Countering Fraud in Social Benefit Programmes

2. Fraud prevention measures in Social Benefit Programmes

2.1. Adopting a holistic approach to prevention

Typically, governments can prevent external fraud with a focus on (a) recipients of social security assistance and (b) private providers of goods and services, who are entitled to receive subsidies. The following focus areas for preventing external fraud are key:

• Strengthening strategies, goals and objectives for combating fraud – This includes public organisations incorporating anti-fraud prevention into their strategies, objectives and procedures, and making sure that they strike the right balance between prevention, detection and prosecution measures.

• Targeting prevention measures at the registration phase – To minimise the chance of fraudulent tactics succeeding at this high-risk phase, governments can ensure that they put in place adequate policies, controls and measures to verify identities and data submitted during the registration process.

• Tailoring communication campaigns and messaging to improve fraud deterrence – By integrating behavioural perspectives into their prevention approaches, public organisations can develop nuanced communication campaigns that include a range of messages to deter fraud in SBPs, for example by including soft messages and reminders, as well as outlining the penalties for committing fraud.

• Focusing on the highest risks – Risk management and assessments can contribute to savings and promote efficiency by targeting the application of preventive controls, as well as identifying areas that are susceptible to false claimants and fake registrations.

2.2. Strengthening strategies, goals and objectives for combating fraud

Fraud prevention measures in SBPs need to be supported by a strategic vision for how public organisations approach the problem, as well as objectives for preventing and detecting fraud. To ensure a harmonised, whole-of-government approach, the Centre of Government (CoG) may have in place policies and standards that require public organisations to combat fraud, while entities themselves can ensure that they incorporate fraud prevention into their strategies, objectives and procedures. A lack of clarity from the central level about how to implement and undertake fraud prevention measures can lead to a perception that fraud prevention objectives, and the activities that support them, are separate from other strategic and operational objectives. The CoG, as well as other bodies with government-wide responsibilities, can play a critical role in helping public organisations to overcome this challenge by providing unified standards, policies and guidance. Anti-fraud strategies or initiatives set expectations on how the government and individual entities will deal with fraud. In France, for example, the National Committee for the Fight against Fraud (Le Comité national de lutte contre la fraude, CNLF) co-ordinates measures to counter different types of fraud across public organisations and state agencies. The CNLF devises multiannual national plans to combat fraud, with fraud in SBPs being one of its focus areas (Le portail de l'Économie, des Finances, de l'Action et des Comptes publics, 2016[10]). These national plans include guidance on implementing control measures, strengthening fraud risk management, and enhancing the use of data.

Despite the need for unified standards to guide fraud prevention and detection measures, the nature of SBPs requires public organisations to tailor their strategies or policies. For example, a programme aimed at providing subsidies for housing involves different processes to those that deal with unemployment benefits or support for children. As such, SBPs may be characterised by a range of fraud risks, requiring alternative measures or responses to mitigate them. In Korea, the government has responded to fraud in SBPs, and specifically those using false identities or making fake claims, by establishing the Welfare and Subsidy Fraud Reporting Center. The body is responsible for collecting and analysing fraud reports from public organisations responsible for SBPs. Its activities have resulted in savings for the government, as well as the recovery of funds obtained fraudulently within SBPs.

As part of their strategic orientation, a critical consideration for governments in the area of SBP delivery is the proportionality of anti-fraud measures to the likely impact of the risks materialising. Fraud prevention measures can affect a large number of claimants. As public organisations update and revise their anti-fraud strategies, they can consider the implications of their approach as well as the impact that new tools and technologies for fraud prevention and detection may have on society. Practically, this means taking into account the impact of strategies and detection mechanisms that cast a wide net on the public, and that focus heavily on prosecutions. It also means investing in the skills, methodologies and technology to distinguish between cases of fraud and error to ensure that well-meaning individuals receive the correct benefits. Box ‎2.1 provides examples of anti-fraud strategies and initiatives that seek to advance proportional approaches, and balance fraud prevention, detection and prosecution measures appropriately.

Box ‎2.1. Striking the right balance when fighting fraud in social benefit programmes

Since 2014, Ireland’s Department of Employment Affairs and Social Protection (DEASP) has been fine-tuning its approach to fraud prevention in social benefits programmes (SBP). The Department’s updated Compliance and Anti-Fraud strategy for 2019-23 acknowledges that the vast majority of claimants are genuine and that there are no issues concerning their claims. As such, the department’s strategy includes:

• continuing to target serious fraud through the work of the department’s Special Investigations Unit and seconded Gardaí

• investing in predictive analytics technology to improve capacity to detect non-compliant cases

• undertaking control surveys of various schemes, including a new continuous surveying approach for some of the department’s larger schemes

• publicising the department’s hotline and encouraging members of the public to report cases of suspected fraud

• working collaboratively with other departments and entities and with cross-border and international organisations to prevent fraud and non-compliance.

Source: (Irish Government, 2019[11]).

