Chapter 5: Oversight of polling
183.In previous chapters we highlighted the fact that recent high profile polling failures have dented confidence in the industry, that polling is getting harder to do, and that there are incidences of misleading reporting in the media that are undermining the efforts of responsible polling organisations. Therefore we examined the current system of oversight for polling to judge whether its mechanisms were adequate to meet the challenges associated with polling now and in the future.
184.This Chapter focuses on the extent and limitations of current supervisory arrangements for polling, and the arguments for and against further regulation.
Current oversight of polling
185.The Representation of the People Act 1983 prohibits, before the poll for an election has closed, the publication of the following:
“(a) any statement relating to the way in which voters have voted at the election where that statement is (or might reasonably be taken to be) based on information given by voters after they have voted, or
(b) any forecast as to the result of the election which is (or might reasonably be taken to be) based on information so given.”170
186.Aside from these provisions, there is no legal regulation of polling in the UK. Instead, most of the polling organisations belong to one or both of two industry bodies which require them to comply with their codes of conduct: the British Polling Council (BPC) and the Market Research Society (MRS).
British Polling Council
187.The BPC describes itself as “an association of polling organisations that publish polls.”171 Membership of the BPC is voluntary and is open to “any organisation that conducts for multiple clients polls or surveys designed to ascertain the views of a representative sample of a specified population, such as all voters in Great Britain.” The BPC described its activities as forming the “current system of self-regulation” for the polling industry.172
188.Self-regulation is described by the National Audit Office in the following terms:
“An industry or a profession can self-regulate, for example through the use of codes of conduct, customer charters, standards or accreditation. In many cases rules and codes of conduct will be formulated by a trade association, or other industry representative under their own initiative. In other cases, an industry or profession self-regulate in response to delivering a stated government objective. In self-regulation, the industry is solely responsible for monitoring and enforcing members’ compliance. This enforcement can be achieved either first hand or through other bodies set up by the industry.”173
The BPC broadly meets this definition—its first stated objective is to “promote transparency in the publication of polls.” It told us that it does this by “providing and enforcing a detailed specification of the information that should routinely be made available whenever a poll is published”.174
189.In August 2017, the BPC told us that it had 16 companies registered as members, including all of the UK-based organisations that conduct polls of voting intention across the UK, with the exception of Lord Ashcroft Polls. The BPC is run by a team of three officers: a President, a Secretary/Treasurer, and a Committee Member. By convention, the President is someone not currently involved in commercial polling and who thus can act as a neutral chair. All three perform their duties in a voluntary capacity. The BPC also maintains a Committee of Disclosure, which can be required to make judgements as part of the BPC’s complaints procedure, and which consists of a mixture of those working in polling organisations, journalists, and academics.175
190.Members of the BPC are required to abide by the Council’s rules on disclosure (which apply to all polls and surveys conducted, not just voting intention polls) and to publish a range of information at publication of a poll (as outlined in paragraphs 113 and 114 of this report). The rules of disclosure are aimed at facilitating transparency—providing the information necessary to allow anyone to evaluate whether a poll and its claims are robust. The BPC can investigate complaints against member companies in respect of its rules of disclosure and has the power to suspend or expel the organisation from the BPC. However, this enforcement power has not been used extensively. The BPC told us that:
“During the last five years, the council has received under this procedure only four complaints that fell within its remit. Of these, two concerned polls that the member incorrectly thought were not covered by the BPC’s rules, and one involved an administrative error by the member. In all four cases the complaint was either partially or wholly upheld by the officers and each time the member organisation made the relevant data available immediately. Since the BPC was established 13 years ago, an investigating sub-committee has only had to be convened on one occasion since the BPC was established; it upheld the initial judgement of the officers that the details of a poll should be released and the member company duly complied.”176
191.The BPC told us that it does not (and practically, cannot) express a view on “the merits or otherwise of a particular poll.” It was clear that this constraint was perceived as being due to limited resources, as the BPC stated that “it lacks the financial resource to sustain any legal action to which any such expression might give rise.”177 Professor Sir John Curtice, President of the BPC, told us:
“There is undoubtedly one constraint on what the British Polling Council can do. This is not an organisation with thousands of pounds of resources. It has a few thousand quid. One of the simple reasons why we as a council are not in a position to say, ‘This poll is good. This poll is bad’, is that we cannot afford to fight a court case brought by a company that says it is bad.”178
192.We also noted that the BPC’s stipulations around transparency of funding only extended to printing the name of the client commissioning the survey, rather than requiring the publication of the source of funding which might lie behind the commissioner of the poll. When asked whether information on who commissioned and financed the polls was readily available, Professor Susan Banducci, Professor and Director of the Exeter Q-Step Centre at the University of Exeter, said:
“I have to say no. That information could be improved, and it might improve public understanding of polls. Of course, it relies on journalists reporting on and investigating those sorts of questions. It is an area where there can be greater transparency about who is commissioning polls and who is paying for them.”179
Market Research Society
193.Standards for the industry are also set by the MRS, of which the major polling companies are also members. The MRS “promotes, develops, supports and regulates standards and innovation across market, opinion and social research and data analytics.”180 Its standards for the research sector are contained in its Code of Conduct, which covers commissioning and design, client confidentiality, informed consent, participant anonymity, data collection, analysis and reporting of findings, and data security.
