Your Dog is the Worst Dog. Ever.

Do you think that your dedication to finding fraud could ever interfere with your ability to properly conduct a fraud investigation?

October 2013 

 

He barks too loud. Jumps on people. He's the worst dog. But wait – doesn't your dog do the same things? Perhaps. Does that make him the worst dog ever? Maybe your neighbors think so. You're both biased. But do you leave your bias at home? Or, as British politician Lord Molson once said, "I will look at any additional evidence to confirm the opinion to which I have already come." 

  

Do you think that your dedication to finding fraud could ever interfere with your ability to properly conduct a fraud investigation?  

  

Watchdog Seeks Fraud Claim  

Take the recent case of Locks of Love and the charity watchdog group Nonprofit Investor. Nonprofit Investor accused Locks of Love of failing to account for $6 million worth of in-kind (noncash) contributions – namely, hair used to make wigs - a pretty strong assertion. I was contacted for comment and, upon review of their IRS Form 990, concluded that while the accounting could be improved, there probably wasn't malfeasance (read the article on ABCnews.com). Nonprofit Investor's methods were shaky at best. Nonprofit Investor used statistics cited from four different sources and from four different years to back into the allegedly "missing" income. 

  

Could it be that the watchdog group, still just two years old, may have wanted to find a discrepancy, to make a splash in a crowded market, to make for a flashy headline suggesting fraud?  We'll probably never know. But the point is this: no matter how much we try to be unbiased, some degree of bias is virtually always present. This is not a matter of opinion. A substantial body of research supports it; below are three examples.  

  

Even Experienced Investigators Biased 

Two psychology researchers, Christian A. Meissner of Florida International University and Saul M. Kassin from Williams College, put 44 North American law enforcement investigators to the test to assess whether there is a natural bias when it comes to assessing whether someone is being deceptive in an interview setting ("He's guilty!: Investigator Bias in Judgments of Truth and Deception," published in Law and Human Behavior, Vol. 26, No. 5, October 2002). The investigators averaged 13.7 years of experience and had undergone professional training in interviewing and the detection of deception. 

  

All 44 investigators were shown a series of crime interview videotapes. The tapes involved four guilty suspects who had committed one of four mock crimes and four innocent suspects who were instructed to merely "appear at the scenes of these crimes." The interviewer was informed about the mock crime, but did not know whether any suspect was guilty or innocent before interviewing them. 

  

The 44 investigators were asked to determine, after viewing the taped interviews, whether each of the eight suspects was lying or truthful regarding his involvement in the crime. The investigators were told that between 25 percent and 75 percent of the suspects were lying. They were also asked to rate their level of confidence in their conclusions that each was lying or being truthful. 

  

The results were amazing. The investigators accurately identified the lying suspects 62 percent of the time – not bad. But they also falsely accused the truthful suspects 67 percent of the time! And they rated their level of confidence in their conclusions at 7.05 on a scale of 1 to 10. These investigators were pretty convinced that these innocent people were lying. 

  

Trained Fooled Worse Than Untrained  

Meissner and Kassin's study built upon a 1999 study in which Kassin and another researcher, C.T. Fong, showed the very same videotapes to 40 students – 20 who had received training in the detection of truth and deceit and 20 who had received no such training ("I'm innocent!: Effects of training on judgments of truth and deception in the interrogation room," published in Law and Human Behavior, 23, 499–516.). The trained students did worse than the untrained students in respect to both types of suspects. The trained students correctly identified the lying suspects 45 percent of the time, compared to 56 percent for the untrained students. But just like with the trained, experienced investigators, the false alarm rate was even more striking. The trained students wrongly accused the truthful suspects of lying 60 percent of the time, compared with just 42 percent of the time for the untrained students. 

  

Could it be that training on what to look for in people being deceptive, as well as experience, leads to a bias? Numerous researchers appear to think so. 

  

Are the trained, experienced investigators looking for deception so hard that they think that they see it? 

  

Studies Show Our Bias 

One other scientific research project suggests that the answer to this question may indeed be, "yes." In 2005 a team of researchers led by neuroscientist Gregory Berns of Emory University School of Medicine's Department of Psychiatry and Behavioral Sciences conducted an interesting experiment aimed at assessing how the brain works when bias is involved ("Neurobiological Correlates of Social Conformity and Independence During Mental Rotation," published in Biological Psychiatry 2005; 58:245–253). While the experiment did not involve fraud or an investigation, it nonetheless applies to how we deal with bias. 

  

In the experiment, 32 people were asked to compare three-dimensional objects. In one version of the test, they were asked to decide by themselves which two objects were the same. But in another version, they were asked to decide after being told what conclusions had been drawn by their fellow participants. And in a third version of the experiment, participants were asked to decide after being told what a computer had determined. 

  

The fact that participants clearly conformed to the decisions made by others was obvious and completely expected. Once people know what others have concluded, they tend to conclude similarly themselves. The overall error rate when participants made their own decisions, without knowing what the computer or other participants had decided, was 13.8 percent. But, when participants knew what their fellow participants had concluded, and that conclusion was incorrect, the error rate spiked to 41 percent. When the computer provided the incorrect conclusion, the error rate also jumped, but not as dramatically, to 32 percent. 

  

But that is not what is most interesting about this research. The researchers also studied what type of brain activity was involved in conforming behavior (particularly when conforming to an incorrect conclusion). 

  

All participants performed their comparisons while they were in a functional magnetic resonance imaging (fMRI) brain scanner. Researchers wanted to determine whether when participants conformed to the incorrect conclusions drawn by other participants the part of the brain most active was the area dealing with conscious decision-making (the prefrontal cortex). 

