Tomorrow, we launch our annual report on the rebelliousness of MPs, packed full of data about the behaviour of MPs in the last session – who’s rebelled, how often, over what. And whether the record-breaking behaviour seen in the first session of the Parliament continued into the second. But how do the public react to MPs who are rebellious?
To test this, Rosie Campbell and I used a technique we’ve employed in a couple of other pieces of research (see, for example, here (£)), which is to split a polling sample, showing slightly altered information to respondents, and to see what differences this produces. In this case, we showed all the respondents of a standard YouGov online survey two fake biographies of politicians, like this:
Politician A is 48 years old. After university, where he studied physics, he trained as an accountant, and set up a successful company. He is married with three children. He is an avid cricket fan, and a keen player in his youth; he is now a passionate advocate for sporting facilities for young people. He also has interests in the health service and pensions. He became an MP in 2010 and is a member of the Heath Select Committee and is known to be a hard-working constituency MP and a party loyalist.
Politician B is 45 years old and studied business at University. Before entering politics, he was a solicitor who ran a busy local practice. He is passionate about the environment and education. His wife is a primary school teacher and they have two children and he is a trustee of an educational charity that supports apprenticeships. He has been an MP since 2005 and he is known for his focus on education policy, and he regularly votes against his party line.
And we then asked them which of the two they would prefer as their MP. However, we changed the final words of profile B, to compare five different ways of presenting an MP’s voting behaviour. One fifth of the sample (chosen randomly) saw the text as above. The remaining respondents saw one of the following four variants:
and is one of the most rebellious MPs at Westminster
and he votes against his party in 10% of votes.
and he votes with his party in 90% of votes.
and he has voted against his party 23 times in the last year.
These are all different ways of saying exactly the same thing. Indeed, they are all factually true statements about the most rebellious Conservative MP in the last session of Parliament, who voted against his party line on whipped votes 23 times, some 10% of the total number of 227 votes. One could perhaps argue about the definition of ‘regularly’ – is one rebellion in ten votes ‘regular’, or is it ‘occasional’ or maybe even ‘rare’? – but suffice to say that to the party whips, his rebellions certainly feel regular.
Shown the full text above (‘he regularly votes against his party line’), the first sub-sample answered: A: 30%, B: 48%, with 23% saying neither, a lead of B over A of 18 percentage points. The wrong way to interpret this is to think that this proves that the public prefer a rebellious MP to a loyalist. They may well do – and there is certainly lots of other evidence to suggest this – but there are lots of other reasons why the public might (on balance) prefer politician B to politician A (some might, for example, prefer his support for apprenticeships to A’s support for sporting facilities, others may dislike cricket or accounts or people who studied physics…). The correct way to interpret it is merely as a base line to compare with the other ways we expressed the same level of rebellion.
These produced rather different results, down from this 18 points lead of B over A to a lead of three points for A over B, as shown in the figure below.
Preference for politician B over politician A
The most popular way to present this particular level of rebellion then is to say that one does it ‘regularly’, with the results being very similar if the actual number (23) or the percentage share of rebellion (10%) is presented. Being labelled as ‘one of the most rebellious’ appears to be slightly less popular. But these are all statistically insignificant differences.
But the least popular way of expressing it – and a difference that is statistically significant – is as a percentage of votes with the party. When the public were shown a candidate who voted against their party in 10% of votes, they preferred that candidate by 15 percentage points. Yet when exactly the same information was presented as 90% loyalism, they preferred the other candidate by three points. That is an 18 point change in the relative position of the candidates, or a hardly insubstantial 9% swing.
In part, this is just evidence of a low level innumeracy amongst much of the public, similar to that found in lots of other studies. But it is also, we suspect, evidence of the dislike that many people have for political parties, and of their desire to see MPs behaving independently. (It is worth remembering that these statistics were of the most rebellious Conservative MP from the last session; most other MPs were much, much more loyal). This has implications for how we present data about parliamentary behaviour – including by organisations like the Commons themselves. And when it comes to campaigning, incumbent MP who have rebelled, therefore, might be wise to stress those occasions on which he or she had deviated from the party line. But even with a very rebellious MPs, all their opponent needs to do is to stress an incumbent’s loyalty to their party.
Note: we split a YouGov sample (of 1871 respondents), polled on 8-9 May 2013 into five sub-samples; findings were weighted according to YouGov’s usual weighting.
This blog post also appeared on the YouGov-Cambridge blog.