2.3. Targeting prevention measures at the registration phase

In SBPs, registration is the first point of entry for potential claimants where governments seek to ensure that individuals receive the correct entitlements. However, this is a critical juncture for fraud prevention measures as fraudsters often target the registration process, using false data to create fake claims and identities. Furthermore, if governments fail to detect fake claimants at this stage, the perpetrators are able to continue defrauding SBPs indefinitely. A number of OECD countries experience these fraudulent tactics during the SBP registration process. For example, in 2019, French and Romanian authorities arrested several individuals charged with fraudulently obtaining welfare benefits. The perpetrators initially sought out women who claimed they were self-employed in scrap-metal recovery and convinced them to declare that they were pregnant to claim parental benefits. None of the women were in fact pregnant at the time, but the criminal group paid them to make fraudulent claims. Eventually, the perpetrators began creating fake identity cards and birth certificates of the women. In total, the syndicate registered 1 200 false identities, and in 2017 alone, collected EUR 1.7 million in benefits. Authorities found documents suggesting that the perpetrators ran similar schemes in other parts of France and Germany (Constant, 2019[12]).

This example highlights the importance of targeting prevention measures and controls at the claimant registration phase. This phase may be the only time that face-to-face interactions occur between officials and claimants, which provides an opportunity to detect identity theft. Once claimants are registered, the government may face more challenges to detect false identities for claimants with distinct names, birthdates, and addresses. Doing so often requires the use of techniques such as data-matching and data‑mining, which can be resource-intensive and require special skills and expertise. To minimise the chance of fraudulent tactics succeeding at this phase, governments could ensure that they put in place adequate policies, controls and measures to verify identities and data submitted during the registration process.

At the policy level, effective fraud prevention relies on certain preconditions, such as legal and policy frameworks that facilitate data sharing between public organisations as well as with the private sector. Such frameworks are critical for effective and efficient checks of claimant information, and ensuring that public organisations respect privacy laws. Data-sharing agreements that support real-time checks can also help to reduce the burden of control on well-meaning individuals. In addition, technology can support anti-fraud measures during the registration phase. For instance, governments are employing biometrics (e.g. scanning fingers, hands, iris or face) to authenticate identities. This can serve as a control against repeated registration attempts, and help to address the challenges related to creating unique identifiers for claimants. For example, SAFE registration is the process used by Ireland’s Department of Employment Affairs and Social Protection (DEASP) to establish and verify a person’s identity. Completion of registration to SAFE Level 2 (i.e. substantial level of assurance of identity) is the minimum requirement for a Public Services card (PSC) to be issued. Since 2013, the DEASP has been using facial matching software to strengthen the identity authentication process by detecting and deterring duplicate registration attempts. The biometric processing of the photograph produces an arithmetic template which allows precise comparison of the photograph in question with others held by the Department. This biometric processing is performed by the Department and the arithmetic template produced is not shared with any other specified body nor is it stored on the PSC. The purpose of the PSC is to enable individuals to gain access to public services more efficiently. Furthermore, awareness and vigilance is critical. Workshops and briefings for officials on the possibility of identity fraud, as well as educational campaigns targeted at the public, can help to increase awareness and improve detection.

2.4. Tailoring communication campaigns and messaging to improve fraud deterrence

In order to develop appropriate measures and responses to fraud in SBPs, governments can benefit from understanding how people behave in different contexts and in response to certain incentives. Traditional attempts to tackle fraud in SBPs have included communication campaigns that focus heavily on penalties, and while this has benefits, it is not wholly effective as a deterrent. Recent studies have shown that messages focusing solely on penalties may provoke a negative emotional response that leads to them being poorly received or ignored by the very claimants they are intended to reach (Lloyd and Wilson, 2019[13]).Such approaches tend to rely on the notion that individuals make rational cost-benefit analyses before committing fraud, when in reality such decision-making processes are not so straightforward. For example, studies have shown that in some cases, individuals commit fraud because they perceive the correct procedure as overly cumbersome (i.e. lengthy application forms and checks). As a result, they choose what appears to be the easier option and bypass a certain procedure, inadvertently committing fraud (Cabinet Office Behavioural Insights Team, 2015[14]).

In recent years, a number of governments have invested in applying behavioural insights in their tax administrations to increase compliance. Public organisations can draw lessons from these experiences to inform their fraud deterrence interventions in SBPs. For example, a recent study found that deterrence methods, in this case, letters using either harsh or soft language, had a positive impact on compliance when combined with other insights such as reinforcing social norms (Irish Government Economic and Evaluation Service, 2017[15]). These findings have implications for how public organisations may consider conveying anti-fraud messages in the delivery of SBPs. For example, communication campaigns can use a combination of ‘harsh’ and ‘soft’ messages, i.e. stating penalties and consequences, while reinforcing messages around social norms. This could include reminding individuals that the majority of claimants choose not to commit fraud, or combining reminders regarding changes of circumstances with messages about the potential penalties for fraud (See Box ‎2.2).

