number of developments combined to motivate a conceptual change in this account. Rawls began to provide a new interpretation of the “ambiguous” phrases “everyone’s advantage” and “equally open to all.” He looked to those in the “lowest positions,” whom equality of opportunity alone could not benefit. In this, Rawls’s focus was in line with a growing emphasis among Anglophone social scientists on “the poor.”¹⁷¹ On the British left, poverty became central. It was necessary, the theorist of the welfare state Richard Titmuss argued, to extend “the welfare state to the poor.”¹⁷² In the United States, where the civil rights movement was transforming the landscape of American politics and would soon shape liberal ideas of equality and justice, the liberal response was initially to develop a social scientific discourse about poverty. Worries among American pluralists about middle-class alienation and the cultural and psychological consequences of affluence gave way to claims that poverty persisted in its midst.¹⁷³ Optimism about the erosion of class boundaries in post-industrial society was replaced by concerns about “status anxiety” and self-respect.¹⁷⁴ Research into psychological deprivation, delinquency, and “lower-class culture” that focused on poor communities—African Americans and other racial minorities in particular—characterized poverty through the lens of culture and family.¹⁷⁵ Rawls came to emphasize the poor too, though not in the terms of the “culture of poverty” thesis. In lectures in 1959 and then in 1962, he argued that equality of opportunity had to be joined to the difference principle, which arranged inequalities so that they worked to the benefit of the “least-advantaged” members of society.¹⁷⁶ He came again to evoke Tocqueville: social conditions bound together the lowest with the highest via those in between (later Rawls formalized this in what he called “the chain-connectedness” of expectations).¹⁷⁷

This focus on the “lowest positions” followed from Rawls’s preoccupation with moral psychology. In the mid-1950s, his earlier concern with concentrated power, stability, moral worth, and character took a psychological turn as he used Wittgensteinian philosophical psychology to explore morality. After 1958, he taught a class on the “moral feelings,” first as a visitor at Harvard and then at MIT (which he would leave for his permanent professorship at Harvard in 1962). This reflected his interest in the eighteenth-century moral philosophers—Hume, Smith, and increasingly Rousseau—who explored the nature and foundations of human reasoning, sociability, and ethical life in ways that many Anglophone twentieth-century philosophers and economists had ceased to do (with the exception of figures like Keynes, Knight, and Gunnar Myrdal, to whom Rawls was drawn).¹⁷⁸ Rawls used the ideas of these earlier thinkers to develop his own account of moral development, particularly the development of the natural attitudes of love and sympathy, into the “sense of justice” he thought was required for persons to live in an equal and just society.¹⁷⁹

Out of this account came Rawls’s concern with the “special psychologies.” These were dispositions that Rawls thought interfered with a proper analysis of the concept of justice, and that an “open” society should mitigate: envy, shame, and humiliation, and the anxieties caused by risk and uncertainty. Rawls tried to describe what mechanisms and how much equality would be necessary to prevent or mute the psychological harms these dispositions caused. For a society to be stable, for the game to carry on, it had to make psychological sense for the players to play. Properly arranged open societies could deal with envy, which was caused by social failures. The failure of equality of opportunity allowed

the upper classes an unfair advantage in the essential matter of education, and the existence of various restrictive barriers to advance; the existence of various forms of monopoly and economic exploitation, the fruits of which are passed down and perpetuated by inheritance and invested in a class ownership of capital and land, and the like; and a failure to maintain certain forms of a social minimum which are needed to set a floor to the standard of life and to provide insurance against hardship falling on the least fortunate.¹⁸⁰

Those who occupy the “lower positions,” Rawls thought, will inevitably experience envy, which was produced by displays of “good fortune.” For society to be stable, the effects of envy had to be constrained.¹⁸¹ Like many Cold War liberals, Rawls thought some anxiety and uncertainty were the costs of an open society.¹⁸² But he wanted anxiety to be borne proportionally to “voluntary” risk-taking. Those who took risks should be those in high places—“politicians” and “entrepreneurs” in capitalist regimes, “managers” in socialist ones.¹⁸³ What now mattered to Rawls was that whatever inequalities existed should be psychologically acceptable to those in the lower positions, who were most liable to suffer from the special psychologies. An open society would sufficiently reduce both the absolute and relative differences between people so as to eliminate the unacceptable effects of envy and the liability to humiliation. Institutions needed to manage both the sources of humiliation—disparities between talents and achievements—and the shame people feel at their failures to make good on “natural gifts” (or the shame of having no natural gifts to begin with).¹⁸⁴

