The Conservatives now have a consistent lead in the opinion polls. ICM, Populus and YouGov have all shown a Conservative lead since back in April. The response to each poll from the Labour party (and from more sage and cautious Conservatives) is to wonder whether the Conservatives have achieved the sort of lead that a party needs at this stage to go on to win an election.

Questions like this always rankle with me to an extent becuase there is a danger of taking a too deterministic view of politics, so first let me get this off my chest: obviously there is always the possibility of “events, dear boy”. A government could have stonking great leads in the polls throughout a Parliament, and then get caught collectively shooting kittens on the eve of election night and lose. Obviously things like wars, mass strikes, economic disaster and so on can change the whole picture. Assuming none of these things happen though, what sort of lead should a party on its way to a victory be enjoying?

Throughout the 1980s the Labour party used to stack up huge double figure leads…and yet lost. Before the eventual Labour party victory in 1997 the party had enjoyed polling leads of over 40 points at some points. The single figure Conservative leads we see today look anaemic in comparison; if they are going to stand a chance of winning, surely they would be in a better position than this?

Perhaps not. Firstly the pollsters themselves have changed since the 1990s, as have the methods they use. The 40 point leads that Labour enjoyed were reported by Gallup, who no longer produce political polls in the UK. MORI produced 30 point leads for Labour in the 1992-1997 Parliament, but that was before their polls prompted by party name or accounted for people’s likelihood of actually voting. Neither YouGov nor Populus were around back then, in fact the only vaguely comparable figures from the 1992-1997 Parliament are ICM’s, who had already learnt the lessons of the 1992 election and adapted their methodology (and it’s worth noting than while ICM adapted their methods after 1992, they didn’t switch to phone polling until 1995).

While MORI were showing double digit Labour leads in 1993 and Gallup were showing leads of over 20 points, ICM was producing figures showing Labour around 5, 6, 7 points ahead (overall ICM’s Labour lead that year ranged from 14 points ahead to a 2 point deficit). In 1994 Labour moved further ahead, but ICM were still showing leads in the mid-teens range, as opposed to Gallup and MORI showing leads in the 20s and 30s. With modern polling methods the days of the opposition leading the government by 30-odd percentage points are probably history. A reasonable yardstick to measure the Conservative opposition’s performance against should be the ICM figures from 1992-1997, and while they have recently started to fall short of the double-digit leads Labour started to record in 1994, they are certainly still in the same sort of ball-park as Labour’s 1993/4 performance.

So, when people say that, up against a government on its way out the Conservatives should be doing X amount better at this stage in the Parliament they are playing politics. Go back a decade or two and polling techniques were different, they aren’t comparable and if you look at ICM’s figures in 1993/4 they show similar pictures. Unfortunately for the Conservatives, this certainly doesn’t mean they are doing well enough to win the nextelection.

At the beginning of 1994 ICM showed Labour with a 12 point lead over the Conservatives. Three years later they went on to win with a lead of 13 points, giving them a majority of 177. An equivalent Conservative lead would give them a meagre 34 seat majority. Back in 1997 Labour could have got an overall majority with a lead of about 2 points over the Tories on a uniform swing, so their poll leads of 6 or 7 percent in 1993 were actually enough to win an election. These days, again assuming a uniform swing, the Conservatives would need a lead of around eleven points to get an overall majority.

In practice I expect the Conservatives would gain a majority on a slighter lower lead as a result of the shifts in the pattern of votes and tactical voting that such a large change in public opinion would imply. The fact remains however that the Conservatives need a much bigger lead in the polls to win an election than Labour do (the reasons for the apparant “bias” in the electoral system are explained here). The Tories aren’t doing much worse than Labour were at a similar period in 1993/4, and haven’t had Black Wednesday to help them. The difference is that in 1993 a six or seven point lead was easily enough support for Labour to win an election even if they had ended up losing a point or two before polling day. For the Conservatives today, even if they maintain their present support, it’s unlikely to be enough to secure a majority.

A YouGov poll in today’s Sunday Times has the parties largely static – the topline voting intentions with changes from the last YouGov poll in the Telegraph are CON 37% (-1), LAB 32% (+1), LDEM 18% (nc). The poll was conducted between the 8th and 9th February.

