Skip to content
Coordinates of ThoughtHome of Philosophy
Join

Thinker

John Harsanyi

FollowStart a topic about John Harsanyi

Harsányi János

Also known as John C. Harsanyi

1920 – 2000 · Budapest; Sydney; Berkeley

Traditions: Analytic

Schools: Utilitarianism

In brief

John C. Harsanyi (1920–2000) was a Hungarian-born economist and philosopher who worked mainly in Australia and the United States. In ethics he gave two arguments for utilitarianism. In the first, a person chooses a society without knowing which position in it he will occupy and treats each position as equally likely; a rational chooser of this kind maximises average welfare. In the second, his aggregation theorem, rational individual and social preferences under risk imply that social welfare is a weighted sum of individual utilities. He used the first argument against John Rawls’s maximin principle. In game theory he showed how to analyse games in which players lack information about one another, the basis for his share of the 1994 Nobel Prize in economics.

In depth Last reviewed

Connections

Tradition
AnalyticUtilitarianism
Influences
6 sourced relationships
Key terms
Veil of ignorance

Life

Harsanyi was born in Budapest on 29 May 1920 and attended the Lutheran Gymnasium there. In 1937, following his parents’ wishes, he began to study pharmacy; he says his own preference would have been philosophy and mathematics (Nobel autobiography). In 1944 he escaped from a railway station in Budapest just before his train left for Austria. He received a PhD in philosophy from the University of Budapest in June 1947, with minors in sociology and psychology, and later taught sociology there (Britannica). In 1950 he left Hungary for Austria and reached Sydney on 30 December 1950. For most of his first three years in Australia he did factory work while studying economics at the University of Sydney, which gave him a master’s degree in 1953. He met Anne Klauber, a psychology student, who became his wife. After a lectureship at the University of Queensland he took a doctorate in economics at Stanford, where he worked with Kenneth Arrow; Britannica dates it to 1959. He then held a research post at the Australian National University and a professorship at Wayne State University in Detroit, and from 1964 taught at the University of California, Berkeley. He shared the 1994 Nobel Prize in economic sciences with John Nash and Reinhard Selten. He died in Berkeley on 9 August 2000.

Problems

  • What would an impartial person choose, and why should that be what morality requires?
  • Can rational choice under uncertainty settle questions of justice?
  • Is it rational to choose as if the worst outcome will happen, as Rawls’s maximin rule recommends?
  • Can one person’s welfare be compared with another’s, and on what basis?
  • How should we model strategic choice when players do not know each other’s preferences?

Core ideas

The impartial chooser

In papers of 1953 and 1955 Harsanyi asked what a person would choose if he had to pick a social arrangement without knowing which position in it he would occupy. Ignorance of one’s own identity secures impartiality (SEP, ‘Original Position’). Harsanyi held that under such uncertainty it is rational ‘always to choose according to the principle of insufficient reason’, that is, to treat every position as equally likely. A chooser who maximises expected utility on that basis will prefer the arrangement with the highest average utility. Morality, on this view, is the choice an impartial and rational person would make, and that choice is average utilitarian. The device resembles Rawls’s later original position, but the two reach opposite conclusions.

The aggregation theorem

Harsanyi’s second argument does not use a veil of ignorance. It starts from three assumptions: individuals’ preferences satisfy the von Neumann–Morgenstern axioms of rational choice under risk; social preferences satisfy the same axioms; and society prefers one prospect to another when every individual does. From these he proved that the social welfare function must be a positive weighted sum of individual utilities. Kenneth Arrow, in his National Academy of Sciences memoir, says this ‘leads to a justification of classical utilitarianism from a remarkably new point of view’ and calls the two papers ‘startlingly original’. The result is formal: it shows that a certain structure follows from assumptions about rationality, leaving open what the individual utilities measure.

