Monitoring the Performance of the Department for Work and Pensions in 2012-13 - Work and Pensions Committee Contents

7  Use of DWP statistics

115. We have commented previously in this Parliament on the negative way in which DWP benefit statistics are sometimes covered in press stories, and on the role which DWP's processes for releasing statistics might play in this.

116. We became concerned during spring 2013 about the way in which a number of releases of DWP benefit statistics had been reported in the media. The UK Statistics Authority investigated two of these cases and highlighted changes it wished to see both in DWP practice, and more widely across Government, in the way benefit statistics are released to the media. We decided to take oral evidence from DWP officials and Ministers to discuss our concerns with them.[118]

Previous comment by the Committee on DWP release of statistics

117. In our 2011 report on the migration of incapacity benefits to ESA, we concluded that:

    [...] more care is needed in the way the Government engages with the media and in particular the way in which it releases and provides its commentary on official statistics [...] In the end, the media will choose its own angle, but the Government should take great care with the language it itself uses and take all possible steps to ensure that context is provided when information [...] is released, so that unhelpful and inaccurate stories can be shown to have no basis.[119]

In response, the Government emphasised that it "takes great care" when publishing statistics to ensure that the information is used in an appropriate manner. However, it also stated that:

    The Committee and Government need to be mindful of the widespread public unease about the number of people claiming incapacity benefits and it is therefore unsurprising that this is reflected in the media. [...] However, it is important to stress that it is not the Department's role to dictate what can appear in stories in the media.[120]

118. In our 2012 report on reform of DLA, we referred back to our 2011 comments on media coverage of benefit claimants and the Government's response and said:

    The Government's view seems to be that the negative tone of press coverage of benefit claimants is unsurprising since it merely reflects the public mood about the integrity of the benefits system. However, the Government should not ignore the fact that public opinion can also be positively influenced by the media and we believe it should take the necessary steps to ensure that its own contribution to media stories about benefits is accurate and contextualised.

We also highlighted that "direct quotations from Ministers can give undue credence to inaccurate or misleading reports" and recommended that DWP "ensure that significant statistical releases are accompanied by a press release setting out the context and providing background explanatory notes".[121] The Government's view was that it already had a "robust process for releasing new statistics into the public domain which conform with the Code of Practice for Official Statistics". [122]

DWP processes for approving press releases and public statements accompanying statistical releases

119. In advance of our oral evidence session in July 2013 with DWP officials, the Department provided us with more details about the processes it has in place for releasing statistics. Press releases are cleared by the relevant policy official before release. Where they contain statistics or data, they are also cleared by "the relevant analyst". Statistics are sometimes released to back up information in press releases rather than the other way round: "Whenever these statistics are not already in the public domain, an ad hoc statistics release will be published alongside any departmental use".

120. Press officers and other departmental officials are given "pre-release access" to national and "high profile" official statistics. This allows them to prepare the Department's public response, which may include quotes from Ministers or the Department, 24 hours before release. These responses are also cleared by the relevant analyst and policy officials before release.

121. DWP states that "press officers are aware of their responsibility to abide by the Code of Practice for Official Statistics". It highlights that the UK Statistics Authority and DWP's Deputy Head of Profession for Statistics have met DWP press officers "to educate them on the requirements of the Code".[123]

UK Statistics Authority investigations into DWP statistics during 2013

122. In May 2013 the UK Statistics Authority (UKSA) reported on its investigations into two separate complaints about benefit statistics.


123. Sheila Gilmore MP (a member of this Committee) asked UKSA to investigate a Sunday Telegraph article in March 2013 which stated that "900,000 choose to come off sickness benefit ahead of tests".[124]

124. This related to DWP statistics on claims for Employment and Support Allowance (ESA), which has replaced Incapacity Benefit (IB) as the income replacement benefit for people with health conditions and disabilities which prevent them from working. People claiming ESA are required to undergo the Work Capability Assessment (WCA) to establish eligibility for the benefit. A process started in 2011 to migrate existing IB claimants to ESA over the period to 2014. The migration process involves a WCA.

125. Grant Shapps MP, the Conservative Party Chairman, was quoted as saying "nearly a million people have come off incapacity benefit [...] before going for the test." The press article was drawn from a Conservative Party press release which stated that "878,300 people claiming incapacity benefit—more than a third of the total—have chosen to drop their benefit claim entirely rather than face a medical assessment, new figures have revealed."

