Published by the Department of Social Protection 3 October 2016
Household Benefits
Fraud & Error Survey 2016
Household Benefits Package(Only) Recipients
July 2016
Table of Contents
1. Summary ................................................................................................................................................ 1
2. Characteristics of the Household Benefits scheme ................................................................. 2
The Household Benefits package .................................................................................................... 2
Household benefits recipients and primary DSP schemes .............................................................. 2
Age profile of Household Benefits package (only) recipients6 ....................................................... 3
3. Methodology ........................................................................................................................................ 5
Irish Statistical System Code of Practice ......................................................................................... 5
Survey principles ............................................................................................................................. 5
Sample selection ............................................................................................................................. 5
Survey procedures ........................................................................................................................... 5
Categorisation of results ................................................................................................................. 5
‘Gross’ and ‘net’ rates of Fraud and Error ....................................................................................... 6
Valuation of changes in benefit entitlement................................................................................... 6
4. Results ..................................................................................................................................................... 7
Normal Movement .......................................................................................................................... 7
Fraud and Error ............................................................................................................................... 8
5. Risk analysis ........................................................................................................................................... 9
General approach ............................................................................................................................ 9
No statistically significant predictor variables................................................................................. 9
6. Conclusions and recommendations .............................................................................................. 10
Conclusions .................................................................................................................................... 10
1
1. Summary
The Department of Social Protection undertakes fraud and error surveys to establish baseline
fraud and error levels for social welfare schemes. The purpose of such surveys is to identify the
level of risk associated with particular schemes and business areas, with a view to designing
processes and control measures specifically targeted to minimise the level of future risk.
This survey was undertaken on the Household Benefits scheme, which provides recipients with
free Television Licences and contributions towards electricity or gas bills (‘the Household
Benefits package’)
1
. Because over 93% of people who receive the Household Benefits package
also receive a weekly Social Welfare payment, this survey is focused on people who receive the
Household Benefits Package only, i.e. the 7%. Household Benefits package eligibility reviews will
also be included in upcoming Fraud and Error surveys of State Pension (contributory), Carer’s
Allowance, State Pension (non-contributory) and Disability Allowance. Together with this
survey, these reviews will cover some 80% of Household Benefits Package recipients by early
2019 (see section 2 below).
For the present survey, a stratified random sample of 500 Household Benefits package (only)
cases in payment in November 2015 was drawn up, and all these cases were reviewed to assess
recipients’ compliance with the rules of the scheme.
The key results of this assessment are shown in Table 1, and further analysis is given in section 4
below.
Table 1 – Estimated Fraud and Error as a percentage of Household Benefits monthly expenditure
2
Gross rate
Appeals
Transferred
claims
Net rate
Fraud and Error as % of expenditure
6.4%
0.0%
1.0%
5.4%
As % of number of claims
9.0%
1.0%
8.0%
Customer Fraud and Error
1.8%
0.0%
0.4%
1.4%
Customer Fraud
0.6%
0.6%
Customer Error
1.2%
0.4%
0.8%
Official Error
4.6%
0.0%
0.6%
4.0%
Overpayment
4.6%
0.6%
4.0%
Underpayment
0.0%
0.0%
(Source: Department of Social Protection)
In November 2015, there were some 31,000 Household Benefits package (only) recipients.
Extrapolating from the survey results, we estimate the monetary value of total scheme-wide
fraud and error for that group in that month at €0.06 million (net of transfers to other schemes).
This rate of Fraud and Error mainly reflects the difficulty of ensuring timely changes to
Household Benefits package entitlements when recipients move address or die (see sections 2 -
4 for details). The relative remoteness of the Household Benefits package (only) recipients from
the Department means that they constitute a riskier than average subgroup of the 420,000
people receiving Household Benefits. No further statistically significant predictive risk factors
were found within this group, however (see section 5).
1
For further information, please see section 2 and http://www.welfare.ie/en/Pages/Household-Benefits-Package.aspx.
2
Figures may not add due to rounding.
