More on that Marketing Sciences Poll
We won't get any polling reaction to the budget for a couple of days - even if anyone does carry out an instant poll this afternoon for publication tomorrow, I'm not a huge fan of them as the fieldwork ends up being done after the budget speech has been given, but before respondents have actually taken it in or read the media's analysis of it. Our first real measure will be YouGov's monthly poll for the Telegraph on Friday - and that's assuming it's not going to be held back a day or three to get a better taste of the budget reaction.
In the meantime I've now got confirmation that the Marketing Sciences poll carried out for the Sunday Telegraph was indeed done using the same methodology as ICM, so the two polls should be directly comparable. Looking again at them...
MS (Apr 15th-16th) - CON 43%, LAB 26%, LDEM 21%
ICM(Apr 17th-19th) - CON 40%, LAB 30%, LDEM 19%
In theory it's possible that there was a 3.5% swing to Labour within just two days, even though the media was still unremittingly horrid for them, but more likely the difference is just down to normal sample error. There's no reason why the real position couldn't be somewhere between the two polls.
It is a good reminder that polls are subject to sample error, that two polls done using exactly the same method, at pretty much the same time, can still show different results, and we should never get too excited over movements of 1 or 2 points in a single poll, unless it supports a trend that has been observed over a number of different polls.
