As the world watched Donald Trump voted in as the next President of the United States of America, two questions emerge: Why did Donald Trump win? Why did the Public Opinion Polls get it so wrong?
By Michele Levine, Gary Morgan and Julian McCrann, Roy Morgan Research
As the world watched Donald Trump voted in as the next President of the United States of America, two questions emerge:
Why did Donald Trump win?
Why did the Public Opinion Polls get it so wrong?
1) Why did Donald Trump win?
The ‘rusted on’ Republicans voted for Donald Trump – or NOT Hillary Clinton. The real question is how did Donald Trump attract such a large vote among those who would traditionally vote Democrat or not vote at all?
“Real unemployment number is 20%” said Donald Trump
It seems that everywhere in the world politicians and Governments are seeking to minimise or hide the ugly truth about massive unemployment (and under-employment). For those who are unemployed, whose family members are unemployed or whose children may never get a job, the ‘great unemployment lie’ as Jim Clifton of the Gallup Organisation called it, is just that – a lie. It creates distrust, and at least 20% of people feel disenfranchised. Donald Trump calling real unemployment at 20% was a voice that recognised these disenfranchised electors.
The language of ‘winning again’
Donald Trump’s carefully crafted language offered those who felt like losers or that they were losing out the hope of ‘winning again’ and being ‘winners’. In the Western World at least, the idea of ‘a win’ or ‘luck’ is one of the few ways to motivate the disengaged, depressed and downtrodden.
Donald Trump defied Political Correctness.
Donald Trump gave voice to the deepest fears and anger of the disenfranchised against ‘other’- people of other or different race, colour, gender, religion, political persuasion and opinion. And he did so in language that defied ‘political correctness’ – that many feel render them inarticulate or unable to speak honestly.
2) Why did the Public Opinion Polls get it so wrong?
There was no dearth of public opinion polls in the lead up to yesterday’s US Election. Most, simply reflected the prevailing view – that Hillary Clinton would win – a view held by 90% of people in the US.
Bias towards the Party people ‘think will win’
The phenomenon of bias towards the party or candidate people think will win is probably the most common bias in political polling. Given 90% thought Hillary Clinton would win, the theory is people who supported Donald Trump would be less likely to agree to be interviewed, and less likely to say how they would vote. We see the same phenomenon in Australia with Pauline Hanson. The most recent example of this phenomenon was Brexit.
A Poll is only as good as its Sample
The aim is to poll a representative sample of the population of interest – whether it be the US or relevant States.
The US population is extraordinarily diverse and much more heterogeneous than Australia. To adequately represent the population requires serious attention to the task and in particular must include polling people may not be easily reached by traditional polling such as landline telephones or increasingly commercial Internet panels. The best practice for sampling and surveying is a multi-mode polling methodology that begins with an address-based probability sample and includes face-to-face, telephone and mobile and may incorporate online surveying depending on the respondent’s preferred method of giving their views. This is expensive and although used regularly by Roy Morgan in Australia is rarely, if ever, used in the US.
Voter Turnout – the ‘wild card’ in this US election.
In Australia voting is compulsory and so the vast majority of electors vote. However in the US voting is not compulsory – so voter turnout can be a critical determinant of the outcome of an election. Early indications are that voter turnout in yesterday’s election was up around 5% to a record, or near record levels, in many States. The additional turnout no doubt from ‘hard to get’, ‘disenfranchised’ voters favoured Donald Trump.
Independence and Provenance matters!
In the US as well as Australia there are many polls reported that are of dubious veracity – leaked private party polling is often selectively reported to ‘ send a message’ or ‘create a perception’. It is important when reading polls to review carefully – where they were conducted, who was interviewed, when the poll was taken, exactly what questions were asked and critically who funded the poll.
For more details contact:
Michele Levine +61 411129093, or
Gary Morgan +61 411129094