194.The MRS stated that its Code “supports those engaged in market, opinion and social research in maintaining professional standards and reassures the general public that research is carried out in a professional and ethical manner.” MRS accredited individuals and organisations must comply with the Code and its associated disciplinary and complaint mechanisms.181
195.Under the MRS Disciplinary Regulations, if a member is found to have breached the Code, a number of disciplinary actions can be recommended, including giving a warning or reprimand, requiring a member to give a written undertaking to refrain from continuing or repeating the unprofessional conduct in question, or ultimately suspending or expelling a member from the MRS.182 Jane Frost CBE, Chief Executive Officer of the MRS, confirmed that the MRS had not yet had cause to expel a company, but that it had found against companies and had had cause to amend its regulations:
“We have not expelled a company. This year we had an issue on which we found against someone. We have only two a year that go to that length. As a result of the way in which previous inquiries were conducted, we have changed our regulations to ensure that nobody can resign before a finding is made against them. They will get an automatic finding against them, which they will carry with them. They cannot resign and get away with it.”183
196.Jane Frost told us that around 80% of research companies were accredited through its system.184 She also confirmed that the MRS worked closely with the BPC, ensuring they were always in contact “whenever there is a big occasion that is likely to generate interest.”185
197.While we were in no doubt that the MRS’s regulatory coverage is extensive, Jane Frost highlighted that not all of the polling organisations that were members of the BPC were covered by the MRS as well: “Most of the members of the British Polling Council—all except three—are accredited by us or include accredited members. We have about 530 accredited companies. The BPC is much smaller”. Given that not all companies are members of both organisations, some polling organisations could therefore be adhering to different standards. When asked whether polling, in particular that done during a General Election, needed more careful monitoring than other surveys, Jane Frost replied:
“Yes, I think that the issue has to be handled properly. Having two organisations for this one area may not be optimal, but the fact that 95% of the area is covered by our wider regulations means that a lot of pollsters have been under regulation and training for lots of issues for a lot of time. Therefore, they are very experienced and well aware of what the rules are.”186
Media regulators
198.Different sectors of the media are regulated in different ways. Traditional print media is self-regulated. Most national newspapers are members of the Independent Press Standards Organisation (IPSO), though some instead choose to regulate themselves (for example, the Guardian has a Readers’ Editor which considers complaints and a review panel which considers appeals). Other publications may be members of the Independent Monitor for the Press (IMPRESS). By contrast, broadcasters are statutorily regulated by the Office of Communications (Ofcom). In the digital sphere, there is a wide variety of outlets which publish what purports to be news. These range from the online arms of reputable sources, such as the BBC News and Sky News websites, to small digital-only platforms, for which there is minimal regulation. While each of the regulatory bodies outlined above produces guidance on fair and accurate reporting, individual media outlets may also have their own codes of conduct on the reporting of polls. This means that there are a variety of codes and guidance which might be relevant in relation to the reporting of polls, and a number of different places towards which complaints might be directed when polls are misreported.
