For years, property and casualty carriers throughout the country have been honing their skills in the fight against fraud. From the gumshoe to the data analyst, carriers have evolved their special investigation units to maximize the tools available while, at the same time, controlling costs and, more importantly, bad faith claims.
Although this is to be applauded, it is necessary to understand that the industry is fighting a battle in which the enemy is well equipped and every weapon available must be engaged. The battle is not for coveted real estate or tactical position, but rather the billions of dollars that are being paid to misguided, dare we say, fraudulent individuals who have learned that insurance fraud is, for the most part, an easy-to-commit and often-unpunished crime.
In reality, as costly as insurance fraud has become, five states have yet to pass insurance fraud statutes and many others find it too costly to pursue. The Coalition Against Insurance Fraud has estimated that spending on fraud prevention in 2003 was nearly $1 billion. According to a McKinsey & Co. survey, the losses due to fraud likely cost property and casualty insurers $40 billion, rather than the $30 billion previously cited. Also noted by McKinsey was the fact that soft fraud seems to have stabilized, while hard fraud continues to increase. Anti-fraud efforts have not kept up with the level of sophistication and frequency of this hard fraud.
Aggressive Action
Although these statistics might be reason for even the most profitable insurance carrier executives to jump from their office windows, there is some hope. The industry is learning that a good defense is a great offense that identifies fraudulent traits at every point in the life of a claim or policy.
At a basic level, these actions work their magic by identifying those characteristics and traits that are most commonly found in the profiles of fraud perpetrators, score the traits, and, much like a screen door, work to keep the pests out. Some measures are home-grown; some require sophisticated technology to mine data, as well as a host of models that are aimed at understanding and identifying fraudulent insureds, claimants, or providers. At a more sophisticated level, so-called neural solutions can be integrated into the policy and claim processes that all but litigate the claim, using a carrier's own experience to derive a model.
Regardless of the process used, the sheer volume of claims with fraudulent indicators leaves many claim executives frustrated, depressed, and in search of additional means to deal with those claims.
External Data
Credit grantors manage their portfolios with models that use not only experience data, such as consumer credit, but external information, as well. Public records, proprietary data, and publicly available information has helped to further segment risks associated with the granting of credit, resulting in the most sophisticated models used today.
Likewise, the effort to stem insurance fraud would be well served with the additional use of external data alongside historical experience. The identification of tendencies that lead to motives to commit fraud is a key element in the detection of fraud at any point. Bankruptcies, liens, and judgments, as well as UCC filings, frequency of change of address, and other data elements, all can be modeled against known fraud artists.
The identification of tendencies that indicate opportunity to commit fraud is another key weapon in the battle. Verifi-cation of professional licensing, fictitious businesses, identity theft, and criminal and civil records can be used in the scoring model along with historical experiences.
Although customized programs offered by software vendors might be costly and somewhat difficult to integrate, those that use external data can be more affordable to the average carrier. Vendor-hosted programs are an inexpensive alternative to custom models and can provide results without integration issues and the costs associated with them.
Vendors that are experienced in neural networking have begun to use historical carrier data to develop models that provide admirable results, as well. Super computers are being progammed with the necessary historical information that, when combined with public records, can better identify fraudulent claims, policies, or service providers early in the relationship.
Collaboration
Despite all this technology and information, a growing problem remains. At a recent CAIF meeting, the 35 insurance organizations represented heard that insurance carriers, as an industry, do not have long-term incentives to invest heavily in anti-fraud strategies.
McKinsey's research also determined that a more collaborative effort is needed to reduce the impact of fraud and that an optimum fraud-fighting model for insurers needs to be developed and adopted widely. This collaboration must go beyond insurance carriers and extend to their partners. Information providers, modeling companies, and other vendors must work with willing insurance carriers in a combined effort to educate senior managers who still do not see a viable return on investment in their anti-fraud efforts.
The benefits of collaboration can be profound. A recent article in the Wall Street Journal ("Insurers Employ Voice-Analysis Software to Detect Fraud," May 17, 2004) reported that one of Britain's largest insurers, Halifax General Insurance, is using lie-detection software to catch policyholders who invent or exaggerate claims when calling to report thefts or accidents. The insurer's three-month test of the software resulted in 12 percent of the claims' being withdrawn, rejected, or referred for fraud investigation.
Although this technology is likely to cause consumer concern here in the United States, other solutions that are used at first notice of loss or at the time of policy application hold similar promise. The good news is that fraud detection and review have garnered much attention from both the industry and its clients. The ability to systematically detect fraudulent traits from the time of inception and mine data to review for fraud at any time during the policy life cycle will improve loss ratios, create more satisfied clients, and maybe keep a few executives from leaping.
Seth Perlmutter is vice president of Insurance Services, LexisNexis.
© Arc, All Rights Reserved. Request academic re-use from www.copyright.com. All other uses, submit a request to TMSalesOperations@arc-network.com. For more information visit Asset & Logo Licensing.