126. UKSA's investigation found that the 878,300 figure resulted from the conflation of official statistics relating to new ESA claims with separate statistics on migration of Incapacity Benefit claimants to ESA. The release also failed to make clear that a number of claims were withdrawn because the person recovered from their illness before the WCA took place. In his letter to Sheila Gilmore setting out the findings, the UKSA Chair, Andrew Dilnot, explained:

    The statistical release does not address the issue of why cases were closed in great depth, but it does point to research undertaken by DWP which suggests that "an important reason why ESA claims in this sample were withdrawn or closed before they were fully assessed was because the person recovered and either returned to work, or claimed a benefit more appropriate to their situation.[125]

In oral evidence on 10 July, we asked the DWP Head of Communications, John Shield, what role his staff had played in the release of this information. He said:

    This is really simple. I knew you would ask, so I have checked with the press office. In no way, shape or form was anyone involved in the production of this. They were not, and I have been assured that this is purely a piece of party output [...] no one in the press office or in communications had any role in that; it is a party matter.[126]

127. When we questioned the Secretary of State about this incident, he told us:

    That was something that they [Conservative Central Office] put together and released themselves. In fact, I was not even aware that they were going out with the comment at the time. [...] They had pulled two things together and conflated them [...] but it is not us. I have had conversations with him [the Conservative Party Chairman] and with others about being careful to check with the Department: if they are going to say something about the statistics that are out there, they should check with us that those are the statistics that can be correctly used. There has been an element of that since then, so there has been less of that going on.[127]


128. The UK Statistics Authority undertook a second investigation following a complaint from Nicola Smith, Head of Economic and Social Affairs at the TUC, about the presentation of statistics on the benefit cap.

129. The Welfare Reform Act 2012 provided for a cap on total household benefits to be phased in from April 2013. The cap limits the total benefit a household can receive to £500 per week for a family and £350 per week for a single person with no children. The cap was initially piloted in four London boroughs from April 2013; national implementation began in July and was completed by September 2013 (except in Northern Ireland).[128] (We have examined the impact of the benefit cap in detail in our inquiry into support for housing costs—because the cap is in practice applied to the level of the housing costs element of total benefits. Our report will be published shortly).

130. The Government's stated intentions for implementing the cap are to improve working incentives for those on benefits, to deliver fiscal savings, and to ensure workless households do not receive more in benefits than the average working household. [129]

131. When the benefit cap pilots began in April 2013, DWP published two sets of related ad hoc statistics: one on Households identified as potentially impacted by the benefit cap; and a second set on Jobcentre Plus activity regarding claimants who have been identified as potentially impacted by the benefit cap.[130] The TUC complaint was that DWP had misrepresented these statistical analyses in a press statement it issued to the Press Association on 12 April.

132. Press reports on the DWP statistics indicated that the estimated number of households affected by the benefit cap had fallen from 56,000 to 40,000. DWP was reported as saying that "part of the reason for the drop is due to 8,000 unemployed people finding work". The Secretary of State was quoted in the Daily Mail on 12 April as saying that "Already we've seen 8,000 people who would have been affected by the cap move into jobs. This clearly demonstrates that the cap is having the desired impact".[131] A media debate followed on whether there was a causal link between the introduction of the cap and people moving into employment.[132]

133. UKSA found that the Secretary of State's statement was "unsupported by the official statistics published by the Department".[133] It highlighted that the Jobcentre Plus Activity Statistics from which the 8,000 figure "appears to be drawn" explicitly states that the figures "are not intended to show the additional numbers entering work as a direct result of the contact [with JCP]". The other set of statistics notes that the reduction in the estimate of the number of households affected by the benefit cap from 56,000 to 40,000 was "due to normal caseload churn, reducing those potentially in scope for the cap" and that both figures "assume no behavioural change".

134. In his letter to the Secretary of State, the UKSA Chair said:

    In the manner and form published, the statistics do not comply fully with the principles of the Code of Practice, particularly in respect of accessibility to the sources of the data, information about the methodology and quality of the statistics, and the suggestion that the statistics were shared with the media in advance of their publication.

