WELFARE FRAUD PREVENTION

AND INVESTIGATIVE FUNCTIONS

OF THE

DEPARTMENT OF PUBLIC SOCIAL SERVICES

INTRODUCTION

The Research and Follow-up Committee of the 1998-1999 Grand Jury reviewed selected topics from the Final Reports of the previous five Grand Juries, noting specifically the responses given to the audited agencies to the recommendations made in those Final Reports.

One of those subjects was an audit by the 1996-1997 Grand Jury, pointing out changes needed in the DPSS to eradicate recipient fraud in the Los Angeles County Department of Public Social Services (DPSS), also called the Welfare Department. It appeared from the Research and Follow-up Committees' study that many of the previous Grand Juries' recommendations were not taken seriously, and in some instances were strongly disputed, or even vehemently rejected. Therefore, that Committee felt those responses to be unsatisfactory, inappropriate, or inadequate, and referred the subject to the Health and Human Services Committee for review and consideration for a more extensive overview audit of the DPSS's operation.

Almost concurrently the Los Angeles Times, in an August 14, 1998, article, reported an egregious instance of large-scale internal fraud by an employee of the DPSS. Subsequent newspaper articles and television documentaries further highlighted that instance and other potential and actual problems in the operation of the DPSS. We also learned from preliminary interviews that the potential for fraud, both external (recipient) and internal (staff), that exists in the Los Angeles County Department of Public Social Services is enormous. Further, based on extrapolated audit data obtained from two other Southern California counties, the potential annual savings that may be realized by Los Angeles County could exceed $500 million.

As a result, the Health and Human Services Committee of the 1998-1999 Grand Jury

determined that a wide-ranging management audit of the DPSS was necessary to accomplish the following goals:

As the Committee studied the magnitude, extent, and objectives of such an audit, it was found that some, but by no means all, of the recommendations of the past audit had, in fact, subsequently been implemented. That, along with the budgetary limitations of the Grand Jury, made it necessary to perform only a Limited Management Audit.

Methodology:

The scope of this audit includes both recipient and staff fraud. The functions and activities of the Department's Welfare Fraud Prevention & Investigation (WFP&I) group, Internal Affairs Section, and Fraud Prevention Unit of two district offices (Metro Family, and Metro Park) were reviewed. Group and individual interviews were conducted with the Director and numerous department representatives, including the Director of the Bureau of Administrative Services; the Chief of the Specialized Services Division; the Chief of the Human Resources Division; the Director of Welfare Fraud Prevention and Investigation Section (WFP&I); the Human Services Manager of the Management Information & Evaluation Section; supervisors and selected staff of the WFP&I, and the Internal Affairs Section; directors and key staff of the Metro Family and Rancho Park district offices; and the Chief of the LEADER project; as well as other selected staff from that project. Representatives of the Fraud Division of the Office of the District Attorney were also interviewed.

There were briefings by the Department's Internal Security Task Force regarding Department plans to strengthen its internal controls. A personal visit was also made to San Diego County's Health and Human Services Agency and the District Attorney's Office to review their internal security, and fraud prevention and detection safeguards.

A sample of Internal Affairs and WFP&I case files was reviewed and 33 internal investigation files were selected at random from 187 cases opened between February 1997 and November 1998. In addition, 165 WFP&I files, which included both early fraud and full field investigation cases, were selected for review at random from the Ranch Park and Metro Family Offices.

Data and documents were collected and reviewed from the Department including caseload statistics, budgetary information, fraud related policies and procedures, previous Department, County, and other reports on fraud, overpayments, and investigative methods and findings.

A glossary of frequently used acronyms is provided at the end of this report.

OVERVIEW OF FRAUD FUNCTION

In Los Angeles County, welfare fraud detection and prevention are performed by a combination of Department of Public Social Services (DPSS) and District Attorney staff. The function is divided into two components: 1) external or recipient fraud, and 2) internal or employee fraud. Recipient fraud cases are investigated by the Department's Welfare Fraud Prevention & Investigation (WFP&I) section. Internal (employee) fraud cases are investigated by the Internal Affairs section of the Department's Human Resources division. WFP&I is part of the Department's Bureau of Health Adult, and Specialized Services. Human Resources is part of the Department's Bureau of Administrative Services.

DPSS Background

DPSS administers Los Angeles County's social services programs. Its major programs are CalWORKs (formerly Aid to Families with Dependent Children or AFDC), Food Stamps, General Relief, Medi-Cal, GAIN, and Adult Services. In the first nine months of 1998, the Department issued benefits to approximately 1.4 million persons, or over 800,000 cases, and received approximately 58,000 applications for aid each month. In September 1998, the value of the benefits issued by the Department totaled approximately $164.5 million for all aid programs. The Department has approximately 10,000 employees. The eligibility and case work staff is housed in 30 offices located throughout the county.

The DPSS has recently begun to implement welfare reform in response to changes in federal program regulations. In addition, the Department is in the midst of implementing a major department-wide new computer system.

The Welfare Fraud Prevention & Investigation Section (WFP&I).

This section has 321 employees, primarily investigators and clerical and administrative support staff. The investigative staff is divided into two groups: 1) Early Fraud/Early Action investigators; and, 2) centralized field investigators. The Early Fraud/Early Action investigators are all assigned to one of the 30 District offices. Investigators are in the District offices and are readily accessible and available to eligibility workers and other District office staff to quickly conduct investigations at the point of intake (Early Fraud) or after aid has been granted (Early Action). They are distinguished from the central field investigators, who are located at the WFP&I headquarters and are responsible for investigations where historical fraud is suspected.

Of the 321 total positions in WFP&I, 120, or 37 percent of the total, are Early Fraud/Early Action investigators and supervisors, and 107, or 33 percent, are central field investigators and supervisors. The other 94 positions, or 30 percent, are administrativesupport and management positions, or staff in the Section's Automation and Prosecution unit.

By design, the caseload for Early Fraud is much higher than that of the field investigators, as the cases assigned to the latter group are more complex and involve more research into case history. Early Fraud and Early Action investigations, on the other hand, are intended for cases that can be resolved very quickly at the District offices, with information that is fairly readily available.

All cases begin as a referral to either the Early Fraud/Early Action or central field investigation staff where they are assigned and processed. If fraud is found and if overpayments have been made, aid can be denied for applicants or discontinued or reduced for ongoing cases (where fraud is discovered at some point after aid has been granted and received). After a case is closed, the Department initiates a collection process to recoup the overpayments.

Caseload and results for WFP&I for FY 1997-98 (the most recent year provided by the Department) were as shown in Table A:

Table A

WFP&I Investigation Caseload Results: FY 1997-98

As can be seen, 53,476, or 72.3 percent of the investigations completed by WFP&I in FY 1997-98, were Early Fraud cases. The other 27.7 percent, or 20,475 completed investigations, were full field investigations. Of the 73,951 total investigations completed, fraud was found and aid was denied or discontinued or court action was taken in approximately 16,358, or 22.2 percent of the cases (combining the 15.2 percent and 7.0 percent as shown in Table A). In 21,169 cases, or 28.6 percent of the investigations completed, fraud or misrepresentation was found but it had no impact on the recipient's eligibility. This means that the recipient provided inaccurate information in some area such as their income or household composition, but the misrepresentation did not change their eligibility or the amount of aid they received.

Finally, in 36,424 cases, or 49.3 percent of the total, fraud charges were not substantiated. Sometimes this conclusion was reached based on lack of credible evidence; in other instances it is reached by default due to the absence of needed evidence such as missing case file documentation. It should be noted that, though it may sound high, this no- fraud-found rate of 49.3 percent is not unusual for most counties. In the third quarter of calendar year 1998, for example, the statewide rate for the number of completed investigations with allegations unfounded or insufficient evidence was 52.3 percent.

Internal Affairs Investigations

Investigations of suspected internal fraud, where a Department employee is suspected of conducting illegal or fraudulent activity, are conducted by a separate section of the Department's Human Resources Division. This is known as the Internal Affairs section and is presently comprised of four investigators and a supervising manager, all overseen by an Administrative Services Manager II. Like WFP&I, the section conducts investigations based on referrals it receives from a variety of sources including other Department employees, aid recipients, and members of the community. The Internal Affairs section's cases do not exclusively pertain to fraud; it is also responsible for referrals and investigations pertaining to any violations of the Department's personnel policies such as misconduct on the job, violation of recipient confidentiality, and others.

In FY 1997-98, the Internal Affairs section opened 65 investigations. Of those, 13 concerned allegations of employee fraud. Neither the Internal Affairs section or any other unit of the Department formally monitors employee activity or transactions for suspicious indicators. If the unit does not receive a referral about an employee's suspected fraudulent activity, an investigation will not be conducted. Referrals come from employees and external sources including anonymous tipsters.

Role of District Attorney's Office

The Los Angeles County District Attorney's Office becomes involved in fraud cases in three ways. Prosecution of recipient fraud cases is conducted by a Welfare Fraud division of the District Attorney's Office. Cases where there is suspected criminal activity, and cases which meet certain other criteria are referred to this division by the WFP&I section of DPSS, by District Attorney investigators, and by outside sources including law enforcement agencies, and anonymous sources.

The decision to prosecute a case is discretionary, but generally is based on the following criteria:

An interagency agreement between DPSS and the District Attorney's Welfare Fraud division calls for the DPSS to refer cases of suspected fraud in the AFDC program, its replacement program CalWORKs, or Food Stamps cases to the D.A. and to coordinate the review and prosecution of these cases. The following staffing for this function is stipulated in the agreement: 1 Head Deputy District Attorney, 10 Deputy District Attorney positions, 11 paralegal and support positions.

A second interagency agreement between DPSS and the District Attorney's Office covers investigation of major welfare fraud cases. The agreement calls for the District Attorney to provide an investigative staff of 31 to fully investigate allegations of welfare fraud and other related crimes. These investigators work side by side with WFP&I investigators in office space provided by DPSS for D.A. investigators at its WFP&I headquarters office on Imperial Highway.

A third way the District Attorney's Office is involved in fraud investigations is through a new internal fraud unit of its Special Investigations division. This unit was created in September 1998 on a pilot basis. Its agreement with the DPSS requires investigation and prosecution of internal fraud cases at selected District offices. The unit is staffed by one attorney, five investigators, and one supervising investigator. Since October, this new unit has received 14 case referrals that it is pursuing. Referrals come from the Internal Affairs section of DPSS, law enforcement agencies, and outside sources.

