How much can we really trust polls?

Posted by on Aug 2, 2013 in uncategorized, Blogs, Philip Berman

Not a lot based on a comical experience I had talking to an ICM researcher this evening.  A young chap from the company phoned about an hour ago asking if he could talk to me about my past general election voting behaviour and voting intentions in Brighton and Hove.

He asked me a number of questions about who I had voted for in the 2010 general election, whether I would vote in the next general election, who I would vote for and so on. After five minutes he asked if I knew the name my local MP. I said that I didn’t think I did although I thought he or she was a Tory. (I’m a little bit embarrassed about not being sure as I used to series produce a programme about politics for the BBC. However, in my partial defence, I left the job before the last general election and was put off most politicians for life after covering the MPs’ expenses scandal). He then asked if I would vote for Caroline Lucas in the next election. I said that I thought that she was in a different constitunecy from where I live (Hove) and that I couldn’t vote for her even if I wanted to!

He seemed rather baffled  by my response. It suddenly became clear to me that he had been asking me all these questions based on an assumption that I lived in the Pavilion constituency. I explained that there were three constituencies in Brighton and Hove and that Caroline Lucas’ was an MP in just one of them – Pavilion – and that I thought I lived in a different one. As I’ve only recently moved to my current address I wasn’t sure what constiuency I lived in but I was presuming he knew, otherwise why was he asking me such specific questions about my intentions for a consttuency I didn’t live in?  I asked whether he knew for sure which constituency I was in. He said he didn’t know and would have to go away and check.

He came back two minutes later and said to me: “I think we have the wrong candidate.” I think what he meant to say was that my answers weren’t relevant to this research because I didn’t live in the constituency for the candidate they were polling. He apologised for wasting my time and made a swift exit.

Odd mistake on the part of  ICM. Why did the junior guy – he sounded very young – assume everyone in Brighton and Hove voted in one constituency? It makes you wonder how accurate these polls really are if they can’t get these basics right.

If I hadn’t picked up on one of his questions, all my answers would have been used in  a misleading way – in a poll  for a constituency I wasn’t entitled to vote in!

Also, I’m quite intrigued as to who is paying for this poll? A newspaper? A political party? There’s evidently great interest already in the Pavilion constitunecy and so no doubt this is just the beginning of a big fuss coming to a street near you very soon.

Leave a Reply