As the federal government reported an estimated $115 billion in improper payments in Fiscal Year 2011, the impetus to eliminate and recover these funds continues to mount. State governments also struggle with mounting and often embarrassing improper payments with estimated totals approaching $125 billion.
So what classifies as an improper payment? An improper payment could be an incorrect payment, an over- or under- payment and could include a payment to an ineligible recipient, a payment for an ineligible service, a duplicate payment or a payment for a service not received. Improper payments can occur for many reasons including inadequate record keeping, inaccurate eligibility determinations, inadvertent processing errors, the lack of timely and reliable information to confirm payment accuracy, and of course fraud. The data required to address these issues is fragmented across more disparate systems and agencies than ever before. The quantity of that data is increasing at an enormous rate. More data in more disparate silos mean more error and variation in the identity information associated with the records; all this while fraudulent activities and capabilities are getting more prevalent and more sophisticated.
Even in the last few weeks, the breadth of issues has been highlighted by recent news stories. While all of these stories are different, each highlights the complexity of combating the problem.
- Michigan woman who won $1M lottery but kept using food stamps
- Texas At-Home Health Services Accused of $374 Million Fraud
- Arizona Inmates Paid $1.1 Million in Unemployment Benefits While in Jail
While there is no easy fix, organizations are bound to employ the most advanced technologies available to curb these improper and/or fraudulent payments. Fundamentally, prevention is based on a two-step process:
1) Person/Identity Resolution – Proper identification of the payee will answer: who is this person? Where is verification data? Answers to those questions alone will often be enough to identify suspicious activity and payments and flag them for review.
2) Complex Event Processing – Even if the payment clears this first test, examination of the payment event compared to other known events could detect error or fraud. For example, duplicate invoices against different vendor codes, company codes, POs, or payable systems will be detected.
Given the economic uncertainty, continuing budget deficits, mounting debt, and high unemployment, government agencies need to good stewards of the precious resources that they do have. By reducing error rates and detecting fraudulent activities before payments are made, government organizations stand to save billions of dollars and at the same time avoid often embarrassing headlines.