There is a lots of coverage for a YouGov poll conducted for the Open Rights Group, which suggests widespread opposition to the government’s proposals to cut off the internet connection of people caught filesharing (or, more specifically, to cut them off without the requirement to get a court order – the poll concentrated on what the legal requirements would be to cut people off rather the principle of whether or not they should be cut off.)

Personally I’d have worded the question differently, but nevertheless the public response was pretty overwhelming, 16% thought an ISP should be able to disconnect people having been informed of several infringements, 68% thought the evidence should have to go before a court.

As a caveat, while online polling can work, certainly it works for political polling as YouGov have proved time and again, there are some issues where it is going to mislead – to take an extreme example, you couldn’t use an online poll to find what proportion of people have internet access (Wow! It’s 100%!). In this poll the first two questions obviously have to be looked at in the light that there are another 38% or so of the population who don’t have home internet access, and for whom loosing internet access would obviously have no impact whatsover.

Finally the poll found that 31% of people said they would be much less likely to vote for a party that followed this policy. As ever, I’d advise people to spurn all “would it make you more or less likely to vote” for a party questions – except just possibly when they are asked only of a group of floating voters, when they should only be compared against similar questions.

It’s worth dwelling a bit longer on this type of question – those which attempt to measure how important an issue is to the public, or how much impact it might actually have on an election. It’s a type of question that often turn up in polls for many other organisations (and for that matter, newspaper polls), and it’s right that they should – after all, whether an issue has an impact is not just down to whether people agree with it, but also how much they care about it. Pressure groups and charities often commission them, because they create the impression that people think their issue is jolly important, and that MPs should take it seriously because it might cost them votes. Unfortunately, the way they are asked often suggests issues are more salient than they are.

Looking at the two ways they are normally asked, more or less likely to vote questions are misleading since more or less likely are very undemanding terms, voting intention is actually quite sticky but it’s very easy to say in a poll you might consider changing your vote. It’s even worse when the question asks specifically about a party (e.g. would it make you more likely to vote Labour?) since it will normally include lots of people who in reality already support that party, or would never do so anyway. Generally speaking respondents seem to use questions like this just to indicate whether they like or dislike a policy.

The alternative approach of asking people directly if they think issue X is important is even worse, as unless the issue is obviously petty and insignificant respondents will almost always say yes. It doesn’t actually cost anything to say “yes” to a survey, and when people are asked about these things by pollsters they will naturally think “Yes, that does sound like an important thing I really should care about” rather than look like an uncaring swine. In really however people will not be forced to consider that issue in isolation – it will have to compete for attention against all those other issues like the economy, health, crime, taxation, immigration. The question that matters is whether people think it is important compared to other issues.

The best way of judging how salient issues really are is Ipsos MORI’s monthly poll on issues. These ask people what they think the most important issue facing the country is, and then any other important issues they think there are out there. Importantly, it does not prompt people at all, there is no list of options to choose from, they are not forced into the pollster’s or the client’s predetermined agenda. The overwhelming majority said things like economy, crime, health, unemployment, immigration and so on. I don’t think issues actually make a massive difference to voting intention, more people vote on broad party image rather than the specifics of policy, but if they do vote on issues, these are the ones.

Even this question isn’t perfect. To take an example, I’m sure there are people who vote on the single issue of fox hunting, but who wouldn’t necessarily claim that fox hunting is one of the most important issues facing the country. The ideal would be an unprompted question asking what issues would matter to people in deciding how to vote at the next election (which MORI also ask, albeit less frequently, the most recent was last month.)

Of course this is oversimplying things a bit. The issues people care about can easily change, and sometimes issues people don’t care about can matter. Low salience issues can still have a wider impact on how parties are perceived. For example, under David Cameron the Conservatives have talked a lot about the environment, an issue of comparatively low salience, as a way of making themselves look like a more caring and progressive party. They have pretty much ignored the issue of immigration, an issue of high salience, because it would have reinforced negative perceptions of the party as being being insular and bigoted.

This has drifted a long way from digital rights, so coming back to it, respondents clearly don’t like the idea of people being disconnected for filesharing without the approval of the courts. Would it really have a large impact on how a third of people vote? Well, do you see any civil liberty issues up there when MORI asked people what issues would decide their vote?


The full tables for the Yougov/Sunday Times poll are now up on YouGov’s website. I’ll start with the standard leaders approval ratings – which we can compared to those in the last YouGov Sunday Times poll just before conference season – Cameron’s approval rating is virtually unchanged, from +28 to +27, Brown’s is slightly up from -44 to -39, Clegg is slightly down from +17 to +13.

Moving on the more varied topics, as usual the Sunday Times asked topics on a range of issues. Asked about whether MPs should be made to pay back expenses that Sir Thomas Legg has ruled are unreasonable, even if they were within the rules at the time, 65% of YouGov’s respondents thought they should, with 28% thinking that retrospective rule changes are wrong. 69% also thought David Cameron was right to say Tory MPs who do not pay up will not be allowed to stand as Conservatives at the next election.

