Written evidence from The Royal Statistical Society (EUR 106)
Public Administration and Constitutional Affairs Committee
Lessons Learned from the EU Referendum inquiry
Summary response
The Royal Statistical Society (RSS) is very concerned about the extent to which official statistics were used by politicians and campaign managers in unadvised and misleading ways during the referendum campaign period. Our response to this inquiry focuses on arrangements that could be made to ensure prominent statistical claims are accurate and are addressed with appropriate advice in future. This includes some consideration of recent recommendations of the Electoral Reform Society (ERS).
Full response and supporting evidence (organised under inquiry questions)
1.1. There are aspects of the conduct of campaigns that The Political Parties, Elections and Referendums Act 2000 does not cover, and which are addressed under the purview of other regulations and regulators. The House of Commons’ Treasury Select Committee noted in May that ‘the public debate is being poorly served by inconsistent, unqualified and, in some cases, misleading claims and counter-claims’ [6]. We are very concerned about ways that official statistics were misused by politicians during the referendum and believe that a strengthened regulatory approach is needed in this area.
1.2. There has been a call from a number of quarters, including from the Electoral Reform Society, for an official body to be empowered to intervene over misleading information. The UK Statistics Authority exists to regulate statistics and should be part of any new arrangement, as one of its roles is to uphold good practice in the use of official statistics by politicians and by officials. The UKSA retained this remit throughout the duration of the referendum campaigns and corresponded with regard to statistical claims [4]. The UKSA also promotes the Code of Practice for Official Statistics [7], which ministers, MPs and other public officials are guided by their Ministerial Code to follow at all times. The ordinary provisions for independent oversight of official statistics were therefore in force. However there was no extraordinary provision made with regard to the referendum.
1.3. Factual information for the public was fractured and uncoordinated. As regulator, the UK Statistics Authority made some interventions regarding campaigns’ presentation of official statistics, but more timeliness and impact was required. To address this we suggest that the UK Statistics Authority, and the Office for National Statistics, need to work with other organisations such as the House of Commons Library and Full Fact to produce a set of statistical fact sheets before referendum campaigns begin, on all the topics of public interest that will be addressed by campaigners. The possibility of a minimum agreed dataset, as recommended by the Electoral Reform Society [1], should be explored. More coordinated factual resources could not only inform the public, but could also form an effective source of guidance for the referendum campaigns, by making clear in advance of the debate what the appropriate figures are to address key campaign topics and questions.
1.4. To uphold normal standards of statistical good practice by campaigners during an exceptionally high-stakes campaigning period, it is apparent that ordinary regulatory action was not sufficient. The UK Statistics Authority should take part in forming new arrangements to strengthen the provision and assurance of key public information for referendum campaigns. A stronger commitment for officially designated campaigns to sign up to in exchange for their access to public funds, such as an enforceable code of conduct, could be helpful as a mechanism for assurance.
2. What role did the machinery of Government play during the referendum campaign and what were the consequences for Civil Service impartiality?
2.1. A period of ‘purdah’ preceded the referendum vote [3], which raises issues for impartial assessment of statistical claims. The UK Statistics Authority needs to be able to continue in its regulatory function, to address misleading uses of official statistics during purdah. The ONS obtains exemptions during such periods for publication of pre-announced statistics, and the UK Statistics Authority continued to operate its public regulatory function. We are pleased that the UKSA was able to act in a regulatory capacity. However it is important that the position of UKSA and the ONS is made clear for such periods and that they are not constrained from meeting their statutory obligations, including (for UKSA) the duty to ensure good practice with regard to official statistics.
2.2. The Royal Statistical Society is opposed to the government’s current policy of pre-release access to statistics whereby government ministers and their officials are able to view figures 24 hours before they are made available to Parliament or to the general public [5]. In the interests of impartiality in referendum campaigns, it is regrettable that pre-release access continued, as this was perceived as giving an advantage to the campaign that was supported by the government, who had pre-release access to statistics. Pre-release access should be ended, with the exception of market-sensitive figures which should be accessible a maximum of two hours in advance.
September 2016
3. References
[1] P. 10 in Electoral Reform Society (2016) It’s good to talk: doing referendums differently after the EU vote (PDF) Available from: https://www.electoral-reform.org.uk/sites/default/files/files/publication/Its-good-to-talk-2016-EU-Referendum-Report.pdf
[2] The UK Statistics Authority was established in 2008 as an independent body operating at arm’s length from government, with a statutory role to promote and safeguard official statistics including upholding good practice in relation to official statistics. It is a non-ministerial department which is accountable to Parliament. https://www.statisticsauthority.gov.uk/about-the-authority/
[3] EU referendum: Guidance for UK Government Departments on Activities During the Period 27 May to 23 June (PDF). Available from: https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/526451/eu-referendum-guidance-for-uk-government-departments-on-activities-during-the-period-27-may-to-23-june.pdf
[4] Eg https://www.statisticsauthority.gov.uk/wp-content/uploads/2016/05/Letter-from-Sir-Andrew-Dilnot-to-Jack-Doyle.pdf and https://www.statisticsauthority.gov.uk/wp-content/uploads/2016/05/Letter-from-Sir-Andrew-Dilnot-to-Dominic-Cummings.pdf. Further examples of correspondence from the UKSA chairman in 2016 can be found at https://www.statisticsauthority.gov.uk/correspondence-list/?keyword=Dilnot&date=2016
[5] Royal Statistical Society (2015) Public trust and pre-release access of statistics (PDF), available from: http://www.rss.org.uk/Images/PDF/influencing-change/2016/manifest-briefing-note4-trust-pre-release-access-of-statistics-11-May-2015.pdf
[6] P. 4, House of Commons Treasury Select Committee (2016) The economic and financial costs and benefits of the UK’s EU membership (PDF), available from http://www.publications.parliament.uk/pa/cm201617/cmselect/cmtreasy/122/122.pdf
[7] UK Statistics Authority (2009) Code of Practice for Official Statistics: Edition 1.0 (PDF). London: UK Statistics Authority, available from http://www.statisticsauthority.gov.uk/assessment/code-of-practice/code-of-practice-for-official-statistics.pdf
2