Skip to main content
News

People are unpredictable at predicting what they will do when it comes to politics

Professor Philip Cowley and Professor Tim Bale from Queen Mary's School of Politics and International Relations have written a joint opinion piece for The Times. Using the example of Donald Trump's recent visit to the UK, and drawing on data from their recent YouGov polls, they argue that people are "hopeless" at predicting their own behaviour.

Published on:

How good are you at predicting what you’ll do in the future? Or even remembering what you’ve recently done — particularly when it comes to politics? Not always that good, if our research is anything to go by.

Prior to Donald Trump’s state visit to the UK, we surveyed Londoners on their attitudes to the visit. This was the latest in a series of polls, conducted by YouGov, which Queen Mary University of London’s Mile End Institute has been running on the politics of London.

Opposition does not mean action

More than half of respondents (54 per cent) said they opposed the US president coming to the UK, compared with just 24 per cent who were supportive. We also asked people whether or not they were intending to go to any of the organised protests against his visit. Some 13 per cent said that they were likely to take part — of whom 4 per cent said “very likely” and 9 per cent said “fairly likely”.

If all of those saying their participation was likely had turned up, attendance at the anti-Trump protests would have involved close to a million Londoners (916,000), in addition to anyone who might have travelled in from outside the capital.

Nobody — apart from Mr Trump himself perhaps — is denying that those protests went ahead in June but almost a million Londoners? Hardly.

While crowd sizes are always difficult to estimate, even the organisers of the main London rally — who have an incentive to accentuate the positive — claimed a figure of about 75,000 people and some news organisations were sceptical it was this high. Yet even 75,000 is a lot less than 916,000.

What explains this gap? It is worth noting that the survey didn’t ask people if they were definitely going, only if they were “very” or “fairly” likely to do so, and we wouldn’t expect every single one of these people to turn up. Yet even if we discount entirely all of those who said they were “fairly” likely to attend, even the figure for those who claimed they were “very likely” to take part in the demo would have yielded a crowd of more than a quarter of a million London adults, plus any out of towners.

This is, then, yet further proof that people can be pretty hopeless at predicting their own behaviour – especially those who say they’re going to something rather than nothing. Some people may have been genuinely keen to go, only for something – work, childcare, whatever – to prevent them.

People do not always tell the truth

Some may have a very low bar for what “very” or “fairly” likely may mean. Some may be using the question more to signify that they did not approve of the trip rather than actually indicating they intended to go, whether they knew this or not.

So in a subsequent survey, this month and again with YouGov, we asked Londoners if they had been on the demo, asking them, in other words, not to predict what they might do, but what they actually had done. Only 2 per cent claimed they’d protested against Mr Trump. This compares with 94 per cent who said they did not go, along with 4 per cent who – for reasons we can only guess at – said they were “not sure”.

That 2 per cent would still, however, mean that 140,000 Londoners joined the protests — roughly double the number that did take part, even assuming that not a single participant came from outside of London. So even this figure is way too high.

Again, there are various possible reasons for this. All surveys have an element of sampling error (of about plus or minus 3 percentage points on any one figure), so it is possible that all of those who responded to the survey by saying they attended did in fact do so. Yet equally we do know that people lie about these sort of things. Some of us – whisper it softly – tell fibs that fit with our self-image as right-on and radical.

One study of voter recall found that people both claimed to have voted when they did not and vice versa, but the former was more likely and it was especially common among those who thought voting was important. As the study’s author, Paul Whiteley, noted: “People only mislead when it matters.”

One clue that this might be going on here can be found in those who said they were “not sure”. While there may be genuine reasons why you could be unsure whether you went on a demonstration (perhaps you go on so many demos you can no longer tell which is which?) the “not sure” figure was as high as 10 per cent among those aged 18-24, which was exactly the same age group who had said they were most likely to attend in the first place. Even anonymously, and as part of an online survey, some people just can’t bring themselves to admit they have fallen short.

This opinion piece was originally published in The Times on 27 June 2019.

Back to top