2. Characteristics of the Household Benefits scheme
The Household Benefits package
The Household Benefits package is a non-statutory scheme which entitles qualifying recipients to
two benefits:
1. Electricity or Gas allowance worth €35 per month – this may be paid either as a credit
on the recipient’s gas or electricity bill, or as a cash allowance (by electronic funds
transfer or through the recipient’s local post office).
2. Free Television Licence – the Department notifies An Post that the recipient is entitled
to a Free Television Licence as long as he or she remains entitled to the Household
Benefits Package.
To qualify for the Household Benefits package, a person must be:
Permanently resident in the State
Living in self-contained accommodation (not in a nursing home or retirement home)
The only person in a household claiming the Household Benefits package (and the
registered consumer of electricity or gas if claiming an Electricity or Gas Allowance).
Once these conditions are satisfied, the Household Benefits package is available to:
Everyone aged 70 or over (regardless of means or pension status)
People aged from 66 to 69 who are receiving one of a number of specified pensions
and allowances
3
, or who satisfy a means test
4
, and who are living alone or with
‘excepted people’
People aged under 66 who are receiving one of a number of qualifying working age
payments
5
and who are living alone or with ‘excepted people’.
Household benefits recipients and primary DSP schemes
The qualifying conditions for receipt of the Household Benefits package mean that a very large
majority of recipients also receive a weekly ‘primary’ payment from the Department:
Most people aged 66 or over receive either State Pension (contributory or non-
contributory), or a Widow(er)’s or surviving civil partner’s pension (contributory or
non-contributory).
The qualifying payments that enable people under 66 to access the Household
Benefits scheme are overwhelmingly illness, disability and carer supports from the
Department of Social Protection, principally Disability Allowance, Carer’s Allowance
and Invalidity Pension.
Overall, this means that some 93% of all Household Benefits package recipients are also
recipients of a weekly working age or pension payment from the Department of Social
Protection (see Table 3 below).
3
See http://www.welfare.ie/en/Pages/Household-Benefits.aspx
4
See http://www.welfare.ie/en/Pages/Means-Assessment.aspx for details
5
See http://www.welfare.ie/en/Pages/Household-Benefits.aspx
Table 3 - Household Benefits package recipients by primary scheme
Number of Recipients
Percent
179,510
43%
68,018
16%
54,826
13%
41,577
10%
26,362
6%
19,252
5%
31,299
7%
420,844
100%
(Source: Department of Social Protection)
Because of this, the Department’s plan in relation to assessing the Fraud and Error risk arising
from the Household Benefits Package scheme has two components:
Dedicated survey of Household Benefits package (only) recipients – a dedicated survey
is the only way to ensure that Fraud and Error risk is assessed for this subgroup which
have higher-risk profile that other recipients of the Household Benefits package.
Inclusion of Household Benefits entitlement reviews in other Fraud and Error reviews
– Fraud and Error reviews of a number of relevant schemes are planned over the next
three years. These will yield cumulative survey coverage of approximately 50% of all
Household Benefits package recipients by end 2016 and 80% by early 2019.
Table 4 - Fraud and Error surveys incorporating Household Benefits entitlement reviews
Scheme to be surveyed
Survey
commencement
Survey completion
Cumulative
coverage
Household Benefits (only)
Q1 2016
Q3 2016
7%
State Pension (contributory)
Q1 2016
Q4 2016
50%
Carer’s Allowance
Q1 2016
Q2 2017
56%
State Pension (non-contributory)
Q1 2017
Q1 2018
69%
Disability Allowance
Q4 2017
Q1 2019
79%
(Source: Department of Social Protection)
Age profile of Household Benefits package (only) recipients6
The age profile of the Household Benefits package (only) recipient population is summarised in
Figure 1. The high average age means that the availability of timely information in relation to
recipients who move to a nursing home or die is important to managing control risk for this
cohort
7
.
6
Strictly, some of these clients are also recipients of one of a group of other, smaller, DSP schemes.
7
Data matches are carried out with General Register Office (GRO), the central civil repository for records relating to births,
stillbirths, adoptions, marriages, civil partnerships and deaths in Ireland. Data matches are also undertaken with the HSE
(Fair Deal) as part of the control strategy for the scheme.
