Opinion polling: democracy’s friend – or foe?

Polling used to be reliable, accurate and uniform – but now it’s anybody’s guess. With headlines shifting from ‘Nigel Farage more unpopular than ever, new poll shows’ to ‘Reform could be the biggest party if Burnham calls an early election, leading polling expert says’ within 48 hours, is polling still useful for tracking shifts in voter intent and attitude toward political parties in the UK? Or is it a sign of a rise in mysterious and unfair methods to sway public opinion?

The rise of new pollsters signals a shift in the landscape, but is this happening because the traditional pollsters no longer know how to poll? The old models were built for a predictable, ‘safe seat’ electorate. But Britain now is more fractured, and politics has become more divisive. We have long had the phenomenon of the ‘shy Tory’, but is there now the ‘shy Reform’ voter, too, unwilling to admit their voting intentions to a stranger?

But polls are powerful, so is there a hidden method for tilting the odds? Does hearing a shift in political attitudes result in swaying public opinion? The UK already restricts the publication of exit polling until 10 pm on the day of elections when voting closes, while South Korea bans polling publications for 21 days prior to the election. If being ahead in the polls shapes how people vote, then publishing polls is not a neutral act; it is an intervention in the outcome. Smaller parties can be declared as irrelevant by pollsters, which becomes a self-perpetuating cycle keeping them irrelevant, feeding the ‘wasted vote’ narrative and suppressing their real support.

And after the furore over the Channel 4 / Verbatim documentary on Reform, the question of who is commissioning a poll seems more pertinent than ever. The British Polling Council and the Market Research Society require that the identity of the client commissioning a poll is disclosed once the results are made public. But that often doesn’t make the headline, nor does it prevent the interested parties from shaping which questions get asked and which results get published.

Is polling still the best method to gauge voting intentions, particularly when people might not be honest about their voting intentions? Are polls simply neutral attempts at measuring the public mood, or is there more going on?