Rawls explored what social mechanisms would reduce the risks facing those in the lowest positions, and what could make envy tolerable enough that inequalities would be justifiable to them. He thought raising the social minimum or reducing the general level of competitive rewards might do the trick. Dulling the threat posed by the special psychologies might require redistribution, but it could take other forms. A good moral education—in the family and in smaller associations—would limit the predisposition to “neurotic anxiety” in conditions of uncertainty.¹⁸⁵ Just as Rawls had suggested that it was important that the players of the game think they have equal starting places, he now also implied that “displays” of fortune were a large part of the problem. The cloaking of visible inequality, even ignorance of relative status, might do additional work in justifying inequality. The experience of relative inequality was enhanced and constituted by the frustration of expectations, which were raised in an open society.¹⁸⁶ When Rawls shifted his argument that inequalities were justified if they benefited every party, he settled on the claim that they were justified if they improved the “expectations” of the “least advantaged.”¹⁸⁷

These ideas placed Rawls at the end of a generation of thinkers who were preoccupied with uncertainty and the task of “counteracting” its effects through different conventions, habits, and techniques.¹⁸⁸ He saw potential solutions everywhere. Limiting inequality was not the only mechanism to keep the game worth playing. He noted other “systems of control” that could stabilize it. Hierarchies, forms of bargaining between interest groups and price systems, and what the political scientist Robert Dahl called “polyarchy,” all provided alternative routes to stability. So did John Kenneth Galbraith’s theory of countervailing powers and the critique of oligopoly that gave a role to government in preventing the formation of power blocs and securing decentralized decision-making in the industrial sector.¹⁸⁹ Organization theorists claimed to provide accounts of how to reach stability through organizational equilibrium. Psychoanalytic ideas claimed to help maintain “the stability and effects of the moral system.” As anthropology provided ways of thinking about the conflicts of values across cultures, it could help smooth over the conflicts that arise when differences of value are recognized.¹⁹⁰ In the end, however, none of these, in Rawls’s view, could secure a “just” stability. Only a set of common rules to which all, including the least advantaged, could agree could do that.

By the close of the 1950s, Rawls thought these rules required much more in the way of government intervention than the maintenance of lighthouses and the enforcement of the rules of the road. The powers of the state in Rawls’s vision were still restricted, but he moved closer to a defense of active redistributive government. The policies of an open society he listed were those of the welfare state: competitive efficiency, full employment, an appropriate rate of growth, a decent social minimum, and redistribution of income and wealth. He now provided a neo-Keynesian account of the role of government and its instruments of economic management.¹⁹¹ Rawls’s idea of a property-owning democracy acquired a new ideological valence, adapted from the British center left, as he grafted their commitment to equality onto his early barebones liberalism.¹⁹²

Rawls also began to explore the question of whether the theory he was developing could fit different political regimes. Later he would argue that justice could be achieved in both capitalist and socialist systems. At one level, the choice seemed unimportant to Rawls. He was more concerned with stability and the fate of democratic publics in an age of expanding states than the choice between capitalism and socialism, particularly compared to those who pointedly abandoned common ownership or argued explicitly that ownership was no longer definitive of power.¹⁹³ If the optimistic sociologists of post-industrial society were to be believed, the line between these two systems was disintegrating anyway. Indeed, Rawls was increasingly optimistic that a just stability might be achievable with minor reforms. The threat of the expansionary administrative state that preoccupied his early thought had been dulled by the promise of the affluent society and the concern over its extension to the poor. “It may be,” Rawls wrote in 1959, that “in a country like the United States at mid-century, in which the reforms of the thirties have set going changes which bring the social system much more in conformity with the requirements of a just social system than it was before, social justice is not the most pressing issue.”¹⁹⁴ Despite the influence on his ideas of liberal socialist theories of equality, Rawls still had a quasi-Hegelian or Wittgensteinian commitment to what was already there.¹⁹⁵ And what was already there was not socialism but a liberal society moving, he thought, toward justice.