Figure 1 - Household Benefits package (only) recipients by age group
(Source: Department of Social Protection)
0
2,000
4,000
6,000
8,000
10,000
12,000
14,000
16,000
18,000
20-<30 30-<40 40-<50 50-<60 60-<70 70-<80 80-<90 90-<100 Over 100
Number of Recipients
Age group
3. Methodology
Irish Statistical System Code of Practice
The Statistics and Business Intelligence Unit of the Department, a part of the Irish Statistical
System, oversees the design, sample selection, analysis, and reporting of the Department’s fraud
and error surveys. The Chief Statistician ensures that these statistics are produced in an
objective, transparent and independent manner, in line with the requirements of the Irish
Statistical System Code of Practice
8
.
Survey principles
The Department, in agreement with the Comptroller & Auditor General (C&AG), applies the
following principles to the design and implementation of Fraud & Error surveys:
All cases for inclusion in the survey must be selected randomly from the population of
cases in payment at a specific time;
The sample size must be sufficiently large to yield reasonably reliable estimates;
The reviews should be carried out as promptly as possible;
Cases should be tested fully for all possible breaches of regulations;
The monetary values of any changes as a result of the review together with the
monetary value of the sample should be captured so that the results can be
extrapolated to draw conclusions about the estimated value of the loss; and
The results of the survey should be capable of being audited.
Sample selection
For the Household Benefits survey, the Statistics and Business Intelligence Unit matched the
Household Benefits recipient database against the recipient databases for the six primary
Department of Social Protection weekly payment schemes, resulting in matches for some 93% of
recipients (see Table 3 for details).
From the remaining cases, the Unit selected a stratified random sample of 500 Household
Benefits claims in payment at the start of November 2015 (‘the Fraud and Error sample’). It
verified that this sample was representative of the Household Benefits package (only) recipient
population profile.
This sample size was chosen in order to strike a right balance between the statistical power of
the survey and the resource-intensive nature of the inspections required.
Survey procedures
Each sample case was assigned to an Investigating Officer to conduct a full review of the claim.
Completed survey material and customer files were submitted for examination, assessment and
revised decision by a Deciding Officer in the Household Benefits Unit.
Categorisation of results
Fraud and Error rates were calculated based on the decisions of the Deciding Officer in each case
included in the survey sample.
Fraud or suspected fraud arises where it appears to the Deciding Officer that the
claimant knowingly gave false or misleading information or wilfully concealed relevant
information.
Error cases arise due to inadvertent customer, third party or official error.
These decisions were carefully reviewed by the Chief Statistician before arriving at a final
classification for each case.
8
http://www.isscop.ie/codeofpractice/
‘Gross’ and ‘net’ rates of Fraud and Error
Fraud and error rates may be quoted either as a percentage of total scheme expenditure (in the
week the survey was initiated) or as a percentage of the number of claims in payment on that
date.
The gross rate refers to the position after account is taken of decreases or increases in
monthly rate (including terminations of payment), and before transfers to other
Department of Social Protection schemes and the position post appeals of any cases
affected. (Cases with an unchanged weekly rate but where a historical overpayment is
identified only affect the number of claims rate.)
The net rate of Fraud and Error is the rate after taking account of transfers to other
Department of Social Protection payments and the post-appeals position of affected
cases.
Valuation of changes in benefit entitlement
The two components of the Household Benefits package have the following values:
1. Electricity or Gas allowance
− €35 per month
2. Free Television Licence
− A saving to the customer of the annual cost of a television licence (currently €160
per year)
There is no marginal cost or saving to the Department in relation to the addition or removal of a
Free Television Licence entitlement because payment for this benefit is made in the form of
monthly block grants from the Department and Social Insurance Fund to the Department of
Communications, Energy and Natural Resources.
Accordingly, we have valued the monthly cost of Fraud or Error for the purposes of this review at
€35 for the Electricity or Gas allowance, and zero for the Free Television Licence.
Cases of Fraud or Error where only a Free Television Licence entitlement was affected are still
included in the number of cases totals for this survey.
4. Results
Review outcomes were established for all 500 cases in the survey sample, and the results of
these reviews are detailed in Table 5 below.