Department Efforts to Combat Fraud

The Department has a number of initiatives underway or planned to address both external and internal fraud. Chief among these are two new information systems that will either establish new internal controls or allow for greater centralized monitoring of unusual or suspicious employee and participant activity. These systems are the new Los Angeles Eligibility, Automated Determination, Evaluation and Reporting System (LEADER) and the Welfare Fraud Linkage Analysis Database System (WFLADS). LEADER will automate the Department's manual, paper-intensive eligibility-determination and case-processing functions. A number of key internal controls will be built into LEADER including passwords for all employees, a record of all case transactions, a requirement that supervisors authorize new cases and changes to existing cases by entering a personal identification number, and others. Information in the LEADER database will be much more accessible and flexible, enabling analysis and investigations that cannot be readily performed at present with manual records and less flexible computer systems.

The primary purpose of WFLADS will be external, or participant, welfare fraud investigations. The system will download data from LEADER, enabling WFP&I investigators to manipulate and analyze the date in many ways, such as by participant, by office, by case number, etc. This will replace limited electronic searches and manual reviews of files that are current techniques used by investigators.

Another step that the Department has recently taken as a more aggressive stance in fighting fraud is the interagency agreement entered into in 1998 which established a new internal fraud investigation unit in the District Attorney's Office. This pilot project establishes a unit in the District Attorneys Office dedicated to prosecution and investigation of DPSS staff engaging in fraud at selected departmental District offices.

In response to the arrest of an eligibility worker accused of defrauding the Department of an estimated $700,000 in benefits, the Department established an Internal Security Task Force which has developed recommendations for eliminating some of the weaknesses in internal controls that allowed this level of fraud to occur. The Task Force will continue to operate for some time in the future to oversee implementation of their recommendations for improvement.

All of the Department's initiatives have, or will improve its efforts at combating fraud. To the extent they are applicable, the Department's initiatives identified above are integrated into the findings and recommendations in this report.

Recommendation

It is recommended that the DPSS develop and implement an incentive program that will motivate employees to report cases of suspected employee fraud.

RECIPIENT FRAUD INVESTIGATION RESULTS

A comparison of Los Angeles County welfare fraud investigation statistics to other Southern California counties and the State was conducted as part of this audit. The results showed that fewer investigation referrals are made to DPSS' Welfare Fraud Prevention & Investigation section (WFP&I) relative to caseload than is the case in other counties. As shown in Table 1.1 for the Aid to Families with Dependent Children (now CalWORKs) program, the Los Angeles County referral rate is lower than that of neighboring Southern California counties and the statewide average.

Finding 1.: The rate of investigation referrals to DPSS's Welfare Fraud Prevention & Investigation section for AFDC/CalWORKs cases was five percent of caseload compared to a median rate of 13.5 percent for other Southern California counties and 7.9 percent for the total State in 1998. Assuming that the rate of welfare fraud is probably not lower in Los Angeles County than neighboring counties, this lower referral rate can be interpreted to mean that many fraud cases are not being prevented or detected resulting in unstopped overpayments.

AFDC/CalWORKs Case Referrals to Investigations Units Compared to Caseload: Los Angeles County, Other Southern California Counties, and Statewide Total; July -September Quarter 1998

Table 1

Though the difference between the Department's 5 percent and the statewide total of 7.9 percent of caseload may not appear significant, the difference represents a significant amount of potential overpayment and cost avoidance. Using the Department's own experience from 1998, the additional referrals that would be generated at 7.9 percent ofcaseload instead of 5 percent represent approximately 27,842 additional referrals per year. Of these additional investigations, approximately 15 percent, or 4,176, would result in aid being denied or discontinued. Based on the Department's own FY 1997-98 average overpayments detected and costs avoided of $9,695 per case for all types of cases, these 4,176 cases would generate an additional $40,486,320 in overpayments detected and costs avoided (4,176 cases x $9,695 per case).

The Department's lower rate of referrals can possibly be attributed to several factors:

The Department has pointed out that its Automated Fingerprint Image Reporting and Match System (AFIRM) could be reducing the Department's referral rate by discouraging some applicants who would otherwise submit a fraudulent application. However, the fingerprint system is designed to detect and stop fraudulent duplicate aid applications. Investigation referrals in other counties that do not have fingerprint systems (which is all other counties) do not contain more duplicate aid investigations than Los Angeles. Rather, their case referrals are concentrated in the same two areas investigated most often in Los Angeles County: unreported income and household composition. The presence of the AFIRM system in Los Angeles County does not appear to explain the County's lower rate of case referrals.

With the exception of duplicate aid cases, there is no reason to assume that the rate of case referrals should be lower in Los Angeles County than elsewhere and one would expect that, with proper protocols and staff training, the investigation referral rate should mirror that found in other comparable counties or the statewide average. To the extent that referrals are not being made due to lack of knowledge on the part of eligibility workers or their not taking time to review cases for potential fraud, District managers should be accountable for these lapses. Management reviews of referral rates by District office to determine where there are lower than average rates, or higher than average rates but with inappropriate referrals, should be flagged and brought to the attention of these District managers.

Recommendations

Finding 2: The proportion of Department investigations where fraud allegations are determined to be unfounded, or where fraud is found but without financial impact, is higher in Los Angeles County than other counties. The number of investigations resulting in benefits being denied is substantially lower for Los Angeles County than for other Southern California counties and the State as a whole. There is a need to make better use of WFP&I investigator's time so that the results of their work are more comparable to other Southern California counties and the statewide average. Achieving the statewide Early Fraud denial rate of 22.8 percent in Los Angeles County would eliminate estimated ineligible benefits payments of approximately $46.6 million per year.

Table 2

Investigation Results: Los Angeles, Other Selected Counties and Statewide Total

APRIL - SEPTEMBER 1998

-

In an analysis of county and state data, it was found that DPSS has a lower rate of denying or reducing aid after concluding its investigations and a higher number of cases where the allegations prove unfounded, evidence is insufficient to substantiate the allegations, or misrepresentation or fraud is found, but it does not affect the recipient's eligibility. Table 2 presents the results of this analysis.

As the Table shows, the percentage of cases where benefits were denied, reduced or discontinued was lower in Los Angeles County at 34.2 percent of completed investigations compared to the 42.9 percent median rate in other Southern California counties and the 41.1 percent statewide rate. This phenomenon is also evidenced by the high rate of completed investigations where allegations proved unfounded. As shown in Table 1.2, 60.4 percent of all referrals statewide resulted in allegations being unfounded and 53.9 percent of all cases had such results in other Southern California counties. In Los Angeles County, a much higher 72.7 percent of all welfare fraud investigations resulted in the allegations being unfounded.

The data in Table 2 suggests two possibilities: 1) that DPSS is getting a higher proportion of inappropriate referrals, or 2) the Department's investigators are not conducting the investigations with the same thoroughness as their counterparts in other counties. Both areas need to be addressed by Department management.

A further analysis of investigations resulting in denials, reductions and discontinued benefits revealed that the major area where Los Angeles County is lower compared to other counties is Early Fraud investigations that result in benefit denials. Table 3 shows comparative Early Fraud investigation results for Early Fraud cases. As can be seen, only 8 percent of Early Fraud cases in Los Angeles County resulted in denied benefits compared to a 36.3 percent median for other Southern California counties and 22.8 percent of all cases statewide. While it has higher rates of reducing and discontinuing benefits than the other counties, the overall Early Fraud effectiveness rate (combining all three categories) is still substantially lower than the other counties and the statewide total.

Early Fraud denials are a much more effective means of eliminating overpayments as aid is cut off at the point of intake. Cases get more complicated, require more investigation time, and have a reduced probability of recovery when overpayments are discovered after a recipient is already on aid. The number of Early Fraud investigations resulting in discontinuances and benefits reductions are generally not going to result in the same level of avoidance of fraudulent aid payments as an Early Fraud denial.

Table 3

Comparative Early Fraud/Early Action Investigation Case Results: Southern California Counties, Los Angeles County, and Statewide Total, April - September 1998

These data show that DPSS' review of its investigation referral procedures and investigation processes mentioned above should be particularly focused on Early Fraud cases. Third party reviews of case work and techniques used by WFP&I investigators are needed to ensure that all cases are being thoroughly investigated. Similarly, ongoing training for eligibility worker staff is needed on signs of potential recipient fraud to ensure that all appropriate referrals are being made at District offices.

Assembling a task force comprised of senior WFP&I investigators and investigators from the District Attorney's welfare fraud unit to review samples of WFP&I investigation case files, assess the techniques used and the thoroughness of investigations conducted would provide assurances that the high rate of investigations resulting in unfounded allegations is not due to less than thorough investigations by WFP&I investigators. Alternatively, if such a review is conducted and less than thorough investigations are found, it will provide an opportunity to improve upon the Department's investigation policies, procedures, and approaches.

Achieving the State Early Fraud denial rate of 22.8 percent as shown in Table 3 would increase the amount by which the Department avoids ineligible benefits payments. The net change would amount to cost avoidance of an estimated $46.6 million, as detailed in Table 4. This assumes that an increase in Early Fraud denials would be partially offset by a decrease in Early Fraud discontinuances and grant reductions. Even with those reductions, achieving the state rate of Early Fraud denials would still result in a significant decrease in ineligible benefits payments. Any costs associated with achieving the state rate are not included in these calculations but they are assumed to be significantly less than the overpayments avoided.

Table 4

Impact of Applying Statewide Early Fraud Denial Rate to LA County

Impact of
Applying State Rate to LA

  # % # % #
       
Early Fraud Denials 22.8% 7,628 8.0%
Early Fraud Discontinued 7.6% 11,932 12.4%
Early Fraud Reductions 3.6% 6,388 6.7%
Total   25,948  
x FY 1997-98 average overpayment avoided by DPSS     $ 7,004
Total overpayments avoided         $46,580,242

Recommendations

Finding 3: The Department's investigations identify higher average overpayment amounts than other counties, indicating that fraudulent recipients receive relatively more aid in Los Angeles County before they are detected. The longer recipients are on aid and the higher the amount to be collected, the lower the probability of full overpayment recovery. The Department currently has an outstanding balance of approximately $325 million in outstanding overpayments to collect for the Aid to Families with Dependent Children program, of which $225 million is estimated by the Department to be uncollectible due to no payments on those accounts for the last three years.

As another means of assessing the impact of the Department's welfare fraud investigations, a review of the Department's overpayment amounts and collections was conducted. As shown in Table 5, total grant reduction and cash collection amounts comprised only 5.2 percent of the Department's overall balance of overpayments in calendaryear 1998. And $26.4 million worth of new overpayments identified through investigations exceeded the approximately $16.3 million of collections and grant reductions (the total of $11.4 million in grant reductions and $4.8 million in cash collections in Table 5) by approximately $10.2 million. Even for active cases, new overpayments identified exceeded collections by $4.9 million due to limits on the amounts active case grants can be reduced (and because some recipients protest their overpayment finding which gives them a right to a hearing before their grant amount is reduced).

