The politics of polling Contents

Chapter 1: Introduction

Political polling in the United Kingdom

1.Political opinion polls have become an increasingly prominent feature of election campaigns in recent decades. In the period between the 2010 and 2017 General Elections, over 2,200 voting intention polls were conducted.1

2.During the last three years, the United Kingdom (UK) has faced two General Elections and a referendum on the UK’s membership of the European Union (EU). For each of those events, albeit to varying degrees, the polls ‘called it wrong’. In 2015, most polls predicted a hung parliament, when in fact the Conservative party won an overall majority. Before the referendum on the UK’s membership of the EU, a majority of the final polls suggested that there was a majority in favour of ‘Remain’, whereas ‘Leave’ won. Then in 2017, most polls suggested that the Conservative party would secure an overall majority, whereas the election actually resulted in a hung Parliament.

3.It is worth noting that each of these elections were close-run contests which are harder to call than less competitive races, and that polling companies made every effort to ensure their polls were as accurate as possible. However, these events have led to a widespread loss of confidence in polling. They also raised concerns about the extent to which inaccurate polls might be shaping the ‘narrative’ during election campaigns, and therefore how they might be affecting the democratic process.

The Committee’s remit

4.It was against this background that this Committee was established. The Committee was appointed by the House on 29 June 2017 “to consider the effects of political polling and digital media on politics”.2 This followed a proposal put forward by Lord Foulkes of Cumnock (a Member of this Committee) to the House of Lords Liaison Committee. In recommending that this Committee be established, the Liaison Committee suggested that we might wish to consider the following themes:

5.In this report, we have made recommendations which address the first three of these themes. However, as our inquiry progressed, it became clear that the fourth strand was simply too large and complex a topic to cover as part of this inquiry.

6.The rise of digital media has diverse implications for politics, far wider than just its association with political polling. It is not possible to summarise effectively the full extent of this issue here, but to take just a few recent examples, politics can and is being informed and influenced by digital activity in the following ways:

7.Furthermore, social media are being utilised to influence the democratic process, in both positive and negative ways. Some examples include:

8.While all types of media could be accused of creating some sort of ‘echo chamber’, the difference between traditional media and social media is the scale and intensity of social media’s reach, and the lack of context. For example, if a person reads a particular newspaper every day then they will be exposed to articles and comments written with a particular political bias, but the reader will generally be aware of the fact that the newspaper is written from a particular political standpoint and can bear this in mind when reading the articles. However, on social media, the writers of online content are often unknown to the reader, who has no way of knowing the political leanings of the author and no other information on which to judge or scrutinise the information they are reading. Furthermore, if the person checks social media regularly throughout the day, they might be bombarded continually with subtle political messaging without even realising it.

9.The other important point to note here is that digital and social media are neither regulated like broadcast media nor carry the responsibilities of publishers like the print media. Political messages can be spread rapidly around the world with few checks or balances, and without anybody having to verify their veracity. Sometimes, an article or a website might note that the information has been written by a particular lobbying or interest group (though this is not required), and if an advertisement has been authorised by a political party then its cost must be included within the campaign spending that they declare to the Electoral Commission. However, many other types of political messages can be posted online by individuals, private companies, lobbying groups, foreign governments and many others with no oversight or regulation. A member of the public might find it impossible to know who exactly has posted a particular comment or article, or who funded it.

10.During the course of our inquiry, we have only been able to touch upon some of these issues in a very limited way. Assessing the overall impact of these developments on politics would be a vast and highly complex task, and would need to include analysis of state responsibilities, corporate financing, media reporting, developments in artificial intelligence and a whole host of social issues, amongst other things. Such an inquiry would also need to keep pace with the rapidly changing developments which are happening in this area every day.

11.The issues outlined above stretched beyond our remit and could not have been covered in sufficient detail within our reporting timeframe. Furthermore, there are already a number of other bodies investigating some of these issues. To take a few examples, during the course of our inquiry, the House of Commons Digital, Culture, Media and Sport Committee were undertaking an inquiry into ‘fake news’,15 the European Commission established a high-level expert group which has now produced a report on fake news and online disinformation,16 and social media companies have announced new initiatives aiming to restore public trust in the content published on their websites17 (although the effectiveness of efforts to address tensions between commercial and public interests remains to be seen).

12.Given the challenges outlined above, we therefore limited our inquiry to the ways in which digital media impacts upon political polling, and vice versa.

