The return of Shy Tories?
A couple of articles by James Forsyth and Gary Gibbon at the weekend mentioned a Conservative belief that they would do better because there were "shy Tories" out there. My initial reaction was that it smelt a bit "straw-clutchy", but equally James cited Andrew Cooper and Stephen Gilbert* who are both wise heads who wouldn't go in for that sort of thing.
Having spoken to some journalists up in Birmingham, it actually seems more sensible than that. The belief in shy Tories seems to be no more than people in CCHQ thinking that the polls that re-adjust to account for "shy voters" (that is, mainly ICM and Populus) are a better steer on the actual position than polls that don't - and indeed, for four months or so ICM were showing a 5 point Labour lead compared to a 10 point Labour lead in YouGov (though in ICM's most recent poll the lead went up to 10).
Given that YouGov don't reallocate don't knows, you might expect me to think reallocation was a bad thing. I really don't - there is good evidence from recall surveys at past elections that don't knows are disproportionately likely to end up voting for the party they voted for at the previous election. I regard it more of a philosophical question, the question of whether a poll should reflect only what respondents say they'd do, or what the pollsters believes they might actually do. I don't think there is necessarily a right answer to the question.
The point is, however, that whether or not pollsters include them in their figures those don't knows are still there in the population. Around about 14% or so of people who voted Tory last time normally tell pollsters they don't know what they would do next time (around about 20% of people who voted Lib Dem say don't know!). Past evidence suggests those people are more likely to go back to the Tories at the next election... this is already factored into polls by ICM and Populus (so don't go double counting!), but not for most other companies - hence one of the reasons they tend to show lower leads than YouGov.
The other things mentioned by James and Gary as the Conservative reasons for some optimism are incumbency and targeting. Incumbency is certainly correct - while MPs don't have as much of a personal vote as they probably think they do, it is there, and it is most evident for first time MPs who have won a seat from an opposing party. They, in effect, have a double incumbency bonus - the party they defeated will have lost the incumbency bonus they had at the last election while the new MP will gain an incumbency bonus he did not have the previous time round. For MPs who are not new it is less of a factor, as they already had an incumbency bonus at the previous election... most of it is already factored into the price as it were. Most Conservative marginals at the moment, however, are seats they won at the last election with new MPs.
Dan Paskini has linked to some of the attempts to measure incumbency bonus in a piece for Liberal Conspiracy here, but the effect of incumbency is probably about 1.2% - 2% of the vote (or double that for those MPs in seats won from an opposing incumbent!). This will obviously be a big advantage to many 2010 Conservative MPs defending their seats for the first time... though obviously wont particularly help the Conservatives to get a majority as, by definition, that involves winning seats they don't already hold!
Targeting is rather more worrisome. Parties always say they will target better and perform better in marginal seats and so on, but it is very easy to say and hard to do. What the Conservatives do have in their favour is that in 2010 they needed to win over a hundred seats for a majority, in 2015 they only need to win 20. Clearly it is easier to concentrate resources on a smaller number of seats. What is less in their control is how many seats they need to defend, and at what level of majority they feel comfortable that a seat will be held and put their resources elsewhere.
A final point they are making is whether the disconnect between Labour's substantial lead in voting intention polls and their less positive polling on Miliband as PM or best party on the economy (where Lab & Con tend to be equal) suggests the Labour lead is fairly soft. I've discussed this at length before in posts on Ed Miliband - suffice to say, it is possible that it means that, or it is possible that people will vote Labour despite not thinking that much of Miliband. Only time will tell.
In conclusion, there isn't really anything here that qualifies as straw clutching - it is all fairly straightforward. Neither, however, is there anything particularly new - reallocation of don't knows, incumbency advantage, whether parties can do better in marginals (something we've avoided since the election but will no doubt return to at some point!), whether Labour's lead is soft - all things we've discussed here before.
(*in case you were wondering, yes, the photo in the Daily Mail article is of the wrong Stephen Gilbert. Their photo is of the Lib Dem MP, not the Prime Minister's Political Secretary of the same name)
