I don’t like talking about polls much. Looking at monthly polls is like looking at the essays of Oxford students in the three years before their finals. They tell you lots of useful things about the weaknesses, strengths and abilities of the tyro gown wearers, but looking at any one piece of work, or even a series of essays, can’t tell you how they’ll do in Exam Schools.
Maybe the kid revises hard, or discovers girls, or becomes complacent. I should know. I was destined for a third, and emerged with a 2:1. Revision FTW.
Interim information helps you see trends, and know what to do about them, but until the exam or election, nothing is settled. (In both cases, the predictive capabilities of interim tests get a lot more useful the closer you get to the finals, obviously).
Over the last two years we’ve seen enough dramatic polling changes to show that very little is permanent. We had consistent Conservative leads for a year, then three months of Labour leads when Gordon Brown took over, then a huge collapse in Labour support for nine months, then a Labour recovery to within striking distance followed by another Labour collapse in support. It’s been emotional.
Here’s all that recent history in graph form:
Within the next year we could see equally striking and dramatic changes, or we could not.
Looking at that graph, there appear to be two periods where an informed observer could see a settled will of the people emerging. In both cases, a disruptive event transformed the political situation. The emerging trends became irrelevant as peoples views of the parties changed. In other words, people change their minds about you when you do things they like or hate.
Few people view polls as much more than a sort of running scoreboard for the Government of the day – a way for us to understand how well the government is doing, how appealing the opposition is and whether our leaders have the confidence of the populace.
All of which is interesting enough, but not helpful in crafting an electoral strategy.
Polls tell you how hungry people feel right now, not how they’re going to feel after they eat tonight. A wise chef uses what they say to find out what they’re looking for from their next meal, and what they hated about the last lunch you cooked. Don’t assume that because someone craves a McDonalds now, they’ll feel the same after you feed them Steak frites.
Having used that inspiring metaphor, I now want to spend an inordinate of time talking about how to approach polls. Then in another post, I’ll try to use all that to explain why I think the views I already hold are the best strategic option for the Labour party. If you’re not interested in polls, best avoid the rest of this and wait for the fun stuff.
You hate to accept reality to change it
The polls are really bad right now.
A combination of a decline in the trust of the Government to handle the economy, a series of embarrassing scandals that have tarnished the parties reputation, voter concerns about the economy and the direction of the country and a more acceptable opposition alternative have combined to reduce both Labour vote share and the likelihood to vote of Labour supporters.
Although there’s been little detailed polling, I expect we’ll see a significant decline in the PMs personal rating (after a rebound from Oct 08-March 09) as well.
These are really, really bad numbers. We’re losing supporters to both apathy and other parties in disturbing numbers. So should we just give up, maybe play for a respectable defeat or some such?
Don’t think it can’t be changed.
Nothing’s inevitable. Nothing. We’ve caused Labour supporters to feel less likely to vote and lost the support of a number of key voters. To win the next election we have to get significant numbers of both of those groups back.
We’ve done that successfully once, and semi-successfully twice in the last two years alone.
When you think about what you need to do, put the voters concerns first and last
In politics, it’s tempting to get caught up in debates about leadership, strategy, initiatives, personality and so on. Yet when voters are asked about these things they’re less important to them than big hairy issues. Economy, Crime. Public Services (though interest in the latter has declined as the economy has worried people).
So in thinking about what political strategy to follow, it’s important to focus on what voters are telling you they think matters, giving them a powerful argument as to why they need to support us on election day.
Right now, one issue bestrides everything
70% of the population think the economy is a key issue facing the country. That’s more than the number who thought defence was important after September 11. It’s more than the number that thought the economy was important after Black Wednesday. 23% say unemployment is a key issue. That’s more than say education.
It shouldn’t be any suprise therefore that when the government looks like it’s in command of the economy, it benefits, and when it looks like it’s struggling to cope it falls behind.
Here’s a chart that plots Labour’s lead (or deficit) against the number of people who think the economy will improve or stay the same. The trends are pretty clear. The more people feel economically secure, the smaller Labour’s defecit.
Think about Labour identifiers first
As the recent comres poll shows, there’s a fairly narrow gap in Labour and Conservative party identification. 27% of voters think of themselves as Conservative, 25% as Labour. When looked at by identifiers alone, the election looks pretty close in every age and social class group other than the over 65s.
We get slaughtered by two things. A number of our identifiers say they’re unlikely to vote or are uncertain about it and a significant number are voting Tory. Motivating them to vote and peeling them back are crucial.
There’s another group I’d link to these – 2005 Labour voters who now identify as conservatives. My hunch is that this group is very similar in attitude to those who identify as Labour but are voting conservative, but come from less traditional Labour backgrounds.*
Pay more attention to voters than to non-voters
This might seem really stupidly obvious, but it’s a useful corrective. Date shows that older people are more likely to vote. It shows that people who voted last time are more likely to vote this time. It shows that people who say they’re going to vote are amzingly enough, actually going to vote. So pay careful attention to trends in these categories and think about how your policy agenda is going to apeal to them.
Speaking of which: Don’t use polls to justify your own pre-existing policy prejudices.
We all do this, me included. But it’s a useful reminder. Use polls to challenge your beliefs about public preferences not re-inforce them.
Is there evidence that scrapping Trident would be popular? Well, you’d need to factor that public opinion is divided on Nuclear weapons, with 49% saying we don’t need them now, and another 20% calling for a delay. You’ also remember the salience of the issue. Defence is not regarded as a topline issue for many people. You also need to balance between the fact that people are receptive to spending cuts arguments in general, but also that there’s significant concern about unemployment.
To me at least, cutting a programme that will create tangible manufacturing jobs amongst skilled manual workers strikes me as one of the less likely spending cuts to appeal to the public, right now. But that’s just an opinion, not a data point.
You might also look at the fact that despite a sustained media onslaught, the proposal for a 50% Tax rate is very popular, with nearly 60% of voters saying they support it. Considering the reaction of virtually every major newsapper and most of the TV coverage, that’s a pretty remarkable number. Those on the right of the party who are uncomfortable about 50% tax rates for electoral reasons need to take that on board.
There’s a good example of this in the Populus post budget presentation that shows that a plurality of people think that the 50% tax rate is both the end of New Labour and a good thing for the country.
Another example: Some people, many on the right of the party, have occassionally murmured about Gordon Brown’s leadership, just as people on the left of the party murmured about Tony Blair’s. Yet satisfaction ratings for Gordon Brown have been significantly above that of the Government for some time.
*This is just a guess as the numbers in polls are too small to be meaningful.
Which reminds me. When looking at relatively small groups of people (weak labour identifiers, Labour identifiers who are considering voting Conservative, etc), Qual work (like focus qroups) is more helpful than quantitative – an opinion poll that finds 20 or 30 Labour weak identifiers isn’t a great way to understand their concerns and worries. You need the quant to tell you how important a group they are and the qual to understand them.
The value of qualititaive work, done well, is that voters are not voting Labour for their own reasons, not the reasons we wish to project upon them. Understanding those reasons is key to changing them. Bad Qualitative work leads people opinions, and is downright dangerous.