How did the pollsters get it so badly wrong? Why did the mainstream media miss out on the issues that shaped these votes?
My new novel, Pendulum, emerged from the idea that the digital world connects all of us in random ways, exposing us to a previously unimaginable network of information and people. The Internet is changing the foundations of society, affecting how we think about the world and the way in which we interact.
I'm currently adapting Pendulum for NBC Universal and am simultaneously writing the sequel as a novel. I'm immersed in the Dark Web, fascinated by social media, and studiously track any research into how technology is changing behaviour. One doesn't need a PhD to know that western society is becoming increasingly polarised. The US presidential election is a prime example of that polarisation in full effect. Donald Trump, a man who seemingly has no qualifications to make him suitable for office, has just trounced Hillary Clinton, a Secretary of State with a career in politics, winning states that had polled to swing for her. How did the pollsters get it so badly wrong?
Donald Trump - The President We Didn't See Coming?
Photo by Gage Skidmore via Wiki Commons
It's very simple: their methodologies didn't reflect reality, but instead reinforced their prejudices. Pollsters don't tend to publish raw data, instead they apply methodology that is supposed to better extrapolate the sample to the state or national vote. In this election, many pollsters were giving Hillary Clinton a six to ten percent advantage in their methodology, meaning that if a sample of 1,000 people showed a 510/490 advantage to Trump, the final published poll would show somewhere between a 52% to 54% advantage to Clinton.
This inherent advantage was based on the assumption that Clinton would be able to mobilise the Democrat vote the way Barack Obama did in 2008. But the primary results suggested more Republicans would be voting this time, and when they did release raw data, poll after poll showed more people saying they'd vote for Trump, and a greater number of respondents identifying themselves as Republicans. If pollsters had simply published raw data, or corrected their methodology to reflect the outcome of the primaries, a Trump victory would not have come as a shock to anyone. Tracking attendance at candidates' rallies was another good gauge of the likely result: if people can be bothered to come to a rally, they're likely to get out and vote.
Hillary Clinton - Let Down By The Pollsters?
Photo by Gage Skidmore via Wiki Commons
I've heard a number of people say that in light of the Brexit and Trump victories, we should treat polls with caution. I'd argue that the polls have been accurate, but the methodology applied to the data has been flawed. In future, pollsters might want to consider sense-checking their methodologies with a wider range of people, or publish the raw data to enable individuals to judge for themselves.
The EU referendum saw thousands of conversations happen outside the mainstream, with opinions formed, challenged and crystallised on social media. I repeatedly saw issues being discussed which were never addressed by mass broadcasters or newspapers, and the same thing has just happened in the US election. Alt-right figures such as Paul Joseph Watson, Mike Cernovich and Alex Jones have built up huge followings, influencing the views of millions of people through extended social media networks that retweet and share their videos, articles and opinions.
As with the Brexit vote, these alternative media figures have been sharing information and engaging in conversations that simply haven't hit the mainstream. Among other things, these pundits have been fuelled by the Podesta Emails released by Wikileaks, and have been suggesting that Saudi Arabia has donated more than $50m to the Clinton Foundation, that Hillary Clinton was given advance notice of CNN debate questions, that the DNC conspired to deny Bernie Sanders the Democratic nomination, and that Wall Street had a hand in selecting Barack Obama's cabinet. These are just a few of the serious allegations that have remained largely unchallenged and unaddressed by mainstream media, but which have informed the opinions of hundreds of thousands, if not millions of voters.
If these allegations are true, mainstream outlets have been remiss in not investigating, and if they're untrue, their dissemination highlights a serious problem for politicians. In a world where individuals can reach millions of people, how does one combat the spread of false information? Reputations and campaigns can be ruined far below the mainstream radar.
In theory technology has opened us up to the world, but in practice it's actually making us myopic. Our tailored news and social media is self-selecting. We tend to watch and listen to networks we agree with, or follow like-minded people on Twitter, Facebook and Instagram. Our opinions are reinforced because nearly everyone we interact with shares the same views, and when we are faced with someone who has reached the contrary position, we simply cannot understand how a reasonable person can have done so, leading us to dismiss them as stupid, irrational, or irrelevant.
If we're not seeing the same things, or experiencing the same journey, our conversations are robbed of context. Our decisions can seem unreasonable or irrational and the natural response is to deride them as stupid or bigoted, which only polarises and drives people even further apart. The constrained format of social media means that messages have to be delivered in short tweets or posts, and nuance is lost. Discussions quickly degenerate into shouting matches or abuse. Anonymity facilitates that degeneration, giving unidentifiable trolls licence to say things they wouldn't dare put their own names to. And the sheer volume of material pumped out by people means that it is almost impossible for anyone to keep track of the ins and outs of an intricate argument. Doing something shocking, shouting the loudest, and holding the most extreme views gets you heard above the noise, and those kind of messages tend to be simple enough for people to digest quickly.
If we don't make a concerted effort to establish context, either by sharing experiences or going beyond our comfort zones and engaging with people who may not naturally be part of our cohort, then this polarisation will only get worse, and we may one day look back at the 2016 US presidential election as a paragon of moderation and reasoned debate.