Research and analysis

Survey of public perceptions of fraud, error and debt: research report

Published 31 October 2023

DWP research report no. 1045

A report of research carried out by Ipsos on behalf of the Department for Work and Pensions.

Crown copyright 2023.

You may re-use this information (not including logos) free of charge in any format or medium, under the terms of the Open Government Licence.

To view this licence, visit the National Archives

Or write to the:

Information Policy Team
The National Archives
Kew
London
TW9 4DU

Email: [email protected]

This document/publication is also available on our website at: Research at DWP

If you would like to know more about DWP research or you’d like a copy of this report in Welsh, email: [email protected]

First published October 2023

ISBN 978-1-78659-585-0

Views expressed in this report are not necessarily those of the Department for Work and Pensions or any other government department.

Voluntary statement of compliance with the Code of Practice for Statistics

The Code of Practice for Statistics (the Code) is built around 3 main concepts, or pillars, trustworthiness, quality and value:

• trustworthiness – is about having confidence in the people and organisations that publish statistics
• quality – is about using data and methods that produce assured statistics
• value – is about publishing statistics that support society’s needs for information The following explains how we have applied the pillars of the Code in a proportionate way.

Trustworthiness

This research was carried out by Ipsos, who worked with Department for Work and Pensions (DWP) to understand the aims of the research. The design, delivery and analysis were carried out impartially and in compliance with the Market Research Society Code of Conduct, the Government Social Research code of practice and the international quality standard for market research, ISO20252.

Although research findings are shared with ministers and other officials before publication, this is done to promote the value of the research to the department. Ministers have no editorial role.

Quality

The survey was carried out using established quantitative research methodology and statistical methods. Details of these methods are provided in an accompanying technical note. The research has been quality assured using Ipsos’s internal quality checking processes, which have been shared with the Department for Work and Pensions. The analysis of findings and report writing has been quality assured by analysts at the Department for Work and Pensions.

Value

This survey provides insight into the public’s views on fraud and error in the welfare benefits system and on proposed new measures to tackle this. Findings from the survey will inform the development of these new measures, how they are brought forward and the way they are communicated.

Executive summary

This report details the findings from the Public Perceptions Survey, which surveyed UK adults to understand perceptions of fraud, error and debt (FED) in the welfare benefits system, and to gauge attitudes toward a range of proposed new powers to reduce it. These include providing greater third-party access to data, modernising information-gathering powers, seeking law-enforcement powers, developing debt recovery, and reforming the current penalty regime.

Overall, fraud and error in the welfare benefit system was seen as a concern: 62% of respondents saw fraud and error in the system as a big problem; 39% thought it was on the increase (compared to 5% who thought it was decreasing); and around a third (31%) believed “most” incorrect benefit claims are a result of dishonesty.

There is some lack of confidence in the current government response to fraud and error detection and prevention: 59% said they believe it is “not likely” that people who falsely claim benefits will be caught (32% believe it is likely they are caught). A similar proportion (56%) said they believe that the government is not doing enough to reduce levels of fraud and error in the welfare system (conversely, 17% believed the government is doing “about the right amount” and 7% that they are doing “too much”). However, there is a high degree of uncertainty about the government’s response, with 1 in 5 (21%) saying they did not know whether the government was doing a good or bad job or whether their response was adequate.

DWP’s existing powers are fairly well-known: the most well-known power was DWP’s ability to prosecute fraudulent cases, which 81% of the public either knew about or would have assumed to be the case; similarly, 80% of the public were aware that DWP could fine claimants for fraud. There were also some common misconceptions: one-third of the public (32%) believe that specially trained DWP investigators already have arrest, search and seizure powers, and over half (52%) believe DWP can access information from someone’s bank account when that person applies for benefits.

Overall, respondents were more likely to feel positive about the new powers than negative. Half (50%) of respondents said they felt positive about the new powers overall, while 21% of respondents felt negatively and 25% felt neutral. Benefit claimants were less likely than non-claimants to feel positive (38% of claimants felt positively about the new powers), but there were still more people among this group who felt positive than negative (27% of claimants felt negatively). Claimants were overall also less likely to see each of the proposed powers as acceptable.

Each of the proposed new powers described in the survey saw more respondents rating them as “acceptable” than “unacceptable”. In particular, there was widespread acceptance of government organisations sharing information with DWP about claimants and for asking banks to flag potentially fraudulent activity.

Perceptions of the remaining proposed powers were more mixed: whilst around half felt the powers were acceptable, each had more than 1 in 5 respondents saying they found the suggestion unacceptable. The most sensitive propositions were specially trained DWP investigators having the power to make arrests in cases of fraudulent activity (51% thought this acceptable, 32% would find this unacceptable), and DWP collecting information about where claimants are spending money (52% acceptable, 27% unacceptable).

Respondents who said they thought fraud and error was a big problem were consistently more likely to believe that each of the proposed powers were acceptable, as were those who thought the prevalence of fraud and error was increasing and those who believed most incorrect claims are a result of dishonesty. Benefit claimants were overall less likely to see each of the proposed powers as acceptable than non-claimants.

There was a clear pattern of response by age, with older respondents more receptive to each of the proposed new measures. Older respondents were also most likely to believe fraud and error in the system is a big problem, is increasing, and is caused by dishonesty. Younger respondents, on the other hand, showed more uncertainty across the board.

Half of the sample was shown additional or different wording in three of the survey questions, to test whether a brief explanation of the positive impact the powers could elicit a more favourable response. In 2 of 3 cases, the different wording had an impact on the immediate reaction to the specific situation being described. However, seeing the different wording throughout the questionnaire had no impact on feelings about the proposed new powers when respondents were asked to give an overall view at the end of the questionnaire.

Chapter 1: Introduction

The Government is exploring potential legislative measures to help DWP reduce levels of fraud, error and debt (FED) in the benefits system^{[footnote 1]}. The proposed legislative measures include providing greater third-party access to data, modernising information-gathering powers, seeking law-enforcement powers, developing debt recovery, and reforming the current penalty regime. The research aimed to explore the acceptability of these proposed powers with the public in general, and with DWP benefit claimants in particular to help inform future policy development, assess communication approaches and build the evidence base on fraud, error and debt.

To contextualise the proposed legislative measures, the research also sought to understand public perceptions of fraud, error and debt in the welfare benefits system. DWP estimates that in 2022, there was an estimated £6.5 billion of welfare fraud, up from £6.2 billion from the year before. Together with £2.2 billion of error, the combined loss as a result of fraud and error was £8.6 billion, or 4.0% of benefit expenditure^{[footnote 2]}. Although the issue of perceptions of fraud has been researched previously (for example in the British Social Attitudes survey), data on a full range of questions had not been gathered prior to the pandemic, and at the time of the survey’s inception it was not clear which questions would be asked in British Social Attitudes survey 2022. Exploring general public attitudes to fraud and error also allowed for the research to investigate whether attitudes affected claimants’ views on the proposed powers or not.

The research also has a role in informing communications messages, so it was of interest to understand how variations in messaging around the proposed set of powers influenced attitudes towards them. It was also important to undertake sub-group analysis of the findings to explore whether communications approaches should be targeted at, and/or tailored to, certain groups.

For the purposes of this research, we did not seek to distinguish fraud from error, and defined both as “when someone is paid too much, or not enough benefits. This may be because of intentionally claiming benefits they are not entitled to, or because of a mistake. This might be their mistake or the government’s mistake.” The questionnaire also explained to respondents that “if someone has been paid too much benefit, they may have a debt to the government.”

