Analysis: Population change, turnout & the election

Population change and developments in patterns of turnout are making a substantial contribution to the Conservatives' lead in the polls, analysis of Ipsos's 2009 voting intention aggregate shows. The parties are competing for votes among an older and more middle class electorate than when the Tories last won in 1992, and what is more both the age and class gaps in turnout are bigger than they were then, giving the party strongest among these groups a double boost. The Conservatives are weaker than they used to be both among the old and the middle class, the groups that most strongly support them, and although they have widened their base of support by finding more votes where they have traditionally been weaker, that in itself would not be enough to compensate. It is the increased voting power of the groups where they are strongest which is making up the shortfall.

Last year, voting intention support for the Conservatives in our polls stood on average at 42% - almost the same as the 43% share of the vote the party gained under John Major's leadership in 1992. The table below shows the findings of our aggregate of polls for 2009, combining roughly 12,000 interviews conducted during the year as part of our Political Monitor surveys, and enabling us to examine the demographic characteristics of Conservative support in some depth. We can compare these findings directly with our analysis of the vote in 1992 (based on nearly 15,000 interviews during the election campaign, weighted to reflect the final result).

Source: Ipsos Base: 14,764 British adults 18+, interviewed March-April 1992, data weighted to final election result; 6,289 British adults 18+ interviewed January-December 2009 and "absolutely certain to vote"

The immediate thing that jumps out of the table is that Britain is less politically-polarised than was the case in 1992. David Cameron's appeal is wider than was John Major's, reaching further into those groups of the public who are not naturally Tories, but it is also shallower in its hold on the groups that make up the Tory core vote: among the Conservatives' strongest groups - older voters and the middle class - they have lost ground. Much of this, no doubt, reflects the movement of the political battle to the middle ground, and perhaps Tony Blair's building of a Labour coalition less class-dependant than Neil Kinnock's. On the other hand, note how one of the most-discussed trends of the Blair years, the luring of female support away from the Tories, has melted away - Conservative support among men and among women is an identical single point lower now than in 1992.

But this is not to say that male and female voters have behaved identically since 1992, only that their different swings happen to have cancelled each other out. If we look at the analysis of gender by age, for example, the youngest group of men are substantially less Tory than were their counterparts in 1992, while the youngest women have become more Tory. Among the next age group, on the other hand, the 25-34 year olds, the difference is in the opposite direction, with men in this age group becoming more Conservative and women less so. There are similarly intriguing patterns in the changes of class support when broken down by gender.

But in fact you can't see the biggest change since 1992 in the table, which is that far fewer people vote these days. That fall in turnout has not been evenly spread - the differentials in turnout have widened, so that older voters and the middle class make up a bigger proportion of the vote than they used to. For that matter, even disregarding turnout, society itself is evolving: we are an older and more middle class country than we were 18 years ago. Both groups, of course, are predominantly Conservative.

This development works very much in favour of the Conservatives. Because, in purely proportionate terms Tory losses since 1992 among their strong groups have been far bigger than their gains among weak groups. For example, their share of the AB vote is down 11 points and of the C1 vote down 8 points, with only a 5-point gain among DEs to offset it. But because the Tories are still strongest among ABC1s, the growth of the middle class vote (now 62% of those who say they are certain to vote, compared to 43% of those who voted in 1992) has shored up their overall strength. The same applies to the comparison between older and younger voters. These age and class trends combine to produce perhaps the most startling figure in the table - among ABC1 adults aged 55 and over, the veritable Tory core vote, Conservative support is 11 percentage points lower than in 1992; and for ABC1s aged 35-54 it is even worse, a 14-point fall.

So how can the Conservatives have made big losses among their strongest groups, offset only by much smaller gains among their weaker groups, and yet be breaking even with roughly the same share of the vote as before? Simply because their strong groups now make up a higher proportion of the vote. Despite their loss of vote share among ABC1s and 55+s, the Tories are still the biggest party in both groups, and the increasing voting power of these groups has increased the value of any given lead - a small lead among a large proportion of voters is as valuable as a large lead among a small proportion. Consequently, David Cameron's share of the vote is almost as high as was John Major's despite having a less secure grip on the loyalties of ABC1s.

Of course, the 2009 figures are based only on those certain to vote, just over half the electorate. Turnout will certainly be higher than that by the time the election comes. But we found in 2005 that as determination to vote rose in the months up to the election, this only slightly eroded the gap between old and young or middle class and working class. Most of the Tories' relative advantage in voting power is likely to survive to the ballot boxes unless turnout is much higher than the 61% achieved last time.