Kohlberg's Stages of Moral Development

In brief

Lawrence Kohlberg (1927 to 1987), building on Jean Piaget's account of children's moral judgment, argued in his 1958 University of Chicago dissertation that moral reasoning develops through six stages grouped into three levels of two. He tracked development not by what a person decided someone should do when faced with a dilemma, but by the structure of the reasoning behind it, using hypothetical dilemmas, above all the Heinz dilemma, to elicit that structure. Two decades later, Carol Gilligan, who had worked as his research assistant, argued in In a Different Voice (1982) that the scale itself carried a bias: built and first validated on boys, it treated one moral orientation, abstract justice reasoning, as the peak of maturity, and scored a different, relationally oriented voice as underdeveloped.

The full treatment

The problem it answers

Kohlberg wanted to describe moral growth the way Piaget described cognitive growth: an invariant sequence of structures a person builds by reasoning through them, each stage an advance in the kind of thinking available, rather than the slow accumulation of a culture's rules through reward and punishment. This broke with psychoanalytic and behaviorist pictures of conscience as internalized parental authority, and argued moral reasoning matures the way logical reasoning does, culminating, he initially claimed, in reasoning from freely chosen universal principles rather than any particular code.

How it works

Kohlberg interviewed subjects using hypothetical dilemmas and scored not their verdict but the justification behind it, since two people can reach the same verdict for structurally different reasons, then grouped the stages into three levels of two.

Level one, preconventional morality, judges right and wrong by direct consequences to oneself: stage 1 is punishment and obedience (an act is wrong if it gets you punished), stage 2 is instrumental exchange (an act is right if it serves your interests, permitting simple reciprocity).

Level two, conventional morality, judges right and wrong by conformity to the expectations of one's group: stage 3, the "good boy, good girl" orientation, holds an act right if it earns approval and preserves relationships; stage 4, the social order orientation, holds an act right if it upholds law, duty, and the functioning of the larger system.

Level three, postconventional or principled morality, judges laws and conventions against prior principles instead of the reverse. Stage 5, the social contract orientation, treats laws as legitimate because they arise from an agreement serving the general welfare, changeable by fair democratic process when they fail to. Stage 6, the universal ethical principle orientation, holds that right action follows from self-chosen abstract principles, such as the equal worth of persons, honored even against a law or a majority. Kohlberg pointed to Kant's test of universalizability and, after 1971, to Rawls's veil of ignorance as paradigm cases. He later doubted stage 6 was a distinct, empirically observed stage rather than a theoretical limit, and it dropped out of the standardized scoring manual finalized in 1987.

What it claims

The central empirical claim is that the stages form an invariant sequence: everyone who develops moves through them in the same order without skipping, though people differ in how far and how fast they progress, and most adults never reach the postconventional level. Each stage is also a structured whole, so a person's reasoning should hold together across content rather than varying case by case. Movement between stages, following Piaget, comes from exposure to reasoning one stage above one's own, producing a disequilibrium that pulls thought upward.

The key study or demonstration

Kohlberg's central instrument was the Heinz dilemma. A woman is dying of a rare cancer. A druggist has discovered a drug that could save her but charges ten times what it costs him to make, 2,000 dollars against 200. Heinz, her husband, can raise only half the price and asks the druggist to sell it cheaper or let him pay later; the druggist refuses. Heinz breaks into the store and steals the drug. Should he have done it? Kohlberg's 1958 dissertation put this and similar dilemmas to 72 boys aged 10, 13, and 16 in the Chicago area, scoring their reasoning rather than their verdict: a stage 4 subject might say Heinz should not steal because laws protect everyone, while a stage 6 subject might say he should, because a human life outweighs a property claim regardless of what the law says. Kohlberg later followed the same cohort for more than two decades, reinterviewing them roughly every three years. The results, published as A Longitudinal Study of Moral Judgment (1983) with Ann Colby, John Gibbs, and Marcus Lieberman, found the predicted order held for nearly all subjects, with no one skipping a stage, though final stages reached varied and most adults settled around stage 3 or 4.

