Non-Environmental Cost-Benefit News

In its aggressive pursuit of fraud, the North Dakota workers' compensation bureau has increasingly turned to private investigators and subpoenaed bank records to detect cheaters. But a consultants' report concludes the bureau has little to show for skyrocketing investigation costs that doubled over the past three years, topping $1 million last year, according to figures obtained by The Forum.


The report, by Octagon Risk Services, estimates the average successful fraud case against an injured worker cost almost $37,000 last year. That compares to an average claim of $40,000 paid out to a disabled worker to replace lost wages.


"We were unable to justify the increased costs," the consultants said, noting the bureau's investigative caseload remained stable while costs soared. The Octagon report provides the most thorough independent evaluation of the bureau's fraud unit since it was established a decade ago.


Released Sept. 22, the report has triggered a major overhaul in the fraud unit, which includes three staff investigators and a paralegal in addition to the private investigators hired at an hourly rate.


Sandy Blunt, the executive director of Workforce Safety and Insurance, the state workers' compensation bureau, removed the manager of the fraud unit last month. He also has hired the Octagon consultant who made the critical assessments about the fraud unit's effectiveness.


Dave Aberle, who had managed the bureau's fraud investigations unit almost since its inception, was reassigned and his salary was reduced from $70,890 to $60,264, effective Nov. 1


Blunt, who previously held an executive position in the Ohio workers' compensation program, took over the bureau in May.The 131-page Octagon report, which also reviewed the bureau's legal department, claims department and ombudsman program for injured workers, cost about $150,000.


By law, the bureau must commission a performance evaluation every two years.


The consultants concluded that private detectives often were called upon to investigate workers' claim cases that "likely do not involve fraud." In almost a quarter of last year's fraud cases, the bureau subpoenaed workers' bank records. Those administrative subpoenas were issued by staff lawyers, not judges, with no notice provided to the suspected worker. The consulting firm sharply criticized the value of private investigators' surveillance tapes and reports, which it found "generally lacking." In addition, the report noted the insertion of private investigators in claims lacking grounds to suspect fraud. "In evaluating the referral process, it is clear the Claims Department is making referrals to the Special Investigations Unit that likely do not involve fraud," the Octagon report said. Private investigators, who are paid $45 an hour, often were called upon to help claims analysts verify a worker's physical capabilities to determine eligibility for disability benefits, Blunt said.


During the course of an fraud investigation, private detectives sometimes interview injured workers' neighbors. They routinely conduct secret surveillance, videotaping their subjects from a discrete distance, sometimes using a long lens to peer inside windows. At times, investigators pose as someone else to elicit incriminating statements from workers suspected of fraud. Workers' lawyers have argued this practice is an invasion of privacy. But bureau officials say posing undercover is a common method of investigating insurance fraud.


Blunt said the bureau's use of investigators was due, in part, to proximity: Because the claims analysts are located at WSI's headquarters in Bismarck, it became normal procedure to use local private detectives to observe or question workers elsewhere in the state.


In the future, Blunt said, private investigators will no longer be called upon to help claims analysts in routine claims management tasks. Private investigators will only be used when the bureau has good cause to suspect fraud, he said.


The Octagon report also found that the bureau's focus on fraud is targeted overwhelmingly at workers, with less scrutiny given to employers or medical providers.
Last year, for example, 89 percent of the fraud unit's caseload involved injured worker fraud; the remaining 11 percent involved employer fraud.


But the results of that lopsided ratio are questionable, according to Octagon's cost-benefit analysis for investigations.


During 2003, 24 of 317 injured worker fraud cases resulted in actions to discontinue benefits, including two criminal prosecutions.


That translates into a "success" rate of 8 percent, at an average cost per "successful" investigation of $36,922. The true average cost for an investigation actually is probably less, since investigators were involved in non-fraud claims cases, but still is too high, consultants said.


By contrast, the bureau had seven fraud actions against employers in 2003, including five criminal prosecutions, or a "success" rate of 18 percent. That yielded an average cost per "successful" investigation of $4,980, according to Octagon's analysis.


The bureau's fraud unit had long claimed high returns for its investigations.


Over the 10-year history of the program, it claimed it saved $26.5 million, a sum based on actuarial estimates of future benefits for disabled workers. The bureau calculated the net savings from its fraud program was more than $21 million.


Those estimates now are under review.


"There's no way to beat around the bush - that was a high number," said consultant Jim Wesson, an insurance fraud specialist who wrote the section of the Octagon report evaluating the bureau's investigations and later was hired to revamp the program.


Any new estimate will be "soft," or subject to some speculation, but will become "harder," or more accurate, over time.


Investigation staff will no longer have a hand in estimating savings from fraud investigations, Blunt said.


"We want a number that is staunchly conservative and staunchly defendable," he said.


Criminal prosecution is relatively rare. According to the bureau, the fraud unit's investigations have resulted in 49 criminal fraud convictions, involving both workers and employers, since the first prosecution in 1995.


Figures show that workers are five times more likely to lose their benefits for suspicion of fraud than to be prosecuted and convicted.