Newspaper regulators
199.As the media’s reporting of polling was a key concern for this inquiry, we considered the role of the press regulators. We took evidence from both IPSO and IMPRESS.
200.IPSO is a regulator for the newspaper and magazine industry in the UK. It is paid for by its members but carries out its work independently from them. IPSO provides a free-to-use complaints service regarding possible breaches of its Editors’ Code of Practice. Adjudications are made by IPSO’s Complaints Committee, a panel of twelve with expertise in journalism and a lay majority. If a complaint is upheld, the Committee can require publications to publish a correction or its adjudication, of which it can determine the nature and placement. IPSO told us that it also monitors complaints for thematic issues (such as the misrepresentation of statistics) and works with publishers to improve their compliance with the Code.187
201.IPSO’s Editors’ Code of Practice sets out the rules that newspapers and magazines regulated by IPSO have agreed to follow. The Code’s Clause 1 has the most relevance for the reporting of polling, as it sets out requirements for journalists to ensure the accuracy of their reporting and to correct inaccurate information promptly with appropriate prominence. This includes considerations such as whether the data are presented in the proper context or whether basic information about methodology has been published.188
202.IMPRESS is an independent press regulator, established by press reform campaigners, that is recognised by the Press Recognition Panel (the body established by Royal Charter as a result of the Leveson Inquiry into press standards). Up until 24 July 2017, IMPRESS used the Editors’ Code of Practice, before switching to its own Standards Code. The Standards Code has a clause on accuracy and IMPRESS can accept complaints up to 12 months from the date of the publication or act complained of. If a complaint is upheld, IMPRESS can impose fines of up to 1% of a publisher’s annual turnover, up to a maximum of £1 million; order corrections; and order apologies.189
203.We recognise that press regulation is still a controversial topic and that many people believe that the current system of press self-regulation is not strong enough. Both regulators operate via a complaint-driven approach. Matt Tee, Chief Executive of IPSO, told us that it did “not receive a lot of complaints about the coverage of political polling.”190 Similarly Jonathan Heawood, Chief Executive Officer of IMPRESS, told us that they had not received any complaints relating to polling.191 We did not find this reassuring as, given the complexity of polling methodology, members of the public may not be sufficiently well informed to question whether polling findings, data and methodological details are reported accurately. However, as we noted in paragraphs 133 and 134, both IPSO and IMPRESS were supportive of the idea of further training for journalists on the reporting of polling, which could help to improve broader understanding of what polling findings mean and how they should be presented accurately.
Broadcaster regulator
204.In the UK, Ofcom’s Code requires broadcasters to refrain from publishing the results of any polls on election day itself, until the voting period for the election or referendum closes. It also requires that broadcasters preserve impartiality in their coverage of elections.192
Digital media
205.When it comes to digital media, the Rt Hon Matt Hancock MP, then the Minister for Digital, said:
“Whereas, historically, we had a highly regulated broadcast sector and a self-regulated press, we now have a highly regulated broadcast sector and, essentially, a self-regulated press, with a small part under Impress with the royal charter, but mostly self-regulated through IPSO, and then huge, vibrant and largely unregulated social media.
The analysis of that is difficult, because there is a mix between the three. You have only to follow the BBC news on Twitter to be engaging in two. If you watch a video on the Telegraph feed on Facebook, you are looking across all three. The separate forms of regulation in the two that have a regulatory structure have grown up separately.”193
The overall landscape
206.Together with the BPC and the MRS, the press and broadcaster regulators make up the broad supervisory framework, such as it is, that currently governs polling.
207.During the course of the inquiry, the Government made it clear that it did not see the regulation of the polling industry as an issue for the state. Chris Skidmore MP, then Minister for the Constitution, told us that “polling methods and their impact on accuracy is a technical area, which is primarily a matter of debate for academics rather than Government.” He added: “We have no plans for intervention in the private polling industry and no current view, for example, on questions of minimum standards required to operate in the polling industry.”194
208.Matt Hancock felt that it was not for Government to interfere in media reporting either. In his view, “saying that newspapers make something of an opinion poll and that newspapers have a political view is a perfectly reasonable complaint, but if you ask what the government action is it is hard to see it. I believe in a free press, and politics is robust.”195
209.One of the important considerations for this Committee was whether more formal—in particular statutory—regulation would offer a more appropriate system of governance. To help answer this question, we looked at the example of France, which does have a system of statutory regulation for polling.
