Is food stamp fraud declining? USDA hones tools to find out.

Even as the numbers of those receiving food stamps grows, error payment rates for the program have dropped, according to a recent report by The Pew Charitable Trusts.

That would seem like good news, except oversight agencies don’t have a clear grasp on the reasons behind the drop in errors. And that means duplicating whatever is causing the decline in mistakes will be difficult.

According to the Pew analysis of U.S. Department of Agriculture data for the Supplemental Nutrition Assistance Program (SNAP), commonly known as food stamps, the error rate in 37 states in 2008 was 8 percent; in fiscal year 2013, that number dropped dramatically to 3.2 percent.

At the same time, enrollment rates have skyrocketed. In 2008 the monthly average number of Americans enrolled in SNAP was 28.2 million, and the cost of the program that year totaled $37 billion. By 2013, enrollment reached 47.6 million, with spending nearly $80 billion. (The recession officially began in December 2007.)

One possible cause for the inconsistency? Bad data.

In a September 11 Iowa Watchdog article , Wendy Dishman, administrator of the Investigative Division of the Iowa Department of Inspections and Appeals who has investigated SNAP fraud for years, said the Food and Nutrition Service (FNS)  does not have reliable data.

“There’s a lot of inconsistencies in the data they collect,” she said. FNS’ data is based on information collected from the states. “The challenge is having 99 different county attorneys and all of them have different standards. Some of them aren’t interested if the fraud is less than $5,000. Some don’t want to see it if it’s less than $10,000,” Dishman added. Thus, “we don’t ultimately know what’s being reported to FNS by the various states.”

While inconsistent standards might produce inaccurate error rates, it’s a certainty that fraud is a major factor.

FNS defines SNAP fraud as participants exchanging SNAP for cash (aka trafficking), applicants lying to get extra benefits or a disqualified retailer lying on an application to get into the program.

FNS uses a range of technologies to reduce SNAP fraud, including issuing food stamps via electronic benefits transfer (EBT), which automatically applies benefits to a card similar to a pre-paid debit or credit card. EBT gives the USDA the ability to better track, identify and take action against possible fraudulent activity. And those cards will soon be issued with chip and PIN technology to better secure them.

The USDA’s Antifraud Locator using EBT Retailer Transactions (ALERT) system monitors electronic transaction activity and identifies suspicious stores for analysis and investigation. The system receives daily transaction records from EBT processors and conducts analysis of patterns in the data, which indicate potential fraudulent activity by stores. Steps that states took to lower their error rates include streamlining paperwork and simplifying the way potential recipients are screened for eligibility. States were also given the OK to crosscheck applications with asset tests for other safety-net programs, cutting administrative burden and potential errors. Other strategies include penalizing states with persistently high error rates.

Still, there’s no clear explanation for why some states have been so successful in reducing error rates while others have not, noted the report. Many think simple manpower is the reason.

Survey respondents in an August Government Accountability Office report on food stamps noted that the biggest hurdle to cutting fraud was a lack of staff to conduct fraud investigations.

States reported either reduced or maintained staff levels as the number of SNAP recipients greatly increased. Additionally, according to some state officials, “investigative agencies are not rewarded for cost-effective, anti-fraud efforts.” The report found “mixed success investigating and pursuing” food stamp fraud in 11 states using a range of detection tools.

According to the report, existing tools were mostly not up to the task of accurate fraud detection. “GAO found limitations to the effectiveness of recommended replacement card data and website monitoring tools for fraud detection,” it said.

“GAO also found the FNS recommended e-commerce website monitoring tool to be less effective than manual searches in detecting posts indicative of SNAP trafficking. GAO found the recommended tool for monitoring social media to be impractical due to the volume of irrelevant data,” it added.

Despite the uncertainties of working with the tools, GAO recommended that USDA keep trying to improve them. Among other suggestions, GAO recommended FNS, “reassess current financial incentives and detection tools and issue guidance to help states better detect fraud and report on their anti-fraud efforts.”

Separately, FNS issued a set of guidelines for states on how to combat fraud called Keys to Payment Accuracy. The guidelines focused on improving processes – including leadership commitment, good communications, data analysis, error reviews, corrective action plans and training.

The Department has responded to the recommendations to improve its technology to improve fraud detection.  Last month it awarded seven states a share of a $5 million in grant funds to identify, track and prevent the misuse of  SNAP benefits by program recipients. Grants were made to Florida, Kentucky, Missouri, Ohio, Oregon, Tennessee and Washington state. Additionally, Agriculture Under Secretary Kevin Concannon announced the availability of up to $7 million in grant funds to prevent SNAP recipient trafficking.

“SNAP fraud is rare, but no level of abuse is acceptable,” said Concannon in announcing the awards. “USDA continues to enhance our efforts to combat retailer fraud and partner with states to improve recipient-focused investigations.”