As to whether these institutions are better seen in a liberal capitalist framework or under a liberal socialist regime, this question turns on many historical and psychological and other questions (e.g. economic efficiency). Since we are a liberal (relatively) capitalist society rational conservationism suggests that we try to work these institutions in a capitalist framework. We always have to begin from where we are and prima facie our obligation is to attempt to reduce the discrepancy between actual and just institutions in a rational way.¹⁹⁶

With his principles consolidated, Rawls’s theory itself, and the vision of society on which it relied, was starting to exert its own force. Its logic now placed a greater hold on the political positions he was willing to defend.

For Rawls, political philosophy could not be satisfied with a “conception of society and the human person,” nor with “a system of evaluative principles.” He also wanted to provide the vision, the “total picture of man and society,” that tied it all together. In the 1950s, at the same time that Rawls read widely in debates about equality, justice, and ownership, he rendered a variety of disparate ideas philosophically coherent by building them into the architecture of his theory. He now extended his framework for ethical evaluation beyond the realm of individual conduct and particular distributions to the law and institutions of state and society. As Rawls developed the apparatus to house his principles of justice, he began the remaking of political philosophy.

It was a slow process. Rawls’s first challenge was to designate the subject matter of his theory in philosophical terms. He got there by a circuitous route. In the early 1950s, Rawls had engaged with utilitarian ethics and economics. He looked to the debates in ethical reasoning that preoccupied postwar philosophers: the nature of rule-following and the logical status of rules in the context of the moral justification for actions. What kind of reasons could individuals appeal to in explaining their actions? Traditionally, utilitarians saw actions as justified by their utility or consequences. Deontologists argued that right acts should conform to moral rules. When asked when it was permissible to break rules, utilitarians had an easy answer: when the consequences justified it. But what if everyone broke the rules? Classical utilitarianism found it hard to respond. Some tried to rescue a form of utilitarianism by distinguishing between “act” and “rule” utilitarianism.¹⁹⁷ Rule-utilitarians did not decide on a case-by-case basis whether promises were worth keeping. Instead, they justified promises by appeal to the rule that promises are to be kept—the practice of promising.

In Rawls’s contribution to these debates, he introduced a fundamental piece of his theory. In “Two Concepts of Rules” (1955), which explored how different kinds of ethical reasoning related to social forms, Rawls followed a distinction made by Toulmin between the justification of “specific actions” by appeals to “socially recognized” rules and the justification of “the practices which the acceptance of the rules involves by appeal to consequences, fecundity, social welfare etc.”¹⁹⁸ When it came to giving reasons for a particular act, one appealed only to the general rule, the practice, under which it falls—not to general principles. In a chapter entitled “Reasoning about the Justice of Social Practices,” Toulmin drew from Wittgenstein: “Within the framework of a scientific theory, one can ask of most things, ‘Is this really straight?,’ but the criterion of straightness cannot be questioned: within the framework of a particular moral code, one can ask of most individual actions, ‘Is this really right?,’ but the standards of rightness cannot be questioned.” There was a difference between questioning the rightness of a particular action and questioning the “justice of a practice as a practice.”¹⁹⁹

Rawls initially criticized this argument but changed his mind. Rules should not be conceived as “summaries of past decisions arrived at by the direct application of the utilitarian principle to particular cases.”²⁰⁰ As he put it in a remark reminiscent of Wittgenstein: “arguing as if one regarded rules in this way is a mistake one makes while doing philosophy.”²⁰¹ There was an alternative “practice view” of rules, in which rules defined a practice just as the rules of games constituted the game.²⁰² Acting under a practice meant giving up appeals to general principles. A game might be justified on utilitarian grounds, but the players of the game played by the rules. Rawls made a distinction, common to J. L. Austin, Stanley Cavell, John Searle, and Anscombe, who distinguished between “brute facts” recording sense experiences and “institutional” facts that required a set of rules to be understood.²⁰³ Rules were “constitutive” of actions. The rules of an institution defined a class of actions that presupposed the acceptance of the existence of these rules.