International comparison: France—Commission des Sondages
210.In France the Commission des Sondages is an independent body tasked with reviewing published opinion polls to ensure that companies use a reputable methodology and that the publication of results conforms to existing regulations.196 It is a statutory body funded from State funds, through the French Ministry of Justice. The Commission des Sondages comprises members of the Conseil d’État (Council of State), the Cour de cassation (Court of Cassation) and the Cour des comptes (Court of Auditors), overseen by a president, who is elected by other members of the Commission.197
211.The Commission des Sondages said that its main function is “to ensure that polls on the electoral debate which are made public are not tainted by any methodological error or manipulation which may affect the fairness of the election to which they relate.” The Commission does not set the methodological approaches that polling organisations should take, rather: “The Commission merely satisfies itself that those methods are not inherently biased, and that the samples are sufficiently numerous and representative. It also satisfies itself as to the traceability of the methods used.”198
212.The Commission des Sondages told us that its main power is to issue notices—known as mises au point (clarifications)—in the press when it feels that a poll does not meet an acceptable standard. The Commission said that: “Sometimes, a mise au point might lead the concerned polling organisation, or the director of the study, to leave the poll sector. It might also lead the media to terminate contracts with targeted polling organisations.” The Commission told us that it had issued 7 mises au point in the 2012 presidential election but none in 2017.199
213.Despite the fact that the Commission des Sondages is a statutory body, it was not clear that this meant it necessarily provided a stricter system of oversight than exists in the UK. Professor Nicolas Sauger, Professor of Political Science at the Sciences Po in Paris, expressed a degree of scepticism about the Commission and suggested that the Commission’s approach was not very “rigid” and “less effective than we might wish in monitoring surveys in France.”200
214.Professor Sauger informed the Committee:
“There is no survey specialist within the commission itself. In fact, most members of the commission are either professors of law or members of the highest court of justice in France, the Conseil d’État. They are not specialists in surveys. They are specialists in two major things. One is conflict of interest, so that regulations about how to commission a survey are clear enough. For instance, we cannot use surveys commissioned by private persons from a party as public results, if that was not the intention in the first place. The second is political appraisal of whether it is good or bad to have this kind of practice. The commission then gets experts—mostly from the public statistics institute, the INSEE—to provide reports on any cases that are seen as potentially problematic.”201
215.Although the Commission des Sondages was a useful example for our inquiry, it did not convince us that statutory regulation based on this model offered any distinct advantages for the UK. It appeared that its powers were not extensive in practice.
The case for action on regulation
216.The concerns outlined in the preceding chapters—the incidence of inaccurate polling, the poor reporting of polling findings and the potential impact of this on the democratic process—mean that different regulatory arrangements should be considered.
217.In terms of the general oversight of polling in its totality, the evidence we received was consistent in its rejection of introducing statutory regulation. There were a variety of reasons for this, including the need to protect freedom of speech and the desirability of ensuring that polling information is available as freely and transparently as other types of information.
218.It was also argued by a number of professional witnesses that there is a diversity of methodological approaches within the industry and that statutory regulation of polling would be likely to stifle methodological experimentation and innovation.202
219.Respondents were generally supportive of the system of self-regulation provided by the BPC,203 with Dr Lauderdale arguing that the “UK polling industry is more transparent regarding methodology, and methodology changes, than the industry elsewhere, precisely because of the BPC’s disclosure rules.”204
220.Given the evidence received, our broad observations are that:
- The present framework for the oversight of polling is characterised by voluntary membership and adherence to voluntary codes of practice, reliant on a complaint-driven system. This means it has limited disciplinary strength.
- Experts, academics and the industry itself have told us that conventional polling is getting harder to do and there is a growing mistrust of polling amongst the public and the media.