38% of people agreed there should be an immediate election to “purge” Parliament of MPs who abused the expenses system, 53% disagreed. I expect even these answers are more about partisan opinions of the government than expenses – 63% of Tory voters think there need to be an election, 84% of Labour voters think there shouldn’t.

YouGov also asked if people might consider voting for a minor party or abstaining over the expenses issue. I don’t like questions like this, since it is psychologically a lot easier to say in a survey that you might switch than it is to actually do it, and the question sets very low thresholds indeed (“might consider switching”). However, the party splits are interesting – amongst Conservative and Labour voters 7% say they would definitely consider switching to a minor party or abstaining, but amongst Lib Dem voters it is 12%.

As I mentioned yesterday, YouGov also asked if the BBC was right to invite Nick Griffin onto Question Time. 63% agreed with the statement “The BBC was RIGHT to invite him, as the BNP has two members of the European Parliament”, 23% that “The BBC was WRONG to give him, as it should not provide a platform for someone with such extreme and objectionable views”.

There were also some questions on the Lisbon Treaty. Voting intention in a referendum on the treaty stands at 18% YES, 41% NO and 41% don’t know or wouldn’t vote. Opinion on Tony Blair’s candidancy as President of the European Union remains split, 38% support to 38% opposed.

While I’m on the issue of Sunday polls, the Mail on Sunday also claimed to have a poll conducted in Redditch asking about Jacqui Smith. As fas as I can tell it was conducted by 3 Mail on Sunday journalists going up there with clipboards, so I’d be gobsmacked if there was any attempt at representative sampling or weighting. Ignore (if you weren’t already).

I haven’t had any confirmation, but on normal timings we should be getting ICM in the Guardian today or tomorrow.


-->

There is also a new YouGov poll in the Sunday Times. I haven’t had full confirmation of the numbers yet, but various sources have them as CON 41%(-1), LAB 30%(+2), LDEM 17%(-2). As with ComRes’s poll earlier tonight, it suggests an advance for Labour – go back before the conference season and YouGov were normally showing Labour around about 27%.

The Sunday Times normally commission questions on a wide range of topics, so I’ll update later on tonight or tomorrow with anything else interesting. One other thing YouGov asked abour was whether the BBC were right to invite the BNP onto Question Time – 63% said yes, 23% no.


I’m expecting at least two new polls tonight. First up is ComRes for the Independent on Sunday. The topline figures, with changes from their last poll, are CON 40%(nc), LAB 28%(nc), LDEM 19%(nc) – so literally no change there.

Polls at the end of last week showed a boost from the Conservative conference, but it’s no great surprise that this doesn’t: almost by definition the conference boosts are transitory things that disappear after a few days. More interesting though is that the previous ComRes poll was showing a Labour boost, and that hasn’t receded. Before the Labour conference ComRes had been pretty consistently showing Labour at around 23% or 24% – now we have them up at 28% (mainly at the expense of others, the Conservatives are pretty steady).

Of course, it’s possible to offer a thousand reasons why that might be, and there is scant evidence to actually tell us the real reason. We can’t even be certain that it’s anything to do with the conference season, since we also have the issue of expenses rising up the agenda. My own guess is that is indeed one factor – Brown’s response to the re-emerge of the expenses scandal, firmly stating that complaining MPs should do what their Legg letters say has made him look something like the old decisive, strong (and popular!) Gordon Brown. That said ComRes did ask in the survey whether people agreed with the statement “David Cameron has dealt with the issue of MPs’ expenses better than Gordon Brown” and while the proportion of people agreeing with the statement had fallen, it wasn’t a huge turnaround (in May it was 59%, now 53%)

On the subject of expenses, it’s also worth commenting on the dog that didn’t bark. This is the first poll conducted entirely since the expenses issues re-emerged and the level of support for “others” has not increased.


One of the unanswered questions I mentioned in my round up before the conference was what the effect of support for others might be. When support for them increased earlier this year it impacted upon Labour and the Conservatives pretty equally, but what if it declined in a less even handed manner? Were those “others” disillusioned Labour supporters who wanted to protest against the government, but wouldn’t vote elsewhere and, presumably, would therefore be more likely to go back to Labour? Were they anyone but Labour voters who were more likely to eventually coalesce around the Conservatives? Would they decline at all?

Well, that last question is perhaps a live one again now the expenses scandal has reared up again, but up to now the answer was certainly yes. The graph below shows the decline of “others” since the European election, compared to the fortunes of the main political parties. There are various ups and downs along the way (particularly the Lib Dem conference!) so I’ve added some trend lines to highlight the bigger picture.

Basically, support for the others has steadily fallen and as it has divided pretty evenly between Labour and the Conservatives, with little benefit for the Liberal Democrats. Labour in fact do very slightly better (their trendline has gone from 23% to 27%, the Conservatives from 38% to 41%), but on the whole their support has moved in parallel. The other vote appears was a plague on both the main parties houses when it appeared, and they seem to be recovering at an equal rate as it recedes.