Elections satisfy both the practical and the theoretical requirements of classical democratic theory if they answer one question only: Who shall rule? Judged by this test the American elections of 7 November 1972 returned as clear and unequivocal an answer as the United States Constitution permits – crystal-clear as to individuals, equivocal as to parties and political forces. But the student of politics and society cannot resist treating elections as data-gathering devices on a wide range of other questions, on the state of the public mind, on the relative potency of pressure groups, on the internal health of the political parties, and, of course, on the shape of things to come. In this ancillary role American elections, despite the generous wealth of statistical material which they throw up – so much more detailed and categorized (though often less precise) than our own – Suffer in most years from one severe limitation, a limitation which in 1972 was particularly conspicuous; they do not engage the interest of more than a moderate percentage of the American citizenry. In 1972 that percentage was as low as 55 per cent, i.e. out of an estimated eligible population of 139,642,000 only 77,000,000 went to the polls. Since this circumscribes the conclusions which can be drawn from the results themselves, as well as constituting a phenomenon of considerable intrinsic interest, it seems worthwhile to begin any examination of the 1972 elections by an analysis not of the votes counted but of those which were never cast.