The work and life of Robin Farquharson

Abstract

The first phase in the history of modern social choice theory comprised two fundamental pieces of work: first, the articles on the theory of voting by Duncan Black, published in 1948 and 1949, and incorporated into a more systematic treatment in his book The Theory of Committees and Elections of 1958; and, secondly, the celebrated short book Social Choice and Individual Values of 1 95 1 by Kenneth Arrow, of whose publications this conference is, a little belatedly, celebrating the 50th anniversary. Duncan Black's work was of course strictly confined to the theory of voting. If it be taken as essential to social choice theory that it generalises the theory of voting to cover all circumstances in which the preferences of many individuals as between several possible outcomes are by some means or other aggregated into a resultant outcome, or social decision, then Arrow's treatise was the founding text of modern social choice theory as such. The second phase in the history of modern social choice theory, which closely followed the first, and has been largely neglected by subsequent contributors to the field, consisted of the work of Robin Farquharson. Robin Farquharson was an exceptionally brilliant young man, of South African origin, who was born in 1930 and came to Oxford University in 1950, having already qualified for a B.A. in a South African university. He became a member of Brasenose College, Oxford, and undertook an undergraduate course in Politics, Philiosophy and Economics; he sat for his Final examination in 1953. While studying politics as an undergraduate, he was greatly interested by the voting in the U.S. Senate on the League of Nations Covenant after the first World War. It struck him that the outcome was a result desired by very few of those taking part in it, and he started devising examples to see how often this happened; in doing so, he discovered for himself Condorcet's voting paradox that majority preference can be intrasitive. His tutor, Norman Leyland, directed him to the articles of Duncan Black and the book of Kenneth Arrow. He discovered the pamphlets of C. L. Dodgson (Lewis Carroll) for himself in Christ Church Library, and studied the classic work of von Neumann and Morgenstern on the theory of games. All this led him to become interested, indeed obsessively interested, in applying the theory of games to voting procedures, as von Neumann and Morgenstern had not done, and to strate-