As one would expect, collections on closed cases is worse than on active cases, with collections in 1998 amounting to only 1.8 percent of the outstanding balance of overpayments on closed cases. The $9.8 million in new overpayments identified on closed cases was almost double the amount collected.

It should be noted that State regulations permit "write-offs" of overpayment accounts on closed cases with no repayment activity within the last three years. The Department estimates that all but $100 million of the outstanding $325 million overpayment balance meet this criteria, further reinforcing the observation that the potential for recovery decreases over time.

Table 5

Collections on Overpayments, Calendar Year 1998

  Active Cases Closed Cases All Cases
Starting Overpayment Balance $63,637,929 $251,658,110 $315,496,039
New Overpayments Identified $16,639,534 $9,813,333 $26,452,867
Grant Reduction Amounts $11,431,392 - $11,431,392
Cash Collections $255,618 $4,569,926 $4,825,544
Total Reductions & Collections $11,687,010 $4,569,926 $16,256,936
C as % Overpayments Identified 70.2% 46.6% 61.5%
C as % Starting Balance 18.3% 1.8% 5.2%
Gross Ending Overpayment Balance $68,590,453 $256,901,517 $325,491,970

The data in Table 5 shows that the prospects for collecting all overpayments are poor, even for active cases where grants can be reduced. The prospects get much worse for closed cases. All of this demonstrates the importance of stopping as many ineligible cases as possible at the point of intake, before they are started and where potential overpayments are completely prevented instead of partially recovered through the actions that are taken after aid is awarded.

Recommendation

WFP&I CENTRAL OFFICE OPERATIONS

According to State regulations, public assistance fraud exists when a person, on behalf of himself or others, has:

To examine the Welfare Fraud Prevention and Investigation section's (WFP&I) procedures for investigating aid recipients suspected of committing fraud, a group of cases was randomly selected from the fraud investigation case files referred to WFP&I by two of the County's 29 District offices, Metro Family (District 13) and Rancho Park (District 60). A total of 165 suspected fraud cases referred to WFP&I from the District offices between 1996 and 1998 were reviewed. These two District offices were selected in order to provide a mix of programs, geographic location, and populations served. The Metro Family Office is located near downtown Los Angeles and provides services to families receiving CalWORKs, Medi-Cal, and/or Food Stamps. The Rancho Park Office is located on the Westside of Los Angeles and serves adults with no dependent children, the majority of which receive General Relief, Medi-Cal, and/or Food Stamps. The Rancho Park Office is also responsible for a satellite office in the San Fernando Valley which was not part of the scope of this review.

The majority of WFP&I investigators are divided into two broad categories: (1) those stationed in the District offices and charged with investigating current suspected fraud with the goal of preventing or halting the issuance of benefits to ineligible individuals, called Early Fraud/Early Action Investigators; and (2) those located in the central WFP&I office who investigate cases of past fraud in order to recoup overpayments and assemble evidence for possible prosecution, as well as investigate cases of current fraud which are not resolved by investigators and staff located in the District offices. The bulk of WFP&I investigationsare performed by the Early Fraud/Early Action Investigators in the District offices. For example, between July and September 1998, nearly 73 percent of all fraud investigations were performed by Early Fraud/Early Action Investigators stationed in the District offices.

Nearly all of the referrals received by investigators in the central WFP&I office from the District offices are from two sources: (1) the Early Fraud/Early Action Investigators; and (2) Eligibility Workers in the District's unit responsible for reviewing alerts of unreported income and aid, called the Income and Eligibility Verification System (IEVS) Unit. IEVS is a statewide system required by federal law which identifies persons who are receiving unreported earnings, other government benefits, or duplicate aid from another county. For a further discussion and findings related to the IEVS system, see the section of this report dealing with Recipient Fraud Referrals and the IEVS system.

In addition to the referrals received from the District offices, WFP&I investigators handle investigations of fraud referrals from the WE-TIP fraud reporting hotline, Central Fraud Reporting Line, high profile investigations, and cases already found that need prosecution.

WFP&I Experiencing a Large Backlog of Fraud Investigations

Finding 4: Between July and September of 1998, there was a monthly average of approximately 31,600 cases pending fraud investigation, nearly six times the average number of investigations completed each month. The large backlog of cases appears to be causing significant delays in beginning fraud investigations. A review of a sample of fraud referrals found that the median amount of time between the date a high priority fraud referral was received in the central WFP&I office and the date it was assigned to an investigator to begin an investigation was over one year.

According to statistics reported to the State, from July to September of 1998, the Department had an average of approximately 63,200 cases per month pending completion of fraud investigation. As shown in the table below, Los Angeles County's average monthly number of cases pending investigation during that period was nearly six times, or 575 percent, the average monthly number of investigations which were completed during that quarter. Comparing new cases accepted to the number of investigations completed, Los Angeles County was able to complete investigations on only slightly more cases per month than it received, making only a small impact on the backlog of cases. As shown in Table 6, all of the other southern California counties listed are also experiencing a backlog, with more cases pending each month than they are able to complete. However, compared to other southern California counties, Los Angeles County's number of pending investigations as a proportion of the number of investigations completed is significantly higher.

Table 6

Fraud Investigation Workload, Selected Counties, July - September1998

County
Average Monthly # of New Cases Accepted for Investigation
Average Monthly # of Completed Investigations
Average Monthly # of Investigations Pending
Pending Investigations as a % of Average Monthly Completed Investigations
Riverside
5,057
5,160
2,864
55.5%
San Diego
3,552
3,449
3,829
111.0%
Ventura
261
231
453
195.7%
Orange
2,727
2,492
6,756
271.1%
San Bernardino
3,317
3,395
14,356
422.9%
Median      
195.7%
Los Angeles
10,858
10,988
63,210
575.3%

The large backlog of potential fraud cases pending investigation in Los Angeles County is reflected in the length of time it takes from the date a case is received by WFP&I until an investigator is available to be assigned the case. When a fraud referral is received by WFP&I, an intake investigator reviews each file and assigns a priority code to the case. The case is then grouped with other similar cases and assigned to the next available unit of investigators who specialize in the type of investigation required. To determine the amount of time that elapsed between receipt of a case in WFP&I and the date that the case was assigned to an investigator, a sample of fraud referrals from the Metro Family District Office to WFP&I was examined. The results are shown in Table 7 on the next page.

Table 7

Metro Family District Office Fraud Referrals to WFP&I

Number of Days Between Date Case Received and Date Assigned to an Investigator, By Assignment Priority

* Table includes only 94 of 121 Metro Family District Office fraud referral cases in sample due to omitted information and files which had not yet been assigned to an investigator.

** All referrals from Beneficiary Earnings Exchange Record, Payment Verification System, Franchise Tax Board, Internal Revenue Service, Central Fraud Reporting Line, State Hotline, We-Tip, Multiple Aid, EAU, AFIRM, Border Crossing, Child Only, PA 230, High Risk, Fictitious Persons or Documents, Questionable Current Eligibility, or Administrative Assignments.

As shown in Table 7, a review of a sample of fraud files found that half of all fraud referrals assigned the highest priority (Priority "A" took more than one year from the date that they were referred to WFP&I to the date that they were assigned to an investigator to begin an investigation. It should be noted that cases are also prioritized by investigators within each priority code based on the amount of overpayments involved, therefore a Priority "A" case with a potential overpayment in the thousands of dollars will be assigned to an investigator before a Priority "A" case with a lower estimated dollar value. The median length of time between receipt and assignment of Priority "B" cases was shorter than for Priority "A" cases, at 188 days.

All except one of the Priority "C" and "D" cases were closed upon receipt by the WFP&I intake unit without being assigned to an investigator. These referrals fell under the monetary threshold for investigation by WFP&I which was $2,000 at the time the cases were reviewed. Currently, cases with overpayments estimated at under $300 are not considered cost effective for WFP&I investigators to pursue. Further discussion and recommendationsconcerning referrals which fall under the monetary threshold appear in IEVS section of this report.

Missing Case Documents Result in Dismissed Fraud Investigations

Finding 5: A sampling of fraud cases found that 15 percent of investigations had to be suspended and the cases closed because the case documents required to investigate the allegations were missing.

In order to investigate potential fraud cases, an investigator must first obtain from the Department's record storage system the recipient's case file and documents pertaining to the period when the alleged fraud was committed. A review of randomly selected fraud referral files from the Metro Family and Rancho Park District offices found that out of 114 cases in which the investigation had been concluded, 19 of the cases, or 17 percent, were closed with negative fraud findings because the investigator was unable to locate the pertinent case information required for the investigation. Of the 19 cases, 11 were closed because the investigator was unable to locate the entire case file for the alleged fraud period and 8 were closed because, although the proper case file was located, the pertinent documents which should have been contained in the file were missing.

Investigators and attorneys from the District Attorney's office, WFP&I and Internal Affairs all cited missing case files and/or case file documents as a recurring and frustrating problem in performing their investigations. Due to time limitations, the Grand Jury was not able to pinpoint the source of the problem for missing files. However, possible explanations are: (1) District eligibility workers and staff are misplacing files; (2) the Department's contracted records storage vendor is misplacing files; or (3) files are being removed to cover up fraudulent activities.

Closed case files and open case historical records are sent by the District offices to a vendor for long term storage. Staff in the District offices prepare a transmittal letter documenting each file or box of files before transferring it to the vendor for storage. The information is then entered into the Department's computerized file management system which records and tracks the location of case files. In addition, the vendor provides the District with a transmittal log which records which files have been transferred to the vendor. Therefore the District office should have a complete record of where the file is and if it was received by the vendor.

When an investigation begins the computerized file management system is consulted to determine the location of the file. If the system indicates that the file is stored with the vendor a request is made to the vendor for the case files. The vendor then either delivers the case file or reports that the case is not in storage. Investigators follow a policy of making upto three requests to the vendor for a case file. This allows for the possibility that the District office may not have yet sent the file to the vendor for storage. If, after three requests, the vendor still reports that it does not have the file, the investigator closes the case with negative fraud findings because the investigation can not proceed without the evidence contained in the case files.

As noted above, eight of the cases reviewed were closed with negative fraud findings because, although the case file was located, the pertinent documents which were supposed to be contained in the case file were missing. This may in part be attributed to workers in the District offices misfiling or neglecting to file the documents. However, missing case files and documents raise the possibility that the files may have been removed to cover up fraudulent activity. Currently, the Department does not track missing case files by office or worker as a potential source for internal investigation. A caseworker with a number of cases which can not be investigated because documents are missing should be a target for internal fraud investigators.

When a case must be closed with negative fraud findings due to missing files, the investigator should contact the District office to determine if the file was lost by the District or by the vendor. This information should be compiled by the Department to identify the source of the problem and hold the vendor or District office accountable for losing files. Additionally, in such cases WFP&I should keep a record of the caseworker assigned to the case in order to alert internal fraud investigators to potential fraud.

