To the old maxim that nothing is sure but death and taxes,insurance companies might justifiably add another category: fraud.But while it is true that fraud always will be with us, insurersare making a serious mistake if they treat the problem as apredictable cost of doing business to be baked into theircompanies' economic assumptions. In this respect, fraud is morelike taxes than death; while undeniably inevitable, there may bemore or less of it depending on what one does about it.Unfortunately, too many insurers are insufficiently focused onfighting fraud, resulting in unnecessary costs to themselves, tothe industry at large, and to consumers who are forced to payhigher premiums.

World of Outlaws

Far from being a steady-state economic consideration, fraud is agrowing problem. Analysts have cited a 63 percent increase in fraudbetween 2001 and 2005. According to the Coalition Against InsuranceFraud, the total cost of fraud in the U.S. is approaching $100billion annually — $25.8 billion of which occurs in theproperty/casualty industry. This gratuitous cost factor isoccur-ring at a time when growth in direct premium written in theP&C industry is slowing and profitability has been hurt bysuccessive record hurricane seasons in 2004 and 2005.

The one thing that does stay the same in the world of fraud isnot the incidence of the crime but the determination of criminalsto stay ahead of the game. Much like computer hackers, professionalfraudsters keep abreast of fraud-detection technologies,techniques, and procedures and they often count on insideinformation about how insurance companies identify and escalatesuspected cases of fraud. Insurers simply are not keeping up.According to Gartner Research, deployment of insurancefraud-detection technologies will fail to keep pace withopportunistic perpetrators through at least 2007.

As the 2006 hurricane season reaches its peak months, insurersshould reflect not only upon the overall impact of storm-relatedcatastrophes on the bottom line, but also on the opportunity theseevents create for fraud perpetrators. The sheer scale ofhurricane-related losses not only increases the difficulty foradjusters to handle claims, it also magnifies the sympathy factorfor putative victims. Professional fraudsters are adept atexploiting all of these factors and the greater the catastrophe,the greater the opportunity. The overall fraud impact of Katrina isyet to be officially determined, but the Government AccountabilityOffice estimates that 16 percent of FEMA payouts on 2.5 millionhurricane relief claims were for improper or fraudulent claims.

Natural catastrophes also underscore the need for insurers tofight fraud as competently and fairly as today's technologies andtechniques permit. False positive identifications of fraud alwaysput insurers in danger of bad-faith judgments. The perceivedpersecution of the victims of tragedy can result in a publicrelations debacle. Even lesser failures endanger customerretention.

Round Up Your Posse

State-of-the-art fraud fighting begins with a strategiccommitment to the endeavor. That means treating fraud as a specificproblem and dedicating resources toward a well-defined future goal.It means re-engineering internal fraud-related processes, includingrestructuring incentives for adjusters and investigators to rewardsuccess, not volume of work. And it means investing in the righttechnology.

The trend of hiring former law enforcement officers to staffinsurers' special investigative units (SIUs) also is a step in theright direction. These professionals bring indispensable experiencein conducting in-depth evaluations, surveillance, andinvestigations to the companies that employ them. But investigatorsonly can be spread so thin; as with most activities in theinformation-intensive insurance industry, automation isindispensable for fraud detection. Failure to put the right toolsin investigators' hands will only limit their successes.

Chief among those tools are rules-based or “smart” systemscapable of identifying potential instances of fraud at any time inthe claim cycle. Anti-fraud measures ideally are built withconfigurable business rules that can be established without the ITdepartment's intervention. This permits the monitoring of resultsand iterative modification of rules to refine procedures towardde-sired outcomes and accommodate differential handling ofcustomers when needed.

Event-based technologies — used successfully in marketinginsurance products to existing customers based upon life events,such as the birth of a child or reaching retirement age — cantrigger escalation of claims for review whenever the appearance (orchange) of significant data points might indicate the possibilityof fraud. SIUs can then monitor the flagged claim to gauge whetherit is following a course toward becoming an exaggerated orfraudulent loss. Early flagging of suspect claims not only eases anSIU's workload, it also helps avoid redundant processes andanticipate future measures, such as addressing specific questionsto claimants when key data elements converge in the case history.In short, companies will continue to depend on the individualskills of SIU adjusters and investigators to execute plans to provecases of fraud, but technology can deliver the right claims to themat the right times.