  

But what they found was that other regions of the brain were the dominant activity centers when conforming behaviors took place – regions associated with perception rather than decision-making. Participants actually saw what their fellow participants saw, rather than evaluating information and making the same decision. 

  

Think of this now in terms of how an investigator responds to an allegation of fraud or a whistleblower call aimed at a particular person. The implications are significant. Knowing what someone (a whistleblower, for instance) has concluded may inherently establish a bias on the part of the investigator. Add to that all of the training in finding fraud, and a job responsibility that includes finding fraud, and one can quickly see that an investigator may commence an investigation determined to find fraud, rather than to conduct a fair and thorough investigation. 

  

Bias in Fraud Investigations – A Case Study 

Being overly eager to prove fraud can have damaging consequences for both the investigator and the company who employed the investigator.  

  

Take the 2010 case of the Canadian Pacific Railway Company (CPRC) as an example. This case stemmed from CPRC's firing of an employee after 27 years of service, on the grounds that the employee submitted duplicate expense reimbursement requests, clearly a policy violation that could result in termination from many companies. 

  

The investigation began after CPRC's internal auditor discovered six irregularities in the employee's expense reports. This was reported to two vice presidents, who instructed the internal auditor to audit all expense reports associated with the project that this employee had been working on. The matter was also turned over the CPRC's Director of Human Resources, who was instructed to hold a "clarification meeting" with the employee. The HR director also met with the internal auditor.  

  

At the clarification meeting, the HR director informed the employee that the internal auditor had found a total of 37 irregularities. These findings totaled approximately $9,000. A second interview was conducted about three weeks later, after which the employee was terminated for cause – fraudulently submitting duplicate expense reimbursements. 

  

The Court found that the employee was wrongfully dismissed, concluding that CPRC's investigation failed to prove that the employee intended to defraud the company. The court's conclusion was based on CPRC's failure to consider several factors that would have been in the employee's favor, such as: 

 

The employee was hand-picked to lead a special training project that required him to travel extensively throughout Canada.

During this time period, his expense reporting clearly showed a gradual deterioration as the project absorbed more and more of his time.

The vast majority of the discrepancies occurred during the latter stages of the project, when the employee's schedule was the most hectic.

  

Related to the above, witnesses who worked with the employee (none of whom were interviewed by the HR Director as part of her investigation) stated that he was very disorganized in his record-keeping for expenses, instead focusing his attention on the critical and time-consuming training project, the topic of which the employee was a recognized expert in. 

  

The HR director knew from the beginning of the project that the employee would have a difficult time doing all of the accounting for the numerous expenses incurred, recognizing that this was not his area of skill and that CPRC was not providing him with any administrative support (the HR director admitted this during cross examination); It was clear that the employee was selected for this project based on his technical skills, not his administrative abilities. 

  

In addition to the $9,000 of erroneous expense claims, the employee also failed to submit requests for reimbursement for $2,800 of other costs he incurred. 

  

In connection with this last factor, the court asked why would someone intent on defrauding a company leave $2,800 behind that he was legitimately entitled to? 

  

Another piece of evidence that hurt CPRC was that the HR director attempted to access confidential information from the company's Employee Assistance Program that the employee had previously participated in arising from an earlier altercation with another employee that was alcohol-related. Her awareness of the employee's previous incident with another employee may have impaired her ability to be objective in the investigation, in spite of the fact that this appears to be a single blemish on the employee's 27-year record with CPRC, a record that included six promotions, a leadership role in management, and regular kudos and performance bonuses. 

  

The judge noted that the HR director's investigation was not a "fact-finding mission" as she asserted, but "an exercise in case building against (the employee)." The court went on to state that CPRC never seriously considered whether the employee's claim that the discrepancies were honest mistakes made during a period in which he was over-worked could be true. To this point, the judge noted that "failing to consider the total picture and turning a blind eye to any kind of context or surrounding circumstances was unfair to say the least." 

  

Needless to say, the terminated employee was awarded a hefty sum for damages in lieu of notice, lost pension and other benefits, and an additional amount for bad faith damages (based on the unduly insensitive manner in which the employee was dismissed). All of this really stemmed from the HR director's investigation bias. 

  

Becoming Unbiased 

So, the old adage that an investigator "must remain unbiased" is really not very accurate. Such a statement suggests that we start out unbiased. Yet, as we've seen, it appears that we often begin an investigation already biased to some degree. Therefore, all investigators should vow to fight to "become unbiased" in order to perform their jobs in as objective a manner as possible, being fair to all parties involved. A few critical steps in achieving this goal: 

 

Consider all relevant records and information, not just that which supports an assertion of fraud.

Consider all the possible non-fraud-related explanations for suspicious evidence and data anomalies (just as we are instructed to brainstorm about the ways in which fraud could occur, we should brainstorm about other explanations for the red flags of fraud we are aware of).

Extend the scope of your interviews beyond those who can corroborate facts which support the assertion to include others who may possess pertinent data.

Include all relevant information in your report, including information you uncovered that could clear someone or that otherwise may not support the case.

  

Remember this: if someone is guilty of committing fraud, our role as investigator should neither be to prove fraud nor even to conclude that it is fraud, but to discover the facts and present them in a manner that would lead others to conclude that the only logical explanation is that fraud occurred. 

  

 

Gerard Zack, CFE, CIA, CPA is CEO of Zack, P.C., a fraud risk advisory, investigation and internal audit firm based in Washington, D.C. and serving clients internationally. Contact him at Gerry@zackpc.com.