Recommendations

WFP&I Unable to Account for All Fraud Investigation Files

Finding 6: WFP&I is not able to account for all of its files documenting fraud investigation - 15 percent of such files requested for this audit's sample could not be located by the Department. These files are needed for any subsequent fraud investigations and to provide information to eligibility workers and investigators on past fraudulent activity when a former recipient reapplies for aid.

WFP&I policy is to keep files with positive fraud findings closed for ten years and those with negative fraud findings for five years. These files are stored in case they are needed for subsequent fraud investigations, for reference and documentation in overpayment cases, or in case a District office requests background information on a former recipient's case who was found to have committed fraud and is now reapplying for aid.

For the sample requested for study, the WFP&I could not locate 25 of the 165 requested fraud investigation files, or 15 percent. The requested fraud files were cases referred for investigation within the past four years and, therefore, according to WFP&I's file management policy, all should have been on hand. WFP&I provided a computer printout for each file containing information which indicated that all of the missing fraud files had been processed by WFP&I and received a final disposition although the actual files documenting the investigation could not be produced. This is a red flag for possible internal fraud.

Recommendation

RECIPIENT FRAUD REFERRALS AND THE INTEGRATED ELIGIBILITY VERIFICATION SYSTEM (IEVS)

From July to September 1998, nearly 20 percent of all fraud referrals to the Welfare Fraud Prevention and Investigation Division (WFP&I) were cases of potential fraud identified by the Income and Eligibility Verification System (IEVS) and referred to WFP&I by the Department's District offices. IEVS is a statewide system required by federal law, which identifies persons who:

Four times a year (quarterly), Los Angeles County sends a record of all CalWORKs and Food Stamp recipients, who were eligible, to the California Department of Social Services. If there is a discrepancy between the employer-reported earnings on record with the State Employment Development Department and earnings reported by the recipient to the County, a report detailing the discrepancy is generated. This report, called an Abstract, is also generated periodically if the system finds that the recipient is receiving unreported aid or aid in another county. The primary drawback of the IEVS system is that, due to the time it takes for the State to process earnings and aid information, the information which appears on the Abstract is typically seven to nine months old by the time it is received by the County. Therefore overpayments may go undetected for quite some time.

The Abstracts are processed in each District by a specialized unit of IEVS Eligibility Workers. The IEVS Eligibility Workers are responsible for ensuring that the name and social security number on the Abstract match that of the aid recipient. In the case of unreported earnings, the amount of earnings reported to DPSS by the recipient is compared against the amount of earnings reported by their employers to the Employment Development Department. If a discrepancy is found, the IEVS Eligibility Worker notifies the current or former participant in writing who then has 10 days to respond. If a recipient who is currently receiving aid does not respond, aid is terminated. If the recipient disputes the information, the income or benefit source is contacted to verify the information. For any month(s) where the unreported income or benefits can be verified, the IEVS Eligibility Worker takes the necessary corrective case action, including computing overpayments, overissuances, and reducing grant amounts to recoup the overpayments. In all cases, if a confirmed discrepancyexists, or if it can not be confirmed that a discrepancy does not exist, a referral to WFP&I for further investigation is to be initiated. Only if an employer indicates that a recipient was not employed is no action is taken.

DPSS policy mandates that Abstracts on currently active cases receiving CalWORKs or General Relief be processed within 45 calendar days from the run date of the report. Terminated cases are to be processed within 90 days, however processing of Abstracts may be, and often are delayed beyond the time limits if the recipient disputes the IEVS information. Should the IEVS Eligibility Worker find that a recipient has unreported earnings or benefits, the worker must also compute the overpayment or overissuance and initiate a referral to WFP&I.

IEVS Abstracts Not Processed Timely by District Offices

Finding 7: As of November 1988, DPSS had 184,203 computer generated reports from the State waiting to be reviewed. These reports, called IEVS Abstracts (Integrated Eligibility Verification System)< alert the County that aid recipients had unreported earnings, received other government benefits, or received duplicate aid in another county. Currently the Department is only able to process 30 percent of its pending IEVS Abstracts each quarter.

Finding 8: DPSS is not in compliance with County regulations which require processing of such reports within 45 to 90 days of receipt. In a sample of files, the median amount of time between the run date of the report and the date that a case was referred for fraud investigation was 570 days (19 months) and ranged from 30 days to over seven years. The delay in processing and referring such reports can result in fraud going undetected for a long period of time. With the passage of time, the amount of overpayments and the likelihood of collecting overpayments is reduced.

In 1993 there was a considerable backlog of Abstracts waiting to be processed by the District offices. According to the Department's Research and Statistics Section, the backlog for all 30 Districts combined as of January 1993 was over 91,000 IEVS Abstracts. As of November 1998 (the most recent quarter available) the backlog had doubled to 184,203 Abstracts waiting to be processed, while the Department's total caseload decreased by 17 percent during that same period. For the fourth quarter of 1997, the Districts received a total of 45,788 new Abstracts but processed only 36,343. Table 8 shows the volume of Abstracts processed from that reporting period.

Table 8

IEVS Abstract Processing, Los Angeles County

October-December 1997 Reporting Wage Match Quarter

As shown in Table 8, the Department received more cases than it was able to process during the quarter, resulting in an ending balance which was nearly 10,000 cases higher than at the beginning. The Department's ending balance of 184,203 cases was over five times the number of cases (36,343 cases) it was able to process during that quarter.

A review of reports provided by the Department which detail IEVS Abstract processing by District office show that, of the IEVS reports generated for the October through December 1997 quarter (the most recent available), only 6 of 29 Districts processed more than 50 percent of the IEVS Abstract workload assigned to them. On average, the District offices only processed 17 percent of their pending unreported income Abstracts. Some District offices are able to consistently process over 70 percent of their pending Abstracts while other process under 10 percent, indicating that there is a problem with the current allocation of IEVS Eligibility Workers among the District offices and/or varying Abstract processing procedures in different offices.

DPSS reports showing the number of IEVS Abstracts processed for active cases (individuals currently receiving aid) and inactive cases (individuals no longer receiving aid) were also reviewed. Although IEVS matches on active cases are assigned the highest priority, less than 20 percent of pending Abstracts on active cases were processed during the fourth quarter of 1997. As noted above, DPSS policy mandates that IEVS Abstracts on currently active cases receiving CalWORKs or General Relief be processed within 45 calendar days from the run date of the report and terminated cases are to be processed within 90 days, unless the recipient disputes the IEVS information. It is evident from the number of IEVS matches on both active and inactive cases which remain unprocessed at the end of each quarter that DPSS is not in compliance with County regulations for processing matches on CalWORKs, Food Stamp, and General Relief cases.

In order to determine how long it takes for an IEVS Abstract to be processed, a sample of 37 IEVS Abstract fraud referrals from the Metro Family District Office were examined. It was found that the median amount of time between the run date of an Abstract (the date the State generated the report) and the date that a case was referred by the Districtto WFP&I for fraud investigation was 19 months and ranged from one month to over seven years.

If IEVS match reports are not processed in a timely fashion, there is a risk that thousands of individuals may be receiving overpayments or overissuances in Los Angeles County. Of the total number of IEVS Abstracts processed over four quarters in 1997, 18 percent of the Abstracts were found to lead to the detection or prevention of an overpayment or overissuance. This represents an enormous number of CalWORKs and General Relief recipients who are continuing to receive aid while earning unreported income or, to a lesser degree, recipients receiving duplicate aid in another county. Furthermore, the longer the elapsed time between when fraud was committed and collection activities are initiated, the lower the probability of collecting overpayments.

There are currently 150 IEVS worker positions in the Department. Based on processing of Abstracts from the October through December 1997 reporting wage match quarter, the total number of Abstracts processed was 38,956, or 86 per IEVS worker per month. In order to process the total workload of 212,150 Abstracts during the quarter, each worker would have to process five times as many, or 471 Abstracts per month.

Recommendation

New Hire Registry

Effective July 1, 1998, all employers in the State of California are mandated to report to the State Employment Development Department those individuals hired or rehired within 20 calendar days of employment, known as the New Hire or New Employee Registry. This includes all business, state and local government, and not-for-profit organization employers, regardless of the number of employees. Every newly hired or rehired employee must be reported regardless of age or projected wages and includes employees who work less than a full day, part-time employees, seasonal employees, or those who discontinue their employment prior to the 20th day of employment.

Like the IEVS system, the State will match New Hire Registry Reports against County aid records. The New Hire Registry is an improvement over the IEVS System since the quarterly earnings information reported by the IEVS system is six to nine months out ofdate before the County even receives it. By detecting employment early, the goal of the New Hire Registry is to reduce the time frame in which fraudulent unreported earnings or employment is undetected and avoid large aid overpayments.

The Department anticipates that the New Hire Registry will create a reduction in the currently quarterly volume of IEVS match workload, by alerting the County to changes in a recipient's employment status before earnings are reported. However, the New Hire Registry Reports are handled by the same IEVS workers responsible for processing the IEVS Abstracts. As described above, IEVS workers are not able to keep up with the current volume of those reports. A further complication is that the State now generates a New Hire Registry Report for all persons receiving aid in Los Angeles County and does not sort out those who were previously so reported to the DPSS. The State and County are working to resolve this problem and only generate reports for individuals with unreported employment. However, the Department is presently receiving an estimated 30,000 New Hire Registry Reports per month in addition to its IEVS Abstracts. The Department reports that it hopes to receive approval from the State to hire additional workers to offset the increased workload created by the New Hire Registry in its FY 1999-2000 budget.

Finally, the success of the New Hire Registry is dependent upon employers reporting new hires. State officials estimate that the majority of employers in the State are reporting such information and report that they are awaiting implementation of a new Federal program which will identify those employers not in compliance who may then be fined.

IEVS Abstract Fraud Referrals to WFP&I

Finding 9: In a sample of fraud referrals, 36 percent of cases referred for investigation were closed immediately upon receipt because they were determined to be not cost effective to pursue. Many of these cases had already been reviewed and overpayment collection action taken through grand reductions by the District office. Processing of those cases by investigator staff when they have already been resolved by the District offices is unnecessary and consumes resources which could instead be redirected to processing and investigating other potential fraud cases.

Over 70 percent of all investigations referred to WFP&I are Early Action/Early Fraud investigations, over 20 percent are a result of IEVS Abstracts, and 10 percent are from other sources. Because all Early Action/Early Fraud investigations are handled by WFP&I investigators in the District offices (except in those cases where historical fraud is suspected), a large proportion of fraud referrals received by the central WFP&I office are the result of IEVS Abstracts.

