Viewpoint: Is opinion polling broken?

This is not something that has suddenly happened this year, this has been something brewing for a number of years, and I'm really beginning to more than question the accuracy of the polls.  I'm actually beginning to think that the whole process of opinion polling may now be fatally flawed, or at best compromised badly.

Now, there have been some changes to what is prompted by the questioners, but in my view, that is not the big problem.  My feeling is that the biggest issue right now with opinion polling is the way it is conducted right now, and how they select the people they interview.

Some pollsters now use a regular group of volunteers polling them on the internet.  Then they weight the results based on demographic information.  This I feel is seriously flawed, because it is open to manipulation.  Anybody can volunteer to give their time to the polling companies, and often, I suspect that some parties might try to get their operatives to volunteer to load the system up with representatives of a particular party.  The weighting aspect of this method is also flawed, because it's based on a lot of assumptions that don't actually always play out as you think they would. 

Those that still rely on telephone calls to canvas potential voters, are also somewhat flawed.  Calls from polling companies are often cold calls, and as such are not treated well by respondents.  The idea that you can get an accurate idea of opinions from a bunch of random cold calls, is a seriously flawed idea.

The net result is reflected in the actual results of the polls themselves.  Since Boris Johnson took over as Conservative Party leader and Prime Minister, he has stumbled so many times that he makes his predecessor Theresa May look actually competent.  He lost almost all of the key votes in Parliament in the short time he was in there as PM, he lost a battle in the courts over an illegal prorogation of Parliament., and he's been losing and poorly performing in TV debates already.  Yet, this week, a poll by Opinionum for The Observer, put Boris Johnson's Conservatives 19% ahead of everybody else. 

This does not add up.  Public perception of Boris Johnson is not a good one.  He's seen as an incompetent blunderer, full of bluster and Brexit, full of sound and fury signifying absolutely nothing.  Even within his own party, he's not seen as a great politician.  Yet somehow, his party has a massive lead in the polls?  How on earth do you square that circle of hell?  Nothing about that adds up at all.

Jeremy Corbyn hasn't been viewed as a competent leader of the Opposition.  This is despite him inflicting more government defeats in his time as leader of the Opposition than any previous opposition leader in the modern era.  And whilst some of his decisions have been questionable, any lack of competence he has shown pales into insignificance behind the incompetence of both Boris Johnson, and his predecessor Theresa May as PM.

To my practiced eyes and mind, something isn't adding up here, and not all of this can be blamed entirely on the media, as much as many might like to believe that to be the case.  Sure some of it can be blamed on media bias, but the Tory press is a dying medium right now, they are losing readers, they are losing money, and they just can't be the only factor.  A contributing factor yes, but not the main one.

And as Sherlock Holmes would say, when you have eliminated the impossible, whatever remains, however unlikely, must be the truth.  So, we must look to the polling methodologies themselves, and how flawed they are, and as I've already stated, they are seriously flawed, compromised beyond reasonable expectations.

Even the Exit Polls of the two previous elections in 2015 and 2017, which are done face to face, rather than over the phone or by internet, have been a bit off the mark when it came to actually predicting the result.  Granted they weren't as far off the mark as they were back in 1992, when they completely screwed up, but it's far enough for me to question their reliability.  In other countries, exit polling has been much more accurate in predicting the result.  It does leave me wondering how we on the UK aren't able to get it right when other countries pollsters can.

That's why I question the weighting that is given to polls.  It seems to be at least a bit out of sync with actual reality.

Overall, I have to believe that something is serious wrong in the world of opinion polling, and yes, it might be broken, maybe even broken beyond repair.