Chapter 2: Method

The survey was carried out using KnowledgePanel, Ipsos’ random probability online panel. This panel is highly representative, and the profile of panellists is similar to national statistics on key demographics. Panellists are recruited via a random probability unclustered address-based sampling method. This means that every household in the UK has a known chance of being selected to join the panel. Letters are sent to selected addresses in the UK (using the Postcode Address File), inviting them to become members of the panel. Invited members can sign up to the panel by completing a short online questionnaire or by returning a paper form. Up to 2 members of the household may sign up to the panel. Members of the public who are digitally excluded are able to register to the KnowledgePanel either by post or by telephone, and are given a tablet, an email address, and basic internet access which allows them to complete surveys online.

This method was chosen for a number of reasons:

• the sample is highly representative,
• Ipsos already held detailed demographic and geographic information on panel members, therefore freeing up time within the questionnaire for other questions of interest,
• the survey could avoid mode effects by using a single, accessible data collection method,
• fieldwork could be conducted within one week

Fieldwork took place between 15 and 21 June 2023.

Sample and weighting

It was important to include a large proportion of DWP benefit claimants in the sample, to be able to conduct robust sub-group analysis. This was important for exploring whether their views may have differed from those of the general population. The research team therefore decided to interview a nationally representative sample plus an additional boost of DWP benefit claimants, to add to the claimants that would naturally be present in a representative sample.

Ultimately, 2,127 people completed the survey, comprising a nationally representative sample of 1,782 people and the boost of 345 additional DWP benefit claimants. Overall, including the boost sample and those in the nationally representative sample, 618 respondents were DWP benefit claimants, in terms of the definition used for this research. This was defined as people claiming at least one of the following benefits:

• Universal Credit
• Jobseeker’s Allowance/New Style Jobseeker’s Allowance
• Income Support
• Employment and Support Allowance (ESA)/New Style Employment and Support Allowance
• Pension Credit
• Carer’s Allowance
• Attendance Allowance
• Personal Independence Payment
• Disability Living Allowance (adult or child)

People claiming Tax Credits, Housing Benefit or the State Pension were not counted as DWP benefits claimants unless they also claimed one of the benefits listed above. This definition was discussed and agreed with DWP at the outset of the research.

In the figures for the overall population, the DWP benefits claimants from the boost sample have been included, with the group of all DWP benefits claimants weighted in proportion to their prevalence in the overall UK population (24.5%). Some additional weighting was carried out, as standard, to correct for imbalances in the achieved sample on other characteristics. Details of this are provided in the accompanying technical note.

Questionnaire design

The questionnaire was designed to be 15 minutes long^{[footnote 3]}. Ipsos developed the survey questionnaire based on an initial draft supplied by DWP. Ipsos then held a stakeholder workshop to feed into questionnaire development as well as attending meetings with DWP policy and strategy teams.

The questionnaire was cognitively tested^{[footnote 4]} with 10 members of the public, including five DWP benefit claimants. This testing resulted in changes to the questionnaire to improve understanding and reduce cognitive effort for participants. This included removing some of the more complex questions altogether, as participants in testing consistently struggled to answer these. The questions that were removed related to changes to penalties for fraud and error, and questions around safeguards for the powers.

Following these changes, the questionnaire was also reviewed for best practice standards by a member of Ipsos’ internal panel of polling experts.

As a general principle, the questionnaire was worded to provide only the information essential to understanding the proposed powers and to describe the powers in a neutral, factual way. Some powers were also described in the form of an example scenario, to check whether this alternative method of presentation affected respondents’ views.

It was also of interest to test alternative wording to understand whether this would affect attitudes towards the powers. To do this, three of the questions had alternative versions which were presented to half of the sample, selected at random. One half of respondents (sample A) saw neutrally-worded versions of these questions, while the other half (sample B) saw versions with wording intended to elicit a more favourable response.

Interpretation of the data

Unless otherwise stated, figures given in this report are for the UK population as a whole. All differences between sub-groups noted in this report are statistically significant at the 95% level^{[footnote 5]}.

Chapter 3: Overall views on fraud and error

In order to understand reactions to the proposed measures relating to FED, it is important to first understand public perceptions of fraud, error and debt in general. This includes perceptions of how prevalent fraud and error is in the welfare benefits system, and views on how big a problem this is.

Responses suggest that there is concern about fraud and error in the welfare benefit system. Overall, 3 in 5 people (62%) saw it as a big problem; this included 1 in 5 (18%) who thought it was a very big problem and 44% who thought it was a fairly big problem. This compares with 24% who said it was not a big problem and a handful (3%) who thought it was not a problem at all (Figure 3.1).

Figure 3.1: People were twice as likely to believe fraud and error was a big problem in the welfare benefits system, as believe it is not a big problem

In general, how big a problem, if at all, do you think fraud and error in the welfare benefits system is in the UK?

Base: 2,127 UK adults 18+

The views of DWP benefit claimants on this issue were the same as those of the general population, with 64% saying it was a big problem and 27% that it was not.

Women, older people, white people and non-graduates were more likely to see fraud and error as a big problem compared to the overall population (see Table 3.1 in annexe). Non-graduates were more likely to see it as a big problem than graduates (67% compared to 49%).

Table 3.1: Differences in how big a problem FED was perceived as, by subgroup

Subgroup (base in brackets)	% Very/fairly big problem	% Not a big problem / Not a problem at all
Overall (2,127)	62%	27%
DWP benefit claimants (618)	64%	27%
Male (979)	58%	31%*
Female (1,131)	65%*	23%
16-24 (85)	40%*	50%*
25-34 (210)	55%*	31%
35-44 (310)	62%	28%
45-54 (363)	58%	26%
55-64 (495)	72%*	18%*
65-74 (449)	73%*	19%*
75+ (215)	72%*	19%*
White (inc. white minorities) (1,930)	63%*	25%*
Ethnic minorities (exc. white minorities) (164)	51%*	37%*

*denotes a number significantly different to the “overall” score.

Figures do not sum to 100% due to those saying “don’t know”.

People were fairly evenly split on whether levels of fraud and error were increasing or staying the same. Those who believed that fraud and error in the welfare benefits system was a big problem were also likely to believe that levels of fraud and error were increasing. Whilst 4 in 10 people overall (39%) said they thought levels were going up, this increased to nearly 6 in 10 (57%) among the group who thought it was a “big problem”.

Only a minority of respondents (5%) thought levels of fraud were going down, while 34% believed they were staying about the same. A relatively high proportion of respondents, over 1 in 5 (21%), said they did not know.

DWP benefit claimants were just as likely as the general population to believe levels of fraud and error were on the increase (38%); however, a higher proportion of this group believed levels were decreasing (10%, compared to 5% overall).

Older people were, again, more negative about the situation. A higher proportion of older respondents believed levels of fraud and error were going up, with the highest proportions falling in the 55-64 age group (50%) and 65-74 (48%). This compares with 25% in the 16-24 age group and 31% among 25-34s.

Figure 3.2: The majority of respondents were split between whether levels of fraud and error were increasing or staying the same

Do you think that levels of fraud and error in the welfare benefits system are generally going up, going down or staying about the same?