Against maximin

Rawls argued that parties in the original position, who do not know their place in society, would use the maximin rule: choose the option whose worst outcome is best. This leads them to his two principles of justice rather than to utilitarianism. Harsanyi replied, notably in a 1975 critique in the American Political Science Review, that maximin is irrational. The SEP reports that ‘Harsanyi and other orthodox Bayesians contend that maximin is an irrational decision rule’, giving examples in which it rejects an option with a high chance of a good outcome because of a tiny chance of a bad one. Under the equiprobability assumption, the parties would choose average utility instead. Rawls answered that such examples show only that maximin is usually irrational, not that it is irrational in the special conditions of the original position, and he denied that the parties are assumed to be averse to risk (SEP).

Rule utilitarianism

Harsanyi defended rule utilitarianism rather than act utilitarianism, in papers such as ‘Rule Utilitarianism and Decision Theory’ (Erkenntnis, 1977) and ‘Rule Utilitarianism, Rights, Obligations, and the Theory of Rational Behavior’ (1980) (Arrow memoir bibliography). The titles indicate his approach: he treated the choice of moral rules as a problem in decision theory and connected it to rights and obligations. The details of his arguments were not checked in an opened source.

Games of incomplete information

Standard game theory assumed that players know each other’s preferences. In three articles of 1967–68, ‘Games with Incomplete Information Played by ‘Bayesian’ Players’, Harsanyi showed how to drop that assumption. He postulated that ‘every player is one of several ‘types’’, each type with its own preferences and its own probabilities over the other players’ types (Nobel press release). Under a consistency requirement on these probabilities, any game with incomplete information can be turned into an equivalent game with complete but imperfect information, which standard methods can handle. The Nobel press release says that this approach ‘may be viewed as the foundation for nearly all economic analysis involving information’. The approach also models rational agents as Bayesian reasoners, a link to his views on rational choice in ethics.

Works

  • Cardinal Utility in Welfare Economics and in the Theory of Risk-Taking 1953Short note in the Journal of Political Economy introducing the impartial choice argument.
  • Cardinal Welfare, Individualistic Ethics, and Interpersonal Comparisons of Utility 1955Journal of Political Economy 63; contains the aggregation theorem.
  • Games with Incomplete Information Played by ‘Bayesian’ Players, I–III 1967–68Management Science 14; the work cited by the Nobel committee.
  • Can the Maximin Principle Serve as a Basis for Morality? A Critique of John Rawls’s Theory 1975American Political Science Review 69; his main reply to Rawls.
  • Essays on Ethics, Social Behavior, and Scientific Explanation 1976Collected papers on ethics and philosophy of science.
  • Rational Behavior and Bargaining Equilibrium in Games and Social Situations 1977Book on rational choice and bargaining.
  • A General Theory of Equilibrium Selection in Games 1988With Reinhard Selten.

Influences and relationships

6 sourced relationships. Open one to read why, and the source.

Shaped or answered John Harsanyi 5

  • Criticised byJohn Rawls1921 – 2002

    Rawls replied that maximin can be rational in the special conditions of the original position even if it is usually irrational, and denied assuming that the parties are risk-averse (SEP).

    Source: Stanford Encyclopedia of Philosophy · stated on the profile of John Harsanyi

    John Rawls’s relationships

  • Criticised byAmartya Senb. 1933

    Sen, with John Weymark, argued that Harsanyi’s theorems show only that overall good is a sum of von Neumann–Morgenstern utilities, not that it is a sum of well-being, so utilitarianism does not follow (Greaves).

    Source: philarchive.org · stated on the profile of John Harsanyi

    Amartya Sen’s relationships

  • Influenced byJohn Nashno profile yet

    Harsanyi says his interest in game theory was first aroused by Nash’s four papers; Britannica says he extended Nash’s foundations of game theory.

    Source: nobelprize.org · stated on the profile of John Harsanyi

  • Criticised byJohn Weymarkno profile yet

    Weymark pressed the same objection as Sen to Harsanyi’s claim that his theorems justify utilitarianism (Greaves).

