Market research asks: What's the point in political polls?
According to recent market research, the reasons behind 2016’s political polls being wildly out was mostly down to shy voters. Or was it?
First came Brexit, when the political polls were fairly steady in predicting a ‘remain’ win by approximately 51% versus 49%. Our own market research gathered through Opinion Outpost’s paid online survey appeared to concur similar results of other market research and political polls before the referendum, yet it showed that the key ‘don’t know’s’ would call it. Then came polling day and the general pre-referendum market research polls did have the percentages mostly correct – though it flipped the other way.
Then came Trump’s surprise victory. Again, market research and political polls predicted a clear win by Clinton, with some online surveys even predicting a 99% chance that she would win. Again, they couldn’t have been more wrong. But why was this?
How market research got it wrong
In the case of Brexit, market research analysts have cited the problems of using telephone polling alongside online survey taking
. Overall, those who answer market research polls over the phone were more likely to vote for Brexit, but with a reduction in this method of polling (and a huge rise in online survey-taking, which generally gave a more pro-remain result) the results were tallied together in a misleading way. What’s more, you cannot vote online in a real election – and the young people who found it easier to vote online in a paid market research poll did not necessarily turn out at the ballot box.
In the case of Trump, market research analysts have cited shy Trump supporters as the reason why the polls were incorrect. Apparently the angry hard-done-by Republicans who voted for Trump were more likely to hang up the phone on a pollster, and many may have been anxious to openly show their support. A similar ‘shy Tory’ model could also have been at play with Brexit, as it was with the general election in the UK the year before. Further market research also considers whether some polling analysists themselves weight the market research polls and online surveys, seeking to confirm their own models of how they want the election to go.
This begs the question – is there really any point in political polls, or should we listen to the market research reports and have a re-think?
Certainly, following this year’s polling catastrophes we will all be taking future political market research polls with a large pinch of salt. The problem is that the market research polls themselves can skew the result. If reluctant voters simply relied on the polling figures, many may have hoped that the ‘remain’ victory was a sure thing, so wouldn’t bother to show up at the polling station that day. Some journalists even went as far as to say that some rogue pollsters’ market research
deliberately lied to us to ‘herd’ their results into the ‘safe’ zone.
So, whether the political market research is still of value remains to be seen. It perhaps it revealed the void in under-represented groups with strong views whom the pollsters (and many politicians) seem to miss. The political polls did not serve their purpose in predicting the result, and in fact they may have contributed to the result swinging in the opposite direction. Next time there is a big election I am sure these factors will be taken into consideration, but nothing gives sure-fire results quite like the actual result. Maybe we will just have to close our ears in the run-up and simply vote privately and however we please.
What do you think about the political polls? Does market research have a place in politics anymore? Join the discussion on our Facebook page, and don’t forget to earn money by sharing your views online with paid market research surveys for cash – on all sorts of subjects – at Opinion Outpost UK.