Related distinctions

Kohlberg separated the structure of moral reasoning from its content: two people who disagree about whether Heinz should steal can be reasoning at the same stage, and two who agree can be reasoning at different ones. He also separated moral judgment from moral action, since later work on the judgment-action link found stage alone predicts behavior only loosely.

Lineage

Kohlberg's direct forerunner was Jean Piaget (1896 to 1980), whose The Moral Judgment of the Child (1932) described children moving from a rigid, punishment-oriented "heteronomous" morality to a flexible, cooperative "autonomous" one. Kohlberg expanded Piaget's two stages into six, extended the range from childhood into adulthood, and gave the postconventional level a philosophical debt to Kant's universalizability test and to the social contract tradition, which he treated, in its Rawlsian form, as the clearest real-world example of stage 6 reasoning.

The strongest case for it

Kohlberg's model gave moral psychology something it had lacked: a testable, stage-based account with a genuine empirical signature, the invariant sequence, that could have failed and largely did not. The 20-year longitudinal study is a rare thing in psychology, a multi-decade panel confirming a directional prediction made in advance. John Snarey's 1985 review of roughly 45 studies across 27 countries found the first four stages held up reasonably well cross-culturally. The theory also reshaped practice: Kohlberg's "just community" schools, run as small democratic institutions where students and staff jointly set and enforced rules, produced measurable stage gains and remain a cited model for moral education.

The strongest case against it

The most influential objection is Carol Gilligan's. She observed that Kohlberg's founding sample was all boys, and that when the instrument was later applied to women and girls, they scored disproportionately at stage 3, oriented toward relationships and approval, while men more often reached stage 4 or 5. Gilligan argued this was not evidence of women's moral immaturity but evidence that the scale, calibrated on one sex, misread a legitimate alternative orientation, an ethic of care reasoning from responsiveness to particular others, as a lesser stage of justice reasoning. Her retelling of the Heinz dilemma through two eleven-year-olds, a boy who treats it as a logical proof and a girl who searches for a way to keep everyone talking, became the emblem of the critique. The full case for the ethic of care built from this observation is treated on its own terms in care-ethics.

The critique did not go unanswered on its own empirical ground. Lawrence Walker's 1984 meta-analysis of more than 100 studies using Kohlberg's own scoring found no reliable sex difference once education and occupation were controlled, suggesting the gap Gilligan reported may reflect her particular samples more than a built-in bias, a point still debated rather than settled.

Two further critiques matter. Richard Shweder, working with Manamohan Mahapatra and Joan Miller on moral reasoning in Bhubaneswar, India, argued that Kohlberg's postconventional stages, built around individual rights and autonomy, encode a Western liberal ethic rather than a culture-neutral peak, since other societies organize serious moral concern around community role and the sacred. And John Gibbs, a longtime Kohlberg collaborator, argued that stages 5 and 6 are a philosophical layer grafted onto a more basic four-stage structure, since unlike stages 1 through 4 they do not appear as a maturational default and instead require exposure to formal philosophy.

Where it stands now

Kohlberg's six stages are no longer a live research program in the exact form he left them. Stage 6 is treated as largely theoretical, the strict invariant-sequence claim has been narrowed by cross-cultural findings like Shweder's, and Gibbs's four-stage revision is common in developmental textbooks as the empirically defensible core. But the basic apparatus, that moral reasoning has a structure separable from its content, developing through identifiable phases that dilemmas make visible, still organizes how developmental and moral psychology frame the question. Gilligan's counter-model did not so much refute Kohlberg as widen what counts as mature moral thought, and the two are now taught together, as an open debate rather than a settled verdict.

Test yourself

Think of a recent decision you judged right or wrong, not the decision itself but the reason you gave for it. Did you appeal to a rule, a consequence for yourself, a consequence for everyone, a principle you would hold even at a cost, or the wish not to hurt someone close to you? Kohlberg would ask which of his six structures that reason fits. Gilligan would ask whether his scale has room for it at all.