Consultation outcome

Response to proposals for changes to fraud and error statistics

Updated 6 December 2018

Overview

The consultation sought user’s views on the:

• frequency of publication
• benefits measured
• breakdowns within the publication

It also asked about how our users use the Fraud and error in the benefit system statistics and their views on presentation.

The consultation closed on 16 September 2018, having received 5 responses:

• one from an individual
• one from an organisation
• 2 from Department for Work and Pensions (DWP) colleagues who use our statistics
• a separate request from the Public Accounts Committee that we are treating as a consultation response

The consultation invited responses to the following areas:

• what you use our statistics for
• would stopping the November publication be acceptable? If not, why not?
• which benefit or benefits you need us to measure that we’ve not measured recently and why?
• if stopping producing age and gender breakdowns would be acceptable? If not, why not?
• what you think about the presentation of our statistics

Responses

The following summarises the main points made by respondents and sets out the department’s response to them. It also sets out how we will take these proposals forward.

What and when we publish

We asked you about the following.

To help us understand your needs, please tell us about what you use the statistics for, including any decisions the statistics will inform and questions you are looking to answer.

Would stopping the November publication be acceptable? If not, why not?

You said

The organisation told us that they use our statistics to help stress the importance of keeping DWP informed of change of circumstances and to demonstrate the value of their organisation.

Our DWP colleagues used the information to:

• identify trends and links to other management information
• assess progress against the departmental objective to reduce fraud and error
• inform decision making on putting activities in place to advance progress against this objective

Almost all respondents said that it would be acceptable to stop the November publication. One respondent said that they felt this was not acceptable, however they were not a user of our statistics and their view was based on this being an important issue for government.

What we are doing

Following the release of our 2017 to 2018 final estimates statistical publication on 6 December 2018, we will stop producing the publications in autumn of each year (usually November).

We will still publish a release in spring each year. These statistics will be used in the department’s annual report and accounts and, as now, will be the statistics that the National Audit Office comment on.

Management information, data and analysis will be regularly provided to officials within the department, allowing trends to be identified and issues addressed, ensuring a focus on the issue of fraud and error.

Which benefits we should measure

We asked you about the following.

Which benefit or benefits you need us to measure that we’ve not measured recently and why. These are the options:

• Attendance Allowance
• Carer’s Allowance
• Disability Living Allowance for children or pensioners
• State Pension measured for fraud and claimant error (we measure it every year but only for official error)

You said

Responses provided reasons for measuring all of the above benefits. Reasons included:

• ensuring that these benefits were not being claimed fraudulently, in particular Carer’s Allowance and Disability Living Allowance for children
• understanding errors within Attendance Allowance and Disability Living Allowance due to their impact on entitlement to other benefits
• understanding error in claiming State Pension

What we are doing

There was no consensus from respondents on which benefit we should measure first. We have carried out an assessment of the options using Multi Criteria Decision Analysis (MCDA) to help us decide the order in which we should measure these additional benefits. This technique allows the comparison of different options. In this case the benefits that could be measured, against a variety of relevant criteria.

We will review Carer’s Allowance cases between April 2019 and February 2020, and hope to publish information on this within our statistical release in May 2020.

The results of the MCDA showed that Carer’s Allowance should be the next benefit we measure due to following factors:

• the time since the last review – we haven’t reviewed Carer’s Allowance since March 1997
• the level of overpayments for Carer’s Allowance when the benefit was last reviewed
• the interest due to this benefit being devolved to Scotland

Based on lessons learnt from measuring Carer’s Allowance we will agree an approach and timing for measuring the other benefits.

Demographic breakdowns

We asked you about the following.

Do you use the age and gender breakdowns on Tables 6 and 12? What do you use them for? Would removing the age and gender breakdowns be acceptable? If not, why not?

You said

Most respondents did not use the age and gender breakdowns and therefore felt that removing them would be acceptable. The organisation did use the breakdowns as they enabled them to focus projects on certain demographics.

What we are doing

Due to their being an expression of user demand we will currently continue to publish the age and gender breakdowns within Tables 6 and 12 of the publication.

Additional information

We asked you about the following.

Do you have any other thoughts or comments?

You said

The organisation asked if it would be possible for a geographical breakdown to be added to support their remit, which is focussed almost entirely at local authority and region.

The Public Accounts Committee recommended that we publish statistics on how many claimants are affected by under and over payments.

What we are doing

Due to the sample sizes involved it is not possible for us to break down the data to local authority or regional level.

We already publish information on the proportion of claims that have under or over payments on them for our continuously measured benefits:

• Employment and Support Allowance
• Housing Benefit
• Jobseeker’s Allowance
• Pension Credit
• Personal Independence Payment
• Universal Credit

Producing a figure on the number of claimants affected would be extremely complex due to:

• people being able to claim more than one benefit at a time
• caseload changes throughout the year
• the fact that we carry out benefit reviews for fraud and error independently for each benefit
• the fact that we do not continuously measure all benefits

We are therefore not currently able to produce robust figures on the number of claimants affected. Once Universal Credit is fully rolled out we will consider whether it would be feasible to publish these figures.

Next steps

We published our 2017 to 2018 final estimates publication on 6 December 2018.

From 2019 we will publish annually in spring.

Following a one year review beginning in April 2019, we hope to publish statistics on Carer’s Allowance for the first time in May 2020.

Learning lessons from this we will develop a timetable for measuring and reporting on:

• Attendance Allowance
• Disability Living Allowance
• State Pension

Report authors: Alexis Marlborough and Catherine Hope

Email contact: [email protected]