Base: 2,127 UK adults 18+

In terms of why that might be, dishonesty was seen as a major cause of the problem, with error also playing a part: the most common view was that around half of incorrect claims were due to dishonesty and half to mistakes (39%), but around a third (31%) believed that most incorrect benefit claims are a result of dishonesty. Half of this number (17%) believed most incorrect claims result from mistakes.

In terms of attitudes towards claimants’ honesty, this figure reflects a finding from the 2022 British Social Attitudes survey (Butt, Clery and Curtice (eds.), 2022)^{[footnote 6]}, in which 27% of people agreed that “most people on the dole are fiddling one way or another”. The British Social Attitudes survey demonstrates that these attitudes have been relatively stable since 2016, with the pandemic having little impact on shifting public attitudes to welfare.

DWP benefit claimants were also more likely to identify dishonesty as being behind most incorrect claims than they were to blame mistakes. However, this group’s views were more evenly split between these 2 viewpoints compared to the general public: 24% of benefit claimants (compared to 31%) said most incorrect claims were a result of dishonesty and 21% (compared to 17%) identified mistakes as the main reason.

People who believed that fraud and error is a big problem in the welfare benefits system were far more likely to say most incorrect claims were due to dishonesty (41%) than people who believed it was not a big problem (15%), and likewise the reverse was true for the belief that most claims are a result of mistakes (9% compared to 37%). Similarly, people who believed fraud and error is on the increase were far more likely to blame dishonesty for the majority of incorrect claims (46%, compared to 24% of people who did not believe it was increasing).

Figure 3.3: Respondents were more likely to blame dishonesty than mistakes for the majority of incorrect claims

Which of the following statements is closest to your view?

Base: 2,127 UK adults 18+

In terms of their own views on whether knowingly giving false information to support a benefits claim is acceptable, people overwhelmingly agreed that doing so was wrong. The majority (85%) said it was “always” or “usually” wrong, with only 1% saying that they did not think doing this would ever be wrong. These results are similar to those seen in the 2019 British Social Attitudes survey, in which 87% of respondents believed this was “always” or “usually” wrong.

DWP benefit claimants were less likely to say doing this was “always” wrong, and more likely to say it was rarely or never wrong. Older respondents were more likely to take the view that it was “always” wrong to give false information, as were people from white ethnic backgrounds (see Table 3.2).

Table 3.2: Differences in views on whether knowingly giving false information to support a benefit claim is wrong, by subgroup

Subgroup (base in brackets)	% “Always” wrong	% “Rarely” or “Never” wrong
Overall (2,127)	64%	3%
DWP benefit claimants (618)	58%*	5%*
Male (979)	63%	3%
Female (1,131)	64%	3%
16-24 (85)	23%*	7%
25-34 (210)	54%*	6%*
35-44 (310)	67%	2%
45-54 (363)	67%	2%
55-64 (495)	72%*	2%
65-74 (449)	73%*	1%*
75+ (215)	85%*	2%
White (inc. white minorities) (1,930)	65%*	2%*
Ethnic minorities (exc. white minorities) (164)	54%*	6%

*denotes a number significantly different to the “overall” score.

Figures do not sum to 100% due to those saying “don’t know”.

Chapter 4: Views on Government response to fraud and error

In order to frame responses to the proposed new measures, the survey sought to understand perceptions of the government’s current response to the issue.

Compared to other questions in this research, there was a high degree of uncertainty about questions relating to the government response, with 1 in 5 people (21%) saying they did not know whether the government is doing a good job or a bad job of reducing levels of fraud and error in the welfare benefits system, and the same proportion saying they did not know whether the government is doing too much, not enough or about the right amount to address this. Cognitive testing of these questions suggested that these “don’t know” responses resulted from participants believing they were not well-informed enough to comment.

Furthermore, there was a lack of confidence in the current response to fraud and error. 3 in 5 respondents (59%) said they believed it is “not likely” that people who falsely claim benefits will be caught, nearly twice as many as the 32% who believed it is likely. These findings are similar to those seen for the same question in the 2022 British Social Attitudes survey, in which 61% of people thought it was unlikely that people who falsely claim benefits will be caught. A majority believed the government was not doing enough to reduce levels of fraud and error in the welfare system (56%); and that the government is doing a “bad job” at reducing levels of fraud and error in the welfare benefits system (58%). In the British Social Attitudes survey 2022, 60% felt the government was not doing enough on this issue.

DWP benefit claimants were more positive about the likelihood of people who falsely claim benefits being caught, with 52% saying they thought this was “likely” (compared to 32% overall) and 40% “not likely” (compared to 59% overall). This group was also more likely to think that DWP were doing “about the right amount” to reduce levels of fraud and error (22%, compared to 17% overall); indeed, 1 in 11 (9%) said they believed DWP does “too much” in this area (compared to 7% overall). They were also more likely than the overall population to say they thought the government was doing a “good job” at reducing levels of fraud and error in the welfare benefits system (31% compared to 21% overall).

Figure 4.1: There is not a high degree of confidence in the government’s response to fraud and error, however DWP benefit claimants are more positive on average than the general population as a whole

Base: 2,127 UK adults 18+ including 618 DWP benefit claimants.

Women had more confidence in the government’s handling of false benefits claims than men, with 35% of women saying they believed it was likely people who falsely claim benefits would be caught, compared to 29% of men. They were also less likely to say the government is doing a bad job of reducing levels of fraud and error in the benefits system (55%, compared to 61% of men).

Older respondents were more negative overall about the government response to fraud and error. Respondents falling into older age groups were less likely to believe that the government is doing “enough” to target false benefits claimants, less likely to believe people who falsely claim benefits would be caught, and more likely to say the government is doing a “bad job” on reducing levels of fraud and error in the welfare benefits system. There was a distinct difference in views between those aged under 55 and aged 55 or over (see Figure 4.2), whereby the proportion believing it was not likely people who falsely claim benefits would be caught went sharply up. Similarly, at 55 and above, the proportion who believed the government is not doing enough to reduce levels of fraud and error increased by 13 percentage points (from 52% of 45-54-year-olds to 65% of 55-64-year-olds). This is also demonstrated in Figure 4.2.

The youngest respondents, aged 16-24, were more likely than average to believe that about the right amount is being done (29% compared to 17% overall) or that too much is being done to tackle fraud, error and debt (13% compared to 7% overall). They were also much more likely than older respondents to say they believed the government was doing a good job of reducing levels of fraud and error in the benefits system (36%, compared to 21% overall).

Figure 4.2: Older respondents were overall more negative about the government’s response to fraud and error than younger respondents

How likely or unlikely do you think it is that people who falsely claim benefits will be caught?

Do you think the government is doing too much, not enough, or about the right amount to reduce levels of fraud and error in the welfare benefits system?

Unweighted bases: 16-24 85; 25-34 210; 35-44 310; 45-54 363; 55-64 495; 65-74 449, 75+ 215.

Respondents from ethnic minority groups were also more confident, with 43% believing it was likely people who falsely claim benefits would be caught (compared to 30% of respondents from white backgrounds), and 40% saying they believe the government is doing a good job of reducing levels of fraud and error in the system (compared to 19% of those from white backgrounds).

Chapter 5: Information-gathering

This chapter looks at the powers DWP has to gather information as part of a criminal investigation into benefit fraud. The Government is of the belief that, having not been substantively updated for over 20 years, these powers “are no longer flexible enough to investigate many of the modern types of fraud we see today, let alone be future-proofed for the frauds of the future^{[footnote 7]}”. The survey asked the public about their awareness of current powers and their views on the potential for a new information-gathering power.

