Skip to main content
Coronavirus (COVID-19): Latest updates.


What a Difference a Day Makes…

30 November 2015

A few things for you this week, blog readers.

First of all, an interesting little curiosity that I came across while working through some British Election Study data from their on-line panel survey. This, as I think I’ve mentioned previously on the blog, is a project that has interviewed a very large sample of people across Britain, via the internet, several times over the last few years (for more details, see the BES site here). For the discussion that follows, the key thing to remember is that this is a ‘panel’ study: in other words you are polling the same people every time (bar a very small percentage of people who drop out along the way). This means that if we see any changes in attitudes between two waves of the survey we can be absolutely certain that this doesn’t mean us simply having two very different samples; we are observing change within the same group of people.

Anyway, I was interested in a pair of questions that were asked about all the major party leaders and their performance in the general election campaign. The post-election wave of this project (i.e. a survey that was issued to respondents immediately after the conclusion of the general election and declaration of the results) asked people which leader they thought had performed the best in the election campaign, and which one had performed the worst.

Here are the responses on these questions among respondents in Wales:


Post-Election: ‘Performed Best’ and Performed Worst’ During 2015 General Election Campaign: Party Leaders in Wales

Performed Best Performed Worst Net: Best – Worst
Cameron 33% 7% +26
Miliband 13% 24% -11
Clegg 3% 12% -9
Farage 8% 16% -6
Bennett 2% 15% -13
Wood 9% 4% +5
None/Don’t Know/ Equally Good/Bad 32% 23%

Source: British Election Study On-Line Panel Study, Wave 6 (post-election wave); number of respondents = 1556.


As you can see, these responses seem to affirm strongly the performance of David Cameron in the election, and to be pretty damning about Ed Miliband. Even in Wales, Miliband trails well behind Cameron in ‘best’ ratings, while leading by a long way in the ‘worst’ ratings. The scores are also fairly negative for Nigel Farage, Nick Clegg and Natalie Bennett. Leanne Wood achieves the fewest ‘worst’ ratings of any leader, which is clearly a good thing for her. But she still lags well behind the Conservative and Labour leaders on ‘best’ ratings.

I was going to write up a blog post purely on these results, and also include them in an academic piece I am writing. However, before doing that I thought that it would be a good idea to check the responses on this question against those given in the previous wave of the study. This previous wave had been conducted during the election campaign itself – the survey was run on a ‘rolling’ basis through the last few weeks of the campaign period.

So I went back and checked how the ‘best’ and ‘worst leader ratings came out during the campaign wave. Here is the equivalent table for that above, but this time for when people were asked the pair of questions during the election campaign:


Campaign Wave: ‘Performed Best’ and Performed Worst’ During 2015 General Election Campaign: Party Leaders in Wales

Performed Best Performed Worst Net: Best – Worst
Cameron 20% 16% +4
Miliband 25% 15% +10
Clegg 3% 7% -4
Farage 10% 19% -9
Bennett 2% 13% -11
Wood 8% 4% +4
None/Don’t Know/ Equally Good/Bad 33% 26%

Source: British Election Study On-Line Panel Study, Wave 5 (campaign wave); number of respondents = 1597.


I think you’ll agree that there are pretty significant differences between this table and the one for the post-election wave! These differences are mostly not about the more ‘minor’ parties. Leanne Wood’s ratings are almost identical across the two waves, while those for Clegg, Farage and Bennett are pretty similar. The big differences concern Messrs Cameron and Miliband. We now see the two more or less level pegging on the ‘worst leader’ question – rather than Miliband scoring more than three times as many mentions, as he did in the post-election survey. And on the ‘best’ leader ratings, Miliband actually had a narrow lead in the campaign wave. When we look at the ‘net’ ratings, the situation is transformed: a narrow Miliband edge in the campaign wave, but an overwhelming Cameron advantage when the same respondents were asked straight after the election.

So what do we take from that? One potential explanation might be that the campaign period saw steady increases in positive ratings for Cameron’s performance and negative ones for Miliband. However, that doesn’t seem to be true: if you examine the ratings from the rolling campaign study across the campaign, there are no obvious signs of any such trend (at least among the Welsh respondents). What we are seeing in the two tables above is how the outcome of the election itself changed perceptions. Perhaps the lesson is simply is the rather unoriginal one that hindsight is a wonderful thing and we should beware how much it can shape our thinking. Of course, it is obvious now to everyone that Ed Miliband had a much worse election campaign than David Cameron. But it wasn’t necessarily quite so obvious to everyone at the time.

That leads me onto the second thing I have for you today. Matt Singh, who runs the website Number Cruncher Politics, was just about the only person to accurately predict the ‘failure’ of the opinion polls in Britain before the election. He’s recently published on his blog his interpretation of some of the main reasons for the errors that the polls made. The full piece is here. I’d strongly encourage everyone to read it – it’s a really interesting piece of work. I’d also, though, like to quote Matt’s conclusion:

“Polling has its pitfalls. This year things went badly wrong, as they did in 1992 and in 1970, and as they probably will another generation from now, if not sooner. But most of the time, polls get it right, and provide extremely useful information. If public opinion matters, then understanding it matters too, and measuring it scientifically is the only way to do that. These two facts are equally important, regardless of whether the last major polling failure was six months ago or 23 years ago.”

Coming from someone who actually foretold the errors of the polls prior to the election, I think that’s a particularly important – and wholly correct – statement.

And that in turn leads me onto my final point for today. A few people have asked me when the next Welsh Political Barometer poll will be published. Our current plan is to conduct the poll at the beginning of December, and report the findings in the first full week of that month (i.e. the week beginning Monday 7th December). If that plan changes, I’ll keep you informed. Our previous poll, in September, did suggest that Wales might be the one part of Britain experiencing a clear ‘Corbyn bounce’ – it will be interesting to see if that has been sustained through the Labour leader’s recent difficulties.


1 comment
  1. Christian Schmidt

    What a great post!

    What I think it shows that, if you understand why opinion polls come up with the results that they do, you can be popular by making opinion follow you instead of following opinion. But getting is not easy – though opinion polls can help with that too, if you can read them…

Leave a reply

Your email address will not be published. Required fields are marked *