Skip to main content
Where I Live
A-Z Index

BBC News

BBC Election 2005

Watch the BBC Election News
  • Election news alerts
  • Email services
  • Mobiles/PDAs
  • News for your site
Last Updated: Saturday, 7 May, 2005, 08:53 GMT 09:53 UK
Who deserted Labour?
David Cowling
By David Cowling
Editor, BBC Political Research

Labour has won power - but suffered a 6% drop in the share of its vote. The Liberal Democrats were the main gainers.

But just as Labour has been ejected from some of the Tory-leaning suburban seats it won in 1997, its support among key middle class social groups also shows signs of haemorrhaging.

However, it has done better than expected among women, where it eliminated the "gender gap", which in the past meant women were more likely than men to vote Conservative.

See who's up and who's down in the polls since January

But Labour is no longer ahead of the Conservatives among lower middle class voters (C1), which they cultivated in 1997. And its support among working class voters and council tenants has dropped sharply.

However, it has hung onto much of its support among owner-occupiers with a mortgage, who used to be the bedrock of Conservative support.

The Liberal Democrats made strong gains among young people aged 18-24, where they are now ahead of the Conservatives.

But they did less well in Conservative-Lib Dem marginals, where there was actually a swing to the Tories - possibly because they are now seen as to the left of Labour.

The Conservatives, whose total vote share was unchanged, were the main gainers of the reduced Labour share of the vote, especially in southern England. The biggest swing towards them among voting groups was among voters over 65 years old.

There were strong regional variations, especially between the North and Scotland (where the Liberal Democrats did well) and London and the South East, where the Conservatives were the main gainers.

One word of caution: at present we can only extrapolate the demographic results of this election by using the eve-of-poll opinion polls conducted by the main polling organisations like NOP, ICM and Populus.

Given that they came reasonably close to predicting the actual result, this can give us some guidance on which groups switched their loyalty - but it is the trend, rather than the raw figures, which matters.

We can then look at the social composition and swings in key seats to confirm these trends, and also compare them to the exit poll figures in Labour's landslide victory in 1997.


Labour spent much of its effort in cultivating the votes of "hard-working families," and especially women with children in the middle two social groups.

Father and baby
Labour has targeted the family vote
And they appear to have had some success - both ICM and NOP find that Labour received the same amount of support among men and women.

This is an important result - up until 1997 the Conservatives were boosted by the fact that their support among women was around 5% higher than their support among men.

Social change may be affecting this result - with more women working than in 1997, and fewer male trade unionists.

But it may also be a measure of the fact that Labour enjoyed a very large lead on issues like childcare, the health service, and education, that are particularly important to many women with children.


Labour's appeal in 1997 was across all social classes (and all tenures).

Tony Blair managed to win a greatly increased share of the middle class vote, including a plurality over the Conservatives among lower middle class (C2) voters and home owners, while holding on to Labour's working class vote.

Based on our polling data in the 2005 General Election, Labour's support is eroding at both ends.

Two out of three of the opinion polls (NOP and ICM) suggest that the Conservatives narrowly beat Labour among the C2 group this time.

And Labour's share of the unskilled working class vote has declined sharply. According to ICM, Labour was supported by 58% of these voters in 1997, and just 45% in 2005.

However, Labour has retained support among owner-occupiers with a mortgage - who perhaps are giving Labour credit for low interest rates.

They still have a 39% to 30% lead compared to the Conservatives (in 1992, the Conservatives led in this group by 48% to 30%).

Finally, the Liberal Democrats have increased their share of the upper middle class vote - mainly at the expense of the Conservatives.


The Liberal Democrats have made strong gains among young people during the 2005 election, according to an analysis of opinion polls.

Elderly lady watching tv
Older voters tend to support the Conservatives
ICM suggests that they have overtaken the Conservatives among voters aged between 18-24 and also those between 25-34.

This has mainly been at the expense of Labour, who gained a spectacular 58% of the 18-29 age group in 1997, according to the NOP/BBC exit poll. (They have 42% of the slightly small 18-24 group, according to ICM now).

The fall in Labour's support among the middle age group, the 34-64 year old "hard-working families", exactly parallels its support in the electorate as a whole, according to ICM.

The only social group that has remained solidly Conservative are older voters, with the over 65s (and over 55s) giving them around a 6% to 10% Conservative lead - almost exactly the same as in 1997.


The Liberal Democrats were the main gainers in this election in terms of the popular vote - up from 19% to 23%.

Charles Kennedy
Charles Kennedy's party scored some successes against Labour
But their pattern of gains in seats was highly skewed, reflecting how they had positioned themselves in different parts of the country.

They seemed to have succeeded in positioning themselves as the party to the left of Labour over issues like the Iraq war and student tuition fees.