Awareness of current powers

Currently, specially trained DWP authorised officers are able to seek information from specified named organisations, such as financial institutions, childcare providers, utility companies, landlords and credit reference providers when investigating benefit fraud. The majority of survey respondents either knew that DWP had powers of this kind or would have assumed this to be the case. When asked whether they were aware “that DWP can ask for information about individuals from organisations such as banks, as part of a criminal investigation into benefit fraud”, 28% of survey respondents were aware DWP could do this, and 42% said that they had assumed DWP could do this, equating to 7 in 10 adults believing this to be the case.

Conversely, slightly fewer than 3 in 10 adults believed otherwise, with 20% stating they were not aware DWP could do this and 8% saying they would have assumed not.

DWP benefit claimants were more likely than the population as a whole to be aware of these powers, with 38% of claimants saying they knew DWP could do this.

Figure 5.1 DWP benefit claimants were more likely to be aware of current information gathering powers than the population as a whole

Before taking part in this survey, were you aware that DWP can ask for information about individuals from organisations such as banks, as part of a criminal investigation into benefit fraud?

Base: 2,127 UK adults 18+ including 618 DWP benefit claimants.

There was an overall trend for older people to be more aware of the powers DWP currently holds; 8 in 10 over 75s either knew or assumed DWP could seek information from named organisations (80% compared to 70% overall, and to 55% of those aged 16-24). Males were more likely than females to say they knew about the current powers (31% compared to 26%).

Views on proposed new powers

The government are considering widening the range of organisations from which a DWP authorised officer can obtain information, enabling earlier access to information (as soon as suspicion arises), and extending the information-gathering powers to all forms of investigations conducted by DWP. This could bring their information-gathering powers up to date with the modern world and future-proof them as far as possible^{[footnote 8]}.

Respondents were asked how acceptable they would find each of the proposed new measures if it were to be introduced. Table 5.1 shows the detail.

Table 5.1: Overall, a higher proportion of people found the proposed new powers to be acceptable than unacceptable

Power	% Acceptable	% Unacceptable
Government organisations sharing data with DWP about DWP claimants	70%	12%
Collecting banking information as soon as fraud is suspected, rather than waiting for a criminal investigation	60%	21%
Collecting information about where claimants are spending money	52%	27%
Collecting information about where claimants are spending money (airline scenario)	51%	29%

Figures for acceptable and unacceptable do not sum to 100% due to those saying “neither acceptable nor unacceptable” or “don’t know”.

Of the measures proposed, respondents were most favourable towards other government organisations sharing data with DWP about people who are DWP claimants, with a total of 7 in 10 (70%) saying it was either “acceptable” or “completely acceptable”. Very few were opposed to this measure, with 12% saying they would find this unacceptable.

6 in 10 (60%) found the sharing of banking information acceptable, compared to 21% who found it unacceptable. Roughly half found it acceptable for DWP to collect information about spending (52%), compared to roughly a quarter finding this power “unacceptable” (27%).

Views were similar when this power was presented in the form of a scenario. Respondents were asked: “Imagine someone claims pension credit and decides to go on holiday for 3 months. They don’t tell DWP. If DWP had increased access to information-gathering, for example information from an airline about that person’s travel, it could detect they were abroad, and the fraud could be prevented.” In this scenario, 51% of respondents thought it would be acceptable for DWP to collect information from the person’s airline to see where they are travelling and 29% thought this would be unacceptable.

For each of the proposed new powers discussed above, within the group of DWP benefit claimants, a smaller proportion considered each of the new powers to be acceptable and a higher proportion considered them unacceptable, compared to the population as a whole (see Figure 5.2). The greatest difference, of 15 percentage points, was for the power relating to collecting claimants’ spending data.

The power to access banking data as soon as fraud is suspected was seen as acceptable by 47% of claimants (13 percentage points less than the overall population), and the power relating to other departments sharing claimant information was seen as acceptable by 58% of claimants (12 percentage points less than the overall population).

Figure 5.2: A lower proportion of DWP benefit claimants found each proposed new power acceptable than in the population as a whole

Base: 2,127 UK adults 18+ including 618 DWP benefit claimants. Figures are % saying “acceptable” or “completely acceptable” about each of the proposed powers.

DWP benefit claimants were overall more likely to be positive than negative about most of the proposed new powers. This was not the case for collecting information about spending, which 41% of DWP benefit claimants felt would be unacceptable, compared to 37% who believed it would be acceptable.

In all cases, the proportion finding the new powers acceptable increased with age, to the extent that in most cases, the proportion in the 75+ age band finding the power acceptable was more than double the proportion in the 16-24 age band. This suggests that the focus of communications will benefit most from targeting younger age groups.

A lower proportion of respondents from ethnic minority groups rated each of the proposed measures as “acceptable” when compared to those from a white background (See Table 5.2).

Table 5.2: Differences in how acceptable respondents found each proposed new information-gathering power to be, by subgroup

% Acceptable	Accessing banking data as soon as fraud is suspected	Collecting claimants’ spending data	Other departments sharing claimant information
Overall (2,127)	60%	52%	70%
Male (979)	61%	53%	72%
Female (1,131)	60%	52%	70%
16-24 (85)	38%*	25%*	37%*
25-34 (210)	49%*	46%	65%
35-44 (310)	61%	52%	73%
45-54 (363)	59%	48%	73%
55-64 (495)	67%*	60%*	79%*
65-74 (449)	70%*	61%*	77%*
75+ (215)	75%*	71%*	82%*
White (inc. white minorities) (1,930)	62%*	53%*	72%*
Ethnic minorities (exc. white minorities) (164)	49%*	42%*	62%*

*denotes a number significantly different to the “overall” score

Respondents who had already expressed the view that fraud and error in the welfare benefits system is a “big” problem were more likely to rate each of the proposed new powers acceptable than those who did not see it as a big problem. The magnitude of difference was large, ranging from 21 percentage points to 34 percentage points difference. For example, for the power to collect information about where claimants are spending money, the proportion finding this acceptable was more than twice as high amongst those who saw fraud and error as a big problem compared to those who did not: 63% of respondents who thought fraud and error was a big problem saw this power as acceptable, compared to 29% of the respondents who did not think fraud and error was a big problem. A similar difference was seen in the relevant scenario question (61% compared to 30%).

As a test, half of the sample was given additional wording for two of the proposed new measures explaining the impact expected by the change, to see whether this had an impact on how acceptable they found the proposal.

At the question about collecting banking information as soon as fraud is suspected rather than waiting for a full criminal investigation, half the sample saw the additional wording: “This would mean DWP could identify fraud or error more quickly, put claims right, and reduce overpayments and debt.” The respondents who saw this additional, favourable wording were more likely to say the power was acceptable (63%, compared to 56% of those who did not see the wording) and less likely to say it was unacceptable (17% compared to 25%). This indicates that explaining the anticipated benefits of the power can result in a more favourable response.

Figure 5.3: The additional explanatory wording made some difference to the perceived acceptability of collecting banking information earlier in an investigation

How acceptable do you think it would be for DWP to be able to collect banking information (such as childcare payments information) about its claimants, as soon as fraud is suspected, rather than needing to wait for a full criminal investigation? Additional wording: This would mean DWP could identify fraud or error more quickly, put claims right, and reduce overpayments and debt.

Base: 1,045 without extra wording, 1,082 with extra wording (assigned at random). Individual code labels do not always sum to net totals due to rounding.