- It is clear that polls continue to influence debates within political parties, within the media and amongst the general public.
- Greater transparency is required regarding the sources of funding which lie behind polls.
The Committee is therefore convinced of the need for further action in order to ensure that polls of all types (not just voting intention polls) are monitored more effectively.
221.We had concerns about the financing of polls. The BPC demands a considerable level of transparency on the methodological approaches, and also requires reports of poll findings to include details of the client who commissioned the poll. However, this requirement does not stretch to full transparency on all the sources of finance which might lie behind the client who acts as the commissioner of a poll. Furthermore, these rules do not apply when polls are conducted by non-BPC members, such as those conducted by unregulated online polling companies.
222.We recognise that this is not a problem which is limited to the polling industry. In fact, the challenge of identifying who pays for what, particularly with regard to political advertising, is one of the big issues facing the digital world at the moment (an issue which we touch upon in Chapter 6). This is not an issue which we have been able to consider in depth and more attention should be given to these matters. Nevertheless, as polls form an important part of the political information that is available to the public over the course of an election campaign, we consider it very important to ensure there is as much transparency as possible about all financial transactions which lie behind published polls.
223.We therefore considered whether the Electoral Commission could extend its remit to play a specific role in overseeing voting intention polls during campaign periods.
The Electoral Commission
224.The Electoral Commission was established by the Political Parties, Elections and Referendums Act 2000. Its objectives are to ensure:
- well-run elections, referendums and electoral registration, and
- transparency in party and election finance, with high levels of compliance.205
225.The Political Parties, Elections and Referendums Act 2000 sets out the rules which apply to candidates, political parties and non-party campaigners, including rules on how much can be spent on campaigning during certain election campaigns and rules which provide transparency about the sources of funding and spending by campaigners at elections. As part of its role as a regulator of political party finances, the Electoral Commission provides advice and guidance for political parties and other groups and individuals on how to comply with these rules within the ‘regulated period’.206
226.As we noted in Chapter 2, with respect to polling, spending on polling is only regulated under electoral law if it is undertaken and used for the purpose of promoting electoral success for a political party, parties or candidates, or for promoting a referendum outcome. If it falls within this category, registered campaigners must include the spending details for such polls as part of their spending returns submitted to the Electoral Commission.
227.We asked the Electoral Commission whether it might be able to take on a greater regulatory role relating to polling. Claire Bassett, Chief Executive of the Electoral Commission, said:
“There are two things. First, we have general powers to make recommendations and to look forward, but we also have very specific regulatory powers. Those specific powers are focused almost entirely on campaigning, parties and money. That is where our expertise sits. Our infrastructure is set up to deliver that. The regulation of polling would require quite a different set-up. Although it would be practically feasible, it would be about creating a new bit within the Electoral Commission to do that, rather than building on the particular skills that we have.
I touched on the other point earlier. We are very wary of getting involved in the content of campaigning, or getting anywhere near being a truth commission. That is not what Parliament intended us to be. It would make our regulatory activity in relation to parties very difficult if we were also opining on the content of their campaigning; it would create a conflict. That would be the biggest risk. That would need to be taken seriously into consideration if Parliament was thinking of this.”207
A co-ordinated approach
228.Given these considerations, we believe that there needs to be a co-ordinated approach to the oversight of polling, which takes into account the variety of challenges it poses and the number of stakeholders it affects.
229.First, we recognise the high levels of respect within the industry for the work of the BPC. We note that self-regulatory models, in which fee-based membership societies are also the bodies overseeing the industry, can sometimes be accused of having a conflict of interest. However, we have seen no evidence of this in the case of the BPC. We are of the view that the BPC’s disciplinary and oversight powers are too limited and, in the light of the serious concerns we have about the oversight of polling, we consider that its remit should be expanded to take on a greater oversight role. Together, the BPC and the MRS should continue to be the bodies which are responsible for setting the standards for the polling industry and for monitoring their compliance.
230.Secondly, media regulators and self-regulators need to step up to the plate by taking a stronger approach towards monitoring the reporting of polling, assisted by the BPC and the MRS. When the BPC and MRS identify instances of bad reporting of polling, IPSO, IMPRESS and Ofcom need to ensure that these are investigated and dealt with quickly through their existing regulatory systems.