Each institution had its own principles, norms of reasoning, and “universes of discourse.”²⁰⁴ Individual agents needed to be understood against the background of practices in which they took part. For Wittgenstein, practices and forms of life referred to implicit modes of conduct. No decision procedures were necessary to reach agreement or make judgments: we already have agreement; where we do not, we have ways of getting on with things. The ethical philosophers who drew from him were concerned, however, with the question of what the “correct” procedures for judgment were, and the constraints imposed on them by facts about the world.²⁰⁵ A proceduralist interpretation of Wittgensteinian ideas was taken up by a number of philosophers, including H.L.A. Hart. Rawls wanted to find procedures for judging practices that were more settled, less indeterminate, and far larger than ball games. He put these ideas to work when he listed examples of practices: “games and rituals, trials and parliaments, markets and systems of property.”²⁰⁶ He was interested in institutional forms that required justification and evaluation—what his sociological contemporaries called social “systems” or “institutional equilibriums.”²⁰⁷

Rawls was especially concerned with the practice he came in the late 1950s to call the “basic structure”—the practice to contain all practices, the boundary of a society. Rawls squeezed many postwar ideas about society and the state into this concept. The basic structure included the rules of the game; it was made up of the social institutions that determined the starting place of players and their chance of winning. It also carved out the domain of politics as the ground rules of society, not the plural associations within it. It was not the centralized state as conceived in state theories that described states as agents or as legal persons acting as a corporate entity through its representatives.²⁰⁸ The basic structure was a practice, not an agent, and practices did not act. The conception of the state as a practice precluded a view of the state as an active interventionist agent. This was no coincidence: Rawls was building a vision of the state and society to fit his wariness of intervention.

The idea of the basic structure encompassed the minimalist view of the state Rawls put forward in his dissertation, but he now tempered that view to more closely resemble the complex of laws, courts, agencies, and services, both public and private, that made up the American administrative state.²⁰⁹ Yet Rawls was not much interested in studying what the state was. In the aftermath of a war in which states had caused so much destruction perhaps that seemed obvious. But Rawls’s thought was also shaped by a constitutionalist anxiety about expansionary administrative and executive power, which had trained the institutional imagination of American postwar political scientists in a different direction from their European counterparts, with their long tradition of state theory: they looked away from the state in efforts to study behavior, policy, and “process” that “demystified” the state in the name of the “stability of the democratic system.”²¹⁰ The functionalist view common to pluralist political science had disaggregated state institutions and detached them from the passions, conflicts, and class analysis of depression and war. Rawls went an extra step: he repackaged these disaggregated parts and put them beyond the main frame of his theory. The basic structure was a practice to be regulated, justified, and evaluated as a whole. The depoliticizing assumptions of anti-interventionist pluralism were baked in. The conception of the state both as a site of administrative conflict and as a distinctive institutional agent remained attenuated in Rawls’s vision. It was not interrogated as a quasi-autonomous realm of agency, power, and interests.²¹¹

Like many of his pluralist contemporaries, Rawls was more interested in nonstate associations—in his case, the family. Indeed, he once described his ethics as “familistic.” Yet Rawls downgraded his pluralism when he adopted the idea of the basic structure, which seemed to collapse the state into the public institutions of civil society.²¹² For a time, Rawls’s distinction between the basic structure and other practices was fluid: he experimented with including in it forms of custom and etiquette and relations within the family and smaller associations.²¹³ This fluidity would be seized on by later critics.²¹⁴ But by the close of the 1950s, Rawls had made the basic structure the primary subject of justice and the practice to which his two principles applied.²¹⁵ Unlike other practices, he saw the basic structure as a nonvoluntary cooperative venture. Both justice and fairness were applications of “reciprocity”—the concept that modeled the relations of recognition between persons—but where the concept of fairness applied to “voluntary” practices, only justice applied to the nonvoluntary realm.²¹⁶ Rawls thus included the family, the labor market, and religion in the sphere of the voluntary, defined by a capacity for exit. As in many liberal theories of civil society, associational life was rendered largely private. Yet neither the basic structure nor the private were ontologically distinctive. The basic structure now became the scene in which all social, ethical, and political life took place. It supplemented the game analogy, and the use of constitutive rules made it cohere. The rules were no longer just regulatory but expressed in individual acts and decisions. The power of the state flowed through the relations it constituted. The shape of the rules was fundamental to the ethos of society.²¹⁷