At the question about inter-departmental data sharing across government, half the sample saw an explanation that this would be used “to prevent benefit fraud or error”, while the other half were told that it would “make sure claimants receive the amount of money they are entitled to”. Those seeing the second explanation were less likely to say the power was unacceptable (9%, compared to 15% seeing the “fraud and error” explanation), though the proportion in each sample seeing it as acceptable was similar.

The difference in “unacceptable” responses between the two versions of the wording was particularly pronounced among those who did not believe fraud and error in the system is a big problem. 38% of respondents who saw the “prevention” explanation said that they thought it was unacceptable, whereas 24% of those whose question was framed in terms of the benefit to the claimant found it unacceptable. This suggests that messaging which implies a potential benefit to some claimants from reducing fraud and error may help elicit a more positive response amongst those who are not currently convinced of the need to tackle the problem.

Chapter 6: Third-party data

Currently, DWP authorised officers can only request information from third parties, such as banks, where there is already a suspicion of fraud about an identifiable person. This means that it is likely that significant sums of benefits may be acquired fraudulently before the fraud is detected, and the ability to proactively identify fraud (for example, at the time of application for benefits), is being hindered.

The survey asked the public about their awareness of current powers to request information from third-parties and their views on the proposed new powers.

Awareness of current powers

Many people believed that DWP have more powers than they currently do to access data from third parties. Applicants were asked what types of third-party data DWP could access to check someone’s eligibility for benefits when they apply. 7 out of 10 respondents incorrectly believed that DWP can access information from employers about the applicant’s pay (71%). Around half (52%) incorrectly believed it was possible to access information from the applicant’s bank accounts, and a quarter (23%) incorrectly believed that DWP would be looking at social media posts. A minority (8%) incorrectly thought DWP could access location data from the applicant’s phone.

DWP are currently able to access information from other government departments, and around two-thirds of respondents (67%) were aware of this.

A minority (7%) of the public believed that DWP could not access any of the listed information sources.

Figure 6.1: Perceptions of third-party data DWP can currently access to check new benefit claims

Before taking part in this survey, which types of information did you think DWP could access about its claimants when they apply for benefits, to check whether they are eligible?

Base: 2,127 UK adults 18+.

In general, DWP benefit claimants were in line with the nationally representative sample in terms of their knowledge of what sources of information DWP had access to at the point of application. The exception was regarding information from other government departments, where claimants were less likely to believe DWP would have access (61% compared to 67% overall).

The majority of respondents (79% overall, and 71% DWP benefit claimants) believed it would be acceptable for DWP to access personal information if it was part of a criminal investigation into benefit fraud, which currently is permitted.

Views on proposed new powers

The Government wish to explore the idea that banks can be asked to share information about accounts where account activity suggests that the account holder may be committing benefit fraud. Respondents were therefore asked in which circumstances (if any), from a given list, they thought DWP should be allowed to access someone’s personal data, including data from their bank accounts.

Whilst the majority felt this access to data was acceptable as part of a criminal investigation (as is currently permitted), fewer believed that it was acceptable where fraud was only suspected. Roughly half (47%) believed it would be acceptable in cases where the bank suspected a customer was breaking DWP’s rules, and a similar proportion (46%) if it were suspected by DWP.

A similar number (45%) believed it would be acceptable for DWP to have access to this personal information whenever a claimant first applies for benefits to allow them to check eligibility. 1 in 5 (21%) believed it was acceptable for DWP to access this sort of information about people at any time while they are receiving benefits.

1 in 20 (5%) believed there were no circumstances listed where it would be acceptable for DWP to be allowed to access someone’s personal data, including data from their bank accounts.

DWP benefit claimants were more cautious with regards to what they believed to be acceptable. Within this group, lower proportions believed it would be appropriate for DWP to access someone’s personal data in each of the listed scenarios than across the population as a whole, by around 10 percentage points (see Figure 6.2). The difference between claimants and the general population was largest concerning instances when the bank suspects the person may be breaking DWP’s rules (32% of benefit claimants believed this was acceptable compared to 47% overall).

Figure 6.2: A lower proportion of DWP benefit claimants believed access to personal data such as bank accounts was warranted across the board

In which of these circumstances, if any, do you think DWP should be allowed to access someone’s personal data, including data from their bank accounts?

Base: 2,127 UK adults 18+ including 618 DWP benefit claimants.

Younger respondents were also more cautious regarding the acceptability of accessing personal data in each of the different circumstances. A greater proportion of respondents in the youngest age band (16-24) believed that none of the situations warranted the sharing of personal data including from their bank accounts, with 14% selecting this option, compared to 5% overall. There was also more uncertainty among this group, with 15% saying they did not know whether the sharing of data in any of the listed situations would be acceptable (compared to 4% overall).

The survey asked specifically about how acceptable it would be for DWP to ask banks to share information about accounts which looked like someone may be committing benefit fraud. More than 3 in 5 (64%) believed it would be acceptable, and 1 in 5 (20%) said they believed it to be unacceptable. This difference was more pronounced among DWP benefit claimants, where 51% believed it to be acceptable but 28% said they found it unacceptable.

Respondents showed consistency in these views when the question was posed as a specific example. Respondents were told “If you have more than £16,000 in ‘capital’ (broadly referring to savings), you aren’t entitled to claim Universal Credit. Imagine someone has savings worth £18,000 and did not declare it when they claimed Universal Credit. If their bank could flag their savings to DWP, then cases such as this could be spotted, and the overpayment could be stopped.” Responses were overall in line with the previous question, with 61% saying they thought it was acceptable for this person’s bank to flag their savings to DWP and 22% saying they thought it would be unacceptable. Again, the difference was more pronounced among DWP benefit claimants, with 53% saying it would be acceptable and 28% unacceptable.

Younger people were less likely to find either the general concept of asking banks to share information or the specific situational example to be acceptable when compared to older people. Figure 6.3 illustrates how the proportion of respondents finding the proposal acceptable increased with age, whilst unacceptability decreased with age. This was a steady trend across all age bands with the exception of the 45-54 age band, in which slightly fewer respondents thought this power acceptable and more thought it unacceptable compared to the age bands on either side. The rise in acceptability was particularly steep in the younger age bands, rising from 44% amongst 16-24 year olds to 68% among 35-44 year-olds, and then to 75% amongst the over-75s. Unacceptability decreased from 24% amongst 16-24 year olds to 11% amongst the over-75s.

Figure 6.3: Proportions of respondents who believed asking banks to share information was acceptable increased with age

How acceptable do you think it would be for DWP to ask banks to share information about accounts which look like someone may be committing benefit fraud?

Unweighted bases: 16-24 85; 25-34 210; 35-44 310; 45-54 363; 55-64 495; 65-74 449, 75+ 215. Figures for acceptable refer to % saying “acceptable” or “completely acceptable”; figures for unacceptable refer to % saying “unacceptable” or “completely unacceptable”.

Respondents from ethnic minority groups were less likely than those from white backgrounds to find the proposals to ask banks to volunteer information where they suspect benefit fraud to be acceptable. Two-thirds (66%) of respondents from white backgrounds said they found the proposal acceptable, and half (50%) of those from ethnic minority backgrounds thought the same.