As described earlier in the introduction to this section, even if an IEVS Eligibility Worker can establish that an overpayment occurred and takes action to collect the overpayment by a reduction in benefits or other means, a fraud referral is still generated to WFP&I. In other words, all cases in which a discrepancy was found, whether or not they are resolved by the District, are referred to WFP&I.

All cases referred to WFP&I are received by the Section's Intake Unit. Six investigators and one supervisor staff the WFP&I Intake Unit and are responsible for creating a cover sheet, called a Facts Input Document (FID). The Intake Investigators review each referral and note on the FID the priority level, type of allegation, and other information from the incoming referrals before assigning the case for further investigation.

Although IEVS Eligibility Workers must refer all IEVS Abstract cases with discrepancies to WFP&I, if the value of the benefits or cash earned during a quarter are under a certain monetary threshold, the case is immediately closed without further investigation. The monetary threshold is set by the Department which raises or lowers it depending on the caseload in WFP&I and the level at which it is determined to be cost effective to pursue an investigation. Beginning in January 1999, the WFP&I intake unit was closing without investigation all cases involving unreported income or other allegations of overpayments of $300 or less.

Of the sample of fraud referrals from the Metro Family and Rancho Park District Offices examined, 52 out of 143 fraud referral cases, or 36 percent, were closed immediately upon receipt by the WFP&I Intake Unit and coded as an administrative close. For the cases reviewed, the threshold for investigation was $2,000 or $3,000, depending on when the case was reviewed. Of the sample cases, all 52 of these referrals were generated by IEVS Abstract matches and nearly all concerned unreported income of $999 or less for a quarter or other allegations of andling of cases which are under the threshold could be improved by directing IEVS Eligibilioverpayments under that amount.

Although it does not require a great amount of time for an intake investigator to determine that a referred case is below the monetary threshold, the volume of files referred from the Districts to WFP&I that fall under this category cumulatively represent a significant amount of time spent by the intake investigators simply completing paperwork to close files. In addition, such referrals consume the time of District IEVS Eligibility Workers and clerical staff who must also complete paperwork to initiate a referral. Furthermore, the threshold treats all cases equally, regardless of the disposition of the case and whether or not it could benefit from further investigation. For example, cases where the IEVS Eligibility Worker was unsuccessful in determining that an overpayment occurred (which might be established through further investigation) are closed along with those where the IEVS Eligibility Workers have resolved the case, including determining that an overpayment has occurred and initiating the collection process.

Hty Workers to only refer those cases under the monetary threshold which they believe are likely to benefit from further investigation. In cases where overpayments have been firmly established and collection action has been initiated by the IEVS Eligibility Workers, there is no reason to believe that a case would benefit from further investigation and the case should not be referred to WFP&I. Those cases which fall under the threshold under which the IEVS Eligibility Workers suspect but are not able to establish fraud should be forwarded to WFP&I. The WFP&I Intake Investigators should then randomly select a percentage of such files for further investigation. These changes would free up district and WFP&I staff time, sorely needed to deal with the work backlogs in both of those departments, as previously discussed in this section and in the section on the WFP&I Central Office Operations.

Recommendations

FRAUD INVESTIGATIONS AND SAN DIEGO COUNTY'S PROJECT 100

Introduction

Each District office has one or more WFP&I investigators on site. These investigators are responsible for (1) preventing cash payments or Food Stamp issuances to new applicants who provide fraudulent information (called Early Fraud investigations), and (2) stopping or reducing aid to current recipients who are found to be committing fraud (called Early Action investigations). It is these on-site investigators who perform the bulk of the investigative work conducted by the WFP&I Division. These investigators are primarily concerned with identifying fraud which affects current eligibility and therefore preventing or stopping overpayments.

On-site investigators receive the majority of their referrals from eligibility workers in the District Offices. Other sources of referral include calls received by the County's Confidential Fraud Referral Line and the County's computerized fraud detection systems such as fingerprint matches and address matches. If the on-site investigator determines that fraud has taken place, a recommendation is made to the District staff to either deny the application or, for recipients currently receiving aid, to terminate, reduce, or re-evaluate benefits. If the investigator suspects or determines that an overpayment or over issuance of benefits may have occurred in the past, the case is referred to the central WFP&I office for further investigation and collection action.

Early Fraud and Early Action Investigations, Metro Family and Rancho Park District Offices

The Metro Family Office, which serves CalWORKs recipients, has five on site investigators plus a supervisor. The Rancho Park office, which provides services to a smaller population than Metro Family Office, primarily serves General Relief recipients and has one investigator on site.

Early Fraud

Early Fraud refers to the investigation of suspects at the time of application for aid, before benefits are granted. Referrals for Early Fraud investigations are typically made by an intake worker who suspects that an applicant is providing false information. An investigation may also be initiated if the applicant has applied for or received aid in the past and was found or suspected of providing fraudulent information. Such information appears as an alert code attached to the applicant's case number in the Department's computersystem, WCMIS. Investigations are handled by the on-site investigators located in the District offices.

To evaluate the Early Fraud activities in the District offices, a randomly selected sample of 50 investigation files from the Rancho Park and Metro Family Center offices was examined. Overall, 44 percent of all Early Fraud investigations in the sample resulted in positive fraud findings. However, it should be noted that positive fraud findings do not always mean that an overpayment has been prevented or detected. For instance in some cases it may be determined that a person has unreported earnings but the earnings may be low enough to not impact the recipient's benefit level. The results of the sample are shown below.

Table 9

Early Fraud Investigations Sample

Metro Family and Rancho Park Offices

 
Positive
Negative
Total

#

 
#
%
#
%
  
Metro Family
9
36%
16
64%
25
Rancho Park
13
57%
10
43%
23
Total
22
46%
26
54%
48

Note: Positive cases are defined by DPSS as those where fraud was determined to have taken place. Negative cases are those where no fraud was found or could not be substantiated.

Early Action

In Los Angeles County, Early Action describes the investigation of persons who are already receiving aid from the County and who an eligibility worker suspects may be providing fraudulent information. In such cases, the eligibility worker makes a fraud referral to the on-site WFP&I investigator who is to investigate and report back to the eligibility worker within a short time frame with either positive or negative findings and a recommended action (e.g. re-evaluate benefits, terminate aid).

To evaluate the Early Action activities in the District offices, a randomly selected sample of 50 investigation files from the Rancho Park and Metro Family Center offices was examined. Overall, 66 percent of all Early Action fraud investigations in the sample resulted in positive fraud findings. The findings from the sample of Early Action investigationsconducted in the two offices is shown in the table below:

Table 10

Early Action Fraud Investigations Sample

Metro Family and Rancho Park Offices

Note: Positive cases are defined by DPSS as those where fraud was determined to have taken place. Negative cases are those where no fraud was found or could not be substantiated.

Techniques Used by Investigators

From the sampling of Early Fraud and Early Action files, the techniques which were used by the WFP&I investigators included field calls to the applicant/recipient's reported address or suspected workplace, speaking with neighbors or leaving business cards requesting that they call the investigator, researching and contacting the property owner at the reported address, checking DMV and Social Security Administration records, and setting up appointments to directly interview the applicant/recipient. Investigative activities were conducted during normal business hours, Monday through Friday. In many cases the applicant/recipient refused to cooperate or did not show up for a scheduled appointment, this typically resulted in the termination of benefits. A number of applicants/recipients, once notified that they were being investigated, requested to be terminated from aid or to have their application withdrawn. In several cases, the recipients were asked to complete an affidavit certifying that the information provided to the County to receive benefits was correct and the investigation was subsequently closed.

San Diego County's Project 100

In June of 1997 San Diego County implemented a program aimed at preventing welfare fraud called Project 100. Under Project 100, San Diego District Attorney fraud investigators conduct a field call to the residence of every applicant to verify eligibility. San Diego County reports that 25 percent of applicants who receive a home visit are denied benefits based on findings made during the home visit. Most importantly, San Diego uses home visits to identify individuals who are ineligible for aid before cash payments or other benefits are issued.

The purpose of the home visit is to corroborate the information provided on the aid application. Investigators look for evidence that the applicant is residing at their reported address, that there are no unreported adults living in the home, that children listed on the application are living with the applicant, and for evidence of unreported income or major assets such as vehicles. Home calls are made during normal business hours, Monday through Friday, and some Saturdays if the investigators have a high workload. In addition to meeting with the applicant, an investigator checks school records for children listed on the application, contacts the landlord, and speaks with neighbors if necessary to verify eligibility, etc.

Although applicants are not told the date or time that a home visit will occur, all applicants are informed by the Eligibility Worker at the time of application that a home visit will take place. In addition, each application form has a large printed notice informing applicants that an investigator will make a home call to verify the information reported on the form and notices are posted in each office lobby.

In order to stay within constitutional bounds, Project 100 investigators may not enter a residence without the express voluntary consent of the applicant. Investigators must identify themselves, state the purpose of their visit, and conduct the visit in a professional and respectful manner. Consent to enter must be given freely by the applicant, under no coercion or unlawful assertion of authority or intimidation. During the investigation of the home, the applicant, not the investigator, leads the search through rooms and is the one who opens doors and drawers. The applicant may terminate the search at any time. According to officials in San Diego, the vast majority of applicants consent to an investigator entering their residence. Applicants may refuse to meet with an investigator, but those who refuse are denied aid due to non-cooperation.

Early Fraud Investigations - San Diego and Los Angeles Counties

Finding 10: During July-September 1998 Los Angeles County denied 13 percent of its applicants for CalWORKs/AFDC compared to 44 percent in San Diego County. San Diego County's Project 100 sends a District Attorney public assistance investigator to the residence of every applicant for aid to verify the information provided by the applicant. In six months of 1998, Project 100 investigations resulted in the denial of 18.7 percent of all referrals based on findings made during the home visits, before aid is approved and overpayments occur. In Los Angeles County, only approximately 5.8 percent of all cases investigated are denied before benefits are issued.

Applicants for welfare benefits may be denied aid based on information reported on the application which indicates that the individual is ineligible; or, based on a subsequent investigation, which finds that the person has provided fraudulent information and is, in fact, ineligible for aid. To determine if Los Angeles County and San Diego County had comparable rates for denying aid, the overall denial rates in both counties were compared and it was found that, based on information provided to the State, during July-SEPTEMBER 1998 Los Angeles County denied 13 percent of applicants for CalWORKs/AFDC compared to 44 percent in San Diego County.

As discussed in the section on Recipient Fraud Investigation of this report, Los Angeles County has a low rate of investigation referrals compared to other counties and denies benefits to fewer cases before benefits are issued than other counties. Table 11 below is a comparison of the Early Fraud activities of San Diego and Los Angeles Counties to average monthly caseload. It should be noted that for purposes of both Table 11 and 12, Early Fraud activities in Los Angeles and San Diego counties include both the categories of Early Fraud and Early Action investigations described in the prior section. In other words, investigations of both applicants and individuals already on aid are included; for this reason, the table below presents an investigation rate based upon caseload rather than number of applicants for aid in each County.

