Featured, Features | 6th January 2015

A history of every general election since 1945 in 12 graphs

Since 1945, there have been 18 elections. Labour and the Tories have won exactly […]

Photo: Getty


Since 1945, there have been 18 elections. Labour and the Tories have won exactly nine each, although the Tories have held power for slightly longer. They have governed in turn, with each often dominating one decade before capitulating in the next.

The pundits and psephologists who covered these elections – from the BBC’s David Butler and Robin Mackenzie in the 1950s and ‘60s to their more recent successors – tackled familiar questions.

Who are newly powerful minor parties taking votes from? What will the swing between the two main parties be? Can either of them win a majority? Who has our electoral system historically favoured? And when did turnout dip to apathetic levels?

This election isn’t “unpredictable”, and there are parallels with 2015.

They looked to the past to answer them. Now, the historical precedents remain as helpful as ever. This election isn’t “unpredictable”, and there are parallels with 2015 – the rise of small parties isn’t unprecedented.

We can describe post-war politics in seven periods: Labour-Tories-Labour-1970s-Thatcher-Blair-coalition. First, this is who governed, decade-by-decade.

Screen Shot 2015-01-06 at 15.06.49

The first three periods alternated like a pendulum. After the war, Labour ran what remained of the 1940s under the mythologised Attlee government.

The Tories then won every election in the 1950s and had turned Labour’s 146-seat 1945 majority into a 100-seat Tory majority by 1959. They then lost every election in the 1960s, with Labour’s Harold Wilson winning a near 100-seat majority in 1966.

The 1970s defy easy analysis. The Tories’ Ted Heath beat Wilson in 1970, before losing twice to him in 1974. (The two fought for Number 10 four times.)

But after industrial disputes plagued Wilson’s unelected successor, Jim Callaghan, Margaret Thatcher closed out the decade by winning a modest majority in 1979. (Explore the 1970s in-depth, and 45 years of polling, on May2015.)

After that hesitant victory, the Tories collapsed in the polls. As unemployment doubled, rising from 6 to 12 per cent, Tory support fell from the mid-40s to less than 30 per cent.

But the Falklands War, from April to June 1982, reversed that three year slide in two months, and Thatcher went onto win triple-digit majorities in 1983 and 1987.

Thatcher’s majorities rose and fell, but her popularity scarcely changed. Every time she ran for Number 10 she won between 13-14 million votes, and 42-44 per cent of the vote.

Blair’s success was quite different. By his third election, in 2005, he had lost a third of his voters, but held onto power. First, he was elected by 13.5 million in 1997 – Thatcherite popularity.

But he quickly lost Labour’s long-earned support, while keeping the big majorities. No election was more important than 2001, often ignored by history. Nothing seemed to change. Labour lost just six seats, and the Tories won just one. But Blair had quietly lost 3 million voters.

No election was more important than 2001. Blair quietly lost 3 million voters.

Turnout had collapsed. In the fifteen elections from 1945 to 1997, turnout was always above 70 per cent. It had technically fallen to a post-war low of 71 per cent in 1997, but it had been at 72 per cent four times before (1945, 1970, October 1974, 1983) before recovering at the next election.

Then in 2001 it fell to 59 per cent. Labour’s vote share only fell from 43 to 41 per cent, but it had lost a fifth of its voters. Fewer people voted in 2001 than had at any election since 1945, even though there were now 11 million more eligible voters.

This new low, and Labour’s 1990s recovery, meant the Lib Dems won fewer than 5 million votes in 2001, after winning nearly 8 million in 1983, when they stood as the SDP-Liberal Alliance.

But, despite losing 3 million votes, they now held nearly twice as many seats. Never had so many votes had so little effect as those won by the SDP-Liberals in 1983. The centre-left Alliance polled only 700,000 fewer votes than Labour that year, but won just 23 seats. Labour won 209.

The bias in the election system has punished the Lib Dems for 70 years. Since 1945 they have won 14 per cent of all votes cast – and just 3 per cent of seats.

If seats were awarded under a pure form of “proportional representation” (PR), so that each party won seats in the same proportion as they won votes, the Lib Dems would have won more than 100 seats at nine of the past ten elections.

