What does our survey tell us about claims fraud globally and what might we do to improve the detection and mitigation of fraud:
- 3-4% of all claims are fraudulent; or put another way, about 1 in 30 claims is identified as fraudulent
- Where underwriting controls appear strongest - North America - the incidence of claims fraud is reduced
- 37% of respondents never allege fraud, relying solely on the misrepresentation to deny the claim
- Of the fraud identified less than 2% resulted in a fraud prosecution
- Fraud investigation can extend the end to end time for consumers by up to 8 times the normal processing time
- Use of machine learning and in-force analysis as fraud detection tools are becoming common in regions with high incidence of fraud
- Respondents are split about future prospects for dealing with fraud; some see an improving picture driven by the use of innovation such as machine learning and in-force analysis, while others predict a deteriorating picture with data privacy and regulation limiting insurers’ ability to investigate fraud.
The 27 responding insurers represented the following regions: The Americas – North America and Latin America (19%); EMEA - Europe, Middle East and Africa (37%); Asia Pacific – Asia including Australia and New Zealand) (44%):
Incidence of Fraud
Survey results suggests that the global incidence of claims fraud is 3.58%, with high
claims fraud incidence in the Asia Pacific region. The Americas has the lowest
incidence of claims fraud at 1.47%, less than half the global average.
It is interesting to note the inversion from claims in terms of percentage, with higher proportions of underwriting frauds identified in the Americas. The average underwriting fraud percentage globally is 1.38% as reported by survey respondents.
RGA noted that the incidence of fraud is stable with little change from prior years. In the Americas, none of the respondents reported an increase in claims fraud activity (see Figure 4 below).
Fraud Assisted by Insurance Professionals
Nearly half of the companies (48%) surveyed have identified incidences of fraud assisted by agents, but overall the incidence of agent-assisted fraud is low, at less than 5% of identified fraud cases. Some insurers, however, reported a significantly different picture, with as much as 50% of their total fraud being assisted by agents.
Fraud assisted from within the life insurance companies is low, with only 7% of respondents indicating they have identified any worker-assisted fraud. Unsurprisingly, this is in the organized fraud category.
Spread of Fraud by Benefit Type
Perhaps surprisingly, mortality benefits were identified as most prone to fraud, with the highest incidence (60%). This may be related to the overall volume of mortality benefits written and the higher sums assured available. Health benefits were reported as the next most prone for fraudulent activity at 24%, followed by living benefits comprising just 10% of the overall claims fraud identified. It is important to note that this relates to all fraud, including underwriting fraud.
The Americas region has the lowest level of identified claims fraud but the highest level of identified underwriting fraud. This inversion might suggest that tighter underwriting fraud controls would give rise to better claims outcomes globally. This may also explain why The Americas is the only region not seeing evidence of increasing claims fraud.
Agent-assisted fraud continues to be a problem for some insurers but is in general a low level problem, with only around 5% of total fraud cases identified. Survey results revealed no real indication that fraud is assisted by other insurance professionals.