Would you tell a pollster who you were voting for?
ONE take-away from the events of the last few days in the United States has been that, once again, the dark art of polling seems to be taking a battering. The once-mooted Biden landslide literally slid into a fast-flowing stream of random samples and statistical theory, and just a few hours after the polling stations closed we were momentarily faced with a YMCA-ING Trump as the bookies odds-on favourite.
Why are the polls increasingly getting it wrong? Faced with polarised politics, more access to misinformation and a general feeling of mistrust, has the apparently randomly selected polling respondent suddenly gone a bit rogue? Or have we all become a bit too smart for the binary nature of some of the questioning?
This latest gaping hole between prophecy and reality in elections and referenda tells us a lot about how we are perhaps becoming immune to pollsters. Polling, if you think about it, is a bit of a cheat. Voting in a democracy is meant to be personal. It is about being able to vote without feeling any pressure to do so for a particular party, and to be able to vote secretly.
A bit like religion, it’s between you and your ballot box. So, when some overly inquisitive pollster phones up and starts quizzing us about the inner machinations of our minds, is there not just a wee bit of us that thinks, I’m not giving you the spoiler to this Netflix series.
A few may even deliberately misrepresent just to be awkward. Some respondents apparently apply a so-called ‘social desirability bias’ which sees us respond by lying, to put ourselves in a favourable light with the stranger on the end of the phone line because we would rather not be associated with a candidate who may stand for seemingly negative policies.
A bit like when we are surveyed about whether we think a lockdown is a good idea. Oh yes, say 64% of us, polishing our halos before nipping out for a boozy sesh with a few households that become more numerous as the night goes on.
As was the case during the 2016 US election there seems to be a massive reservoir of shy Trump supporters this time round who are still too embarrassed to admit they supported him. For a multiplicity of reasons their true opinion isn’t being heard by pollsters, with some respondents seemingly fibbing in polls. It’s not new, but does seem to be a product of the increasingly polarised views both in society and on social media.
It is what happened with the Brexit referendum where polls had Leave at a disadvantage, but it went on to win. Because Brexit had been relentlessly painted with the brush of being racist and inward-looking people felt it had a ‘social undesirability factor’ which made them shy about professing support for it. Trump opportunistically saw this as a strategy in the 2016 US Election, even calling himself Mr Brexit in a bid to encourage the silent majority, who he intuitively knew was there, to let their voices be heard.
In 2020 commentators arguably believed this could not happen again because surely the silent majority now had solid evidence to base their decisionmaking on – with nearly 240,000 Americans dead from coronavirus. Or so they thought, until they saw what people were basing their decisions on: 34% of respondents saw the economy as the most important factor in their decision, ahead of the handling of the coronavirus pandemic and racial inequality.
Precisely because of Trump’s emphasis on the economy and apparently on law and order, his appeal is in fact much wider than he is given credit for. There is a hidden seam of Trump voters who are not always accounted for, or connected with. Recently an Arab friend told me that she was shocked that all her friends and family in Michigan – all Arab Muslims – were voting Trump despite what they saw as his anti-muslim stance, but for them a strong economy took precedence. I have relatives State-side who feel the same.
Similarly, despite his misogyny, Trump is a favourite with white women with some 55% voting for him according to a recent exit poll. And in Florida, Trump won 45% of the Latino vote. So despite the Build a Wall calls, his message on strong growth and the entrepreneurial spirit really resonated with these groups who are not always top of the lists that pollsters connect with.
Something does seem to be afoot in the world of pre-election polling. Something that can’t quite be corrected or scientifically weighted one way or the other. But it does make for one hell of a finalé.
Some respondents apply a ‘social desirability bias’ which sees us respond by lying to pollsters