Since 1945 they have won 1,249 fewer seats than they arguably should have.

The failings of the system mattered little in the 1950s or ‘60s. But became meaningful when one in nine voters switched to the Liberals in 1974, and again in 1983.

Other parties have been similarly deprived. At the last election all the parties other than the Tories, Labour and Lib Dems won 28 seats, rather than the 77 PR would have given them.

Their rise has been consistent since the 1970s, when Northern Ireland’s unionists split from the Tories, nationalists parties made their first in-roads, and more independents began to stand.

In 2015, the parties being slighted will change but the issue remains. Ukip will likely be more under-represented than the Lib Dems have ever been.

Farage’s party is set to poll 15-16 per cent, but win just a handful of MPs. Under PR they would win more than 100. And despite losing seven in ten of their voters, the Lib Dems are still likely to be disadvantaged. If they win 7-8 per cent, as they are polling, PR would mean 45 seats, but they are on course for 25-30.

Ukip is set to poll 15-16 per cent, but win just a handful of MPs. Under PR they would win more than 100.

Who has benefitted? The two main parties of course, but none more so than Labour. For fifty years, our system has handed them artificially high majorities (under Blair) or saved them during nadirs (1983 and 2010).

Since 1945 the Tories and Labour have, respectively, won 41 and 40 per cent of all the votes cast across all 18 elections. But our system has given the Tories a slight advantage – they have won 46 per cent of all possible seats – Labour, who won slightly fewer votes, have held more seats (48 per cent).

Here’s how parliament – and history – would have looked if elections had been held under pure PR. Attlee and Heath would have held on to power in 1951 and 1974, but the Liberals would have effectively determined every government.

But the game is the game. And studying it uncovers two surprises.

• First, no leader has ever won more votes than John Major.
• Second, no Tory – or any leader bar Blair – has ever won more seats from one election to another than David Cameron did in 2010.

Could Cameron win more seats this year? It’s unlikely. He will struggle to without winning more votes, but no party has added to its vote share for four elections in a row, and the Tories haven’t meaningfully added to theirs twice in a row since Churchill.

So can the Tories win more seats? The polls imply the opposite: they suggest a 4 per cent Tory-to-Labour swing.

Who’s ahead today? Keep track of all the polls with May2015’s Poll of Polls.

That’s not because Labour are winning over any 2010 Tories – they are winning just 4-5 per cent of those (per The Drilldown) – but because Ukip are still taking more votes from the Tories than Labour, and, most importantly, the Lib Dems collapse has gifted Labour 7 per cent of the vote (around a third of the 23 per cent who voted Lib Dem in 2010 have defected to Labour).

If it wasn’t for the Lib Dem collapse, Labour would currently be languishing in the mid-20s, and Ed Miliband would have been toppled long ago.

Instead, Ashcroft has put Labour ahead in nearly 40 Tory-held seats, and shown the Tories are no threat in almost any Labour seats, which is exactly what the national polls, or swing, implies.

The betting markets expect Labour’s leads to narrow, but they still imply a swing of around 3 per cent. That would be in-line with history. The average swing on election night swingometers has been just over 3 per cent, and, except for 1997, always between 1 and 5 per cent.

Ultimately, the swing is what matters. Ukip, the SNP, Lib Dems and Greens all eventually matter because of their effect on the “Big Two”, however disillusioned voters are with them. This election will be decided by whether Labour can win in places like Ealing and Stevenage, or Colne Valley and Glamorgan. These are Tory seats with majorities of 8-9 per cent that Labour need to win for an overall majority.

This election will ultimately be about whether Labour can win in Tory seats like Chester, Croydon and Stevenage.

To be the largest party, Labour need seats slightly higher up on their target list, like Chester and Croydon, or Wirral West and Worcester, where the Tories won in 2010 by 5-6 per cent. These are the seats Lord Ashcroft has been polling most recently.

His polls are all integrated into Florence, our new election-forecasting machine. She is predicting Labour to be the largest party, but fall short of a majority and 300 seats. For those, like Robin Day, interested in “principles not percentages”, we have been touring the key seats since September, and will be doing so every week until election day.