The basic structure was a development and extension of the idea of society as a game. Rawls had experimented with other metaphors that indicated a self-regulating capacity and limited the need for political intervention, particularly ideas of equilibriums and systems. In this vision, the agential dimensions of the state and other social institutions were attenuated. When Rawls did explore the functioning of institutions, he focused more on the juridical and legislative institutions than the executive or bureaucracy he wanted to constrain. Yet at the same time, the invention of the basic structure gave his theory its institutional casing. Until now, Rawls’s vision of society had flowed from the bottom up, from convention and community. The basic structure made it a political theory and set the institutional character of philosophical debate among Rawls’s followers for decades. It also allowed one way of reconciling the ambiguity of perspective that cut across his thought, from the puzzle of how to understand the general irrelevance of desert alongside rewards for effort to the worry about situating the interpersonal experience of inequality alongside its general character. The principles of justice regulated general institutions, not particular acts. Rawls’s theory would now primarily be read as an institutional one.

What were the philosophical mechanisms by which the society that Rawls envisaged could be judged? Over the course of the 1950s, Rawls extended his “reliable method” to create the conditions for agreement and judgment of the basic structure. He described how Hume had imagined reaching agreement by judging actions from a general standpoint, “a point of view, which, if we take it up, our judgments will be the same.”²¹⁸ For Hume, this had been the standpoint of a representative person, the impartial spectator. Rawls’s contemporaries adapted this idea to produce a “moral point of view” from which to construct an objective ethics.²¹⁹ Roderick Firth provided an ideal observer—a God-like perspective through which to solve ethical disagreement.²²⁰ Others, including Rawls, tried to formulate a moral point of view, “a standpoint from which we all say one thing,” to find standards of social justice.²²¹ Economists experimented with similar hypothetical procedures: in an essay on risk-taking, John Harsanyi provided a procedure for grounding principles (in his case using a group choice situation to justify utilitarian principles).²²² Rawls soon departed from ideal observer theories, with their single decision-maker and religious overtones. While observer theories required one perfect individual, Rawls sought agreement among imperfect ones. While they assumed perfect knowledge, Rawls restricted knowledge. In this, he also departed from general equilibrium theory, with its assumption of perfect knowledge of prices.

Rawls wanted to connect these ethical procedures to his moral psychology and conception of natural persons. He argued that people had a natural “disposition” to act justly and to recognize others. This disposition was “blocked,” or it “broke down,” in certain societies: in “a corporate society,” or in a society grounded in “slavery and aristocracy,” where “justice as reciprocity does not arise.” Other kinds of social circumstances “release the disposition” for equality. “Industrialization and the modern temper has produced social circumstances releasing this disposition and blocking various inhibitors of it,” he wrote, hence the modern “demand for justice” made within “socialism and allied ideas and movements.”²²³ Rawls thought that a decision procedure could simulate the blocking of these inhibitors and thus simulate the conditions of equality that were potential in all moral relations. He searched for “the conditions of considered judgments” to “set the stage” for the natural reactions, of sympathy and compassion—the reaction of recognizing “persons as persons.” These

remove anything that may distort it or inhibit it. They allow free and full play for the natural reactions of recognition; and in this way, the explanation for the agreement which follows may lie. A failure to agree, at least in the clear and concrete cases, would mean that something were missing in this reaction.²²⁴

The point was to model what persons would be like if their relations were unmediated.

This was part of the aim of the hypothetical device that came to be associated with Rawls’s description of the theory of justice as part of “the theory of rational choice.”²²⁵ Into the 1960s, Rawls took on new tools for modeling that perspective, yet the point remained the same: to find a way of imagining what it was like to be ethical and equal. He sought to find a point of view from which agreement could be conjured. Then ethical life could be understood as Rawls thought it actually was, if only we could see it clearly. This was the core idea behind his “analytic construction” and its most famous interpretation, the original position, with its veil of ignorance. Behind the veil, people were blinded to their talents and their starting places in the game of society, which were products of luck and fortune and not themselves deserved. It was from that point of view that the principles of justice to apply to the basic structure of society could be agreed upon and the game of society could be judged to be under control, stable, and worth playing, with its successes fairly divided.