231.Thirdly, we believe that steps need to be taken to ensure that the sources of funding for polls conducted in the run-up to elections are made known publicly as transparently as possible. We recognise that there are commercial sensitivities to consider and do not, therefore, suggest that the actual costs of polls need always be disclosed. We also recognise the Electoral Commission’s reservations about becoming an arbiter of truth. However, we believe that the Electoral Commission should play a greater role in overseeing the publication of voting intention polls during the ‘regulated period’ in the run-up to elections. This should involve a requirement that all published voting intention polls be declared to the Electoral Commission, which should then publish the sources of funding for such polls.
232.We therefore recommend the following steps which we believe should lead to a strengthened, more co-ordinated approach towards oversight of the polling industry.
233.The different bodies involved in the oversight of polling need to respond to the challenges involved in the polling of the modern electorate, and to the misreporting and misrepresentation of polls. There are limitations in the current system of self-regulation for polling, and clear areas where the system could be strengthened.
234.We recommend that the remit of the British Polling Council should be expanded to take on a more substantial oversight function. The British Polling Council should adopt a collaborative approach, involving both industry and independent representation. In taking on this expanded role, it will clearly need to work closely with other regulatory stakeholders in this area, including the MRS, IPSO, IMPRESS, the Society of Editors and Ofcom. Some of the functions of the newly-expanded British Polling Council should include:
- Issuing guidance on best practice for the methodologies used in polling.
- Providing an advisory service for reviewing poll design. This would be a service intended to give companies the assurance that their questions and survey design had been evaluated independently, which could provide a degree of cover when dealing with sensitive or controversial issues.
- Ensuring that its members declare not just the client commissioning the poll, but the full details of all sources of funding for polling (excluding actual costs).
- Conducting a post-election review of the conduct of the polling industry after every General Election and referendum, and publishing its findings.
- Co-ordinating a programme of training opportunities for journalists on how to read, interpret and report on polling data.
- Developing its guidance for journalists on best practice for the reporting of political polls.
- Providing specific advice to the media on how to report on particular polls. This advice could be made public in certain cases.
- Proactively reviewing selected samples of media coverage of polls on an annual basis, in order to monitor standards of media reporting. This should include analysis of print media, broadcaster coverage and digital media, and analysis of polls in general, not just those conducted by its members.
- Continuing to run an effective complaints procedure, with speedy investigations of complaints.
235.It is often difficult for members of the public to recognise when polling results have been taken out of context or misreported, so it is not enough to simply wait for complaints to be made. We therefore recommend that the BPC and MRS should identify and report instances of bad reporting of polling (whether or not the polls have been conducted by their members) and draw them to the attention of IPSO, IMPRESS or Ofcom as appropriate. Such cases should then be investigated and dealt with quickly through the existing regulatory systems for the media.
236.In order to ensure transparency around voting intention polling in the run-up to elections, we also recommend that the Electoral Commission should take on an enhanced role in monitoring voting intention polling conducted and published during the regulated periods which precede UK elections. In particular, there should be a requirement for the details of all published voting intention polls to be declared to the Electoral Commission, regardless of who the poll was commissioned by, what its purpose was, or how much it cost. The details of all the sources of funding for such polls should then be published by the Electoral Commission, although we recognise that, in order to protect commercial confidentiality, the actual sums of money involved need not be made public. We acknowledge that this will require an extension of the Electoral Commission’s existing remit and recommend that action is taken to achieve this.
237.We realise that expanding the Electoral Commission’s role in this way would extend their involvement in the electoral process, but believe that this would be a proportionate response to the issues we have identified above. We do not envisage its role competing with that of the BPC. Instead, we hope that all the relevant bodies would be able to work together and take co-ordinated action to ensure that polls are used to benefit the democratic process, rather than to undermine it. For example, if the BPC identified a voting intention poll which had not been declared to the Electoral Commission, it should proactively flag this up to the Electoral Commission for investigation. Likewise, if the Electoral Commission identified a poll which it suspected of being conducted in a misleading way, it could flag it up to the British Polling Council for detailed examination.