People who had already reported that they felt fraud and error in the benefits system to be a big problem were more likely to say it was acceptable for DWP to ask banks to share information about accounts which look like someone might be committing benefit fraud (72% compared to 48% of those who did not think it was a big problem). Similarly, the proposed change was more popular among people who believed most fraud was down to dishonesty (77%) rather than mistakes (49%).

To test the impact of including favourable messaging, half of respondents at this question were shown the additional wording “This would mean DWP could detect fraud and error more easily, stop fraud more quickly and prevent people from being overpaid and building up debt.” In this instance, there were no differences in levels of acceptability between those who saw this message and those who did not.

Chapter 7: Law enforcement powers

Sometimes, DWP conducts serious criminal investigations into benefit fraud. These investigations are about complex and sophisticated attacks on DWP’s benefits and grants payments – usually by organised crime groups. The department may also investigate individuals who have committed fraud serious enough that they may face going to prison.

Currently, when DWP wishes to conduct an arrest or search and seize evidence as part of an investigation, it needs to involve the police. “Fighting Fraud in the Welfare System” (DWP, 2022)^{[footnote 9]} highlights how this leaves DWP heavily reliant on “police availability and prioritisation”, and puts them at odds with other government departments such as HMRC and the Gangmasters and Labour Abuse Authority (GLAA), who have powers of arrest and powers that allow them to conduct search and seizure.

The proposed measure discussed here considers giving specially trained DWP economic and serious crime investigators the power to make arrests and conduct search and seizures. The survey asked the public about their awareness of current powers and their views on the proposed alternative powers.

Awareness of current powers

A third of the public (32%) believe that specially trained DWP investigators already had the power to conduct arrests and search for and seize evidence as part of the criminal investigations DWP conducts. People from ethnic minority backgrounds are particularly likely to believe this (43%).

However, there is a high degree of uncertainty about this, with roughly two-thirds of the public (63%) reporting that the police were able to conduct arrests and search and seize evidence in these circumstances and around a fifth (22%) saying they did not know who was able to do this. This uncertainty was higher among younger people, with 32% of those in the 16-24 age category saying they did not know the answer.

Figure 7.1: Uncertainty as to where the current powers of arrest and search and seizure lie is high

As far as you know, at the moment, when DWP is conducting a criminal investigation, who is able to conduct arrests and search and seize evidence as part of the investigation?

Base: 2,127 UK adults 18+. Respondents could select both of the “Police” and “DWP investigators” answers.

Views on proposed new powers

The government are considering giving DWP’s own Economic and Serious Organised Crime Investigators the power to make arrests and apply to search and seize evidence in criminal investigations. DWP’s aim is that this will enable them to act in a timely fashion, with less reliance on police availability and prioritisation^{[footnote 10]}.

Powers of arrest

Survey respondents were divided on whether it was acceptable for specially trained DWP investigators, in addition to police, to have the power to arrest people suspected of serious and organised benefit fraud. Half (51%) believed it would be acceptable, and a third (32%) believed this would be unacceptable. Within these totals, 1 in 5 (20% of respondents overall) thought this power would be completely acceptable, whereas 13% of respondents overall thought this power would be completely unacceptable. This was the highest proportion of respondents saying “completely unacceptable” of any of the questions in the survey.

Figure 7.2: People were divided over whether it would be acceptable for DWP investigators to have powers of arrest

How acceptable do you think it would be for trained DWP investigators to have the power to arrest people suspected of serious and organised benefit fraud, rather than only using the police?

Base: 2,127 UK adults 18+. Individual code labels do not always sum to net totals due to rounding.

DWP benefit claimants, male respondents and older respondents were all more negative about this proposed change, with each group more likely to find the idea of specially trained DWP investigators having power or arrest to be not acceptable (see Table 7.1).

Table 7.1: Differences in views on acceptability of specially trained DWP investigators having powers of arrest, by subgroup

Subgroup (base in brackets)	% Acceptable	% Not acceptable
Overall (2,127)	51%	32%
DWP benefit claimants (618)	39%*	37%*
Male (979)	51%	35%*
Female (1,131)	52%	29%*
16-24 (85)	41%	30%
25-34 (210)	40%*	34%
35-44 (310)	49%	35%
45-54 (363)	48%	37%*
55-64 (495)	61%*	28%*
65-74 (449)	63%*	26%*
75+ (215)	62%*	29%
White (inc. white minorities) (1,930)	52%*	32%
Ethnic minorities (exc. white minorities) (164)	47%	30%

*denotes a number significantly different to the “overall” score.

Figures do not sum to 100% due to those saying “fairly acceptable” and “don’t know”.

Respondents who believed fraud and error in the system to be a big problem were more likely to find it acceptable for DWP investigators to have the power of arrest (61%, compared to 32% who did not think it was a big problem).

Respondents who already believed DWP investigators had the power of arrest were more likely than average to say they found it acceptable for this to be the case (66%, compared to 51% overall.)

Respondents’ views were more negative when presented with a scenario to illustrate this power. Respondents were asked “Imagine a situation where someone has committed benefit fraud. The fraud was so large that they may face going to prison. They are invited to an interview with DWP to talk about their benefit claim, but do not attend.” Around half (49%) said either trained DWP investigators or the police should be able to arrest the person for not attending their benefit claim interview, similar to the 51% of those who thought these powers would be acceptable in the more general question. However, 40% said only the police should be able to conduct the arrest in this scenario, a higher proportion than the 32% who saw extended arrest powers as unacceptable in general. This difference suggests that some of the public are supportive of the proposed power in principle but not for its use in a scenario like the one described. In addition to this 40%, a minority (6%) said that no-one should be able to arrest the person for not attending a benefit claim interview.

Powers to search for and seize evidence

Compared to arrest powers, respondents were more accepting overall of the idea of trained DWP investigators having the power to search and seize evidence from people suspected of serious and organised benefit fraud, rather than only using the police. Around 6 in 10 (59%) said they would find this acceptable, and 25% said they would find it unacceptable for trained DWP investigators to have these powers.

Figure 7.3: A sizeable minority found the suggestion that trained DWP investigators get powers to search for and seize evidence unacceptable

And how acceptable do you think it would be for trained DWP investigators to have the power to search for and seize evidence from people suspected of serious and organised benefit fraud, rather than only using the police?

Base: 2,127 UK adults 18+.

Again, DWP benefit claimants were less likely to find the idea of DWP investigators having powers to search and seize evidence acceptable.

Similar demographic differences in opinion were also seen for giving DWP investigators the power of search and seize to those seen for giving them the power of arrest; differences were seen by gender and age, and in this case, there was also a sizeable difference by ethnic background (see Table 7.2).

Table 7.2: Differences in views on acceptability of trained DWP investigators having powers of search and seize, by subgroup

Subgroup (base in brackets)	% Acceptable	% Not acceptable
Overall (2,127)	59%	25%
DWP benefit claimants (618)	44%*	35%*
Male (979)	56%	27%
Female (1,131)	61%*	23%*
16-24 (85)	31%*	30%
25-34 (210)	51%*	28%
35-44 (310)	59%	27%
45-54 (363)	56%	29%
55-64 (495)	68%*	22%*
65-74 (449)	73%*	19%*
75+ (215)	71%*	19%*
White (inc. white minorities) (1,930)	61%	24%*
Ethnic minorities (exc. white minorities) (164)	45%*	30%

*denotes a number significantly different to the “overall” score.

Figures do not sum to 100% due to those saying “neither acceptable nor unacceptable” and “don’t know”.

