Using Benfords Law To Detect Fraud In The Insurance Industry

Abstract

Benford's Law is the mathematical phenomena that states that the first digits or left most digits in a list of numbers will occur with an expected logarithmic frequency. While this method has been used in industries such as oil and gas and manufacturing to identify fraudulent activity, it has not been applied to the health insurance industry. Since health insurance companies process a large number of claims each year and these claims are susceptible to fraud, the use of this method in this industry is appropriate. This paper examines the application of Benford's Law to four health insurance companies located in the Midwest. For each company, analysis was performed on the first digit distribution, the first two-digit distribution, and providers with high volumes of claims. The results show that the populations are similar to the frequencies predicted by Benfords Law. The findings also suggested possible fraudulent activity by specific providers, however, the com-panies determined that these results occurred due to abnormal billing practices and were not frau-dulent. The insurance companies that participated in this study will continue to use this method to further detect fraudulent claims.