238.Furthermore, we recognise that expanding the role and remit of the British Polling Council will require significant additional resources and funding. Part of this additional funding may need to be generated through increased subscription costs for its members. However, we hope that by expanding its remit, the British Polling Council will become the one, clearly identifiable body that can be trusted to ensure standards of accuracy in relation to polls. Any media sources which report on polls would be able to check the accuracy of its sources with the British Polling Council. In turn, the BPC and the MRS could highlight particular concerns about particular polls to either the media regulators, or the Electoral Commission, as appropriate.
239.If the co-ordinated approach we outline above does not prove to be successful, it is likely that politicians and others will return to this issue in order to reassess whether statutory regulation is required instead.
Banning of polling
240.We also considered whether the publication of polls should be banned for specified periods in the run-up to election days.
241.Many countries restrict the publication of the results of polls during the period leading up to an election. Within the EU, 16 out of 28 countries have a ban on the publication of opinion polls, ranging from one day before the election (such as in France) to 15 days (Italy) to one month prior to voting (Luxembourg).208
242.It would also be possible to ban the carrying out of opinion polls in the run-up to elections, rather than just their publication.
243.There were some arguments for the introduction of a ban on polls. The main reason given was that it would allow voters the space to make their decision without undue influence from polls. Professor Sauger suggested that, in the absence of polls: “The assumption is that people have to think about the election, without any interference from anyone.”209 Restrictions on the publication of polls could also limit the focus on the ‘horse race’ and enable a greater emphasis on discussion of policy issues.210 Finally, a clear advantage of restricting the publication of polls would be that it would limit the risk of incorrect poll estimates affecting voters’ choices in elections.
244.The banning of polling is, nonetheless, a controversial approach. In many of the countries that have implemented a ban on polling, legal challenges in recent years have reduced the time period over which the ban applies.211 Bans have been reduced due to concerns around protecting freedom of expression, and specifically the electorate’s right to receive and communicate information.212 For example, in Canada, a Supreme Court decision reduced the previous 72 hour ban on the publication of opinion survey results prior to elections, to 24 hours. The Court claimed that the 72 hour ban violated freedom of expression as protected by the Canadian Charter of Rights and Freedoms, and could not be justified as necessary to protect the integrity of the electoral process.213
245.It is sometimes assumed, by proponents of a ban, that a ban on polling would mean that voters would make electoral choices on the basis of normatively superior criteria, such as policy issues and the quality of candidates. However, it is also maintained that voters will still rely on expectations about the likely outcome of the election and that, in the absence of polls, the accuracy of their expectations will be even worse.
246.That a ban on polling would be a threat to freedom of expression was, in fact, the most prevalent argument we heard against the banning of polling. Several witnesses suggested that banning polls or their publication would be undemocratic and that it would restrict useful information for the public, or risk creating inequalities in access to information. The difficulties of social media were also frequently cited, with many witnesses pointing out that the advance of social and other online media would render entirely nugatory any attempt to ban the publication of polls.
247.Critics cited concerns that the banning of polls might create a black market for polls conducted by actors outside the jurisdiction of the country in question. The World Association for Public Opinion Research claimed that a ban in the UK could create a black market with polls being conducted outside the UK and “no guarantee of the polling companies’ competence or even the provenance of the poll.”214 There is evidence that this does happen. In France it is apparently common for francophone Swiss and Belgian media with websites available in France to start reporting the results of election day exit polls well before the ban is lifted.215 In the most recent presidential elections in France, “Polls show French backing Macron”, was reported by the Associated Press, at a time when millions had yet to cast their votes. This was based on multiple polls cited by the Swiss newspaper La Tribune de Geneve and Belgium’s RTBF and Le Soir.216 A ban on polls in the UK might result in the publication of polls on the websites of Irish or French newspapers and their results would soon be disseminated via social media and other channels.
248.We are not convinced of the case for introducing a ban on the undertaking and publication of voting intention polls in the run-up to elections. In the future, if polls continue to be a poor predictor of the eventual outcomes of elections, and if the media reporting of such polls continues to influence public and political discourse in a misleading way, then arguments by supporters of a ban would be strengthened.