Predictive modeling can play a role in casting scenarioscorresponding to flagged data patterns, among other applications,but past use counsels some caution. Analyses of cases to predictsettlement costs can expose carriers to bad-faith judgments whenthe individual facts of the claim buck tendencies. Thus, a claimidentified by a few stark criteria that typically settles for$30,000 can lead a carrier to a jury award of $1.2 million. Bearingin mind these possibilities, insurers should guide theirdeterminations with the idea that every claim is different andemploy technology to reflect those differences as much as possible.The key is populating the relevant databases with a granularity ofdata sufficient to make significant distinctions betweenclaims.

Available data is, of course, the foundation of anytechnologically driven fraud-detection effort. It presents perhapsthe greatest challenge and the greatest opportunity to gain groundin the battle against fraud. Quality of internal data is a concernwith anti-fraud efforts, as with other information-processing taskswithin complex insurance organizations. But the most urgent needfor insurers is a greater range of external sources of data.Addressing that need will require access to or development ofpublic sources of information, such as law enforcement databases,and an unprecedented degree of information sharing about past claimactivity on the part of insurance carriers.

Unfortunately, many insurers are reluctant to reform theirinternal processes and technologies, let alone work toward a modelof vastly improved industry-wide cooperation on fraud. Somecompanies lack the underlying core systems that provide thecapability necessary to build the superstructure of a rules-based,fraud-fighting technology platform. But even the carriers who enjoysuch capabilities often balk at making the investment. The reasonis that it is hard to make the business case for fraudre-engineering and technology because return on investment is hardto calculate. Everyone knows fraud exists, but it is impossible toknow what it is costing any individual company. Furthermore,industry analysts have no way to judge the relative performance ofcompanies. The only meaningful metric is the number of successfulhits relative to referrals to the SIU, along with the resultingrecoveries. But no one knows how much is missed and how much thatcosts.

Additionally, even though awareness of newer technologies iswidespread, many companies fear the difficulties of integratingthem within their current technology environments, which arealready complex in terms of the application architecture itself, aswell as with regard to the multiplicity of integration points.

Marshaling Resources

Nevertheless, the gap between the current state of fraudfighting at insurance companies and the existing but underutilizedtechnologies and techniques demonstrates clear potential not onlyfor improving a company's bottom line, but also by benefiting endcustomers through reduced premiums. The potential benefitsassociated with greater organization and distribution ofcross-company and other data not currently leveraged only makes thecase stronger.

Companies that lack underlying core systems should consider thefact that improved anti-fraud capability is only one factor injustifying broader system improvements. For example, the coreapplications necessary to support state-of-the-art fraud fightingare also the basis of competitive levels of service to customersand distributors. Also, the short-term costs of configurablebusiness rules are readily justified by the long-term savings thataccrue from eliminating the need to hard code applicationfunctionality across an unlimited spectrum of businessprocesses.

But whatever carriers do individually, major leaps in fraudcontrol will be possible only with the advent of greater access tonon-industry sources of information, greater attention to theproblem on the part of state insurance commissioners, and greatercooperation among insurers, in particular the larger stocks.Existing industry-wide databases are helpful, but they are limitedin the range of information they can provide. For example, ISOClaimSearch is an effective tool to understand prior claim history.However, no central data-bases exist to inform you of prior fraudconvictions, fraud investigations/suspicions, or criminalrecords.

The top-tier P&C companies possess the majority of theavailable customer data but are reluctant to share it because of aperceived strategic advantage. That perception is sound, but onlybecause of the narrowness of its scope. In the big picture, theindustry leaders will gain along with the smaller players on boththe bottom line and in terms of improved market image, which itselfis priceless in dealing with the legal side of fraud fighting. Sofar, few of these companies have shown the vision or courage toface up to fraud by broadcasting their willingness to prosecute andpublicize convictions, sharing the methods and circumstances ofprosecution, and explaining to consumers and investors about thebenefits associated with fighting fraud.

Michael Lucarini is senior executive in charge of globalclaims services in Accenture's Insurance prac-tice. He is based inNew Jersey.