Polls off-centre, US presidential elections
The application of the law of large numbers and long periods to politics or history signifies nothing less than the wilful obliteration of their very subject matter, and it is a hopeless enterprise to search for meaning in politics or significance in history when everything that is not everyday behavior or automatic trends has been ruled out as immaterial.
(Hannah Arendt, The Human Condition)
US Elections have been at the centre of world mass and social media networks for a while. Statistics and opinion polls, days and hours before the elections, gave some “comfortable” lead to Ms. Clinton, former Secretary of State. But presidential electors’ votes went to Mr Trump who is now the president-elect and will be inaugurated on 20th January 2017.
In this site we are not to evaluate programmes and intentions of Mr Trump and Ms. Clinton. Nor are we in the “catastrophism” of mass and social medias and their apocalyptic approach of elections. But polls off-centre are what interests us here as polls are also power and are supposed to show some perspective of what the centre (being seat of power) is.
Opinion polls have failed again. They failed predicting the “return” landslide of Conservatives to the UK Parliament in 2015. They failed predicting Brexit vote, and now they have just failed the US elections. Thus we can say that political polling has recently become off-centre.
These polling methodes were elaborated in the post-Cold-War world of the 1990 and the new generation of computers then. Thus they are still mass-media-era methods mixed with those of new computers and polls began to predict rather accurate elections results The figure (Courtesy of Washington Post).
But with social medias the world has indeed changed: they are now becoming the centre of the mediasphere, it is no more mass medias.
It is time new polling systems emerged: our world has changed, and the mediasphere’s centre has shifted from mass media. A new centre is new ways of thinking, too, thus new polling methods are needful.
In spite of the fact that Ms. Clinton won the popular votes.
The correlation between computer technology and polling accuracy is interesting according to The Washington Post’s chart. Every “new wave” of computer polling leads to more accurate numbers, but with the time the accuracy vanishes away as people moved on to new generations of computer before polling methods did.
Interesting topic and indeed polls are somehow becoming obsolete, well at least its methodology. Four cases in different context confirm this reading. In 2012, polls gave an advantage of more of 10 points to the actual president of Mexico (Enrique Peña). On Election Day, it turned out he only won by less than 2 points. Media went and attack the poll organizations blaming them of their methodology being flawed or that they did in in purpose to favor the current president. Second case as you mention is the Brexit referendum where the polls seemed to favor the NO. Third case, Colombia peace referendum where the “yes” vote to the peace agreement was winning in the polls. And finally, Trump’s victory.
These four cases in very different democratic socio-political make clear something is going on with polls. It is either the methodology is no longer adapted to our changing societies. Or, our democratic systems are no longer representative of the voters. Or even a mixed of both hypothesis. I think we need to do research from different fields to try to understand this phenomenon.
Thank you Jhonnatan for your comment. And thank you more for the examples you have mentioned (I totally forgot about the Colombian one while writing this entry!). In fact, we need to do research to try to understand this “rupture” between opinion polling and “effective” voting results.
You are welcome. Another example to follow is the primary election for the candidate of the Right in France. In all polls candidate Alain Juppé has always been ahead for 6 points or more compared to other candidates. Even though this is a primary election (not the same than a general election) I am curious to know if the polls will be this time right and if so how accurate. On November 20th we will know for sure.
And yes, Dear Jhonnatan, opinion polls (“sondages”) failed again to predict the favourite candidate of the Primaires de la droite et du centre. Mr Fillon’s lead was higher than any polling centre or method had predicted. But it is different from national elections. Let us watch and see the interesting outcome of this almost-dialectic “opinion polls/votes”.
Yes, opinion polls failed again, and by far! It is worth trying to figure out the reason behind this opposition polls/votes.