13.At the outset, we agreed that our inquiry should not conduct a post-mortem on what went wrong with polling over recent elections. This was comprehensively covered by the Report of the Inquiry into the 2015 British general election opinion polls.18 That inquiry was established by the British Polling Council (BPC) and Market Research Society (MRS) but chaired independently by Professor Patrick Sturgis (who also acted as Specialist Adviser to our Committee). Following the 2017 General Election, the BPC asked those of its members who produced final polls for that election to produce a “lessons learned” report, in order to examine the ongoing challenges facing the industry.19 We have not, therefore, attempted to replicate this work by delving in detail into the methodological causes of polling errors. Instead, we have taken a wider look at the challenges facing polling organisations and how their work can influence the political process in the UK.

14.As a committee of the UK Parliament, our recommendations are limited to polling in the UK. We have, however, made recommendations to the UK Government on how it might work with international partners to address some of the wider issues relating to digital media.

The Committee’s work

15.In July 2017, we published a call for evidence, which is reprinted in Appendix 4. Over the course of our inquiry we received 31 submissions of written evidence and heard from 40 witnesses in 23 evidence sessions. We are grateful to all those who took the time to provide us with evidence. A list of all these people and organisations is included in Appendix 2.

16.In order to assist our deliberations, we also received informal briefings from a number of academics and other experts, who are listed in Appendix 3. We thank them for giving up their time in order to help us explore the topics in greater detail.

17.Finally, we are also grateful to Professor Patrick Sturgis, Professor of Research Methodology and Director of the National Centre for Research Methods at the University of Southampton, who served as the Committee’s Specialist Adviser. His knowledge and assistance have been immensely helpful throughout the course of our inquiry, though we stress that the views contained in this report are ours alone.

1 Anthony Wells, UK Polling Report, ‘Voting intention’: [accessed 20 March 2018]

2 HL Deb, 29 June 2017, col 563

3 Liaison Committee, New investigative committees in the 2017–18 Session (2nd Report, Session 2016–17, HL Paper 144)

4 Crowdfunder, ‘From the 48% to Theresa May’: [accessed 20 March 2018]

5, ‘About’: [accessed 20 March 2018]

6 UK Parliament, ‘Find out more about e-petitions’: [accessed 20 March 2018]

7 Perry Stein, ‘The woman who started the Women’s March with a Facebook post reflects: ‘It was mind-boggling’’, The Washington Post (31 January 2017): [accessed 20 March 2018]

8 ‘Corbyn tackles PM with crowdsourced questions’, Sky News (16 September 2015): [accessed 20 March 2018]

9 ‘Russia-Trump inquiry: Russians charged over US 2016 election tampering’, BBC News (17 February 2018): [accessed 20 March 2018]

10 Gordon Corera, ‘Russia ‘will target US mid-term elections’ says CIA chief’, BBC News (29 January 2018): [accessed 20 March 2018]

11 ‘France Macron, alongside Putin, denounces two Russian media for election meddling’, Reuters (29 May 2017): [accessed 20 March 2018]

12 ‘Theresa May accuses Vladimir Putin of election meddling’, BBC News (14 November 2017): [accessed 20 March 2018]

13 Kevin Granville, ‘How Facebook Users’ Data Was Acquired by Cambridge Analytica, Triggering an Outcry’, The New York Times (20 March 2018): [accessed 20 March 2018]

14 ‘Cambridge Analytica: Mark Zuckerberg asked to appear before MPs’, BBC News (20 March 2018): [accessed 20 March 2018]

16 European Commission, ‘Final report of the High Level Expert Group on Fake news and Online Disinformation’, (12 March 2018): [accessed 20 March 2018]

17 Written evidence from Google (PPD0029) and written evidence from Facebook (PPD0030)

18 Professor Patrick Sturgis, Dr Nick Baker, Dr Mario Callegaro, Dr Stephen Fisher, Professor Jane Green, Professor Will Jennings, Dr Jouni Kuha, Dr Ben Lauderdale and Dr Patten Smith, ‘Report of the Inquiry into the 2015 British general election opinion polls’, National Centre for Research Methods, British Polling Council and Market Research Society (March 2016): [accessed 20 March 2018]

19 British Polling Council, Press Release, ‘General Election: 8 June 2017’ (9 June 2016): [accessed 20 March 2018]

© Parliamentary copyright 2018