Every week I get this question from some member of the media, "How much fraud occurs each year in the property and casualty industry?" My response is always the same, "About $30 billion," and then I hope that the reporter doesn't ask how that figure is determined.
The truth is, no one really knows how much fraud is perpetrated each year. We use that familiar figure because it's been used forever. But do we really know what the fraud picture is?
According to research conducted in 1992 by the Battelle Seattle Research Center for the Insurance Information Institute, fraud accounts for about 10 percent of the property/casualty insurance industry's incurred losses and loss adjustment expenses each year. A good number of potentially fraudulent claims are paid each year without being investigated due to the desire and requirements to make payments in a timely fashion and the difficulty in proving fraud. As a result, we only see a small percentage of those claims—and an even smaller percent are ever prosecuted.
The following graph shows incurred losses and loss adjustment expenses since 2006:
P/C Industry, Income Analysis, Source: "The Insurance Fact Book 2012 & 2013" with data from ISO, a Verisk Analytics company. ($ in billions).
So while the average estimated fraud loss over these six years is $31.33 billion, the annual figure swings from $28.39 billion to $34.45 billion—a $6 billion variance.
Taken as a whole, the average fraud loss of $31.33 billion is within parameters, especially when using imprecise language like "about $30 billion." That is made painfully clear whenever a reporter who's been around for a while observes that the number hasn't changed in years.
That is generally followed by another inevitable observation that goes something like this: "It seems odd that for an industry loaded with analysts, data-mining tools, actuaries and claims professionals that it can't develop a more reliable method to determine how much fraud it sees in a given year." Good point.
If we approach this data from the perspective of a fraud-fighting metric, how do we describe the industry's effectiveness? Since we're identifying the scope of fraud as 10 percent of a number that is based on two loss-related variables, then any movement up or down is just a function of those variables and not at all tied to the industry's fraud-fighting prowess.
Yet, we see case after case where multi-million dollar insurance fraud rings are busted all across the nation. The cases that NICB works with its SIU and law enforcement partners are complex. They run the gamut from local, significant loss episodes to national, multi-carrier, multi-claim scams that drain hundreds of millions of dollars from insurance companies and the federal government.
Maybe $30 billion is just the tip of the "fraudberg." Maybe not, but who really knows?
ISO receives more than 235,000 claims a day. With that amount of data and the analytical and investigative talent available to the industry, it does seem awkward that we can't do the research that realistically and reliably answers the question, "How much fraud occurs each year in the property and casualty industry?"