Is Political Polling Losing Its Touch?

In the lead up to every election and referendum, pollsters try to gauge the nation’s voting intention. How? They do it by surveying a relatively small sample of the public – usually in the thousands – who they feel are representative of the nation as a whole. And for the most part, historically, they’ve got it right. However, some elections in the past decade have thrown up surprises for the pollsters. Read on as we explore whether political polling is still the reliable barometer it used to be.

Modern day hiccups

Going into the 21st century, pollsters did pretty well. They correctly predicted a Labour government in 2001 and the same again in 2005, accurately forecasting the reduced majority. In 2010, however, they predicted quite firmly that the Conservatives would win – so it was a shock to most in the UK when they woke to a hung parliament.

The hiccups continued in 2015, when pollsters suggested a narrow lead for the Conservatives, probably requiring another coalition. The result was a Conservative majority, albeit just six seats. Then to the EU referendum in 2016. Not strictly an election, it was predicted to lean towards Remain by an average of around 5 percent in the week before the vote. That didn’t translate on June 24rd, however, when we saw a lead of 3.8 percent for Leave.

Most recently, the polls were – by and large – proven wrong when Theresa May’s snap election ended in a hung parliament. Only Survation and YouGov predicted a similar result, with others suggesting a reduced but sizeable gap of around 8 percent.

Over in the states, the polls have seen a similar hint of unreliability. Hillary Clinton lead in almost every pre-election poll, but it was Trump who triumphed in late 2016.

So, what can we read into this trend of polling inaccuracies?

The end goal for polls

To assess the efficiency of opinion polls, we need to look at their actual purpose. Sometimes, they can’t accurately predict elections – but are they intended to do so in the first place? The answer is no. Opinion polls are a way of gauging public opinion. As long as they’re not taken too seriously, they can give a decent idea of the broad picture.

In weeks leading up to the 2017 General Election, for instance, the polls displayed a surge in support for Labour. Some might argue that it was up to political commentators to use this to predict the hung parliament. Instead, they took the numbers at face value and saw it as an outright Conservative lead.

By reflecting the final results on a number of occasions, opinion polls have gained a reputation as an election forecast. Essentially, they’re a victim of their own success, and have become relied upon quite heavily by the media to predict election outcomes. In the words of many politicians from all sides – there’s only one poll that counts and it’s the one that takes place on election day.

Over to you

We’d love to hear what you think about opinion polls. Are they relied upon too much? Or do you think they need to up their game? The political landscape is always changing, so maybe it’s time they adapted and took on their new role as election forecasters? Whatever you think, be sure to give us your polling opinion in the comments section below.