Table 5 – Household Benefits Fraud and Error survey outcomes by type and number of cases
Outcome
No. of cases
% of total
Fraud & Error (net)
40
8.0%
Fraud & Error (gross)
45
9.0%
Total Customer Fraud & Error
9
1.8%
Customer Fraud
3
0.6%
Not resident in State
3
0.6%
Customer Error
6
1.2%
Disqualifying child(ren) residing with claimant
2
0.4%
Whereabouts not known
1
0.2%
Not resident in State
1
0.2%
Deceased - spouse now eligible
1
0.2%
Nursing home - spouse now eligible
1
0.2%
Official Error
36
7.2%
Overpayment
36
7.2%
Deceased
3
0.6%
Deceased - electricity account changed but TV licence not updated
3
0.6%
Deceased - primary payment stopped but HHB not updated
1
0.2%
Moved
2
0.4%
Moved - TV licence not updated
4
0.8%
Moved - electricity account changed but TV licence not updated
2
0.4%
Moved - primary payment stopped but HHB not updated
7
1.4%
Not resident in State
3
0.6%
Nursing home, single or widowed
7
1.4%
Other DSP claimant at same address
1
0.2%
Claim transferred to spouse
2
0.4%
Prison - HHB not updated
1
0.2%
Underpayment
0
0.0%
Customer Error: Spouse now awarded HHB
-2
-0.4%
Official Error: Spouse now awarded HHB
-3
-0.6%
No Fraud or Error (net)
460
92.0%
No Fraud or Error (gross)
455
91.0%
Benefit Correct
437
87.4%
Normal Movement
18
3.6%
Not resident in State
2
0.4%
Deceased
10
2.0%
Moved
1
0.2%
Nursing home
5
1.0%
Customer Error: Spouse now awarded HHB
2
0.4%
Official Error: Spouse now awarded HHB
3
0.6%
(Source: Department of Social Protection)
Normal Movement
The number of Normal Movement cases found by the survey was at 3.6% of all cases sampled.
These overwhelmingly related to people who had died, left the State, moved house or moved to
a Nursing Home, but where the benefit was correct before the change in circumstances
occurred.
Fraud and Error
The overall Fraud and Error rate observed for Household Benefits package (only) recipients is
shown both as a percentage of expenditure (Table 6 & 7) and as a percentage of the number of
claims affected (Table 8). 95% confidence intervals are provided for all tables, showing that
statistically significant results were obtained for all metrics except Customer Fraud. There is a
very low estimated level of intentional wrongful claiming of the Household Benefits package,
and conversely the large majority of Fraud and Error cases found were Official Error resulting
from the difficulty of ensuring timely changes to Household Benefits package entitlements as
mentioned earlier.
Table 6 – Gross and Net Fraud and Error by percentage of total monthly spending with 95% confidence intervals
9
Lower
bound
(gross)
Gross
Higher
bound
(gross)
Lower
bound
(net)
Net
Higher
bound
(net)
Fraud and Error (% of expenditure)
4.62%
6.40%
8.18%
3.79%
5.40%
7.01%
Customer Fraud and Error
0.63%
1.80%
2.97%
0.37%
1.40%
2.43%
Customer Fraud
-0.08%
0.60%
1.28%
-0.08%
0.60%
1.28%
Customer Error
0.25%
1.20%
2.15%
0.02%
0.80%
1.58%
Official Error
2.76%
4.60%
6.44%
2.28%
4.00%
5.72%
Overpayment
2.76%
4.60%
6.44%
2.28%
4.00%
5.72%
(Source: Department of Social Protection)
Table 7 – Gross and Net Fraud and Error by estimated total monthly overpayment with 95% confidence intervals
9
Lower
bound
(gross)
Gross
Higher
bound
(gross)
Lower
bound
(net)
Net
Higher
bound
(net)
Fraud and Error (monetary value)
€50,569
€70,110
€89,651
€41,572
€59,155
€76,739
Customer Fraud and Error
€6,952
€19,718
€32,484
€4,055
€15,337
€26,618
Customer Fraud
-€843
€6,573
€13,988
-€843
€6,573
€13,988
Customer Error
€2,690
€13,146
€23,601
€210
€8,764
€17,318
Official Error
€30,277
€50,391
€70,506
€25,003
€43,819
€62,635
Overpayment
€30,277
€50,391
€70,506
€25,003
€43,819
€62,635
(Source: Department of Social Protection)
Table 8 – Gross and Net Fraud and Error by percentage of claims affected with 95% confidence intervals
9
Lower
bound
(gross)
Gross
Higher
bound
(gross)
Lower
bound
(net)
Net
Higher
bound
(net)
Fraud and Error (number of claims)
6.39%
9.00%
11.61%
5.52%
8.00%
10.48%
Customer Fraud and Error
0.53%
1.80%
3.07%
0.27%
1.40%
2.53%
Customer Fraud
-0.18%
0.60%
1.38%
-0.18%
0.60%
1.38%