Box ‎2.2. A nuanced communication campaign to deter fraud in employment insurance benefits

Service Canada is responsible for delivering the government’s Employment Insurance (EI) programme, which provides temporary financial assistance to those who are unemployed or unable to work due to health reasons, as well as social assistance to other groups. For the EI programme, Service Canada has devised a communication campaign that utilises a mixture of harsh and soft messages to deter fraud. For example, the campaign includes a section on protecting the EI programme from fraud that reminds claimants that they also play a role in protecting the programme from fraud. Inclusive language is used, for example, ‘We are all responsible for helping to detect and deter EI fraud’. Framing the issue of fraud prevention as a shared responsibility, with a focus on maintaining the integrity of the EI programme, demonstrates a useful example of positive framing to connect with the audience.

Furthermore, the campaign includes messages around error, making it clear to claimants that Service Canada is aware that mistakes may lead to unintended omissions or mistakes. It instructs the claimant to contact the organisation if this is the case. Service Canada combines these softer messages with reminders about the consequences and penalties for committing fraud.

Source: (Service Canada, 2020[16]).

By developing a deeper understanding of what motivates perpetrators and when individuals choose to engage in fraudulent behaviour, public organisations can design effective communication campaigns to deter fraud in SBPs. To achieve this, public organisations may consider testing the efficacy of different fraud prevention messages. This can be done using focus groups, in-depth interviews, and testing different messages in different geographical locations (i.e. local councils).

2.5. Focusing on the highest risks

Risk management consists of the policies, practices and tools that help an organisation to identify and mitigate risks, but more generally, facilitate informed decision-making to advance objectives. In the context of SBP and fraud, risk management is a critical pillar of prevention. Mature risk assessments incorporate both qualitative and quantitative methodologies. They help to identify and understand the effect and likelihood of risks, and then prioritise ways to spend resources on to address control weaknesses before fraud occurs. For instance, in France, the entity that provides welfare benefits, the French Social Security system’s Family Branch (Caisse d’allocations familiales, or CAF) developed an analytical model to perform fraud risk assessments, the results of which are used to direct in-person visits and inspections towards high-risk claimants or areas (OECD, 2019[17]). Moreover, risk assessments can help managers to decide not only when to act, but when controls are sufficient for reducing risk relative to a pre-determined criteria (i.e. within risk tolerances). These criteria can include factors that are relevant for service delivery, such as the amount of time it takes to process a registration (i.e. the investment in controls) versus the likelihood and effect of a fraud risk occurring.

Box ‎2.3. Targeting verifications and controls at high-risk claimants

In France, the Social Security system’s Family Branch (Caisse d’allocations familiales, or CAF) and other social security institutions use data mining and predictive modelling to determine which beneficiaries may be at risk of committing fraud. Data mining was rolled out experimentally in 2011 and implemented in all regional branches of CAF in 2012. This technique targets additional verifications based on risk and permits claimants with a low risk score to submit less documentary evidence than those considered high risk. The predictive models allow CAF to identify cases with similar characteristics to those already identified as fraudulent.

Applications are double-checked internally to make sure that the information reported is consistent and matches the documentary evidence supplied. Furthermore, CAF checks the validity of administrative documents with the issuers (banks, internet and telephone access providers, utility companies, etc.). After review and verification of claimants’ information, which is facilitated by automated exchange of information with the tax authority and a shared national social protection register, CAF targets additional verifications towards claimants with a high-risk score. For applications with the highest risk, this involves sending certified inspectors to the homes of claimants to conduct inspections and face-to-face interviews in order to determine the veracity of their claim.

Source: (Caisse d'allocations familiales, 2015[18]) (La Revue du Digital, 2018[19]).

To understand the impact of fraud risks on the entity’s reputation, finances, and operations within SBPs, there are a number of high-risk areas that governments can focus on in addition to the registration phase, as described above. A high risk for SBPs is the non-declaration of additional income, often received in cash form by a recipient of benefits. Declaration of such additional income would likely reduce the recipient’s benefits. In many cases, individuals may choose to omit this declaration because they do not want their benefits to decrease. However, there are a number of reasons a recipient may avoid declaring additional income. These include:

• The amount of additional income received is small, and the recipient deems it as insignificant.

• Guidelines or information about additional income are unclear or not readily available.

• The job or additional work they are undertaking may be temporary.

• Making a declaration or re-registering may require an in-person visit to the welfare office.

Declaring additional income can be a significant inconvenience, particularly if multiple trips to a welfare office are required. In the case of temporary work, the recipient may lose their right to the benefits entirely and they may be required to re-register multiple times. To deter individuals from engaging in SBP fraud in this area, governments can ensure that the process of declaring additional income is not burdensome for recipients. This could include the option of reporting additional income via email, phone or directly to a web-based personal account. Repeat registrants could be fast-tracked to facilitate the process of re-registering (Tunley, 2010[20]). This might involve a ‘jump the queue’ process, for example.