Uncovering social service fraud saves millions, reinforces public trust

In Los Angeles County, the Department of Public Social Services (DPSS) offers a range of programs to alleviate hardship and promote health, personal responsibility and economic independence. Across the county's many communities, DPSS offers temporary financial assistance, employment services, free/low-cost health insurance, food benefits, in-home services for the elderly and disabled, and other financial assistance.

To assist in program integrity efforts in the CalWORKs Stage 1 Child Care Program, LA County turned to SAS® analytics solutions to identify potential fraud, enhance investigations and prevent improper payments. By doing so, they've helped the most vulnerable members of the community while protecting millions in taxpayer dollars.

The system analyzes social networks to determine if individuals are likely to commit fraud. Social network analysis also helps identify collusive fraud rings companion cases. The bottom line: more than $6.8 million in savings.

Analyzing the data, finding the fraud patterns

Fraud cases can include false employment claims where nonexistent employees are declared. In other cases, businesses are created by the heads of fraud rings who collude with recipients who falsely declare that their children are attending nonexistent child care centers. Sometimes, criminals declare work schedules that are false or shorter than the time amount claimed.

To combat fraud, LA County first needed a data integration solution and a powerful analytical engine to bring together numerous internal and external data sources to build and run predictive models. With social network analysis and analytics, LA County can predict which benefit recipients and service providers are most likely to engage in fraudulent activity and create potentially large fund losses.

Using predictive models and peer group analysis to detect anomalies in the use of child care services, LA County developed high-risk scores to decrease the number of false-positive cases assigned to investigators. The system uses a predictive model to analyze social networks and to assess the likelihood of child care fraud and collusion in fraud networks in Child Care Program. Social network analysis also helped identify collusive fraud rings in companion cases.

LA County uses the SAS Fraud Framework for Government and incorporates SAS data mining technology with social network analysis, predictive analysis, rules management and forecasting techniques. SAS Business Intelligence has also been used to create an information portal where reports are housed and used to monitor and share information on fraud cases. By identifying historical patterns of fraudulent activity, investigators can focus on cases with a higher probability of fraud. These improved process efficiencies mean fraud investigators have more time to review high-risk cases.

Saving millions of dollars

SAS models have enabled DPSS' Welfare Fraud Prevention and Investigations staff to identify and expedite the review of suspicious cases much earlier than waiting on referrals from contracted agencies or other referral sources.

In its first 10 months of operation, the system has produced 197 additional referrals for child care fraud investigations; 143 from front-end triage review, and 54 initiated directly by investigators upon review of other active investigations. Additionally, another 67 non-child care related referrals were initiated as a direct result of triage review.

Separately, SAS Social Network Analysis functionality detected two conspiracy groups (consisting of 16 cases) much earlier, significantly reducing the duration of fraudulent activities. LA County mapped out a network of participants and providers that visually displayed their relationships. They looked at whether any given small network fit into a larger scheme of networks, in which participants are in collusion with other child care providers. They identified strong central nodes and, in one case, found a child care provider serving many nodes of participants colluding in fraudulent activities. The investigation is currently being conducted on six separate conspiracy allegations that were a direct result of a data mining review by investigative staff.

The aspect of the network that proved most valuable for fraud investigators was the social relationship network display. This display shows a web of complex relations linked, for example, by common telephone numbers and addresses. Instant access to this network of child care recipients and providers saved fraud investigators innumerable hours of casework preparation. "It would take me months or years to uncover all of the relations shown," said one investigator.

"On one of my cases, with a single click of my mouse, I saw leads to additional evidence that would have otherwise taken weeks, possibly months to uncover. This included evidence such as addresses and names of potential unreported employers and potential second-residence addresses. The system also showed a connection between my suspect and two other suspects on two other cases."

Also, one investigator, who was nearing the conclusion of a 10-person conspiracy investigation, ran the main suspect's name through the social network analysis and discovered seven potential additional co-conspirators that she would not have otherwise discovered.

In just a few short months, LA County's pilot implementation achieved an 85 percent accuracy rate in detecting collusive fraud rings, which is projected to enable cost avoidance in new fraud referrals of approximately $2.2 million, early detection of fraud of approximately $1.6 million, and other increases in efficiencies of approximately $3 million. Moving forward, LA County hopes to expand its reviews to improve in a variety of areas:

  • Early detection of suspicious cases.
  • Early detection of suspicious child care providers.
  • Detection of previously unknown suspicious participants.
  • Detection of previously unknown suspicious child care providers.
  • Increased investigatory efficiencies.
  • Detection of collusive behavior.
  • Identified $6.8 million in projected savings
  • Interagency collaboration.


One of the nation's largest county governments needed to uncover systemic fraud in its delivery of social services to economically disadvantaged citizens.



Investigators can spot more cases of fraud sooner, resulting in fewer losses, lower investigative costs and greater public confidence. Total cost avoidance projected from the pilot results is expected to exceed $6.8 million.

The results illustrated in this article are specific to the particular situations, business models, data input, and computing environments described herein. Each SAS customer’s experience is unique based on business and technical variables and all statements must be considered non-typical. Actual savings, results, and performance characteristics will vary depending on individual customer configurations and conditions. SAS does not guarantee or represent that every customer will achieve similar results. The only warranties for SAS products and services are those that are set forth in the express warranty statements in the written agreement for such products and services. Nothing herein should be construed as constituting an additional warranty. Customers have shared their successes with SAS as part of an agreed-upon contractual exchange or project success summarization following a successful implementation of SAS software. Brand and product names are trademarks of their respective companies.

Back to Top