The House Opinion Article | The Professor Will See You Now: Records

Illustration by Tracy Worrall

4 min read13 min

The story goes that in early 1945 Ronald McCallum was sat reading in the university library when he suddenly remembered he was supposed to be at a committee meeting to discuss future research projects. Having given it no thought, and with not a single idea ready to propose, he set off at pace and out of nowhere it came to him: a study into the forthcoming general election.

So began what is now the longest-running election series in the world. The latest volume, the 22nd in the series, was published recently.

McCallum’s story says a lot about how academic life has changed. He came up with the idea on the hoof and they just gave him the money. These days, I have about a dozen projects for which I can’t get funding and the process of applying makes Parliament’s renewal and restoration programme look snappy. 

That first study on 1945 begins with the Duke of Wellington’s observation that you could no more describe a battle than you could describe a ballroom. “Still less,” McCallum, and his co-author Alison Readman, remarked “can you describe a general election”. They listed a series of ‘named’ elections: 1874, when the Liberals went down in a flood of gin and beer; the Midlothian election of 1880; the Khaki election of 1900; the Chinese Slavery election of 1906; the People’s Budget election of 1910, the ‘Hang the Kaiser’ election of 1918; and the 1924 Zinoviev letter election. They were sceptical that in reality these issues had ever been so dominant. In 1945, they pointed out, the key issue of the election was housing – yet no one referred to it then (or since) as the Housing election.

Since then, we have had fewer such ‘named’ elections. Even with the 1983 contest, which could easily be known as the Falklands election, there is plenty of evidence that the Falklands conflict was much less significant than people think. The 2005 election could have been labelled the Iraq election, given the extent to which national debate focused on the consequences of the 2003 war, but we know it was fairly low down voters’ priorities. The recent Brexit elections were about a lot more than Brexit.

Still, reading the latest volume (which is a cracker) it occurred to me that we could name 2024 the Roy Castle election – younger readers should ask their parents – because there were record-breakers everywhere. The Conservatives alone managed (i) their lowest vote share, off the back of (ii) the largest fall in vote share of any British party, resulting in (iii) the greatest loss of seats ever. 

The 2005 election could have been labelled the Iraq election, given the extent to which national debate focused on the consequences of the 2003 war, but we know it was fairly low down voters’ priorities

There was (iv) the highest level of vote switching by electors for any contest for which we have data and (v) the most ever seats changing hands, along with (vi) the largest number of incumbents falling from first to third or fourth (all Conservatives). The record for (vii) the largest Conservative-to-Labour swing was broken in 47 separate seats; the largest was in Norfolk South West. 

Labour benefited from (viii) the largest ever winner’s bonus – the gap between the vote and seat share of the winning party – winning a majority (ix) on a lower share of the vote than any other majority government. The combined vote share of the two largest parties (x) was lowest at any election since 1922, with MPs from other than those two parties (xi) at its highest since 1923.

Plus: (xii) the largest Lib Dem number of seats; the best Reform performance in (xiii) seats and (xiv) votes of any radical right party; the best Green performance in (xv) seats and (xvi) votes; more independent (xvii) votes and (xviii) seats than any election since 1945; and (xix) a record number of candidates: 2024 was the first election where every voter had at least five choices on the ballot paper. This is before we get onto the changes in the Commons – a record number of (xx) women, (xxi) graduates, (xxii) those from ethnic minorities, and so on.

And then you look at the current polling and think: what are the odds on some of these records surviving the next election? 

Further reading: R Ford et al, The British General Election of 2024 (2026)