Customer Error
0.15%
1.20%
2.25%
-0.08%
0.80%
1.68%
Official Error
4.83%
7.20%
9.57%
4.32%
6.60%
8.88%
Overpayment
4.83%
7.20%
9.57%
4.32%
6.60%
8.88%
(Source: Department of Social Protection)
9
Negative values (marked in red) mean that we cannot claim to have identified with 95% confidence a population-wide
outcome that is different from the null hypothesis of a zero rate of fraud and error.
5. Risk analysis
General approach
The sample was divided into two groups: those with and without a recorded adverse outcome.
Chi-square tests were used to identify potential predictor variables for a Fraud or Error outcome
from the administrative data for the scheme, then logistic regression models were used to
determine strength of association with these candidate variables.
For this survey, we performed this analysis twice: once with Customer Fraud or Error (yes/no) as
the outcome variable of interest, and once with Official Error (yes/no) as the outcome variable.
No statistically significant predictor variables
The present survey was targeted at the 7% of Household Benefit (only) package claimants who
are not recipients of a primary Department of Social Protection benefit.
Recipients of the Household Benefit package are considered to be a low risk population relative
to other schemes funded and managed by the Department. This survey targeted what was
identified as the riskier sub-group of recipients of the Household Benefits scheme, due to their
greater remoteness from the Department compared to those who are also recipients of another
weekly social welfare payment.
6. Conclusions and recommendations
Conclusions
The survey covers 7% of all Household Benefits package recipients who are not in receipt of a
weekly Social Welfare payment. The survey estimated that the level of Fraud and Error for
recipients is 5.4% of total expenditure for people receiving the Household Benefit package (only)
within the Household Benefits scheme.
Three quarters of the overpayments found were categorised as Official Error. This arises mainly
from the difficulty of ensuring timely changes to Household Benefits package entitlements when
recipients move address, move into a nursing home, or die. This, in turn, reflects the greater
remoteness from the Department overall of the Household Benefits package (only) group
compared to the 93% of Household Benefits package recipients who also receive another weekly
social welfare payment.
As outlined earlier, the Department’s plan in relation to assessing the fraud and error risk for the
Household Benefits Package scheme has two components:
Dedicated survey of household Benefit package (only) recipients – a dedicated survey is
the only way to ensure that Fraud and Error risk is assessed for this subgroup which have
a higher risk profile than other recipients in the Household Benefits package;
Inclusion of household Benefits entitlement reviews in other Fraud and Error surveys.
Household Benefits package entitlements will be included with Fraud and Error surveys
of a number of large schemes over the coming years – State Pension (contributory),
Carer’s Allowance, State Pension (non-contributory) and Disability Allowance. Together
with the present survey, this will allow the Department to assess the Fraud and Error risk
for some 80% of all Household Benefits package recipients.
A number of control actions have been specifically identified as part of the control strategy to
reduce the Fraud and Error risk of both the Household Benefits package (only) group and for
Household Benefits package recipients in general. A key element of this strategy targets the
Household Benefits package (only) group to improve assurances around their continuing
eligibility and the identification of changed circumstances via data matches and random reviews.
Additionally, the Department is in the process of implementing enhanced automated
notifications from internal and external systems to update the Household Benefits package
scheme. These system changes are expected to be completed by the end of 2016.
END