Again, respondents who felt fraud and error in the system was a big problem were more in favour of the measure, with 68% saying they would find it acceptable.

Respondents who had indicated they already believed DWP investigators were permitted to search for and seize evidence were more accepting, with 71% saying it would be acceptable (compared to 59% overall).

For this power, presenting it as a scenario appeared to elicit similar views to the general question. Respondents were told “Imagine a situation where an organised crime group has been producing false identities to make bogus universal credit claims.” Reactions were in line with the previously stated views on acceptability, with 63% saying either trained DWP investigators or police should be able to seize evidence from members of this group about this fraud, and 31% believing only the police should be able to seize this evidence. Only a handful (1%) believed no one should be able to seize evidence from members of the group about this fraud.

Chapter 8: Penalties

DWP currently has a range of penalties in place that can be used when an offence has been committed. These range from a fixed penalty fine, where a claimant has given incorrect information in error, to prosecution and imprisonment for the most serious cases of fraud. The government wish to consider refining the system to ensure fewer people who have committed wrongdoing escape punishment (24% of claimants identified as committing fraud or error in 2019/20 did not face a penalty)^{[footnote 11]}.

Awareness of current powers

Awareness of DWP’s current powers around the penalties that can be applied was relatively high, compared to other powers examined by the survey: around 4 out of 5 people were aware of these powers. Around half (47%) said they knew DWP could and do criminally prosecute some fraudulent cases^{[footnote 12]}, and a further third (34%) said that, whilst they wouldn’t have been certain, they would have assumed DWP could do this. Similarly, almost half (45%) said they knew DWP can fine claimants for fraud^{[footnote 13]}, and 35% said they had not been certain but would have assumed this to be the case.

Figure 8.1: Awareness was high about DWP’s current powers to administer penalties.

Before taking part in this survey, were you aware that DWP criminally prosecutes some fraudulent cases?

Before taking part in this survey, were you aware that DWP can fine claimants for fraud?

Base: 2,127 UK adults 18+.

Overall views on penalties

A large majority of respondents agreed that people who commit benefit fraud should be punished. Eight in ten members of the public (80%) agreed with the sentiment, compared to 4% who disagreed.

The majority of DWP benefit claimants also agreed that people committing benefit fraud should be punished, although the proportion was lower than across the population as a whole (72% compared to 80%).

Young people, in the 16-24 age group, showed more uncertainty over whether they agreed with punishing perpetrators of benefit fraud. Roughly half (51%) stated they agreed with the concept, and respondents in this group were more likely to say they neither agreed nor disagreed (31% compared to 13% overall), or that they did not know whether they agreed or disagreed (12% compared to 3% overall). This compares to 75% of those aged 25-34 agreeing that people who commit benefit fraud should be punished. Overall, the proportion of those agreeing with this statement rises with age, staying around the same across the 35 to 64 age groups and reaching 91% among the over-75s.

Figure 8.2: The majority of respondents agreed that people who commit benefit fraud should be punished, with some variation by age

In general, to what extent do you agree or disagree that people who commit benefit fraud should be punished?

Unweighted bases: 16-24 85; 25-34 210; 35-44 310; 45-54 363; 55-64 495; 65-74 449, 75+ 215.

Similarly, the majority of the population said they believe it is acceptable for DWP to remove some benefits from someone who has deliberately given false information when claiming benefits. Overall, 78% of people said they believed removing benefits was acceptable, including 38% who said this was completely acceptable, with 7% saying they believed it was unacceptable (this rose to 11% among DWP benefit claimants).

Again, the biggest demographic differences came by age group, with the youngest respondents showing uncertainty, being more likely than average to say that they either found it neither acceptable nor unacceptable (21% compared to 11% overall) or that they did not know how acceptable it was (14% compared to 4% overall). The majority of respondents aged 25 and above thought removing benefits in the case of fraud was acceptable (75% for those in the 25-34 age group, rising to 86% for those aged 75 or older).

A lower proportion of respondents from ethnic minority backgrounds believed that removing some benefits from someone who has deliberately given false information was acceptable. Seven in ten (70%) thought this was acceptable, compared to 79% of respondents from white backgrounds.

Respondents who believed fraud and error in the benefits system is a big problem were more likely to be in favour of removing benefits where false information had been deliberately given (87%), as were those who believed the biggest cause of fraud and error to be such acts of dishonesty (90%).

Figure 8.3: The majority of respondents believed it was acceptable to remove some benefits in the instance of fraud

How acceptable do you think it is for DWP to remove some benefits from someone who has deliberately given false information when claiming benefits?

Base: 2,127 UK adults 18+.

When considering incentives for correcting false information, two-thirds (64%) thought that penalties should be reduced for those who correct false information they have previously provided. Respondents were asked “Sometimes, people who have knowingly given false information to claim benefits are given a chance to correct this, and provide the right information, when DWP discover the fraud. If someone admits they have committed fraud and corrects the information they provided, to what extent, if at all, do you think they should still be punished for giving the false information?” This compares to 1 in 5 (19%) who thought the penalty should remain the same. These proportions were consistent between the general population and DWP benefit claimants (63% and 18% respectively).

As before, respondents in the 16-24 age category were less likely to have a clear view, with 17% saying they did not know whether the penalty in this situation should be less (compared to 5% overall). No significant differences were seen by ethnicity.

People who believed fraud and error in the system is mostly caused by dishonesty were more likely to say the punishment should remain the same even if the claimant corrects the information (31%). 

Chapter 9: Debt recovery

Debt recovery is an important area for DWP, helping to collect overpayments, ensure value to the taxpayer and ensuring those given penalties are complying with the punishment. As part of the overall research into public perceptions of fraud, error and debt, it is important to understand views about the current avenues available to DWP. The survey asked the public about their awareness of current powers and their overall views on the principles behind debt recovery.

Awareness of current powers

Awareness of debt recovery powers was relatively low compared to awareness of other current powers. Approximately a third (31%) of respondents said they knew that DWP could recover debt from former claimants by taking money from their salaries, and 16% knew that DWP requires a court order to recover debt in any other way^{[footnote 14]}. In each case, a further third of people (33% and 37% respectively) said they wouldn’t have been certain but had assumed that DWP could do this.

Figure 9.1: Awareness of DWP’s powers of debt recovery was low

Before taking part in this survey, were you aware that DWP can recover debt that former benefit claimants owe DWP by taking money from their salaries?

And before taking part in this survey, were you aware that DWP requires a court order to recover debt in any other way apart from salaries and benefit payments (for example, directly from bank accounts)?

Base: 2,127 UK adults 18+.

Views on debt recovery

Views on the acceptability of debt recovery varied depending on the cause of the overpayment. Respondents were asked their views on debt recovery across three different situations:

• as a result of the claimant deliberately giving false information.
• as a result of the claimant mistakenly giving false information.
• as a result of a DWP mistake

In each instance, respondents were asked the extent to which they agreed or disagreed that DWP should make the claimant pay the money back. As seen in Figure 9.2 below, a majority thought that overpayment as a result of deliberate dishonesty should be paid back and that overpayment as a result of a DWP mistake should not be.

Figure 9.2: Views on DWP enforcing repayment of debts varied significantly depending on who is responsible for the overpayment

To what extent do you agree or disagree with the following statement?

Base: 2,127 UK adults 18+

In the event of the claimant deliberately giving false information, the vast majority of people (88%) said they agreed DWP should make them pay the money back. 1 in 20 (5%) disagreed with this viewpoint.