Table 11

Early Fraud Investigation Requests, AFDC/CalWORKs Cases

San Diego and Los Angeles Counties, July to September 1998

As shown in Table 11 above, when the number of Early Fraud cases accepted by investigators is compared to average monthly caseload, San Diego County investigated a higher proportion of its monthly caseload, 10.4 percent, compared to Los Angeles County which investigated an average of 3.7 percent.

Table 12 compares the results of Early Fraud investigations in Los Angeles County to San Diego County.

Table 12

Early Fraud Investigation Outcomes, Los Angeles and San Diego Counties

All Programs, April to September 1998

As displayed in the table, San Diego and Los Angeles Counties conclude Early Fraud investigations with positive findings and take action at a nearly identical rate of approximately 20 percent of all completed investigations. However, an important difference is that the most common result of San Diego's Early Fraud investigations result in the denial of benefits. In other words, San Diego County is identifying fraudulent aid applications before any aid is granted, therefore preventing overpayments. As discussed previously, increasing the Early Fraud denial rate in Los Angeles County to the state rate of 22.8 percent would result in a reduction in ineligible benefits payments amounting to approximately $46.6 million annually. The impact of Los Angeles County achieving the San Diego County Early Fraud denial rate of 18.7 percent would have an impact of similar, though slightly less, magnitude.

In contrast, the majority of Los Angeles County's fraud cases are detected after aid has been approved and benefits are issued. If an overpayment has been issued, the County must then consume staff time and resources to determine the amount of overpayment and initiate collection activities, with no guarantee that the overpayment will be recovered. At the end of 1998, DPSS had a total of over $325 million in gross outstanding overpayments in the CalWORKs (or AFDC) program alone. During 1998, Los Angeles County was able to recoup only $16 million. Of the $325 million outstanding, the Department estimates that approximately $225 million will be written off as "uncollectible" due to lack of any payment activity on those accounts over the last three years, as permitted by State law. This further indicates the beneficial impact of denying payments at the point of application rather than trying to collect overpayments after benefits are issued and cases are closed.

Recommendation

INTERNAL AFFAIRS INVESTIGATIONS

Investigations of potential employee fraud are conducted by the Internal Affairs section of the DPSS Human Resources division. The section is currently comprised of four investigators and, a supervising manager. The unit is responsible for: 1) investigating allegations of employee wrongdoing; 2) ensuring the internal security and integrity of the Department's payment and ancillary systems; and 3) investigating reports of inappropriate employee conduct in the performance of their duties.

Most referrals for investigations received by the Internal Affairs section come from one of the following sources:

If a referral is accepted by the section, it is assigned to an investigator who then conducts the investigation. Investigative techniques may include interviewing the claimant and witnesses, reviewing the employee's personnel records, case files and computer records, interviewing the accused employee, field surveillance, obtaining records from other agencies such as the Department of Motor Vehicles, and working with other law enforcement agencies. Upon completion of an investigation, the results are forwarded to other DPSS sections or other agencies for follow up action, where warranted. Follow up actions may be performed by DPSS's Discipline Unit, WFP&I, the Auditor-Controller, or the District Attorney's Special Investigations Unit.

Finding 11: With Internal Affairs section staffing currently consisting of four investigators and one supervising manager for approximately 10,000 employees, DPSS has a low ratio of investigators relative to all cases of potential employee fraud. Based on experience elsewhere, it appears that there are more potential employee fraud cases than just those referred to the Department's Internal Affairs investigators.

A review of the Internal Affairs caseload and case timing from FY 1997-98 is as follows. There were three investigators staffing the Internal Affairs section during FY 1997-98 which has since been increased to four.

Table 13

Summary of Internal Affairs Activity: FY 1997-98

As shown in Table 13, the Internal Affairs section accepted 65 referrals for investigation during FY 1997-98. Of those, 25 were closed; and the balance was in progress at the end of the fiscal year. More than 25 cases were closed during the year but some of those had been initiated in previous years. Of the cases opened in FY 1997-98, each Internal Affairs investigator conducted an average of 21.7 cases per year. Each case that was closed took an average of 206 days to complete, or approximately 7 months. Table 13 also shows that, based on a sample of 20 closed case files reviewed, the result of these investigations in 50 percent of the cases was that the allegations were not substantiated.

The Internal Affairs section investigates not only potential employee fraud cases. In fact, only 13 of the 65 cases opened in FY 1997-98 actually involved charges of program fraud by an employee. Most of the rest of the cases involved charges of misconduct, such as sexual harassment of another employee or a client; breaches of confidentiality, such as using client social security numbers; other information for personal purposes; or conflict of interest, such as an eligibility worker failing to report that they themselves were living with an aid recipient. Table 14 shows a breakdown of FY 1997-98 case referrals by subject matter.

Table 14

Internal Fraud Referrals by Subject

FY 1997-98

Estimating the Extent to Which Employee Fraud is Being Detected in the Department

While the current number of cases seem to be manageable for the number of investigators on staff, an assessment of referral rates and Internal Affairs staffing levels requires first estimating the actual level of employee fraud in the Department. While it is not possible to come up with the precise number of actual fraudulent employee activities compared to the number of referrals, there are indications that the fraud rate may be higher in the Department than is captured by the referral system.

In a site visit to San Diego County's Health and Human Services Agency, data was collected on the number of referrals made to and staffing level of their Internal Security Section. Comparing that data to DPSS's Internal Affairs section, it appears that DPSS's referrals and Internal Affairs staffing level is low given the size of the Department. San Diego County's Health and Human Services Agency reports that for Calendar Year 1998, they received 132 referrals for investigation, of which 27 were potential fraud cases. The San Diego County social services division has approximately 3,250 employees, so there are significantly fewer referrals relative to the number of employees in Los Angeles County than in San Diego County.

As another measure, staffing of San Diego County's Internal Security Section consists of eight positions: four investigators; one quality control technician; one eligibility technician; one senior clerk; and, one manager. This level of staffing exceeds that of the DPSS Internal Affairs section even though San Diego County's department and workforce are smaller. Table 15 presents these comparative data.

Table 15

Los Angeles and San Diego County Social Service Agencies

Internal Affairs Case Referrals and Staffing Levels

As shown in Table 15, San Diego County's Internal Security Section receives 40.6 referrals for every 1,000 employees, compared to only 6.5 for the DPSS. When just potential fraud related referrals are counted, the San Diego unit receives 8.3 such referrals for every 1,000 employees compared to 1.3 in Los Angeles County. San Diego County's unit also has a higher level of internal affairs staffing, with 2.5 positions for every 1,000 employees compared to 0.6 in Los Angeles County. It should be noted that the additional staff in San Diego County cannot be considered excessive relative to Los Angeles County as each investigator processes more cases than staff in DPSS's Internal Affairs section.

As another means of assessing the extent of potential employee fraud caseload at the DPSS compared to the number of referrals received by the Internal Affairs section, an inquiry was made of the number of referrals received by the new DPSS employee fraud unit of the Special Investigations Division of the Los Angeles County District Attorney's Office. This new unit has just been staffed since October 1998. Though the new unit's agreement with the Department calls for it to process cases received from the Department, representatives of the unit report that it is also receiving referrals from other sources such as law enforcement agencies. Further, its investigations of some of the cases referred by the Department have generated other leads.

A mechanism in place in San Diego County that has reportedly contributed toward its higher number of employee fraud referrals is extensive staff training at employeeorientations and other agency training sessions in detecting employee fraud and on the agency's conflict of interest policies. The Internal Security section staff provides training in these subjects to all levels of staff: managers, supervisors, and line staff. Related to this is an agency policy that managers, supervisors, and other employees who are found to have known about a potential employee fraud case, but did not report it, are subject to disciplinary action. Increased training in employee fraud and a similar policy regarding employee accountability for reporting potential employee fraud would be beneficial at DPSS.

To further strengthen DPSS employee fraud reporting, District managers should be made accountable for fraudulent activity reporting in their offices in two ways. First, they should be held accountable for not reporting cases of potential fraud if it is determined in an investigation that they should have detected certain suspicious transactions of one of their employees (e.g., if an application processed by an intake eligibility worker remains in the intake unit for more than 60 days instead of being transferred to a continuing case eligibility worker, as is department policy).

As a second means of strengthening employee fraud reporting, repeatedly high numbers of referrals, or the absence of any referrals, from a District office should be flagged in upper management reviews of referrals by District. Unusually high or low numbers of referrals could indicate that certain controls are not being adhered to, or that the manager is not regularly reviewing internal control reports and documents to identify potential fraud cases. Information regarding level of fraudulent activity by District office should be compiled semi-annually by Internal Affairs so that trends can be identified. In addition to upper management reviews of this information, District managers' adherence to internal control policies and procedures should be included as one of the criteria used for District managers' performance evaluations.

Based on San Diego County's experience, and the number of referrals received by the newly created Special Investigations Unit of the District Attorney's office since October 1998, it appears that there could well be more employee fraud occurring than the Department's current system is capturing. It also appears that existing staff may be able to process more cases per investigator, possibly through some streamlining of procedures. Additional staff may be needed if the number of referrals increases. However, before adding staff to this unit, the Department should first determine if there are changes that could be made to its investigation procedures to increase productivity of existing staff from the current average of opening only 16 cases per year.

Recommendations

It is recommended that the DPSS Director:

Need for Department to Pro-actively Monitor Employee Activity for Potential Fraud

Finding 12: The Department's internal investigation function is reactive; investigations of employee fraud are only conducted in response to referrals. Information is not readily available to allow investigators to identify unusual or suspicious transactions by employees. The Castillo employee fraud case in 1998, and other less visible cases before and after, demonstrated that the Department's existing internal controls are not always adhered to by staff. New DPSS policies implemented since the Castillo case represent a start of a more proactive approach to controlling employee fraud.

The Castillo employee fraud case in the summer of 1998, where an Eligibility Worker created over 20 cases for family members and friends who received more than $700,000 in fraudulent payments, proved that DPSS's existing internal controls had weaknesses or were not being adhered to in that employee's District office. The Eligibility Worker had been committing fraud over an extended period unbeknownst to her supervisor and District office director. Through a tip to the County's Central Fraud Reporting Line, Department investigators from WFP&I and Internal Affairs began an investigation which ultimately uncovered her extensive fraudulent activities. She was arrested, tried, found guilty, and has been sentenced. However, had the tip not been provided, it is possible that this worker could have continued to commit fraud, bypassing departmental controls. In addition, there could be other workers doing the same thing, and if tips are never provided about them, they could continue to go undetected.