This has helped them in their Labour- Lib Dem marginals where they made some spectacular gains.

But it also meant that they did worse than expected in the Conservative - Lib Dem marginals, where they actually had more losses than gains.

In the 25 most marginal Labour - Lib Dem seats, the swing to the Lib Dems averaged 6.7%.

Overall, they gained 12 seats from Labour, including big gains in some student seats, including a 17% swing in Manchester Withington, a 15% swing in Cambridge, and a 14% swing in Hornsey and Wood Green.

But they lost five seats to the Conservatives, while only gaining three.

In the top 25 most marginal Conservative - Lib Dem seats, there was actually a swing to the Tories of 1.6%

The Liberal Democrats also did much better in Scotland, the North West and the North East - and worst in Devon and Cornwall.

Iraq was a key issue in explaining Lib Dem gains.

An early MORI poll for the GMB showed that among Labour defectors to the Lib Dems, 33% cited Iraq as a key influence on how they would vote.


Labour lost 47 seats, and their share of the vote went down to 36%, compared to 41% in 2001 and 43% in 1997.

Tony Blair at No 10
Tony Blair must get used to working with a smaller majority
They will govern with only 22% of the electorate having voting for them, when taking turnout into account.

Nevertheless Labour has a substantial, though reduced, working majority in Parliament.

This is because the electoral system still favours Labour.

Labour needs a smaller share of the vote to win power because each of their seats is smaller than the average Conservative seat - and this difference won't be corrected until the boundary commission revises the 1997 boundaries (in 2007).

Labour lost votes to a variety of opposition parties, not just the Conservatives, as well - including big swings in Blaenau Gwent in Wales (where a Labour-imposed women candidate was rejected) and Bethnal and Bow in East London, where the anti-war former Labour MP George Galloway defeated the sitting Labour MP.

And it lost heavily among students where both Iraq and student tuition fees provided a potent combination for the Lib Dems.

But there was a strong regional effect, with the biggest swings against Labour (outside student and ethnic minority seats) concentrated in London and the South East.

However, many Labour voters were backed their party because of negative views about other parties (52%) rather than enthusiasm for the Labour Party (45%), according to a Populus poll - and 53% of Labour voters said they would prefer Labour to be re-elected with a smaller majority.


The Conservatives achieved a rather remarkable result.

Their share of the vote has been stuck at around 33% for the last three elections.

Sandra and Michael Howard in Folkestone
Most Tory gains were made in the South of England
Nevertheless, they managed to gain 33 seats - to take their total in Parliament to around 200.

They did this by concentrating their vote more in the seats that mattered most - particularly in the South, and in Lib Dem/Tory marginals.

Only one of the intended targets of Lib Dem "decapitation" - Tim Collins in Westmorland - lost his seat, while key Tories like David Davies, Oliver Letwin, and Theresa May beat off Lib Dem challenges.

The Tories achieved a swing against Labour - especially in London and the South East - partly through a decline in the Labour vote, rather than an increase in their own.

But they returned to power in suburban London, with big swings seats like Putney, Wimbledon, and Enfield Southgate, as well as Home County seats like Welwyn Hatfield, Peterborough, and St Albans.

And they regained similar seats - like Newbury and Guildford - from the Liberal Democrats, while beating off the Lib Dem challenge in all but three seats.

But it was the negative image of Mr Howard among voters - which increased during the campaign - that was their biggest electoral liability.

Therefore his resignation should not come as a total surprise.NOP interviewed a sample of 1000 on 2-3 May, margin of error +/-3% . Populus interviewed a sample of 2042 between 2-3 May, margin of error +/-2%. ICM interviewed a sample of 937 between 1-3 May, margin of error +/- 3%.

% support in 2005 (% in 1997)
Men 33 (28) 38 (47) 21 (17)
Women 32 (35) 38 (43) 23 (17)
AB (middle class) 37 (43) 32 (30) 24 (21)
C1 (lower middle class) 34 (35) 35 (37) 24 (21)
C2 (skilled workers) 32 (28) 43 (52) 18 (13)
DE (unskilled workers) 28 (21) 45 (58) 19 (15)
Age 18-24 24 (25) 42 (50) 26 (17)
Age 25-34 24 (27) 42 (50) 26 (17)
Age 35-64 33 (31) 38 (43) 22 (18)
Age 65+ 42 (38) 35 (42) 18 (15)
Home Owners with mortgage 30 (32) 39 (41) 23 (19)
Home owners owning outright 43 (42) 30 (36) 20 (16)
Council Tenants 16 (15) 56 (65) 19 (13)
ALL VOTERS 33 (31) 36 (43) 23 (17)
Source: ICM. All campaign polls (sample 13,730 in 2005, 10,000 in 1997 weighted for outcome); GB only