A polling proposal

“ Public disillusionment with the EU is at an all-time high” said David Cameron in his “referendum” speech on Wednesday. It obviously suited his argument to say that, but it is not borne out by the poll published yesterday by The Times/Populus which showed 40 percent in favour of leaving the EU, 37 percent for staying in and 23 per cent “don't knows”. This was then converted to 53-47 by taking into account of “people's likelihood to vote” and “stripping out” the don't knows -- but even a doctored difference of six points can hardly be called “public disillusionment”.

Because of the uncertainties and complexities of Mr Cameron's plans public opinion polls are likely to be very busy indeed sampling voting intentions before the election in 2015 and referendum in 2017, and their results may influence the outcome.. For this reason I think it is important that all opinion polls and their media partners should make full and very clear statements on how their questioning is conducted (person-to-person, or telephone or internet, and the numbers), how their respondents are chosen (randomly or on a pre-selected basis) and, especially, how and why raw data is adjusted. And perhaps the “don't knows” should be asked whether their position is due to apathy, ignorance, uncertainty or responsible caution until nearer polling day when all the issues are on the table.