He pointed out that he had been given an assurance by DWP in March 2013, following a similar incident, that senior officials had reiterated to staff "the seriousness of their obligations under the Code of Practice". His letter asked for a further assurance "that the working arrangements within the Department give sufficient weight to the professional role and public responsibilities of statisticians".[134]

135. We questioned John Shield about the source of the Secretary of State's press comments on the impact of the benefit cap. He told us that it was:

    [...] an opinion piece given to the Daily Mail where the Secretary of State was stating his opinion on the statistics, and not only basing it on that, but basing it on what staff had been telling him about the impact of the cap, the management information that he had been receiving and what claimants had actually said to him. That was a judgment formed by him and, as a politician, obviously he can make those judgments around what he thinks the data are saying in the context of everything else.[135]

136. John Shield made clear that this was "a written article" by the Secretary of State; it was not based on an interview. He also stated that "if a Minister is doing an opinion piece that is about their reflections and views on how policy is working and performing, sometimes they will be produced without press office involvement." He said that such articles were sometimes produced by Ministers themselves or by Special Advisers. However, Mr Shield said that on this occasion, "it did involve the press office".[136]

137. In an interview on the BBC Radio 4 Today programme on 15 July, when the benefit cap began to be rolled out nationally, the Secretary of State defended his use of the 8,000 figure in relation to people likely to be affected by the cap who had sought work as a result. He said that he "believed" that it was having this effect: people on benefits who had not previously been seeking work now were, as a result of the cap. He said it was not possible to "disprove" this claim and he believed it to be right.[137] In oral evidence to us in December, the Secretary of State made a similar point:

    [UKSA] said that whilst we could not prove that they had gone in [to work] or could not prove that they had not, but I should therefore not make the link other than that I believe it to be the case that those people are going back into work is hugely to do with the fact that we introduced the cap. That is my belief, but they said that that should remain as a flat statistic, which we have accepted.[138]

Recent UKSA comment on use of DWP statistics

138. More recently, Sheila Gilmore MP raised with UKSA the use made by a senior DWP official, during one of our evidence sessions, of unpublished statistics on Work Programme support.[139] On 20 November 2013, Neil Couling, the DWP Director of Work Services, gave evidence alongside the Minister for Employment in our inquiry into Jobcentre Plus. During the session, he cited statistics on the number of people who had found work after they had completed two years on the Work Programme.[140] In response to a subsequent parliamentary question, it became clear that the statistics cited were based on unpublished Management Information.[141] (This information was subsequently published as part of the Work Programme statistical summary on 19 December 2013.[142])

139. The Chair of UKSA highlighted in his response that "the National Statistician has issued guidance on the use of Management Information". This states that in "exceptional circumstances", the departmental Head of Profession for Statistics may agree to the use of information of this kind in public statements if its use is "justified and clearly explained". DWP had reported to UKSA that its "normal practice" was for an analyst to "check and sign off the accuracy of the statistical and internal management information to be used in briefings for select committee hearings." However, no analyst was involved in checking the information provided to the Committee on 20 November. The UKSA Chair said that "it is a matter of regret" that DWP's normal practice was not followed.[143]

Ensuring accurate presentation of Government statistics

140. Our colleagues on the Public Administration Select Committee (PASC) have undertaken a major review of Government statistics. In a report published in May 2013, Communicating statistics: not just true but also fair, the Committee concluded that: "in some cases, the story behind the statistics is reduced in its presentation to such an extent that the picture is no longer true and fair".[144] PASC emphasised that "to underpin good policy-making, statistics must be presented in a fair, accurate way, 'unspun'" and suggested that Government press releases "sometimes go too far to create a newsworthy headline". The Chair of the Committee (Bernard Jenkin MP) went on to say that:

    Politicians tend to promote the statistics which best present their case. Finding the whole truth about government statistics is not always easy, and it should be. The numbers may be perfectly true but the act of selecting certain numbers distorts the true picture. This is important when those numbers are being used to justify a particular policy, a particular apportioning of resources. In some cases, spinning reduces the story behind the statistics to such an extent that the picture is no longer true.[145]

141. DWP releases a great deal of statistical information about benefits. We have commented before that it needs to exercise care in the language used in accompanying press releases and ministerial comments in the media. 2013 saw heightened and quite widespread concern—including from the UK Statistics Authority and organisations representing disabled people—about the DWP commentary accompanying releases of benefits statistics.