On loans-for-peerages, 56% of people believe that Tony Blair has given peerages in exhanges for loans and donations to party funds, including 27% of current Labour voters. However, 53% of people think that the Conservatives did the same sort of thing when they were in power. While 53% of people think that Tony Blair should resign immediately if charges are brought against any of his aides, 55% of people say it is time for him to go anyway.

70% of people say they are concerned about loans or donations buying peerages – something that the Sunday Times interprets as meaning it isn’t just a Westminster village story. I wouldn’t be so sure, that 70% was made up of 45% of people saying it worried them, “but there were more important issues” with only 25% saying it worried them a great deal. Plus, there is the nature of the question – if asked directly people will say that issues concern or worry them, since it costs nothing to tell a pollster that the issue concerns you. It doesn’t necessary mean those things genuinely concern voters when put alongside bread and butter issues like health and education.


Populus’ February poll for the Times has topline voting intention figures, with changes from last month, of CON 36% (-3), LAB 33% (+1), LDEM 19% (+1). After unusally good figures for the Conservatives in January this is in line with most of the voting intention figures produced by Populus since last Summer. The poll was conducted between the 2nd and 4th February.

Populus also asked when people thought Tony Blair should stand down – 49% of people said now, including 33% of Labour voters. As I commented at the weekend, this is actually very much in line with people’s stated opinions last year, suggesting that the renewed focus on the loans-for-honours inquiry in the last week hasn’t had much effect upon people’s opinions. Asked specifically about the loans-for-honours inquiry, 56% said it had significantly reduced their trust in the Labour Government. 43% said that the police investigation seems to have been “unnecessarily heavy-handed”.

Populus’s now regular hypothetical question of how people would vote with Gordon Brown as leader produced one of his best ratings so far, with the Conservative lead cut to only 1 point, 35% to 34%. Questions on Gordon Brown’s public image reveal the normal pattern – on the upside he is seen as strong and statesmanlike, on the downside he is seen as uncharismatic, arrogant and dour.

Go now?

The Sunday Express’s front page reports a new ICM poll asking if Tony Blair should now step down. The poll includes questions on the casino decision and the investigation into loans-for-honours which I’ll look at when the actual tables appear on ICM’s website. Until then, the topline figures are that 56% of the public think Tony Blair should stand down now, including 43% of Labour supporters.

For all the Sunday Express’s hyperbole, this isn’t actually much of difference from past polls. Populus asked questions on when Tony Blair should step down quite regularly last year – in September they found that 51% of respondents thought that Tony Blair should step down either “now” or by the end of 2006, including 30% of Labour voters (which might well be somewhat different from Labour supporters). Now 2006 is over, 56% of people thinking that Tony Blair should step down now isn’t much of a surprise, it’s pretty much what they told us last year.

On an unrelated matter, Rallings and Thrasher have release the “official” notional figures for the new electoral boundaries which will be used by the media at the next election. Detailled figures for each seat don’t seem to be available anywhere yet. Meanwhile, you can see my notional figures for each seat on the UK Polling Report Election Guide.

UPDATE: Unlike Martin Baxter’s notional figures, which use a slightly different formula for translating local election votes into notional general election votes, the Rallings and Thrasher figures use a methodology that is almost identical to the one I used (the main differences are that when using local election results they take the average vote of each party, when I use the highest vote, and they give higher notional votes to parties that returned councillors uncontested, and lower notional votes to parties that didn’t contest wards). Hence, while I’m sure there will be lots of seats where the notional majories are different, overall the two sets of notional results are very similar. There are only 6 seats where the notional 2005 winners are different under the Wells or Rallings & Thrasher projections: I have the Conservatives winning Gillingham, Rugby and Portsmouth North, Rallings & Thrasher have these as notional Labour seats; I have the Conservatives winning Somerton & Frome, Rallings & Thrasher have this as a notional Lib Dem seat; I have the Lib Dems winning Rochdale and Oxford East, Rallings and Thrasher have these as notional Labour seats.

MORI January poll

MORI’s January poll has topline figures, with changes from last month, of CON 39%(+2), LAB 35%(-1), LDEM 19%(+1). The poll was conducted between the 19th and 29th January. The unusually long fieldwork period is a hangover from the takeover of MORI by Ipsos last year, the two sets of interviewers were being integrated into one force at the time, which led to some delays.