Food Stamp Program: Better Use of Electronic Data Could Result in
Disqualifying More Recipients Who Traffic Benefits (Letter Report,
03/07/2000, GAO/RCED-00-61).

Pursuant to a congressional request, GAO reviewed the efforts of states
and the Food and Nutrition Service (FNS) to reduce Food Stamp
trafficking, focusing on the: (1) extent to which the states with
statewide electronic benefit transfer (EBT) systems are identifying and
disqualifying recipients who engage in trafficking; and (2) actions FNS
has taken to encourage the states to identify and disqualify recipients
engaged in trafficking.

GAO noted that: (1) of the 29 states with statewide EBT systems, only
4--Florida, Missouri, South Carolina, and Texas--independently and
proactively analyzed their electronic databases to identify suspect
recipients; (2) a fifth state--Maryland--has used a list of suspected
traffickers provided by the Department of Agriculture's Office of
Inspector General since 1994 to provide a basis for follow-up
investigations; (3) all five of these states invested the resources
necessary to investigate suspect recipients, but, for fiscal years 1998
and 1999, only two states--Maryland and Texas--were responsible for
about 87 percent of the 6,873 individuals disqualified nationwide from
the Food Stamp Program for trafficking benefits; (4) in addition to
these five states, nine others investigated suspect
recipients--identified primarily by FNS--and disqualified those who
engaged in trafficking, albeit to a lesser extent; (5) the remaining 15
of the 29 states did not disqualify any recipient for trafficking during
the 2-year period; (6) EBT data have been available since 1993 to
analyze and identify trafficking patterns; (7) however, since only five
states had statewide EBT systems before 1997, FNS has only recently
initiated action to work with the states to ensure that they target
recipients likely to be engaged in trafficking--those identified by FNS
in its successful cases against storeowners found to have engaged in
trafficking; (8) in July 1999, FNS directed its seven regional offices
to develop plans to work with the states to identify suspect recipients,
investigate the suspects, and to disqualify those engaged in
trafficking; (9) as of December 1999, these plans were still in the
preliminary stages; (10) FNS will not be able to determine the
effectiveness of its recent efforts in reducing the overall level of
trafficking because it lacks a current, reliable estimate of the extent
of trafficking; (11) such an estimate would also better permit FNS to
adhere to the principles of the Government Performance and Results Act
of 1993, which state that goals and strategies should be quantifiable
and measurable; and (12) FNS could use EBT data to develop such
estimates and to target its available resources.

--------------------------- Indexing Terms -----------------------------

 REPORTNUM:  RCED-00-61
     TITLE:  Food Stamp Program: Better Use of Electronic Data Could
	     Result in Disqualifying More Recipients Who Traffic
	     Benefits
      DATE:  03/07/2000
   SUBJECT:  Food relief programs
	     Computer matching
	     State-administered programs
	     Fraud
	     Program abuses
	     Welfare recipients
	     Welfare benefits
	     Internal controls
	     Data bases
	     Federal/state relations
IDENTIFIER:  USDA Electronic Benefit Transfer System
	     Food Stamp Program
	     Florida
	     Missouri
	     South Carolina
	     Texas
	     Maryland

******************************************************************
** This file contains an ASCII representation of the text of a  **
** GAO Testimony.                                               **
**                                                              **
** No attempt has been made to display graphic images, although **
** figure captions are reproduced.  Tables are included, but    **
** may not resemble those in the printed version.               **
**                                                              **
** Please see the PDF (Portable Document Format) file, when     **
** available, for a complete electronic file of the printed     **
** document's contents.                                         **
**                                                              **
******************************************************************

GAO/RCED-00-61

Appendix I: Scope and Methodology

16

Appendix II: Twenty-Nine States With Statewide EBT Data, as of April, 1999

17

Appendix III: Key Contacts and Staff Acknowledgments

18

Table 1: Number of Trafficking Storeowners and Recipients Removed From the
Food Stamp Program in the 29 States With Statewide EBT Programs, Fiscal
Years 1998-99 7

EBT Electronic benefit transfer

FNS Food and Nutrition Service

GAO General Accounting Office

OIG Office of Inspector General

USDA U. S. Department of Agriculture

Resources, Community, and
Economic Development Division

B-284230

March 7, 2000

The Honorable Richard G. Lugar
Chairman, Committee on Agriculture,
Nutrition, and Forestry
United States Senate

Dear Mr. Chairman:

In 1999, the U.S. Department of Agriculture (USDA) provided about $16
billion in Food Stamp Program benefits to a total of about 18 million
recipients. Until the mid-1990s, most recipients used benefits provided in
the form of coupons to purchase allowable food, but currently about 70
percent of all benefits are provided electronically. As of November 1999, 35
states and the District of Columbia had statewide electronic benefit
transfer (EBT) systems; all the remaining states are required to have signed
contracts for such systems by October 2002.1 Recipients receiving their
benefits electronically use cards, much like debit cards, to pay for their
groceries at the checkout counter, and the benefits used are deducted from
the recipients' monthly allocation. By providing benefits electronically,
the federal government saves time and money because the process of printing,
safeguarding, distributing, accounting for, and destroying the coupons is
eliminated. Furthermore, an EBT system creates an electronic record of each
food stamp transaction, making it easier to identify and document instances
of food stamp trafficking. Trafficking is a process whereby some recipients
and storeowners illegally exchange food stamp benefits for cash.

