Ed Humpherson to Peter Schofield: Universal Credit management information

Dear Peter

UNIVERSAL CREDIT MANAGEMENT INFORMATION

Thank you for your response to our letter dated 22 April, which sets out the future publication of information on Universal Credit. Our understanding is that you will be publishing a weekly management information series summarising declarations made by individuals and households to Universal Credit and information on advance payments of Universal Credit. Data on geographical breakdowns and other characteristics will continue to be published as part of the monthly Universal Credit Official Statistics.

It is reassuring that you recognise the importance of not using unpublished information in public statements and we hope that the weekly series mitigates the risk of this issue reoccurring. We are aware that the timing of the work and pensions oral session is not aligned with the release of weekly management information and therefore Ministers may occasionally wish to refer to additional data. In these cases, you should look to ensure this additional data is made publicly available to support the statement being made.

We will continue to monitor the evolving interest in data on Universal Credit and work with your department to ensure its statistics remain aligned to the Code of Practice for Statistics

I am copying this letter to Steve Ellerd-Elliott, Chief Statistician for the Department for Work and Pensions.

Yours sincerely

Ed Humpherson

Director General for Regulation

 

Related links

Ed Humpherson to Steve Ellerd-Elliot: Universal Credit management information

Response from Peter Schofield to Ed Humpherson: Universal Credit management information

Response from Peter Schofield to Ed Humpherson: Universal Credit management information

Dear Ed,

UNIVERSAL CREDIT MANAGEMENT INFORMATION

Thank you for your letter of 22 April. I was pleased to read your positive comments about the supplementary information we have published alongside our Official Statistics but I equally recognise your concerns around the use of unpublished information and that the release did not include all of the information that we had preannounced.

I take the matter of handling and using statistics very seriously and I am sorry that on this occasion there were issues in the way in which we have handled this release. This is an exceptional release which sees us publish more timely information than we ever have before on Universal Credit to meet the needs of users and our duty of equality of access at this challenging time.

The release was developed at speed and, whilst we were working with the best intentions, it was a mistake that the pre-announcement was made whilst internal discussions were ongoing and before agreement had been reached on what additional data we would be able to release. I recognise that on this occasion it was not accurate on the information that would be published.

Our Chief Statistician, Steve Ellerd-Elliott, has reviewed what happened and is updating our guidance and processes to ensure appropriate approval and sign-off is in place for this type of release in future.

We will continue to provide the full Universal Credit breakdown in our monthly Official Statistics but during this unprecedented time when Universal Credit claims are at an exceptionally high level we also want to provide more timely information. I am mindful of the need to do this in a way which ensures that any public statements are supported by information which is equally available to all.

The management information we have published around claims for Universal Credit and advances are matters that have been raised by the Work and Pensions Committee and that have been used in the public updates we have given on the Department’s performance. We will now be increasing the frequency to publish this on a weekly basis. The first release will be published on 5 May and then weekly until at least the end of June.

We will continue to review the information available on Universal Credit and other benefits, and aim to provide informative statistics that are trustworthy, high quality and represent good value.

I welcome your organisation’s continuing positive work with the Department and I am happy to discuss ways in which the Department can continue to improve its release of statistics.

I have copied this letter to Steve Ellerd-Elliott, DWP Chief Statistician and will also send a copy to the WPSC.

Yours sincerely,
Peter Schofield CB

 

Related links:

Ed Humpherson to Steve Ellerd-Elliot: Universal Credit management information

Ed Humpherson to Peter Schofield: Universal Credit management information

Ed Humpherson to Steve Ellerd-Elliott: Universal Credit management information

Dear Steve

UNIVERSAL CREDIT MANAGEMENT INFORMATION

 I am writing to welcome the publication of supplementary management information on Universal Credit yesterday by the Department for Work and Pensions (DWP), alongside the regular Official Statistics release, but also to share our concerns about how this differs from the information that was preannounced. Universal Credit continues to be an area of high user interest and has increasingly moved to the foreground of public debate as the UK adjusts to rapid changes in society and the economy as a result of Covid-19.

We recognise the challenging environment in which your department is working in and support the work it is doing to balance the demand for up-to-date information with ensuring relevance to users and good quality data. The statistics teams in your department have shown flexibility and responsiveness by exploring different data sources to meet this demand.

