Political Commentary - In Which Gordon Is Unbounced?

The much-discussed "Brown bounce" persists. Since our last poll (towards the end of July), which found Labour six points ahead of the Conservatives in the voting intentions of those who say they are absolutely certain to vote, five other polls have been published, all putting Labour ahead by between three and nine points — indeed, every published conducted since the last week of June has put the lead at six points, plus-or-minus three; until the middle of June, all the polls had Labour behind the Tories. No polling company has yet found a statistically significant fall in the lead compared to their previous poll, since Gordon Brown entered Number Ten.

Though it should make little difference to the political implications either way, is it really a "Brown bounce" or an "end-of-Blair bounce"? Tony Blair, though still respected by many voters for his government's achievements, is thought by many to have outstayed his welcome. It seems clear that his continued leadership was an obstacle to supporting the Labour party for many floating voters. Not only voting intention, but other key political indicators as well, began to move upwards for the government immediately after Blair announced his retirement date; the most recent month of polls are merely the culmination of that.

Nor, for all the flak that David Cameron has received over the past few weeks, are there many signs that this is an anti-Cameron bounce. His satisfaction ratings as leader have remained steady at around 30% who are 'satisfied' with his performance, even while Labour support has risen. Furthermore, although the increase in Labour support has been draining from the Tories over the last few months, the net changes since the last election show a different story. Though Labour, at 41% in our last poll, has apparently made a five point gain since 2005, the Conservatives are doing better than in 2005 as well. (Some of the other polling companies differ slightly from us on current Tory strength, but none is showing a significant fall in Tory vote share since 2005.) The Brown bounce is apparently not attracting committed Tories, it is attracting voters who supported the Lib Dems in 2005 and quite likely (if they were old enough to vote) were Labour in 2001 and 1997, and have merely flirted with the idea of voting Conservative in the last couple of years without yet having put in into practice. The Brown bounce may be merely the reconvening of the New Labour coalition which was splintered in 2005 by loss of trust in Blair and the Iraq invasion, bringing back the temporary deserters.

If Cameron is failing it is not that he is losing support now, but rather that he had failed to garner enough support while Blair was still in Downing Street. The Brown bounce was virtually inevitable — he has received no more than the boost that is customary for new Prime Ministers. The bounce has become a big story because it has put Labour back ahead, but that could only happen because the Tories had not already established an unassailable lead before Brown took over. To quote Winnie-the-Pooh:

If Rabbit Was bigger And, fatter And stronger, Or bigger Than Tigger, If Tigger was smaller, Then Tigger's bad habit Of bouncing at Rabbit Would matter No longer, If Rabbit was taller [note 1]

If the Tories had established themselves more strongly when they apparently had the opportunity to do so, Brown's bouncing would seem to matter less. (Perhaps this is why more than a quarter of those intending to vote Conservative say they are dissatisfied with Cameron's leadership.)

Questions about how long the bounce can be expected to continue ignore the virtual certainty that it will not get the chance to do so. The bounce consists of a free ride as the voters give a new leader the benefit of the doubt until they have had a chance to make up their mind; but everything he does from the moment he takes office will help more voters decide one way or the other. In our first satisfaction rating measure for Gordon Brown we found 44% saying they did not know whether or not they were satisfied with the job he was doing. While this seems high compared to the 7% of "don't knows" in Blair's last rating, or even the 34% who still don't know what they think of David Cameron, it is already lower than any recent leader of any major party in his first month — and it will surely fall further quite quickly. (And, for the moment at least, those who do have an opinion split strongly in Brown's favour, 36% satisfied and only 20% dissatisfied.) Both John Major and Tony Blair apparently had long honeymoons as PM, but in fact each faced a defining moment only a short time into their first terms — Major the first Gulf war, Blair the death of the Princess of Wales — and were quickly being judged, favourably, on their own merits. Gordon Brown has already had to deal with two terrorist attacks, the flooding crisis, and now over the last weekend (since when there have been no polls published) a foot-and-mouth outbreak. The opinion of the commentariat seems to be that he is coping impressively; whether the public agree we shall soon see, but either way it is unlikely that he will be relying for long on his "bounce" alone. (And we might add that "bounce" is a potentially misleading, because it seems to imply that the upward movement will inevitably be followed by a matching movement downwards. That may not be the case if Brown can use it as a jumping off point from which to consolidate his standing by his own efforts, as both Major and Blair successfully did.)

In any case, how "real" is the bounce? Some speculation in recent weeks has even suggested a snap election this autumn in order to exploit the government's perhaps temporary popularity. But that makes the very questionable assumption that opinion poll popularity at this stage of a Parliament can be translated directly into votes at the ballot box. Almost every new PM since regular polling began has had some sort of a "honeymoon" or bounce, but none of them has ever called an election to test whether it would hold firm through an election campaign. As we consistently tell anybody who will listen, opinion polls are no more than a snapshot of what a representative sample of the electorate say they would do if there were a general election tomorrow. But there is not a general election tomorrow. A poll is certainly not a measure of "what would happen" if there were an election; indeed, since the concept rests on counterfactual assumptions it may not even be intellectually coherent. Members of the public answering opinion polls do not agonise over their answers, debate the pros and cons and come up with a considered opinion; they give a snap reaction, their top-of-the-mind impression, an instant answer. Even after an election has been called and we are asking them "How will you vote?" on a specific day a few weeks in the future, we know that however honest they are with us they may not be able to predict their own behaviour perfectly. How much less reliance, then, should we place on a question about an election that is not going to happen anyway?

In fact there is no mystery about what is going on. You probably do it every day yourself, whether or not you have ever been polled. How often do you get frustrated by bad service or a dodgy product and tell yourself, and anybody else who will listen, that you will never, NEVER, NEVER buy that brand again? And how often do you find when it comes to the point that you choose the same brand as usual after all? But what would you have told a market researcher if you had been interviewed before your anger wore off? It applies just as much to politics, and it can just as well cause upwards as downward bounces. Nor is it confined to opinion polls, of course — by-elections, local elections, European elections, they are all affected by protest voting and the public making a gesture that they would not make at a general election because it exaggerates their real depth of feeling and has much more meaningful consequences.

So we should not be tempted to take mid-term opinion polls as literal "what would happen now" projections, let alone assume a six-point lead now equates to a six-point victory at an election a month away if the PM asks for a dissolution tomorrow. That does not mean that polls are useless. They are a real indicator of public mood, and the best one we have. Gordon Brown is getting a "bounce" because the public are genuinely better disposed to the Labour Party than they were earlier in the year. But solidifying that goodwill into something that will translate into real votes is another task entirely and will still need the public to believe in real success for the Prime Minister, or real failure for his Conservative opponent. Of course, it may be that this is already happening, that Brown's handling of the first few crises of his premiership has made a lasting positive impression — it is too early to tell. But it may also be that nothing lasting has yet happened and that, some time soon, Gordon is going to be unbounced.

Notes

1. A A Milne, The House at Pooh Corner, "In Which Tigger Is Unbounced"