So there are just two weeks left until the results of the US presidential election will be known but already we have a good idea of who will win. But do we really?
It’ll be Clinton, right?…
Most opinion polls, both nationwide and state-by-state, point to a Clinton victory. On a national level some surveys, such as the one carried out by ABC News, even have her winning by up to 12 percentage points.
Even more encouraging for the Democrat candidate, polls in the so-called ‘battleground’ states, those that will actually decide the outcome, give her a net advantage: in “must-win” Florida for example she is 3.8% up, a lead which rises to 6.8% in Pennsylvania. (See the bottom of this page for the latest Battleground State polls)
Clinton can even steal some traditional Republican states. The races in Arizona and Georgia are said to be tight and there is even talk of Bush-lovin’, Guvment-hatin’ Texas turning Democrat for the first time since 1976.
CNN (@CNN) October 25, 2016
Don’t bet on it!
Yet all is far from lost for the Trump faithful. For a start, one poll had him tied with Clinton as recently as the past weekend. Not just any old poll either but the International Business Daily tracking poll (IBD/TIPP), which was the most accurate of all the US opinion polls in the 2012 election, according to the New York Times.
So an opinion poll with a good track record gives Trump a fighting chance even with the whole Pussy-grabbingGate affair still fresh in the memory. That memory may all but have vanished come November 8.
And IBD is not the only poll pointing to Trump.
So which poll to believe?
Opinion polling is far from an exact science. Pollsters take a random sample of the population and then, usually, phone those people at home to ask who they’ll vote for. This methodology poses numerous problems: firstly it is hard to put together a completely representative sample of the electorate, then of course not everybody nowadays owns a landline (the response rate has slipped to less than 10 percent of people called). On top of that, no responder is obliged to tell the truth. Given much of the media treatment of Donald Trump, whether justified or not, people may not admit over the phone that they’ll vote for him before doing just that behind the comfort of the polling curtain. It’s a phenomenon that has names and Wikipedia pages (The Bradley Effect or The Shy Tory Factor).
And opinion polls have been wrong many times before. Pollsters Gallup predicted Mitt Romney would beat Barack Obama in 2012.
Then of course this is the year of the Brexit. Who actually thought that would happen until it did? Not most polls, that’s who.
Here is the polling average for Brexit, vs. the past three months of election polls. pic.twitter.com/PQO3cNkxlA— Ariel Edwards-Levy (@aedwardslevy) October 24, 2016
That is not to say opinion polls are useless and should be ignored. When many different polls are taken together, trends emerge. That is the case with the infographic below, which suggests that, if the vote was today, Hillary Clinton would win the election.
But don’t forget, opinion polls are not gospel truth. And the vote is not today.