In recent weeks, this management information has been provided by ministers to give context in public statements. Following yesterday’s publication, and in line with the Code of Practice for Statistics, the DWP should ensure that any public statements are supported by information which is equally available to all.

It is both regrettable and concerning that some of the breakdowns of management information that your department had preannounced last week and raised with the Work and Pensions Committee were not published yesterday. On 15 April 2020, the DWP preannounced the release of supplementary management information and stated it would include breakdowns by geography and key characteristics. Prior to this Peter Schofield, Permanent Secretary of the DWP, wrote to the chair of the Work and Pensions Committee on the 3rd April 2020 stating the DWP was expecting to publish a set of supplementary data on Universal Credit including the number of declarations, the number of awards, the number of advances and information relating to the busiest times of the day for calls and other relevant performance information. Yesterday, however, the DWP only published information on the number of households and individuals making a Universal Credit declaration and the number of Universal Credit advances by type of advance.

The labour market will continue to feel the effects of Covid-19 over the coming months and therefore user demand for the timely, relevant data this management information provides will probably persist. We hope that the DWP will publish the remaining breakdowns that it had preannounced and we look forward to seeing this information published on a regular basis for the duration it remains relevant to users. It is our understanding that the DWP had already been working on these breakdowns. As management information produced by the DWP is subject to voluntary application of the Code of Practice for Statistics, the DWP should look to enhance its trustworthiness by publishing information it had committed publicly to release. Not doing so creates a risk to both transparency and to public understanding.

The DWP should be transparent about the frequency it intends to publish this management information once it has been determined. We will keep in touch with the statistics team as part of our ongoing monitoring of statistics on Universal Credit.

I am copying this letter to Peter Schofield, Permanent Secretary of the DWP.

Yours sincerely

Ed Humpherson

Director General for Regulation

Compliance Check of DWP Universal Credit Statistics

Dear Steve

UNIVERSAL CREDIT STATISTICS

As you are aware, we recently completed a compliance check of the DWP Universal Credit (UC) statistics. We considered the trustworthiness, quality and value of these statistics in relation to the Code of Practice for Statistics. We have made a number of recommendations to support your continued development of these important experimental statistics.

We appreciate the positive way that the team has engaged with us during the compliance check, and it is clearly committed to delivering improvements. Our Labour Market and Welfare Domain Lead, Catherine Bremner, will continue to engage with the team on progress over the coming months.

We found a range of positive features that demonstrate the value and quality of the statistics. The statistics are easily accessible, and the gov.uk landing page is well laid out, with useful links to previous publications, publication release strategy and interactive maps and data tools. It also highlights related publications, research and analysis papers, evaluation frameworks, and ad hoc statistical publications, all of which add insight and value. The bulletin is easy to follow, the ‘What you need to know’ and ‘About these statistics’ sections explain the key features of the statistics, and informative subheadings explain key trends. We welcome the comprehensive report on the limitations of the data sources and the clear definitions in the Background and Methodology document.

We identified several areas for improvement which would enhance the clarity, value and quality of these statistics. We consider that the commentary in the bulletin could generate more insight for instance by explaining what the different comparisons mean, such as those between male and female claimants. We also encourage you to consider more-informative ways of presenting the data, for example a trend analysis of the four measures (claims made, starts, people on Universal Credit, households on Universal Credit).

We encourage the team to explain technical terms to help all users understand key concepts and methods, such as ‘pathfinder’ and ‘conditionality regime’. In addition, it may be helpful for you to review whether some of the visualisations could be improved to aid interpretation of the statistics; for instance, the pie charts or the map with the breakdown on UC claimants by local authority, which lacks context. The team may find the Good Practice Team’s guidance on effective charts and maps helpful.

We welcome the section on uses and users of the statistics and would look to DWP to proactively engage with external users and stakeholders to encourage wider use of the statistics, and to ensure that users’ needs are understood, and can feed into further developments of the statistics.

To help users understand the limitations of the statistics and the methods, further quality information could be included on the impact of the limitations, and information on methods changes and why the four measures (claims, starts, people, households) were chosen. Further information on comparability with other statistics would help users generate additional insight, by explaining why these comparisons are relevant and important. It would also be useful to signpost users to related statistics and explain why they are relevant, for example, the Northern Ireland Universal Credit statistics produced by the Department for Communities.

