When you think of “insurance crime,” probably the firstthing that comes to mind is some form of third-party fraud, withpolicyholder claims-padding on one end of the spectrum and Russianmob staged auto accident rings on the other.

But insurance companies are increasingly looking at their agencydistribution force as a source of potential fraud, especially infinancially tough times.

Within the industry, associations that track insurance fraud ofall kinds say they've seen instances of producer-perpetrated fraudincrease during the recession. 

Based on statistics from 37 state insurance fraud bureaus,agency fraud jumped 24 percent in 2009 at the height of therecession, said Dennis Jay, president of the Coalition AgainstInsurance Fraud (CAIF). That year, the largest average increases infraud categories were in bogus health plans, drug diversion andfraud by insurance agents. A total of 69 percent of respondentssaid agent cases were up slightly higher or much higher in2009.

And although CAIF data suggest that agency fraud had fallen to2007/2008 levels by the end of last year, other numbers sayotherwise.

Related: Read the article “Bergeron on SIUs andInnovation” by Christina Bramlet.

For example, a February 2012 report by the National InsuranceCrime Bureau (NICB) finds referrals of insurance agent/adjusterfraud were up 33 percent between 2010 and 2011. 

A quick look at the bulletins put out by state insurancedepartments' fraud bureaus make it clear that crooked agents andbrokers are definitely out there. 

For example:

• Earlier this year, federal authorities in New York arrestedthree life insurance agents in a $100 million scheme of conspiringto defraud major insurance companies into issuing life insurancepolicies to straw buyers, when the true owners of the policies werethird-party investors and financiers. 
• A north New Jersey insurance agent pleaded guilty to defraudingnine clients out of more than $1.9 million. The licensed insuranceproducer, securities dealer and certified financial planneradmitted he obtained the money to invest for clients, but usedinstead to fuel his online gambling habit.
• Two Pennsylvania insurance agents defrauded school districts inBerks County of more than $1 million. The agents used the premiummoney for luxury cars, expensive wine and other personalitems—leaving the school districts without coverage.
• Five life insurance agents in New York were involved in ascheme that used identity theft, false business records and forgeryto establish fake insurance policies, earning commissions andguaranteed bonuses based on selling those policies. They werearrested and charged with using the company-issued agent code of aformer employee to write policies and forging his signature to cashthe commission checks.

“Most agents are honest. But there are always a few bad applesand the challenge is in identifying them,” said James D. Ruotolo,principal for insurance fraud in the global fraud and financialcrimes practice at security software company SAS.

For years, insurers and law enforcement alike focused onexternal insurance fraud claims generated by policyholders andthird-party claimants. But as hard times linger, insurers areincreasingly looking at significant exposure from internal fraudscams—not just by agents but by their own employees, Ruotolo said.“They have access to internal systems and very detailed informationabout workflow and protocol and become very educated in how to workthe system.”

New technology is helping insurers stay one step ahead of thecriminals. SAS, a software company specializing in businessintelligence and analytics for financial institutions, has recentlybuilt out a security intelligence practice and created software tohelp detect suspicious activities in insurance, banking, healthcareand government.

Ruotolo, whose background is in the specialinvestigations unit at The Hartford, is familiar with all thescams, from hard fraud to simple sloppy agency processes. “At theHartford, when we examined claim experience reports and saw lots offraudulent claims activity from a particular agency, it's obviousthey're either not doing a good job of qualifying customers orthere is active fraud going on,” he said.

Although producer fraud can be complex, probably the most commonis misappropriation of funds  belonging to thepolicyholder, premium finance company or insurer, Ruotolo said. Inthis case, agencies having financial difficulties collect premiumsfrom customers and don't pass it along to the insurer. Becausethese funds go unreported to an insurer or premium finance company,counterfeit insurance identification cards and certificates ofinsurance are often also involved, according to a 2010 report bythe Louisiana Dept. of Insurance.

Related: Read the article “Big Data is Blowing Up”by Robert Regis Hyle.

“We always see this in tough financial times,” Ruotolo said.“And it's not a victimless crime—many people are negativelyaffected because the policy is not procured and the insured has nocoverage.”

Other common producer frauds include:

1. Double dipping. Crooked property-casualty andlife agents cancel a policy and then rewrite a new policy for thesame insured. Because the coverage is uninterrupted there's no riskto the policyholder, but the agent gets a double commission.
2. Catastrophe fraud. In advance of a big storm,fraudulent agents try to write policies without any intention ofproviding coverage. After the storm hits, crooked agents colludewith contractors and adjusters for kickbacks, promising customersexpedited repair or other preferential treatment with no intentionof delivering.
3. Crop insurance fraud. Because of the recentglut of bad weather, people who never even planted a crop arefiling claims, and in some cases, they're colluding with crookedagents, said Jay of CAIF. 
4. Phony agents. Simply put, it's unlicensedagents selling policies. This scam is especially prevalent amongimmigrants and the disadvantaged who might not speak English andface high premiums. Because the fake agents are migratory, thecrime is difficult to track.
5. Premium finance scams. Crooked agents collect100 percent of the premium, then turn around and finance thepremium, giving themselves an interest-free loan at the financecompany and insured's expense. Because they might be paying off thepremiums every month, this type of fraud may never come tolight.
6. Phony insurance. Agents selling fake coveragefrom a phony insurance company, or bogus coverage using alegitimate company's name. This usually involves the use of fakedocumentation.
7. Worthless investments. Viaticals taken out onsick or terminally ill people can be a legitimate investment, butsome can also be phony or misleading. Another scam involves agentspromising quick, high and certain returns for investing inpromissory notes supposedly backed by insurance. Often thepromissory notes are fakes.

Although insurance crime knows no season, hard times are sure tobring it to the forefront, especially when a lagging economy isaffecting both agencies and the businesses they serve. “There'smore evidence of collusion between agents and businesses when bothof them are desperate,” Jay said. “During the recession, a lot ofpremium went away because there were no new homes or cars andbusinesses weren't expanding. Economic stability has hardly comeback for many businesses, so people are looking to cut corners anyway they can.”

Technology, however, is making it easier for law enforcement andinsurance companies to spot such scams. “The biggest factor overthe last few years is technology,” Jay said. “They're runningsystems with predictive modeling and link analysis, programs thatwere initially designed to spot external fraud. But now they'reputting everything through it, including claims handlers andagents. It's been an eye-opener because a lot of things came tolight.”

During the last 15 years, insurers have added technology totheir arsenal of anti-fraud weapons, which includes the creationand expansion of special investigation units and civil actionsagainst suspected fraudsters. Insurers are using technology-basedanti-fraud tools both inside the company and outsourced tovendors. 

A new study of 74 insurers by CAIF and SAS findsthat: 

• 88 percent are employing anti-fraud technology. Less than half,however, are using technology for non-claims functions, such asunderwriting fraud.
• Slightly more than half of insurers surveyed said they had beenusing anti-fraud technology more than 5 years.
• Three most common technologies employed are red flags/businessrules, claim scoring and link analysis.
• 40 percent use text mining as an anti-fraud tool.
• 52 percent say the major benefit of anti-fraud technology liesin uncovering complex or organized fraud activity.
• Biggest challenges in deploying technology insurers cited werelack of IT resources and determining the cost-benefitanalysis.
• 31 percent report that they expect increases in technologybudgets next year. Predictive modeling and text mining are the toptwo areas in which insurers are looking to invest in thefuture.

Although most analytics do a deep dive on claims because that'swhere insurers are spending money, they are increasingly turning asuspicious eye to agents, a trend that Ruotolo expects willgrow.

Insurers using the services of SAS have access to analyticsmodules specifically designed to detect certain types of fraud, hesaid. For instance, when analyzing all applications processed by acertain agent over a period of time, SAS can compare patterns tothe agent's peer group or other agents in the same area writing thesame business. SAS also can conduct an analysis in which itexamines changes in a book of business over time to weigh whetheran agency is writing more specific lines of business, handlingriskier business, etc.

Related: Read the article “Allstate's Latest FraudSuit Targets 16 New Yorkers” by Chad Hemenway.

A powerful indicator in such modeling lies in the measure of“velocity,” Ruotolo said. “When people go bad, they usually startwith something small and then see that they can get away with it.Often in data analysis, we can see that grown quickly over time. Achange in magnitude or velocity in data patterns is a strongindicator of fraud.”

SAS also analyzes social networks and links at the network levelto sniff out suspicious patterns and relationships that couldindicate fraud, Ruotolo said. By bringing in external data such aspublic records and social media, they can spot suspiciousconnections between an agent and policyholder suggesting collusion.When considering claims, they may examine first and third partiesin an auto accident (“if they're friends on Facebook, it could be astaged accident”).

The trend toward fraud and its detection means that honestagents must take additional steps to protect themselves, Ruotolosaid. Although stricter anti-fraud regulations are primarilyfocused on banking, it will eventually flow into insurance,especially on the life side. He suggests that agents:

1. Follow the rules. If there are guidelines orprocedures, adhere to them. If a rule doesn't make sense in yourarea, contact someone, explain why and suggest alternatives. Don'tjust go your own way and hope no one notices. 
2. Be honest and encourage customers to do thesame. Insurance is a business of trust. A small lie now tosave a few dollars in premium could turn into a big nightmarelater. 
3. When in doubt, document. Often there is alegitimate explanation for making an exception of some kind. Justmake sure you have the information to back up that decision in casethere is a question. 
4. Be on the lookout for fraud. If you suspectsomeone is gaming the system or intentionally trying to obtainbenefits to which they are not entitled, contact the insurancecompanies' special investigation unit or the National InsuranceCrime Bureau hotline at 1-800-TEL-NICB.