What was striking about Rawls’s original position was how many of his concerns it incorporated. The choosing “parties” were representatives, chosen from different parts of society, but with their knowledge of their place in society denied to them. The parties had general facts available to them: facts about the “circumstances of justice” that made the coordination of distribution necessary, that included facts about psychology, pluralism, diversity, and the fact of “moderate scarcity.” They were not “everyday life” persons, but their choices were still “constrained by having a morality.”²²⁶ In seeking a distribution, they advanced their own ends.²²⁷ They were not rational egoists.²²⁸ The tendencies that obstructed the natural moral feelings and attitudes were removed so the parties would not be “anxious or insecure,” liable to humiliation, shame, envy, or the special psychologies that made people turn inwards. Since in a properly just society these psychologies would not be a problem, their removal from the analytic construction was justifiable. Rawls’s reasoning here was deliberately circular. In a just society, moral education in the family and smaller associations would allow for the development in each person of a sense of justice that led them to support the principles of justice. The “aberrant propensities” of childish passions like envy would be cleansed.²²⁹ Envy would be diminished both by just institutions and by cloaking differences in fortune and preventing the consequences of their visibility. In the choice situation, these could thus be cloaked behind the veil of ignorance.

Rawls eventually described the choosing parties, not as persons, but through their position in society: “continuing persons,” or “genetic lines,” best understood as “heads of associations or group or firms.” With a nod to the Jeffersonian tradition in which white male property owners were the primary agents of democratic society (where property often included women and slaves), Rawls would settle on “heads of households.”²³⁰ This was meant to ensure that the parties chose the principles with an eye to the future—to their descendants, at least two generations along. Rawls’s analytical construction thus entrenched the family as both the vehicle for the moral claims of the future and the primary site of meaning, education, and development. The partial and associative nature of society and Rawls’s attachment to the family were simulated all the way down and built into his theory at multiple levels. “He was a firm believer in paternal supremacy,” Judith Shklar would write of Rousseau. It was “the one form of inequality that he did not even recognize as such, because he did not think that it created any of the emotional miseries that every other sort of inequality brought with it.”²³¹ The same would soon be said of the increasingly Rousseauvian Rawls by his feminist critics.²³²

Rawls still aimed at the “rational choice” he had tried to secure by appeal to the “free and uncoerced opinion” of “average men.” He was committed to the idea that the criterion for judgment should be set by a plurality of common men—a democratic “discussion” modeled on a court or jury rather than an expert administrator. But as he developed A Theory of Justice, Rawls gave up on the “discussion” emphasis. He looked instead to a collective decision, or what he came to call the contract. The contract began as the localist, democratic core of Rawls’s theory: it did not model a founding moment but simulated the democratic discussion and formalized Tocqueville’s town hall meeting. There was no multitude; Rawls’s aversion to the idea of atomistic individuals contracting to set up society had not gone away. Yet it did model agreement and his quasi-deliberative procedure. It also therefore modeled his commitment to justice as an alternative to intuitionism and utilitarianism.

In his vision of society as a game, Rawls had experimented with the early utilitarian tradition of political economy insofar as it extended the commitment to commercial society with a minimally interventionist state. Yet he worried a great deal about a different, statist species of utilitarianism. Later, Rawls’s critique of utilitarianism would be summarized by one pithy phrase: that utilitarianism did not take seriously the “distinction between persons.”²³³ By aggregating and maximizing, and by ignoring distributive questions, utility principles failed to protect society’s losers and minorities and threatened what Rawls came to call the “inviolability” of persons, violating the rights of individuals. But initially Rawls was also saying something more. Utilitarianism failed to take seriously what morality was like—that it rested on the recognition of persons and their community. It was a “technocratic,” “administrative” theory that was excessively “individualist.”²³⁴ When utilitarianism took persons in the aggregate, it individualized them and eliminated the plurality of their relations. Members of a society were not isolated “persons who stand as claimants on an administrative or benevolent largesse,” but participants in a mutually advantageous cooperative venture. It was, Rawls wrote, a “fundamental mistake to apply the principles of higher order executive decision to justice.” By treating citizens as “separate lines for the assignments of benefits,” utilitarianism ignored both Rawls’s idea of a consensual moral community and its political correlate: democratic participation through discussion.²³⁵ Alongside other arguments for why his two principles of justice were superior to those of utility, the contract helped express Rawls’s opposition to this species of “individualist” utilitarianism—the maximizing, aggregating tradition of government that decided on social policy by summing the well-being of a given population and that was associated with the bureaucratic executive governance of the midcentury state. Rawls’s ethics was reciprocal and interpersonal. Utilitarian politics neglected the interpersonal.²³⁶ Rawls’s contract was an extension of his discussion model; utilitarianism, in this vision, was the extension of executive power and the administrative state.