142. The Government is doing a great deal to promote a positive image of disabled people, including in the principles behind its Disability Strategy and the Disability Confident campaign to help disabled people into employment. However, this positive action risks being undermined if the language used in DWP press releases and ministerial media comments accompanying releases of benefit statistics adopts a tone which feeds into negative preconceptions and prejudices about people on benefits, including disabled people.

143. We agree with our colleagues on the Public Administration Select Committee (PASC) that Government statistics should be presented in a way that is fair, accurate and "unspun" and that this is especially the case when they are being used to justify a particular policy or a particular allocation of resources. We reiterate our view that DWP should avoid feeding into negative public views about people who receive benefits, and that statistics should be used objectively to shed light on policy implementation, not to prop up established views and preconceptions. We recommend that, in response to this Report, DWP sets out the specific steps it has taken in response to the comments from PASC, the UK Statistics Authority, and this Committee, to ensure that statistics are released in a way which is accurate, and fair to benefit claimants.

118   Oral evidence taken on 10 July 2013 from DWP officials, HC 570; and on 9 December 2013 from the Secretary of State, HC 867 Back

119   Sixth Report of Session 2010-2, The role of incapacity benefit reassessment in helping claimants into employment, HC 1015-I, para 41 Back

120   Seventh Special Report of Session 2010-12, Government Response to the Committee's Sixth Report, HC 1641, response to recommendations in paras 40-41 Back

121   Seventh Report of Session 2010-12, Government support towards the additional living costs of working-age disabled people, HC 1493-I, paras 53-54 Back

122   First Special Report of Session 2012-13, Government Response to the Seventh Report of Session 2010-12, HC 105, response to recommendations in paras 53-54, pp 4-5 Back

123   DWP written evidence published with oral evidence taken on 10 July 2013, HC 570 Back

124   Sunday Telegraph, 30 March 2013 Back

125   Letter from Chair of UK Statistics Authority to Sheila Gilmore MP, 29 May 2013, available on UK Statistics Authority website at: (citing DWP, Unsuccessful ESA claims - qualitative research, 2011Back

126   Oral evidence taken on 10 July 2013, HC 570, Q33 Back

127   Oral evidence taken on 9 December 2013, HC 867, Q6 Back

128   DWP, Impact Assessment- Benefit Cap (Housing Benefit) Regulations 2012: Impact assessment for the benefit cap, July 2012 Back

129   DWP, Impact Assessment- Benefit Cap (Housing Benefit) Regulations 2012: Impact assessment for the benefit cap, July 2012 Back

130   DWP, Ad hoc statistics on Households identified as potentially impacted by the benefit cap, April 2013. The April 2013 document Jobcentre Plus activity regarding claimants who have been identified as potentially impacted by the benefit cap, no longer appears to be available on the DWP website; it has been replaced by an updated version published in December 2013. Back

131   Daily Mail, 12 April 2013 "One in four facing a cut in their benefits has found work" Back

132   See for example, The Guardian, 13 April 2013, "No evidence for Iain Duncan Smith benefit cap claim, says research chief" Back

133   Letter from Andrew Dilnot to Nicola Smith, 9 May 2013, available on UK Statistics Authority website at:  Back

134   Letter from Andrew Dilnot to the Secretary of State for Work and Pensions, 9 May 2013, available on UK Statistics Authority website at: Back

135   Oral evidence taken on 10 July 2013, HC 570, Q25 Back

136   Oral evidence taken on 10 July 2013, HC 570, Qq28 and 32 Back

137   BBC Radio 4 Today Programme, 15 July 2013 Back

138   Oral evidence taken on 9 December 2013, HC 867, Q4 Back

139   Letter from Sheila Gilmore MP to the Chair of the UK Statistics Authority, 21 January 2014, available on UK Statistics Authority website at: Back

140   Oral evidence taken in the inquiry into the role of Jobcentre Plus in the reformed welfare system on 20 November 2013, Q559  Back

141   HC Deb, 9 December 2013, col 52w Back

142   DWP, Work Programme Statistical Summary, December 2013 Back

143   Letter from the Chair of the UK Statistics Authority to Sheila Gilmore MP, 21 February 2014, available on UK Statistics Authority website at: Back

144   First Report from the Public Administration Select Committee, Session 2013-14, Communicating statistics: not just true but also fair, HC 190, Summary Back

145   PASC press release, 29 May 2013, "PASC demands that Government stats are presented with "the whole truth" Back

previous page contents next page

© Parliamentary copyright 2014
Prepared 18 March 2014