The typical profile of a perpetrator of fraud has been that of trained con artist who plans his or her scheme in a calculated fashion. But a new study depicts insurance fraud not only as a sophisticated ring carried out by professionals but as an act often executed by mainstream insurance customers.The study, by Bermuda-based consulting and technology solutions provider Accenture, found that nearly one in four U.S. adults say that overstating the value of claims to insurance companies is acceptable, and more than one in 10 say they approve of submitting insurance claims for items that were never lost or damaged or for treatments that were not provided.
The survey, based on a random sample of more than 1,000 U.S. adults, examined consumer attitudes toward insurance fraud and found that two-thirds of respondents (66%) said that people are more likely to commit insurance fraud during an economic downturn than when the economy is strong. About half the respondents (49%) said that people commit insurance fraud because they can get away with it.
"Fraud is a growing concern for insurers, whose aging technology and inefficient processes often prevent them from detecting fraudulent claims, which in turn hurts their long-term profitability," says Michael Lucarini, a partner in Accenture's Insurance practice. "In addition, increased consumer exaggeration to improve the claims payout is becoming more prevalent in the current weakened economy."
Smoking gun
Accenture's study also found 30% of respondents linked insurance fraud to the offenders' needs for money, while nearly one-fourth (24%) said they believe that the people who commit insurance fraud do so because they believe they pay too much for insurance. Twenty percent said they believe that the offenders commit fraud to compensate for high claims deductibles.
"The burden is clearly on insurance companies to ensure they have the proper tools, technologies and skills to combat fraud," Lucarini says. "Those committing fraud are becoming more sophisticated and advanced in their methods, while many insurance companies still lack the necessary processes and systems to detect and stop fraud."
Some varieties of fraud are easier to detect than others. In fact, it might be the most mundane cases that will be difficult to corroborate. A percentage of those responding to Accenture's survey indicated that fraud was defined as "overstating the value of claims" and as "submitting insurance claims for items that were not lost or damaged."
Without a smoking gun, these incidents of fraud are harder to detect, short of submitting to a lie detector test.
To combat these hard-to-detect fraud cases, leading insurers are implementing Web-based technology that can help re-evaluate claims for fraud during the claims lifecycle and alert the appropriate people when thresholds are exceeded, Lucarini explains.
The industry, for instance, has begun to mine data with neural networks for early identification of fraud patterns. The use of neural network technology-for both simple and complex fraud-can enable an insurer to take a good amount of industry data and train the networks to identify exceptions to rules.
San Rafael, Calif.-based Fair, Isaac & Co., for instance, offers its Claims Advisor for Exception Management, a neural network-driven technology that consists of 50 to 60 variables that interact at any given time. Using a combination of data-mining and advanced analytics, the software scores claims on a scale from 0 to 1,000. The higher the score, the more likely fraud has occurred. Claim scores are posted on an intranet site for claims adjusters to access.