A review of a sample of employee fraud investigation case files show that many of the same characteristics found in the Castillo fraud case appear in other employee fraud cases, both before and since that case was investigated by Internal Affairs investigators. These characteristics include:

In response to the Castillo case, Department management created an Internal Security Task Force that developed recommendations for closing the loopholes used by Castillo and possibly other employees. The Task Force found that existing DPSS policies and procedures should have prevented some aspects of Castillo's operations, but they were not all being implemented. The Task Force also concluded that implementation of two new automated systems, Los Angeles Eligibility, Automated Determination, Evaluation and Reporting System (LEADER) and the Welfare Fraud Linkage Analysis Database System (WFLADS), both scheduled for later in 1999, will render may of the actions taken by Castillo impossible. In the interim, the Task Force has prepared a series of recommendations for enhancing and reinforcing a number of existing DPSS policies and procedures. A number of these recommendations have already been implemented. The Task Force recommendations included:

Additional new controls, policies, and case monitoring are planned for after the new LEADER and WFLADS systems are implemented.

All recommendations of the Internal Security Task Force represent improvements in the Department's internal controls. But it should be noted that a number of them are reinforcements of existing policies and procedures that are not always complied with. Further, many of them are areas of weakness already well known to the Internal Affairs investigators. What has been missing in the Department is regular and ongoing input from Internal Affairs to management regarding the internal control aspect of policy and procedure development and monitoring. Regular and ongoing communications between the Internal Affairs investigative staff and policy development staff is needed.

In San Diego County, where there is a higher rate of employee fraud referrals, there is regular consultation between the manager of Internal Security and the Program management staff before new eligibility and case management policies and procedures are implemented. This coordination is intended to ensure that changes in policies, procedures,and automated systems do not inadvertently also create new opportunities for employee fraud. The San Diego Internal Security manager is a former eligibility worker with an information systems background, who knows program requirements, information systems capabilities, and how an employee might take advantage of both. This type of involvement of DPSS's Internal Affairs staff would be beneficial. The new Internal Security Task Force, created after the Castillo case was uncovered, includes the Chief of Human Resources and the director who oversees the Internal Affairs section. Though it has developed its recommendation, this cross-functional group is planning to remain active to oversee and work out the details for implementation of its recommendations.

The Internal Security section in San Diego County has a Quality Control Technician. That position's primary function is to monitor special fraud detection reports generated by that agency's computer systems. This technician conducts ongoing centralized monitoring of regular fraud detection reports. These reports, generated by the Department's computer system, aggregate suspicious activities such as the following and can be sorted by worker, by district office, by supervisors, by date, and by other key variables.

While much of this information is already available at DPSS, it is not systematically compiled or reviewed by centralized Internal Affairs investigative or other staff. Much of it is provided to District managers for their review or is available to investigator staff as needed, but usually only after a case is referred. In San Diego, if one of the reports revealed some suspicious transactions, investigators would conduct an unrequested and unannounced investigation of the worker.

If the DPSS were to begin regular monitoring of information such as the reports listed above, it would result in a more pro-active approach to deterring and detecting employee fraud. It would detect cases that might otherwise go undetected through the current reactive system. To the extent employees are caught committing fraud, it would also signal to staff that their actions are being monitored and unusual actions investigated. This would have a deterrent effect on potential employee fraud.

Though much of the information described above can be obtained now from the DPSS's current computer systems, the system lacks flexibility and cannot always consolidate certain key variables. The new LEADER system, on the other hand, will reputedly provide much more flexibility in terms of generating custom reports and changing sorting criteria.

Recommendations

Investigative Techniques and Coordination with the District Attorney's Investigators

Finding 13: Internal investigation techniques and results could be enhanced through more training and improved coordination with investigators from the District Attorney's new internal fraud unit. Joint case review sessions with investigators from Internal Affairs and the District Attorney's Office would allow DPSS to take greater advantage of the investigative skills and knowledge of DA Investigators.

In terms of investigation approach and techniques, the Internal Affairs section has historically operated very independently. They have traditionally conducted investigations on their own, without law enforcement agents, mostly using information culled from the claimant, witnesses, Department records, and the charged employee. The Department's new agreement with the District Attorney's office presents an opportunity for new coordinationand input on case approach from the experienced investigators assigned to the new Special Investigations Division unit.

The Internal Affairs investigators are largely former eligibility workers with strong program knowledge but little experience, if any, in law enforcement. The District Attorney's investigators, on the other hand, have law enforcement experience and have spent their careers pursuing potential criminals. They are more likely to know when to conduct a surveillance of an employee, how to quickly access background information about the suspect, and they often have contacts that can facilitate quick collection of needed information. Whereas the role of the Internal Affairs investigators is broad - they receive and investigate more cases pertaining to employee misconduct than fraud - the District Attorney's investigators are specialists in investigating criminal behavior. For all of these reasons, coordination and cooperation between the Internal Affairs and District Attorney's investigators must be mandatory, and will be extremely beneficial.

Since creation of the new internal investigation unit in the District Attorney's office and establishment of the agreement between DPSS and that unit this fiscal year, some of the details of how investigation protocols are to be implemented have yet to be fully developed. There is some uncertainty about exactly how far the Department's Internal Affairs employee fraud investigations should go before they are referred to the District Attorney's office. In some instances, it may be more expedient to turn a case over to the District Attorney's office sooner based on preliminary information collected. In the past, Internal Affairs investigators would conduct a full investigation of the charges made before referring the case to the District Attorney. This approach has some disadvantages as this preliminary investigation could tip off the suspect who would then have an opportunity to destroy potentially damaging records or to make a change in their behavior before the District Attorney's investigators can get started on the case. This could eliminate the possibility of the District Attorney's investigators conducting a more extensive investigation of the employee's past practices based on the charges filed and preliminary information collected.

What is needed is close communications between the Internal Affairs and District Attorney's investigators to determine what constitutes a preliminary investigation and a process for deciding when it is in the best interests of the investigation to turn the case over to the District Attorney's investigators. A regularly scheduled case review session jointly attended by District Attorney and Internal Affairs investigators would be an effective means of working out the details of these protocols.

The outcome of most of DPSS's internal investigations in the recent past has been to discipline the employee and/or give them the option to resign. Until FY 1997-98, very few cases were referred to the District Attorney even if they involved potentially fraudulent activity. Since FY 1997-98, more cases have been referred to the District Attorney. With the creation of the new unit, there is now a need to clarify who should conduct the investigations, at what point they should be referred to the District Attorney if embezzlement appears to beoccurring, and how each set of investigators can be most effective without duplicating efforts or working at cross purposes. By working closely together, the effectiveness of the investigation process, and the skills and knowledge of both sets of investigators, will be maximized.

Recommendations

MANAGEMENT INFORMATION AND REPORTING

Finding 14: The Department is in the process of implementing two new computer systems this year that will make better information available for detecting and investigating recipient and internal fraud. The systems will also provide enhanced internal controls. But some of the information and controls planned for the new systems already exist and, as evidenced by some recent high visibility employee and recipient fraud cases, these controls and detection reports have been circumvented. Though the new systems will make it more difficult to commit some types of fraud, they will provide new opportunities for committing fraud and new controls to be circumvented.

DPSS is in the midst of introducing two new computer systems this year. The first, the Los Angeles Eligibility, Automated Determination, Evaluation and Reporting System (LEADER) will consolidate a number of separate systems into one, and will enable the Department to automate much of its paper-based eligibility and case management system. Key new internal controls will be implemented as part of the LEADER system including employee passwords required every time an employee uses the system (this will provide a record of every transaction performed by each employee), limited access to the system based on an employee's position, a transaction history, and electronic authorizations where a supervisor's password will need to be entered to authorize issuance of benefits to a new recipient or changes in benefits to an existing recipient. These controls will help stem some of the common approaches to employee fraud such as: forgery of supervisors' signatures; missing documentation and case files; and absence of historical records on cases in the current computer systems.

In terms of monitoring reports, LEADER will provide access to the present twenty separate databases, and allow for custom designed queries from the consolidated data base. This will allow for reports listing specific types of transactions for a certain month, by District office, by worker, or other sorting criteria. Specific monitoring reports have not yet been designed but the system will allow for designing and changing such reports after it is fully operational. LEADER is scheduled to begin operating on a pilot basis in May 1999 with a ten month phased implementation scheduled for the entire Department.

The other new system being developed this year is the Welfare Fraud Linkage Analysis Database System (WFLADS). Department representatives report that this new system will provide access to the LEADER database and will allow for custom queries and reports, sorting transactions and information in any way desired. This will enable the Department's Welfare Fraud Prevention & Investigation section (WFP&I) to more pro-actively search for welfare fraud, rather than conducting investigations only on cases referred to them. Nightly matches of case data will be conducted and sorted for suspicious linkages and inconsistencies. For example, WFLADS could run a report identifying and matching allchildren's social security numbers to determine if the same children are being claimed by different recipients, thus entitling them to more aid than they otherwise deserve. The system could run a report listing all changes in payee for the previous month, by address, to determine if a number of new payees from different cases are all listed at the same address, possibly indicating that these cases have been fraudulently altered. Like LEADER, WFLADS will also maintain a history of transactions so that trends and repeat approaches to fraud can be identified and analyzed.

Both of the new systems represent significant improvements in the Department's ability to control and detect employee and recipient fraud. However, the systems alone will not put an end to either type of fraud. Requiring supervisors to enter passwords on the computer each time they authorize benefits will diminish the opportunity for forgery, a common technique in fraud cases. However, once the confidentiality of a password is broken (e.g., an employee writes down their password and inadvertently leaves it on their desk, a new opportunity for fraud is generated.

The Department has many controls and computer based information that are to be used to detect and stop fraud. Some of the key computer based information is as follows:

Computer based information presently used to prevent new applicant fraud:

Computer System/Report How Used for Fraud Control
   
WCMIS (Welfare Case Management and Information System): Database for active and three years worth of closed cases for all programs: CalWORKs, General Relief, Food Stamps, and Medi-Cal. Includes same information for Orange County. Clears all information provided by applicants against existing and past clients to detect applications for duplicate aid. Searches include names, addresses dates of birth, social security numbers, phone numbers, household members, and other data on application face sheet.
AFIRM (Automated Fingerprint Image Reporting & Match System): This is the Department's computerized fingerprint system. Searches electronic database of recipient fingerprints to identify applicants already on aid, or on aid in the past.
Medi-Cal Eligibility Data System (MEDS): State's automated Medi-Cal system Through access to State computer system, DPSS staff can check and see if applicants are currently on Medi-Cal.
Income & Eligibility Verification System (IEVS): State system that tracks income earned by State residents and reports it back to County. Includes other income subsystems. Applicants are checked against this system to verify their reported income.
WISE (Widebased Information System Exchange): Tracks changes in case status and case movement within and between offices. Generates a report identifying cases that have not been moved out of Intake in an appropriate amount of time, which could indicate employee fraud.

