David Butler, who has died aged 98, taught the British to understand elections. He was not the first to try, but he was much the most successful and influential, notably on BBC TV election night coverage from 1950 to 1979, and then on the radio and in the press until 2010.
He did not invent the term psephology (from the Greek word psephos meaning a pebble, since the Greeks used pebbles to vote), but he put it into the language. He did not invent the concept of a “swing” – the shorthand way of expressing the relationship between a party’s performance and that of its rivals – but he popularised it and made it the basis of all subsequent electoral analysis. He did not invent, only rediscovered, the cube law – the formula used to translate a given share of the vote in a first-past-the-post election into the likely number of seats that each party would win – but he brought it into general use.
Above all he was a teacher. At his seminars at Nuffield College, Oxford, on and off the platform at conferences, over lunches and suppers, on buses ferrying political reporters from one press conference to the next, he gossiped, speculated and reminisced; but always, he taught. Having done his research degree at Nuffield (1949-51) he continued as a research fellow, and then a fellow (1954-92).
He published a vast array of books, not all about elections, producing standard texts on the consequences of hung parliaments and who would have to do what if an election failed to give any party a working majority.
Failure in British Government: The Politics of the Poll Tax (1996), written with Andrew Adonis and Tony Travers, detailed a stupendous governmental fiasco. Political Change in Britain (1969), with the US political scientist Donald Stokes, provided the most thorough analysis of its day of what drove people to vote as they did – alongside class, a strong impulse was people’s tendency to vote as their parents did.
The histories of successive elections that came out of Nuffield became an institution. The first was produced after the 1945 election by RB McCallum and Alison Readman. McCallum could not handle statistics, but Butler could, and he furnished the statistical appendix to the British General Election of 1950, by HG Nicholas. Part of McCallum’s initial purpose, which then became Butler’s, was to purge the process of myth and wishful thinking.
In tandem with successive co-authors, Richard Rose (1959), Anthony King (1964 and 1966), Michael Pinto-Duschinsky (1970) and from 1974 until 2005, Dennis Kavanagh, Butler broadened the concept through interviews with the principal players in each election.
The books were often criticised for failing to be more adventurous, but Butler wanted analysis firmly rooted in verifiable fact; as much so as his other great project, British Political Facts, co-written first (1963) with Jennie Freeman, then (1980) with Anne Sloman, and finally, culminating in the end-of-century edition (2000) with his son Gareth.
This allegiance to verifiable fact reflected his upbringing. His first – wholly unaware – taste of elections came while he was still in the womb, when his mother, Margaret (nee Pollard), canvassed for his grandfather, the London University historian AF (Albert) Pollard, standing as a Liberal for the Combined Universities seat (those other than Oxford, Cambridge or London). The addiction to fact, though, may have come more from his father, Harold Edgeworth Butler, a don at New College, Oxford, who kept two copies of the Dictionary of National Biography in the house, ready for use should some dispute emerge.
From St Paul’s school, London, David followed his father to New College. There he studied philosophy, politics and economics, with a break of two years for service as a lieutenant in the Staffordshire Yeomanry, crossing the Rhine as a tank commander. After completing his degree he went to Princeton University as a visiting fellow (1947-48), but apart from a spell as personal assistant to the British ambassador in Washington (1955-56) he remained firmly moored at Oxford thereafter.
However, he was constantly off on the Paddington train to commune with politicians; also with journalists, whose society he seemed to enjoy, whom he constantly probed for information and, rather less blatantly, taught.
If you met him by arrangement, he always came armed with questions. Even if you met him by accident, some kind of interrogation was not long delayed.
In the 1950 and 1951 elections, he sat alongside more eminent pundits on television and radio, keeping them fed with relevant facts and figures. After that he came into his own as a somewhat more organised Patrick Moore of election night, his enthusiasm as boundless as his erudition.
One night a TV political magazine called Gallery was interrupted by news that a Labour MP had died. That, Butler commented, almost before the news was out of the presenter’s mouth, would not be a problem for Labour, since the late MP had an unassailable majority – which he cited, correctly, from memory. It sounded horribly cold and unfeeling: in fact, it was merely an engine of verifiable fact going into characteristic overdrive.
His double act – mostly lively and cheerful, though sometimes a little uneasy – with the ebullient Canadian politics don Robert McKenzie, became an enduring ingredient of every election night. Butler was, in the terms of those days, a telly don; and therefore, to the academic establishment, suspect. Some senior dons resented the activities of colleagues in constant demand from the media, which was said to be one reason why David Butler remained Dr Butler, rather than Professor.
A profile in the mid-1970s described Butler, then 50, as “grey and dishevelled”. There were fears now and then that his energy might be starting to fade. These were confounded. As he entered his 70s, he seemed if anything to be even more insatiably active than before.
He was always just back from Australia, or packing his bags for India; sometimes both.
In 1962 he married Marilyn Evans, daughter of the Fleet Street labour reporter Trevor Evans. She went on from journalism to become a literary scholar, and a professor at Cambridge.
By the 90s a new breed of pundits, some of them Butler’s pupils, were filling the studio chairs which used to be his and McKenzie’s. On election nights, you found him on radio. His other main media outlet was commentary on polls and related matters for the Financial Times. And the changing nature of British elections – the collapse of old tribal loyalties, the rise of tactical voting, the readiness of electors to make decisions pragmatically – had begun to undermine some formerly solid psephological assumptions developed in the very different political world of the 50s and 60s.
Butler’s “cube law” was long dead and buried, and even the concept of swing, with its inherent assumption that local divergencies were likely to cancel each other out, seemed a less dependable tool.
But that did not mean that the maestro, knighted in 2011, had lost his touch. Before the 1997 election, the Reuters news agency asked a panel of experts to predict the result. When it published the findings afterwards, the pundit who had come closest to the actual Labour majority (179 seats) was listed as “unattributable”. Had he been wrong, it might have stayed that way. But in the flush of success, the mask was removed. Behind it was David Butler.
Gareth died in 2008, and Marilyn in 2014. He is survived by his other sons, Daniel and Ed, and seven grandchildren.
• This article was amended on 10 November 2022 to specify in a caption that the image showed the BBC team for the October 1974 general election – there was also an election in February that year.
• David Henry Edgeworth Butler, political scientist and psephologist, born 24 October 1924; died 8 November 2022
