In 2014, more than 85 million Americans received a refund from the Internal Revenue Service. It is money they want quickly, money they count on. The IRS has worked hard to speed up the refund process.
At the same time, criminals try to scam the system by submitting bogus returns. In fact, the nation’s tax refund system is under continuous attack. The challenge: how to catch and stop fraudulent tax refunds—without slowing down the overall refund process?
The IRS turned to Booz Allen to enhance the agency’s Fraud Detection Program and, in particular, manage the “EFDS”—Electronic Fraud Detection System. The agency’s “first line of defense,” the EFDS scans all 145 million plus tax returns within hours of receipt and assigns a risk score. The risk score determines if a return will be processed (and, where appropriate, a refund issued), or requires further review. For Booz Allen, the challenge was twofold: improve the system’s performance and enable ongoing modifications to an incredibly complex system, with minimal disruption. In particular, the Booz Allen team:
- Dramatically strengthened the agency’s ability to detect fraud in areas like the Earned Income Tax Credit program (the federal program with the second largest amount of improper payments next to Medicare)—through new processes, systemic filters, advanced analytics.
- Drove business process improvements that reduced fraud and lightened internal workloads by improving the quality of returns coming in from paid return preparers.
- Created a streamlined process to implement 40+ major software changes each year, required by changes to the tax code; focused both on software upgrades and the “human side” (organization and training for the IRS staff).
- Designed and implemented new data models on a “fast track” basis—in time for each tax filing season to stay ahead of changes in the tax code that could create opportunities for fraud.