If the false information was given by the claimant by mistake, views were more varied, with 61% agreeing the money should be paid back, compared to 20% saying they should not be required to pay it back.

Where DWP was at fault for the overpayments, respondents were more likely to disagree that the money should be paid back (53% thought they should not be required to pay it back) than to agree (33%).

DWP benefit claimants were less likely to feel that the money should be paid back in each scenario, with 82%, 55% and 23% respectively agreeing this should be the case.

In the event of the claimant providing the false information (knowingly or unknowingly), older age groups showed a higher proportion who agreed the money should be paid back. However, when it came to overpayments due to DWP making a mistake, views were fairly consistent across all age groups above 16-24.

Across all three scenarios, the youngest respondents (aged 16-24) showed the highest level of uncertainty (with “don’t know” responses given by 12%, 9% and 9% respectively, see Table 9.1).

Table 9.1: Older respondents were more likely to agree money should be paid back if the claimant was at fault for the overpayment

% Agree DWP should make the claimant pay the money back	If claimant deliberately gave false information	If claimant mistakenly gave false information	Overpaid as a result of DWP error
Overall (2,127)	88%	61%	33%
16-24 (85)	59%*	36%*	16%*
25-34 (210)	85%	55%	33%
35-44 (310)	91%*	66%*	34%
45-54 (363)	90%	65%	35%
55-64 (495)	93%*	66%*	35%
65-74 (449)	93%*	62%	36%
75+ (215)	95%*	67%*	37%

*denotes a number significantly different to the “overall” score.

People who believed fraud and error in the benefits system is a big problem were more likely to believe in each scenario that the money should be paid back (93%, 67% and 38% respectively). People who believed most fraud and error is caused by dishonesty were also more likely to say the money should be paid back in each case (92%, 73% and 44% respectively).

Chapter 10: Overall views on the new powers

After having answered all questions around the proposed new powers regarding fraud and error in the welfare benefits system, respondents were asked how, on balance, they felt about the proposed new powers. The survey had provided only brief information on each of the powers, presented in a neutral manner, except for three questions in which half of the sample were given further wording explaining the rationale behind introducing the new power.

Views overall were more positive than they were negative. Half of all respondents said they were either very of fairly positive about the proposed new powers (50%), compared to 21% who felt very or fairly negative about them. A further quarter (25%) said their feelings were neutral.

Whilst DWP benefit claimants were overall more likely to be positive than negative about the proposed changes, views were more evenly split. A lower proportion than average felt positive about the proposed new measures (38%) and a higher proportion felt negative (27%) (see Figure 10.1).

Figure 10.1: Respondents were overall more positive than negative about the proposed new powers

Having now answered questions around the proposed new powers regarding fraud and error in the welfare benefits system, on balance how do you feel about these proposed new powers?

Base: 2,127 UK adults 18+

Overall, there were no differences in overall favourability towards the powers by gender or ethnicity, however, significant differences were seen by age with a higher proportion of older respondents feeling positive about the changes than younger respondents (see Table 10.1). Younger respondents, particularly those in the 16-24 age group, were less certain how they felt about them, with a higher proportion than average giving a “don’t know” answer (15%, compared to 4% overall).

Table 10.1: Differences in overall positivity towards proposed new powers, by subgroup

Subgroup (base in brackets)	% Positive	% Negative
Overall (2,127)	50%	21%
DWP benefit claimants (618)	38%*	27%*
Male (979)	51%	22%
Female (1,131)	50%	20%
16-24 (85)	30%*	27%
25-34 (210)	40%*	23%
35-44 (310)	52%	22%
45-54 (363)	49%	24%
55-64 (495)	58%*	19%
65-74 (449)	62%*	17%*
75+ (215)	60%*	14%*
White (inc. white minorities) (1,930)	51%	21%
Ethnic minorities (exc. white minorities) (164)	48%	21%

*denotes a number significantly different to the “overall” score.

Figures do not sum to 100% due to those saying “don’t know”.

The majority of respondents who believed fraud and error in the welfare benefits system is a “big” problem were positive about the proposed new powers, with 61% saying they felt positive, compared to 29% who did not feel fraud and error was a big problem. People who were not convinced it was a big problem were more likely to be negative (42%) than positive about the proposed new powers. Similarly, respondents who believed the levels of fraud and error are increasing were more positive (63%).

People who believed most incorrect claims are caused by dishonestly were far more likely to be positive about the proposed new measures than people who believed most incorrect claims are caused by mistakes (68% compared to 29%).

People who had previously stated that they thought the government was doing a “bad job” of addressing fraud and error in the system were equally likely to be as positive as those who thought the government was doing a “good job” (51% and 52% respectively). This suggests that current views of how well the government is doing in this area did not lead to increased scepticism about the proposed new measures or affect positivity in any way. However, views on the Government’s handling of other policy areas did seem to have an impact on positivity towards the proposed measures. Respondents who believed the Government is doing a “good job” of improving the NHS, tackling regional inequality, and reducing crime and anti-social behaviour were all more likely than average to be positive about the proposed new measures (72%, 68% and 63% respectively, compared to 50% overall). Conversely, respondents who were negative about the way the Government is handling those three policy areas were more likely to be negative about the proposed new measures to target benefit fraud and error (24%, 27% and 24% respectively, compared to 21% overall).

The group of respondents who were given additional wording on some of the questions to test how this impacted the response to the descriptions of the new powers were no more likely to rate the proposed changes as positive as those who had not received the additional wording (51% compared to 50%, not a significant difference).

1. See Fighting Fraud in the Welfare System ↩

2. See Fraud and error in the benefit system: financial year 2021 to 2022 estimates ↩

3. The median completion time was in fact 12 minutes and 39 seconds. ↩

4. Cognitive testing is a technique used to test and improve survey questions. During a cognitive interview, survey questions are administered (in this case by being displayed on screen for participants to read, to mimic the online mode of the survey) and participants are asked about the cognitive processes they go through in answering those questions. This helps to uncover problems with the questions and identify improvements. ↩

5. Statistical significance testing is used to determine whether differences in results are likely to be due to a genuine difference between groups, as opposed to chance variation. The threshold used in this research is the 0.05 level, meaning there is less than a 5% chance that results deemed significantly different differ due to chance. This is a standard level of significance used in social sciences ↩

6. Butt, S., Clery, E. and Curtice, J.(eds.) (2022), British Social Attitudes: The 39th Report. London: National Centre for Social Research ↩

7. DWP (2022) “Fighting Fraud in the Welfare System”, Department for Work and Pensions. ↩

8. DWP (2022) “Fighting Fraud in the Welfare System”, Department for Work and Pensions. ↩

9. DWP (2022) “Fighting Fraud in the Welfare System”, Department for Work and Pensions. ↩

10. DWP (2022) “Fighting Fraud in the Welfare System”, Department for Work and Pensions. ↩

11. DWP (2022) “Fighting Fraud in the Welfare System”, Department for Work and Pensions. ↩

12. To be precise, DWP does not prosecute but rather refers cases to prosecuting bodies for consideration of prosecution: however, for ease of understanding this wording was used in the questionnaire. ↩

13. Again, strictly speaking the term “fine” is reserved for a fine issued by a court, but for ease of understanding this term was preferred over “financial penalty”. ↩

14. To be precise, a court order is only required to recover debt when the claimant has not set up a repayment arrangement, but this wording was used for brevity in the questionnaire. ↩