An essential part of assuring yourselves and users about quality, and enhancing the trustworthiness of the statistics, is to provide information about quality assurance. Producers of statistics should explain clearly how the statistics and data are accurate, reliable, coherent and timely. As these statistics are based on administrative data, we encourage the team to apply our Quality Assurance of Administrative Data (QAAD) framework to assure users about the quality assurance arrangements and to help them understand how the Universal Credit data are collected and processed.

These statistics have been published as experimental statistics since they were first introduced in December 2013. We encourage the team to continue to develop these important statistics, to enhance public confidence in their trustworthiness, quality and value. In general, we feel that you could be more ambitious about improvements to the Universal Credit statistics, focusing on how they can enhance value through, for example, linkage with other DWP benefit data and survey data.

Please do not hesitate to contact us if you would like to discuss this further.

Yours sincerely

Mark Pont
Assessment Programme Lead

Compliance Check of DWP Benefit Cap Statistics

Dear Steve

BENEFIT CAP STATISTICS

As you are aware, we recently completed a compliance check of the DWP Benefit Cap statistics. We considered the trustworthiness, quality and value of these statistics in relation to the Code of Practice for Statistics. We have made a number of recommendations to support your continued development of these important experimental statistics.

We appreciate the positive way that the team has engaged with us during the compliance check and it was encouraging to hear of its development plans and desire to maintain contact and discuss improvements. Our Labour Market and Welfare Domain Lead, Catherine Bremner, will continue to engage with the team on progress over the coming months.

We found many examples of good practice. The background information on the statistics is clear and informative, providing a good description of what Benefit Cap is and how it works. The background and methodology document also explains the policy context and contains links to pages on the key benefits. The statistics are easily accessible and the gov.uk landing page is easy to follow, with useful links to previous publications, interactive tools, related publications and research papers, all of which help users better understand the statistics and data. We identified the need to present a clear and coherent narrative about benefit statistics in the bulletin, that is easy to read and supports use by all types of users.

We welcome the helpful summary of key points about the methods at the start of the bulletin and the detailed explanation on the limitations of the statistics in the background and methodology document. The reasons for change and their impact on the statistics are helpfully explained in depth along with how the outcome categories have changed. We have identified a range of detailed improvements that we consider could be made to the presentation of the statistics, which we have passed on to the team separately.

We are encouraged by the recent development of a regular user group and engagement with local authorities, along with embedding user engagement as part of the production process.

We would encourage more engagement with external users and stakeholders to promote wider use of the statistics, to ensure user needs are understood and to maximise public value.

An essential part of assuring yourselves and users about quality, and enhancing the trustworthiness of the statistics, is to provide information about quality assurance. Producers of statistics should explain clearly how the statistics and data are accurate, reliable, coherent and timely. As these statistics are based on administrative data, the team has agreed to produce a quality report using our Quality Assurance of Administrative Data (QAAD) framework to assure users about the quality assurance arrangements and to help them understand how the Benefit Cap data are collected and processed. This is a good development.

Please do not hesitate to contact us if you would like to discuss this further.

Yours sincerely

Mark Pont
Assessment Programme Lead

Statements on ‘new jobs’

Dear Steve

Thank you for responding to my query regarding an article published in the Daily Telegraph on the 21st April 2019, in which the Minister of State for Employment, Alok Sharma MP, was discussing UK employment.

The original article made several statements, such as that “Britons have filled nearly all of the new jobs created in the UK since the 2016 referendum” and that “UK workers have accounted for around nine in ten of new people in work since 2016, compared to half of the people entering work in the two years before”.

You recognised that these statements are not consistent with the relevant official statistics. You explained that the Department for Work and Pension’s internal analytical briefing discourages the use of language which confuses changes in employment with the different and vaguer concept of “new jobs”, in particularly avoiding:

  • The use of the word “job(s)” when referring to headline Office for National Statistics measures of the number of people in employment.
  • The use of the phrase ‘job creation’ where it is claimed or implied that the Government is responsible for directly creating or is the main reason for the creation of “new jobs”.
  • The mixing of statistics on jobs and employment to avoid public confusion.
  • Anything that may confuse net changes in employment with total flows into employment.

This briefing is important and helpful. On this basis, I am reassured that this is an isolated case in which the Department’s analytical advice has not been followed in full. I am also pleased to note that the Telegraph article has been amended and all references to “job creation” and “new people in work” have been removed.

Yours sincerely

Ed Humpherson
Director General for Regulation