Over the past three years, 85 workers received notice their benefits would stop because of suspected fraud, a decision reversed in two cases, bureau spokesman Mark Armstrong said.


Questionable value


Fees collected by private investigators can add up to five or six figures for firms handling multiple cases. In 2003, the bureau paid Rollie Port Investigations of Minot, N.D., $320,622, or almost 30 percent of its $1.1 million total outside investigation costs.


Typically, private investigators would be allowed about 15 hours of surveillance to determine an injured worker's activity level, the Octagon report said.


Often, staff investigators had to spend too much time reviewing private detectives' work.


"If the Special Investigations Unit investigators are required to watch the videotapes and document their own observations, then we are not sure what the private investigators are being paid for other than holding a camera," the Octagon report said.


Investigations usually originate with tips called in to the bureau's hotline or by referrals from the claims analysts who handle injured workers' cases.


"A great number of the referrals are made on claims that have aged and the claims analyst is looking for mitigating factors," the report said.


Defending probes


Lawyers who represent injured workers argue the bureau has aggressively used fraud investigations as a tool to find reasons to cut off workers' disability benefits.


They also maintain inappropriate case referrals subject workers to needless intrusion, including surveillance and administrative subpoenas issued without a judge's order.


"They're going after people because they filed their forms wrong," said Steven Latham, a Bismarck lawyer who represents injured workers.


Disabled workers must disclose work activities and income in monthly status reports they send to the bureau.


Armstrong, the bureau spokesman, denies that workers' comp uses fraud investigations, including the use of surveillance and subpoenas, indiscriminately.


"This is not a willy-nilly organization," he said. "We abide by and enforce the law. The truth is the truth."


Subpoenas, issued by an assistant attorney general who works for the bureau, aren't sought unless investigators have an "articulable suspicion," said Aberle, the former fraud unit chief.


"I don't want you to think I wield some arbitrary power," Aberle told The Forum in an interview last summer before his reassignment.


If the bureau has issued a subpoena, he said, it means investigators have collected evidence "and that evidence has been blessed by an assistant attorney general."


That standard of evidence is lower than the "probable cause" required for a subpoena in a criminal investigation.


But evidence obtained from the bureau's subpoenas and surveillance sometimes supports criminal prosecutions, said Bruce Schoenwald, a Fargo-Moorhead lawyer.


Schoenwald, the unsuccessful Democratic candidate for North Dakota attorney general in the Nov. 2 election, said the bureau's lawyers need more stringent oversight.


Similarly, private investigators aren't held to the same professional standards as licensed law enforcement officers, and are exempt from the anti-stalking law, he said.


The bureau's assistant attorneys general do not report to the attorney general. "There's no accountability for them, that's the problem," Schoenwald said.


Subpoena use rising


Attorney General Wayne Stenehjem said the workers' comp bureau is not the only state agency with assistant attorneys general who are under his limited control. He can remove them only if they are found incompetent or unethical.


Other examples, all exceptions to the general rule that he ultimately supervises state lawyers, include attorneys for the Tax Department, Insurance Department and Public Service Commission.


Also, assistant attorneys general in some other agencies can issue administrative subpoenas. Those in the office of the attorney general, for example, sometimes subpoena records in civil fraud and charitable gambling investigations, Stenehjem said.


To improve oversight, Stenehjem said the bureau should be brought back under the governor's authority. The 1997 Legislature removed the governor's control of the bureau and placed it under an appointed board of directors.


Blunt said he is accountable to his board and ultimately the public. The revamped fraud unit will focus more on serious fraud than mere abuse, he said.


Both can involve misrepresentation, but abuse of benefits is more minor, such as submitting a workers' comp claim for a non-work related injury.


In recent years, the bureau has become more aggressive in its use of subpoenas. The bureau's lawyers subpoenaed bank records of 77 workers in 2003 while investigating 317 cases of suspected fraud by workers collecting disability benefits, up from 21 in 2001.


Wesson, the new fraud consultant, said he didn't consider the bureau's use of subpoenas "inordinately high." However, he is drafting written guidelines for subpoena use.


Fraud studies lacking


One huge unknown still confronts the workers' comp fraud program: the incidence of fraud, nationally and in North Dakota.


When the fraud program was launched in 1994, workers' comp officials cited national estimates that as many as 20 percent to 25 percent of all claims were fraudulent.


Those estimates now are regarded by many as too high. No studies have been done, in North Dakota or nationwide.


The Coalition Against Insurance Fraud, a national advocacy group dominated by the insurance industry but also including regulators and legislators, is considering a study. Dennis Jay, the coalition's executive director, believes between 3 percent and 10 percent of workers' comp claims are fraudulent.


Although fraud estimates range from 10 percent to 25 percent nationally, Wesson regards 15 percent as a realistic estimate nationally. North Dakota's fraud rate is probably low, perhaps less than 10 percent, he said.


Regardless of which estimate is accurate, nobody disputes that fraud is a real problem that hurts honest workers and employers.


"It's a lot of money being lost out there," Jay said. "We don't have any money to lose here, so we have be smart."

by Patrick Springer (701) 241-5522 The Forum - 11/28/2004