For continuing cases:

Computer System/Report How Used for Fraud Control
   
Income & Eligibility Verification System (IEVS): State system that tracks income earned by State residents and reports it back to County. To identify aid recipients who have earned unreported income or received unreported other aid or benefits such as disability insurance. When recipients are identified who have earned income not reported to DPSS, they are supposed to be investigated.
WISE (Widebased Information System Exchange): Tracks changes in case status and its movement within and between offices. Assigns and tracks cases including location and changes in case status. Tracks changes in payees, addresses and household compositions. System produces Intake Control Log.
Food Stamp Automated Issuance & Recording System (FAIR): Automated on-line issuance system for Food Stamps and cash aid. Automatically restricts duplicate aid issuance for month. Produces exception reports and statistical compilations.
New Employee Registry: State database that is supposed to list all persons hired by all employers each month, by social security number. Just recently implemented, this database is to be used for running matches to determine if any of those listed in the New Employee Registry are also receiving aid and have not reported their new income to DPSS.
CAR (Caseload Activity Report): Log of all applications and continuing cases. District offices receive reports to reconcile.

As evidenced by the list above, much of the emphasis of the Department's current information system-based internal controls for applicants is detecting and eliminating duplicate aid. The various application clearances through WCMIS and the computerized fingerprinting system, AFIRM, are aimed at eliminating duplicate applications for aid. The IEVS and New Employee Registry systems are oriented to identifying applicants' unreported income. Employee and recipient fraud could potentially be detected through the WISE-generated report which lists cases that have been in the Intake unit longer than normal instead of being transferred to a continuing eligibility worker and through the Caseload Activity Report, which lists all applications and new cases authorized.

For continuing cases, the computer system-based controls are largely aimed at detecting recipient fraud through identification of unreported income and other information potentially affecting eligibility from the IEVS system, New Employee Registry, Out of County matches, and SSI/SSP Listings (identifies recipients who are receiving Supplemental Security Income). With the exception of the Retroactive Issuances report, there are few regularly produced reports aimed at detecting potential employee fraud. As discussed in the section on Internal Affairs Investigations, department-wide computer generated reports listing transactions often associated with employee fraud, such as simultaneouspayee/address changes, or cases opened and closed in the same month, cannot be produced by the Department's current computer systems. This type of information can and is retrieved after the fact for individual investigations, but not for pro-active department-wide monitoring for employee fraud.

Besides limited data and computer system flexibility, a key problem with the Department's fraud monitoring system is that after some of the reports that are produced are distributed, there is no mechanism for upper management to know that they have been reviewed and acted upon. If a District manager receives a report showing a high number of an intake worker's cases remaining in the intake unit for 90 days or more, it is up to that District manager to probe the matter or make a referral to an investigator. Further, if the same District manager ignored the report, there could be no repercussions from upper management. Or if a report showed that a number of applicants that have begun receiving aid at a certain office had not had their fingerprints taken, it would be up to the individual District manager to act on this suspicious information.

Recommendations

Finding 15: Though the new computer systems will provide greater flexibility and richer information than is available at present, it will not resolve the issues of accountability and responsibility for monitoring, analyzing and acting on employee fraud monitoring information.

Though LEADER will provide greater flexibility and richer information than is available at present, it will not resolve the issues of accountability and responsibility for monitoring, analyzing, and acting on employee fraud monitoring information. The Department needs to address these issues as part of LEADER implementation (though procedures should be established even with the current systems). Other questions to be resolved with LEADER implementation include:

With implementation of LEADER pending, it is critical that systems and procedures be immediately established to ensure that information generated from the Department's computer systems is put in the hands of people who understand how to use the information and that a feedback loop exists so that the impact of distributing this information can be measured and documented. Procedures should be initiated before LEADER is implemented that will hold District managers accountable for reviewing and acting on certain fraud-related reports and information for their District offices; and that will delegate responsibility for some department-wide monitoring and analysis activities to the Department's Internal Affairs and Welfare Fraud Prevention & Investigation (WFP&I) sections. This would be an appropriate role for the Internal Security Task Force created last year in the wake of the Castillo case. The Department plan is to maintain the Task Force to help oversee implementation of new internal controls.

District managers, Internal Affairs and WFP&I should be required to report to upper management that they have reviewed key fraud related data and to indicate what actions they have taken on suspicious transactions in their District. Internal Affairs and WFP&I could complement these managers' reports by reporting their participation in any investigations that grow out of the District office's review.

As with LEADER, the Department, and particularly WFP&I, needs to develop plans detailing what information should be regularly reported from the WFLADS system, who in WFP&I should be responsible for analyzing it, and how follow up actions will be documented and monitored.

Recommendation

Management Information

Finding 16: Management reporting on the effectiveness of Department investigations needs to be improved. It is difficult with the information currently available to track

different types of investigations from referral through disposition in order to measure their effectiveness.

DPSS needs improved management information to better monitor the activities and outcomes of its investigation functions and resources. WFP&I's statistical tracking system is derived from reports that the Department is required to prepare for the State. The State format does not include a useful breakdown of all types of investigations so that WFP&I management can monitor and assess the effectiveness of its efforts. Specifically, the State's Fraud Investigation Activity Report (DPA 266) does not provide:

This information would provide WFP&I management with a basis for assessing the relative effectiveness of the different type of investigations performed, staff productivity, and areas where recipient fraud is regularly recurring. Compilation of these type of statistics does not require a new computer system but could be done by enhancing the existing statistical compilation process.

Though a much smaller operation, the Internal Affairs section maintains statistics on their cases including source of referral, District office, investigator, and nature of case. Because the number of cases is small (65 were opened in FY 1997-98), it is fairly easy for this section management to readily compile and analyze their case statistics from a variety of perspectives. Besides the types of sorting of these cases that is currently prepared (by investigator, by District office, by type of case), the Section should enhance its management reporting system and case work oversight by regularly compiling the following information and reporting it to upper management:

The backlog and time information could be used to monitor staff productivity and identify areas where staff training may be needed. The type of fraud and internal controls circumvented information could be used to help analyze weaknesses in the Department's internal controls, and to identify areas where special reports and pro-active monitoring may be needed (consistent with the recommendations for more pro-active monitoring made previously).

Recommendation

CONCLUDING SUMMARY

The primary mission of the Department of Public Social Services is to allocate, administer, and monitor the economic resources of the County of Los Angeles which are allocated to aid those with demonstrated economic need. In most cases, it is intended that such economic assistance be temporary; however, the secondary duty of the Department is to prevent the distribution of economic resources to undeserving individuals. The third obligation of the Department is to detect and prosecute (or refer for prosecution) those persons who fraudulently obtain DPSS funds or other benefits.

This audit was conducted to evaluate the efficiency of the DPSS in carrying out its responsibilities in the second and third segments of its mission. Specifically:

Under the title Findings, in various sections of the report, there are sixteen deficiencies noted which are in need of revision and improvement. In response, and as a result of these findings, twenty-four major Recommendations have been proposed to accomplish the defined objectives. In general, the Findings reflect the inertia of a bureaucratic monolith separated by both geography and, in some instances, philosophy, among its several divisions. Many of the fraud prevention or investigative functions in place, prior to the public disclosure of a glaring instance of employee fraud and the onset of this audit, were sometimes unduly delayed or even completely ignored. Existing management practices were also lax, not enforced, or sometimes ignored. It is true that the DPSS is now taking remedial actions, especially since the present audit has been taking place; but they appear to be "too little, too late." The new computerization in process of implementation appears promising, assuming the programs are as effective and efficient as described, and assuming that proper management, monitoring and verification policies are prepared and strictly enforced. If these Recommendations are implemented, they should accomplish the following improvements:

As noted in the Introduction, potential annual savings to the Los Angeles County through the implementation of the recommendations in this audit are enormous, and are estimated to possibly exceed $500 million.

The Grand Jury is convinced that implementation of these Recommendations by the Board of Supervisors is vital to the economic survival of the welfare services being provided to worthy citizens of Los Angeles County.

FINAL RECOMMENDATION

GLOSSARY

AFDC - Aid to Families with Dependent Children is the former federally funded welfare program for families with children which was by the CalWORKs program, enacted in 1997 (see below).

CalWORKs - California Work Opportunity and Responsibility to Kids is a new federally funded welfare program for families with children who are in need. Like its predecessor, Aid to Families with Dependent Children (AFDC), the new program provides cash grants and welfare-to-work services to families whose incomes are not adequate to meet their basic needs. Under CalWORKs, able-bodied adult recipients (1) must meet participation mandates, (2) are limited to five years of cash assistance, and (3) must begin community service employment after no more than 24 months on aid. Los Angeles County is responsible for administering the program, although the requirements are set by state and federal laws and regulations.

DPSS - Department of Public Social Services is the agency responsible for administering and distributing welfare in Los Angeles County.

EW - Eligibility Workers are DPSS employees whose primary responsibility is to determine and monitor eligibility for welfare aid.

FID -- Facts Input Document is the investigation cover sheet used by WFP&I to track assignments, fraud allegations, priority and disposition codes, and other information pertinent to each fraud investigation.

GAIN - Greater Avenues for Independence was a program under AFDC which provided employment preparation services for a limited number of AFDC recipients. The GAIN program has been largely replaced under the new CalWORKs program which has much broader work participation requirements for recipients.

GR - General Relief is a County-funded program that provides financial assistance to indigents who are not eligible for federal or state assistance programs, typically adults without children. It also provides emergency assistance to individuals and families in temporary need.

IEVS - Income and Eligibility Verification System is a match system required by federal law designated solely for CalWORKs, Food Stamp, and Medical Assistance programs. The system matches both recipients and applicants. The IEVS system identifies statewide

matches with assets, earnings, employment and disability insurance benefits, and duplicate aid for new applicants and ongoing recipients.

LEADER - Los Angeles Eligibility, Automated Determination, Evaluation and Reporting System is a new computer system which will consolidate a number of separate systems into one and will enable DPSS to automate much of its paper-based eligibility and case management system.

LEADER is scheduled to begin operating on a pilot basis in May 1999 with a ten-month phased implementation scheduled for the entire Department.

WFLADS - Welfare Fraud Linkage Analysis Database System is a new computer system which will provide access to the LEADER database and will allow for custom queries and reports, sorting transactions and information in any way desired. This system will allow for investigators in the WFP&I section to more pro-actively search for welfare fraud.

WFP&I-Welfare Fraud Prevention and Investigation is the Section of the Department of Public Social Services charged with preventing and investigating recipient fraud.

Back to Table of Contents Emergency Medical Services