    Source: philarchive.org · stated on the profile of John Harsanyi

  • Influenced byKenneth Arrowno profile yet

    Harsanyi names Arrow as his adviser at Stanford; Arrow, in his memoir, says he was not strictly the official supervisor but that they had many discussions.

    Source: nobelprize.org · stated on the profile of John Harsanyi

John Harsanyi answered 1

Reception

In economics Harsanyi is best known for the theory of Bayesian games, which the Nobel press release said ‘may be viewed as the foundation for nearly all economic analysis involving information’. His ethical work had a mixed reception. Arrow’s memoir notes that philosophers were not very responsive to it. Nonetheless his impartial choice argument is the standard utilitarian rival to Rawls’s use of the original position, and the Harsanyi–Rawls exchange is a fixed point in teaching on distributive justice. His aggregation theorem is widely discussed in welfare economics and population ethics. The main philosophical criticism, by Sen and Weymark, is that his results concern von Neumann–Morgenstern utilities and so may not establish utilitarianism about well-being; Hilary Greaves has argued that this criticism can be answered.

Disputed interpretations

Would rational parties behind a veil choose average utility or Rawls’s principles?

Harsanyi held that, treating each position as equally likely, they would maximise average utility; Rawls held that the special conditions of the original position make maximin rational. The disagreement turns on whether equiprobability is a reasonable assumption when the parties have no basis for probabilities (SEP).

Do Harsanyi’s theorems prove utilitarianism?

Sen and Weymark argue that the theorems show at most that the good is a weighted sum of von Neumann–Morgenstern utilities, which need not measure well-being. Greaves (Utilitas) argues that Harsanyi went as far towards defending ‘utilitarianism in the original sense’ as could coherently be asked.

Can interpersonal comparisons of welfare be made?

Both of Harsanyi’s arguments need some way to compare or weigh different people’s utilities, as the title of his 1955 paper indicates. Critics doubt that such comparisons have an objective basis; his own account of how they are made was not checked here.

Primary texts

  • Cardinal Welfare, Individualistic Ethics, and Interpersonal Comparisons of Utility (1955)Journal of Political Economy 63: 309–21; no free legal full text located.
  • Can the Maximin Principle Serve as a Basis for Morality? (1975)American Political Science Review 69: 594–606; no free legal full text located.

Sources

Scholarly sources and reference works, free to read online.

Guides

In the conversation

No topics yet. Start one about this thinker.

Cite this page

Written and kept up to date by the Coordinates of Thought editors. Each statement was checked against the sources listed above before publishing; this page is not peer-reviewed. Last updated 30 October 2026. How to cite this site.

APA (7th edition)

Coordinates of Thought Editors. (2026, October 30). John Harsanyi. Coordinates of Thought. Retrieved [date accessed], from https://coordinatesofthought.com/thinker/john-harsanyi/

MLA (9th edition)

“John Harsanyi.” Coordinates of Thought, 30 Oct. 2026, coordinatesofthought.com/thinker/john-harsanyi/. Accessed [date accessed].

Chicago (17th edition)

Coordinates of Thought Editors. “John Harsanyi.” Coordinates of Thought. Last modified October 30, 2026. https://coordinatesofthought.com/thinker/john-harsanyi/.

BibTeX

@misc{cot-john-harsanyi-2026,
  author       = {{Coordinates of Thought Editors}},
  title        = {John Harsanyi},
  howpublished = {Coordinates of Thought},
  year         = {2026},
  month        = oct,
  url          = {https://coordinatesofthought.com/thinker/john-harsanyi/},
  note         = {Last updated 30 October 2026}
}

RIS

TY  - ELEC
AU  - Coordinates of Thought Editors
TI  - John Harsanyi
T2  - Coordinates of Thought
PB  - Coordinates of Thought
PY  - 2026
DA  - 2026/10/30
UR  - https://coordinatesofthought.com/thinker/john-harsanyi/
LA  - en
ER  -

For reference managers (Zotero, Mendeley, EndNote): BibTeX file · RIS file