Opinion polls – are they accurate?

An analysis of some recent polling by our Vice Chairman Anthony Scholefield has thrown up some interesting points.

The poll, by ICM was conducted between 13th and 15th May and consisted of two parts – on-line and by telephone. The former showed a slight lead for “leave” but the latter gave an 11-point lead for “remain”.

A number of commentators have claimed that telephone polling tends to be more accurate, but Anthony’s dissection of the figures has cast doubt on the accuracy of the telephone part of this poll.

Both parts of the poll split the resondents into three separate age categories:  18-34, 35-64 and 65+. Predictably both polls yielded a substantial majorioty for “remain” among younger voters with a slight leaning towards “leave” in the middle category. With older voters, there was a marked discrepancy. The on-line poll, as would be expected, favoured “leave” by a margin of 55% to 31%. The telephone polling, on the other hand, actually gave a lead to “remain” (41% to 35%). Given that support for “leave” has consistently been very high among older voters, Anthony has quite rightly queried this figure. “It looks like this age group’s Remain vote has been over-estimated in the telephone poll by 25%,” he concludes.

Adjusting the data on these lines would make the two campaigns neck and neck in the telephone poll. Further more, anecdotal evidence clearly supports Anthony’s conclusion that the telephone poll has over-estimated  support for “remain”. The number of older voters who have been in touch to say “Virtually all my friends support leave.”

However, perception of the outcome can influence voters’ intentions. If “leave” is seen as a lost cause, some voters may decide not to bother to turn up. While we cannot force pollsters to re-calibrate their polls, one “perception” factor which works in our favour will be billboards and posters outside houses. Far more people are likely to judge the likely result by the relative numbers of these than by poring over the finer detals of opinion polls and while I can only offer a non-scientific study of my own observations, “leave” boards are sprouting up across my neigbourhood but I have yet to see a single supporter of “remain” advertise the fact outside his or her property. When it comes to passion and commitment, the “remain” side cannot touch us. We can but hope that ths will translate into the victory we have long desired.

The data on which Anthony performed his analysis can be found here.