I find the statistics of elections endlessly fascinating.
I have found a story on the internet. I was alerted to it by a satirically fake David Cameron on Twitter, and have seen no other references to it, so I assume that everybody else is ignoring it with the derision it deserves. Except for Witney LibDems, who, without a hint of shame, blazen it on their website as:
I voted in the poll, merely to access the results for the sake of this blog. At 1440 on 26 April, the 'results' read as:
Conservative 12%
Green 29%
Labour 3%
LibDems 52%
Other 5%
The obviously not-very-bright LibDem candidate says:
(although in fairness, she also says she's not planning to hand in her resignation at work any time soon)
If David Cameron were to be unseated next week, there would be huge glee nationwide, not least in some sections of the Tory party, but it's simply not going to happen. Or, in the unlikely event that it does, there is no way that the Oxford Mail can claim to have predicted this.
The simple fact that it allowed to me vote is enough totally to discredit it. I'm not registered to vote in Witney, I haven't even been there! It hasn't allowed me to vote a second time, probably because of cookies, but it would, I assume, allow me to vote from a different machine.
Although it gives the latest picture in terms of percentage, it does not say how many people, registered in Witney or not, have voted. Is it tens of thousands or just a few dozen?
Any such poll or survey which invites people to participate is susceptible to manipulation. Over the years I lose track of the number of times I have been contacted by someone organising a mass-vote in any number of 'polls' from the frivolous to the very serious. They tend to demonstrate who is the most organised partisan.
It is more likely to attract people who regularly interact online. Approximately half the electorate is over 50 and about a quarter of the adult population is of retirement age. A rural Home-Counties constituency such as Witney is probably older on average than, say, where I live, urban Streatham (unless RAF Brize Norton provides an unusual counter to the general rule).
Although I don't have statistics to hand, I think we all know that internet use is much less amongst older people; I'm not sure whether they are more or less likely than younger people to access the local paper proportionate to internet usage. Again, everybody knows older people are much more likely to vote. I would hazard a guess that Witney Conservative Association is very organised at ferrying their elderly supporters to poll**
The only poll that actually matters is the one on 6 May. But, obviously, there are lots of Opinion Polls around that attempt to predict the outcome.
There is a Code of Conduct governing these. Furthermore, they are conducted on the basis of what they believe to be a statistically valid sample. Typically the sample size is about 1,000, and I have problems with this, even though I understand perfectly that this is absolutely valid (+/- 3%) if the sample is random.
Often, they try and get a stratified sample, attempting to reflect the demographic make-up of society. I find it difficult to believe that, considering how much more diverse society has become even since I first studied this subject,let alone since 1945.
I just cannot accept that a sample size of about 1,000 is equally accurate for a larger electorate, with a wider age span, that is generally much better educated, that has so much more ethnic diversity, in a society where women are (generally) less likely to vote as their husband instructs them to, where people are less likely to vote the same way as their parents for that reason, where there is greater access to news, which has faster reaction/ turn-around times, where people's community is less local and more widespread, be it online or with fellow workers who live miles away, and where people are less deferential.
Although I do recognise that irrespective of the population size and individual characteristics, everyone can be grouped into one of a very small group of options - but even then, the options are more numerous than ever before - some commentators proclaiming the move from a two to a three party system fail to see that we are close to a multi-party system: the London Assembly has members from 5 parties, the European Parliament has 8 from Britain, including SNP and Plaid Cymru but excluding Northern Ireland
I can see that a newspaper runs an online poll, hoping that it will attract and keep readers by being interactive and user-driven although I think it.
I realise that I am in danger of disappearing up my own geekiness in this election. A stranger reading this should challenge me to cite sources. I am not writing as an academic; I'm not even writing as a political activist*. I just know this stuff, to me it's common sense and self-evident (but that's because I have specialised over the years, not putting down people whose expert knowledge lies elsewhere)
* I was very active in the Labour Party between 1984 and 2004, was a Labour Councillor in Lambeth 1994-2002, remain a Labour Party member and (small) donor, and have given some time to Streatham CLP in this election
** My partner's late father and my colleague's elderly mother, both Labour voters, habitually got/get lifts to polling stations from Tory workers