169 NatCen, ‘Four in five trust official statistics, but not how they are presented by Government’ (27 February 2017): http://www.natcen.ac.uk/news-media/press-releases/2017/february/four-in-five-trust-official-statistics,-but-not-how-they-are-presented-by-government/ [accessed 20 March 2018]
170 Representation of the People Act 1983, section 66A
171 British Polling Council, ‘About the BPC’: http://www.britishpollingcouncil.org/ [accessed 20 March 2018]
173 National Audit Office, Using alternatives to regulation to achieve policy objectives (June 2014) p 15: https://www.nao.org.uk/wp-content/uploads/2014/06/Using-alternatives-to-regulation-to-achieve-policy-objectives1.pdf [accessed 20 March 2018]
175 Ibid.
176 Ibid.
177 Ibid.
181 Ibid.
182 Market Research Society, MRS Disciplinary Regulations 2017 (June 2017): https://www.mrs.org.uk/pdf/MRS%20Disciplinary%20Regulations%2014.06.2017.pdf [accessed 20 March 2018]
185 Ibid.
188 Ibid.
189 IMPRESS, ‘Complaints FAQ’: http://impress.press/complaints/complaints-faq.html [accessed 20 March 2018]
192 Ofcom, The Ofcom Broadcasting Code, Section six: Elections and referendums (3 April 2017): https://www.ofcom.org.uk/tv-radio-and-on-demand/broadcast-codes/broadcast-code/section-six-elections-referendums [accessed 20 March 2018]
196 Organization for Security and Co-operation in Europe, Office for Democratic Institutions and Human Rights, France, Presidential and Parliamentary Elections 2017 (11 April 2017): http://www.osce.org/odihr/elections/france/337346?download=true [accessed 20 March 2018]
198 Ibid.
199 Ibid.
201 Ibid.
202 Written evidence from ComRes, Opinium, Ipsos MORI, Panelbase, LucidTalk, ORB International, BMG Research and Survation (PPD0014)
203 Q 36 (Professor Jane Green); Q 47 (Will Moy); Written evidence from Professor Alisa Henderson (PPD0012)
205 The Electoral Commission, ‘Roles and responsibilities’: https://www.electoralcommission.org.uk/our-work/roles-and-responsibilities
206 The Electoral Commission, ‘Our role as regulator of political party finances’: https://www.electoralcommission.org.uk/our-work/roles-and-responsibilities/our-role-as-regulator-of-political-party-finances
208 Information taken from Houses of the Oireachtas Library and Research Service note No. 1 2009, Spotlight: Political Opinion Polls (2009), p 16: https://oireachtas.ie/parliament/media/housesoftheoireachtas/libraryresearch/Polling_web.pdf [accessed 20 March 2018]
210 David Cowling, ‘Should polling be banned before and election?’, BBC News (20 January 2016): http://www.bbc.co.uk/news/uk-politics-35350419 [accessed 20 March 2018]
211 The Electoral Knowledge Network, Ace Project, ‘Media and Elections’: https://aceproject.org/ace-en/topics/me/mea/mec03c [accessed 20 March 2018]
212 Article 19, ‘Comparative Study of Law and Regulations Restricting the Publication of Electoral Opinion Polls’, January 2003: https://www.article19.org/data/files/pdfs/publications/opinion-polls-paper.pdf [accessed 20 March 2018]
213 Ibid.
215 ‘Belgian, Swiss media circulate report Macron leads French vote’, Reuters (23 April 2017): https://www.reuters.com/article/us-france-election-belgium/belgian-swiss-media-circulate-reports-macron-leads-french-vote-idUSKBN17P0O5 [accessed 20 March 2018]
216 Caoilfhionn Gallagher QC and Jonathan Price, ‘Breaking the election silence: cross-border reporting of election day polls’, The International Forum for Responsible Media Blog (18 May 2017): https://inforrm.org/2017/05/18/breaking-the-election-silence-cross-border-reporting-of-election-day-polls-caoilfhionn-gallagher-qc-and-jonathan-price/ [accessed 20 March 2018]