Picking the polls: how can business prepare for the UK general election?

Crispin Beale
- Leadership - Jun 02, 2017

As June 8th draws closer and political parties across the UK have shared their manifestos, the pressure is mounting for businesses to prepare for life after the general election.

However, given the unexpected unfurling of political events in 2016 – the Brexit referendum and the US general election – how can companies plan productively in the face of apparently historically unreliable polls? 

With a host of business implications dependent on the outcome – from share prices and investment opportunities, to international trade deals and employment costs – it’s vital for companies to be equipped with the best-possible data in advance of the election to prepare for potential challenges that may arise.

Gaining accurate insight into current opinion, and therefore the probable outcome of the public vote, depends on reliable methods of polling in the run-up to election day, which begs the question: what went so wrong on the previous occasions?

To answer this question, let’s explore some of the research considerations that businesses can look out for to understand which polls to trust.

Establishing a method in the madness

It can be easy for pollsters to become caught up in the moment and focus on extracting quick-fire responses from their audience in a race to reveal public opinion, when in fact garnering more accurate responses should be the top priority. Adopting a robust methodology coupled with a representative sample of the population is key to achieving the most reliable, insightful data.

Firstly, samples need to be truly representative of the audience most likely to turn out to vote on 8th June. For instance, there is little use in basing survey findings simply on the first 1,000 respondents from a demographically random group. Or equally, specifically questioning a group of millennials when the likelihood is a large proportion of voters may fall into the retiree age bracket (or vice versa). Therefore, polling companies should be mindful of choosing a methodology that allows them to determine their audience in advance of carrying out a survey.

Selecting the right profiling methods is also key in identifying a suitable sample. For example, telephone interviews allow the researcher to establish the basic profile of their respondent within minutes, and are therefore often a far quicker method than conducting both online and face-to-face interviews. However, it is important to remember that although it is common for telephone interviews to take place within standard office hours, it is vital to include evening and weekend conversations and calls to mobiles within the survey to form a more representative sample of participants from all walks of life.

All in all, telephone interviews, especially if for a specific geography/constituency, are generally found to produce more accurate – and perhaps surprisingly faster – results in comparison to relying on information supplied through an online questionnaire, especially when online participants are incentivised by a reward. Despite the popularity of online surveys – due to their speed, cost-effectiveness, and scale – the inability for the researcher to gain a background understanding of the respondent often leaves the data open to misinterpretation, thus increasing the risk of unintentionally skewed results. The researcher is also limited by the volume of appropriate targeted samples available and is at the mercy of when a respondent looks at their device and decides to respond (or not). While good online research is hugely powerful, it is important not to think of it as a panacea. With telephone interviewing, 200 interviewers can proactively be ringing numbers in a particular geography and seeking out, say, 18-24 year old employed males.

Removing the margin for error

It’s also important to bear in mind that respondents’ thought processes can vary significantly depending on the environment in which they are interviewed and this is particularly true when discussing a politically sensitive subject. For example, when speaking face-to-face, many participants feel pressure to provide answers that they deem more “socially" acceptable, regardless of their own feelings on the matter, and crucially, the way in which they expect to vote.

On the other hand, online respondents often feel they are more at liberty to express bold or controversial opinions, and may sometimes even alter their responses to challenge the status quo.

In telephone conversations, however, well-trained interviewers will quickly establish a relationship with the participant to ensure they feel comfortable with providing honest, heart-felt opinions, thus forging a path between the two extremes and ultimately, achieving more accurate results. This method is particularly effective if the researcher can explain that they are working independently and without the influence of any other organisations – i.e. political parties – as respondents will be less wary of indirectly offending the candidates they are discussing.

Looking beyond the data

Collecting accurate data on public opinion is a good start, although based on the nature of current electoral systems it is by no means the only factor in predicting the correct results. In the US for example, while many pollsters predicted the presidential election voting proportions within a reasonable margin for error, this proved to be of little consequence when the results were announced, due to the way the US political system operates. Despite outperforming Trump in the popular vote, as quite rightly predicted by the pollsters, Clinton did not become president due to the breakdown of votes into Electoral Colleges, a significant factor that some polling companies simply failed to consider at the time.

To restore voter confidence in the run-up to the UK election, pollsters need to determine public opinion in each constituency and structure their surveys – and therefore predictions – on these results, rather than the wider national voting numbers. Therefore, interviewers will not only need to understand who they are speaking to, but also, where in UK they are located.

Despite the seemingly relentless negative media attention over the past year or so, polls are becoming more important than ever, especially for businesses that need to understand the future implications of the snap election in June.

By carefully identifying how pollsters are carrying out their research, and therefore which polls to trust, business leaders can rest safe in the knowledge they will be prepared – regardless of the outcome.

By Crispin Beale, CEO, Chime Insight & Engagement Group

  • Tags:

Like what you see! Signup for our weekly newsletter