Technology Helping in Battle Against Fraud

Crime and detection normally play a strategic game of one-upmanship—as the good guys get smart to the bad guys' methods, and vice versa—but technology offers the insurance industry a distinct advantage in fighting fraud, and insurance companies are increasingly appreciating the benefits of investing in that technology.

Technology solutions on the market today can result in exponential improvements over manual methods used in the industry today, according to David Rioux, corporate security and investigative services manager, Erie Insurance Group (Erie, PA). "Small incremental changes in the fraud identification and investigation process can have a large impact on a company's fraud mitigation rate," he says. "Anti-fraud technology provides benefits beyond the hard dollars in loss impact that can be realized from electronically enhanced detection and investigation efforts."

According to the results of Insurance Services Office, Inc. (ISO Jersey City, NJ), and the Insurance Research Council's (Malvern, PA) 2002 study "Fighting Insurance Fraud: Survey of Anti-Fraud Efforts," insurers are investing more in fraud-fighting efforts in general. Of those insurers able to track their anti-fraud spending, 53 percent were spending more than they were three years ago, 44 percent were spending about the same and only four percent were spending less, the survey found. Of those spending more, 37 percent were spending 15 percent more, 14 percent were spending from 10 to 15 percent more, 14 percent were spending from five to 10 percent more, and 19 percent were spending from zero to five percent more. The remaining 16 percent of those companies spending more did not report how much that spending had increased.

"On balance, it would seem that insurers as a group are spending more to fight fraud and that some are spending significantly more," says Richard P. Boehning, senior vice president, ISO. While the survey's findings pertain to total fraud-fighting expenditures, "it would seem that a substantial portion of the increased spending was for fraud-detection technology," Boehning adds.

Through a combination of traditional investigative techniques and technology-based approaches, the industry is making good progress in combating flagrant-or "hard"-fraud, in the workers' compensation and private passenger auto product lines, where such fraud is most prevalent, according to John Lucker, senior manager, Deloitte & Touche Advanced Quantitative Services (Hartford). However, less progress has been made in fighting "soft" or opportunistic fraud-e.g., fraud involving exaggeration of legitimate claims-despite widely accepted estimates that such fraud accounts for 15 to 25 percent of every claim dollar, Lucker asserts. "The biggest impediment to the fight against soft fraud is that the investigative costs for each suspected incident are usually greater-and sometimes much greater-than the cost of soft fraud itself," he says.

Nevertheless, predictive modeling technology-such as that used by underwriters to evaluate risk -shows great potential for fighting soft fraud, and techniques employing this technology are gaining traction in the industry, according to Lucker. "By prospectively evaluating an insurer's book of risks, predictive-modeling segmentation techniques allow insurers to identify those risks which are expected to perform worse than average," he says. An extension of this logic will show that worse-than-average risks are also more likely to experience a higher-than-average incidence of soft fraud, Lucker adds. "Based on this investigative resource triage technique, insurers can build operational processes to better review and/or challenge certain claims arising from policies identified as worse-than-average."

For organizations looking to beef-up their investment in anti-fraud systems, the following technologies can help insurers gain ground in the battle:

Predictive Modeling: Uses key data elements in a loss report within the claim system for review of characteristics that fit historical patterns of fraud. Claims are flagged and reviewed more closely for potential fraud and investigations are initiated as needed.

Visual Link Analysis: A software tool is used for analysis across multiple sources of data, both internal and external. Proactive and reactive analysis is done to explore trends, patterns, correlations and relationships, both direct and indirect, in claims and claims-related data. These patterns and relationships emerge from the data and are presented in a graphical representation.

Similarity Search: Sometimes referred to as "fuzzy logic" searching, embedded within the Predictive Modeling and Visual Link Analysis components. It identifies similarities in data to more accurately identify matches. For example, it helps to identify small variations of names and addresses that may very well be the same. It also allows for lookups against industry and company intelligence information.

(Editor's Note: More information about technology's role in fighting insurance fraud will appear in the November issue of Insurance & Technology.)

Anthony O'Donnell has covered technology in the insurance industry since 2000, when he joined the editorial staff of Insurance & Technology. As an editor and reporter for I&T and the InformationWeek Financial Services of TechWeb he has written on all areas of information ... View Full Bio