IT’S not necessary to look far to see why opinion polls fail so often. It’s because their predicate is – if corrupt is too strong a word – biased against accuracy.
Polls produce the answers the people who pay for them want. This is as true in the UK as it is in the US. It is why so many polls give counter-intuitive results on so many topics that are then pushed to the media with pretension to accuracy.
It doesn’t matter when in an election cycle political polls are conducted for the media, the results are always what the clients want in order to sway public opinion.
The media try to hide this by having polls conducted jointly on behalf of opposing political sides; Fox News polled in conjunction with the Associated Press for example. The AP is anti-Trump but Fox News is nominally conservative. Their polling put Biden well ahead.
The news media – even when their op-ed pages are conservative – generally skew left so the polls tell them what they want to hear, which is why so much was heard about the putative Biden landslide.
Another crucial factor is that pollsters and journalists come from the same liberal stable in the education industry. They think alike and tell each other what they want to hear.
The polling organisations’ problem is that they were completely wrong in 2016 when they said Hillary Clinton would win and almost as wrong this year when their findings incited the media and politicians to crow that Biden would win by a landslide.
In hindsight, it was delusional to think that the Democratic machine in the battleground states, where Trump surprised them by winning in 2016, would be caught out again. It’s not that the Trump campaign wasn’t ready for dirty tricks. They were powerless against the ability of savvy Democratic officials and judges to manipulate the system.
Pollster Frank Luntz said beforehand that if the polls got this election as spectacularly wrong as they did in 2016, it would be the end of his industry. There, there, Frank. Your industry will have got over its angst by, say, next Wednesday and pursue exactly the same biases for exactly the same reasons in its polling of the 2022 mid-term elections. It’s impossible to be too cynical about politics.
People who do polling for a living like to pretend that their job is a science and have no intention of allowing something like commonsense reasoning to intrude upon whatever abstruse theory emerges from their process of self-justification. Pace Luntz, polling is a billion-dollar industry and it is not going away. And people have short memories.
Every wrongly called election produces its crop of mea culpas, seasoned with a glimpse of some of the supposedly scientific arcana that makes the job so difficult.
The final Wall Street Journal/NBC poll gave Biden a 10 per cent lead. A story in the Journal yesterday said the pollsters had suffered their second successive ‘black eye’.
It quoted Courtney Kennedy of the Pew Research Centre as saying ‘I readily admit that there were problems this year, but it is too soon to know the extent of the problems, or what caused them . . . there was a widespread overstatement of Democratic support.’ The causes in each instance could be different, she cautioned.
In other words, no matter how refined their methodologies, the pollsters still haven’t found out how to speak to the right people in a way that enables them to predict outcomes with reliable accuracy. But they go ahead anyway.
The mid-terms in two years’ time will be fought as fiercely as this election, as Democrats battle to wrest the Senate from its Republican majority standing in their way. The pollsters and their liberal media allies will be up to the same old tricks. These guys never change and nor will the media. Neither will ever let up while there’s still a single Republican left standing to challenge Democratic party rule.