Rawls would discard some of these ideas as he struggled to make his theory coherent and to meet the objections of early readers. He carried on working at the theory throughout the 1960s. By 1964, he had completed a first full draft of the book he published seven years later.²³⁷ In subsequent years, he gave his lifelong commitment to moral persons and procedures a Kantian form. He instated an interpersonal account of justification he called “reflective equilibrium.” This final test, by which individuals decide whether they would accept the principles of justice as chosen in the original position, consisted in reflecting and revising our “considered judgments” about particular cases, the principles and rules that govern them, and other considerations that bear on our acceptance of these judgments.²³⁸ He refined his account of merit and desert, his equality of opportunity principle, the account of background conditions for justice, and the priority ordering of the principles. He transformed his earlier moral psychology into an account of the right and the good, the ends of justice, and the primary social goods and social bases of self-respect that addressed the questions of envy and the special psychologies via a different route.²³⁹

Yet at the end of the 1950s, the foundations of his theory of justice were set. The basic structure, the original position, the principles of justice, and the challenge to utilitarianism and intuitionism were in place. These ideas reflected Rawls’s efforts to build a new kind of theory out of the resources of postwar philosophy and the politics of postwar liberalism. Liberal anxiety about centralized power found philosophical expression. So did a distributional and institutional theory of equality. Rawls wanted the least fortunate to be compensated, but he was anxious that the midcentury expansion of the administrative state not eliminate the pluralist life he held dear. By the time Rawls had developed the ideas that made his theory “come alive,” his view of the role of government went beyond the control of traffic. Despite the persistence of poverty amid affluence, he shared the optimism of the theorists of post-industrial society. He thought it possible to provide objective standards for judgment and to imagine publics as more than a mere aggregation of individuals. Taking up a general point of view gave Rawls what he believed was already known—an idealized vision of social and ethical life, inseparable from a pluralism in which the firm, the church, and the family were naturalized associations, at the core of which consensus could be revealed by the right philosophical devices.

In the coming years, Rawls would train a generation of moral and political philosophers at Harvard, many of whom went on to teach philosophy elsewhere.²⁴⁰ Encouraged by a novel public discourse of justice and a growing number of philosophical publications on social and distributive justice theory, Rawls argued in his lectures to these students that the study of justice would revive political philosophy.²⁴¹ He also imparted to them his faith that the deepest structures of society were changing for the better. The right kind of structure could protect the human form of life and its associational flourishing. The practices of a constitutional democracy could correct for the effects of luck, contingency, and fortune, without excessive intervention. His two principles could show not only how to understand the fairness of particular distributions but how to judge the justice of a system of rules and procedures: the rules of a social game that, if set up correctly, would allow for fair equality of opportunity, the wide dispersal of capital, and a collective political life in which citizens participated as equals and politics could not be bought.

As Rawls carried on building his theory, a different kind of politics would start to intrude. His consensual vision of ethics and society, set into the foundations of his theory, had been enabled by the postwar ideology of political consensus.²⁴² For Rawls, that consensus with its few “practical distractions,” had made it the perfect time to do “abstract” philosophy and dig down to fundamentals.²⁴³ Such philosophy could seem “of little practical importance” in periods of calm, he admitted. But, Rawls suggested, that view “would dissolve quickly enough, should this consensus, or apparent consensus, break down.”²⁴⁴ If there ever had been a consensus, it was shattering as Rawls wrote.