USDA's Food and Nutrition Service (FNS) administers the Food Stamp Program
in partnership with the states. The Service provides funding for the
benefits and 50 percent of the state program's administrative costs,
develops the program's policies and guidelines, authorizes retail food
stores' participation, and monitors storeowners' compliance with program
requirements. The states administer day-to-day operations, including
certifying households' eligibility and delivering benefits to recipients.
The states are also responsible for investigating and prosecuting
individuals suspected of intentionally misusing their benefits. Storeowners
and recipients found to engage in trafficking could face a number of
penalties, such as temporary or permanent disqualification from the program.

Because of continued congressional concerns about fraud and abuse in the
Food Stamp Program, you asked us to determine (1) the extent to which the
states with statewide EBT systems are identifying and disqualifying
recipients who engage in trafficking and (2) the actions the Service has
taken to encourage the states to identify and disqualify recipients engaged
in trafficking. As agreed with your office, we examined the actions taken by
the Service and the 29 states (including the District of Columbia), that
were delivering food stamp benefits through statewide electronic systems
prior to April 1, 1999. We also visited six of these states--Florida,
Georgia, Illinois, Louisiana, Maryland, and Texas--that have cities with
high levels of trafficking in comparison with other cities. (See app. I for
more details on our scope and methodology.)

Of the 29 states with statewide electronic benefit systems, only 4--Florida,
Missouri, South Carolina, and Texas--independently and proactively analyzed
their electronic databases to identify suspect recipients. A fifth
state--Maryland--has used a list of suspected traffickers provided by USDA's
Office of Inspector General since 1994 to provide a basis for follow-up
investigations. All five of these states invested the resources necessary to
investigate suspect recipients, but, for fiscal years 1998 and 1999, only
two states--Maryland and Texas--were responsible for about 87 percent of the
6,873 individuals disqualified nationwide from the Food Stamp Program for
trafficking benefits. In addition to these five states, nine others
investigated suspect recipients--identified primarily by the Food and
Nutrition Service--and disqualified those who had engaged in trafficking,
albeit to a lesser extent. The remaining 15 of the 29 states did not
disqualify any recipient for trafficking during the 2-year period.

EBT data have been available since 1993 to analyze and identify trafficking
patterns. However, since only five states had statewide EBT systems before
1997, the Food and Nutrition Service has only recently initiated action to
work with the states to ensure that they target recipients likely to be
engaged in trafficking--those identified by the Service in its successful
cases against storeowners found to have engaged in trafficking. In July
1999, the Service directed its seven regional offices to develop plans to
work with the states to identify suspect recipients, investigate the
suspects, and to disqualify those engaged in trafficking. As of December
1999, these plans were still in the preliminary stages. The Service will not
be able to determine the effectiveness of its recent efforts in reducing the
overall level of trafficking because it lacks a current, reliable estimate
of the extent of trafficking. Such an estimate would also better permit the
Service to adhere to the principles of the Government Performance and
Results Act of 1993, which state that goals and strategies should be
quantifiable and measurable. The Service could use EBT data to develop such
estimates and to target its available resources.

We are recommending that the Food and Nutrition Service work with those
states that routinely use electronic data to determine the best techniques
for using the data to identify suspected recipient traffickers and then work
with the other states to implement the best practices, as appropriate. We
are further recommending that FNS use electronic data to routinely develop
reliable estimates of the extent of trafficking and establish goals and
strategies for reducing recipient trafficking on the basis of these
estimates.

Food stamp trafficking is a violation of program regulations and normally
involves two parties--a recipient and a storeowner. Trafficking generally
takes place when a recipient collaborates with a storeowner and exchanges
coupons or electronic benefits for cash. The storeowner gives the recipient
a discounted cash payment for food stamp benefits (often 50 cents on the
dollar) and then redeems the benefits at full face value from the
government. In a previous report,2 we examined USDA's estimate of the extent
of trafficking--$815 million annually, or 4 percent of annual benefits--and
concluded that the estimate was not reliable because it was based on 1993
data, which did not consider the effect of EBT systems.

FNS is generally responsible for monitoring the actions of storeowners, and
the states are responsible for monitoring recipients. The states are
specifically responsible for investigating recipients alleged to be engaged
in trafficking and for disqualifying those found trafficking. Typically, a
recipient found guilty of trafficking is disqualified from the program for 1
year for the first offense, 2 years for the second offense, and permanently
for the third offense or for trafficking an amount that exceeds $500.3 To
disqualify an individual from the program, the states often have to conduct
costly, time-consuming investigations--interviews with the suspect,
undercover observations of transactions, and a more detailed analysis of the
recipient's shopping habits. Under a proposed regulation, the states could
establish claims against recipients for amounts trafficked and collect from
both the trafficker or from the benefits available for remaining household
members.

EBT databases are a new and potentially effective tool in combating fraud
and abuse in the Food Stamp Program. USDA has cited the implementation of
EBT systems as a significant step forward in identifying and combating food
stamp trafficking. In recent years, FNS has analyzed data from state EBT
systems to identify possible cases of food stamp trafficking. These data
include the date, time, and amount of the sale; the store authorization
number; and the recipient's identification number. Both FNS and USDA's
Office of Inspector General analyze the EBT data using a computer program
that identifies transaction patterns indicating possible trafficking. These
data are then used to investigate and take action against suspected
traffickers. The use of EBT data has a particular advantage: Under the
Personal Responsibility and Work Opportunity Reconciliation Act of 1996,
federal agencies may use EBT data alone, without the expense of conducting
an undercover investigation, to take action against storeowners violating
the requirements of the Food Stamp Program. This act also mandates that all
states implement EBT systems by October 1, 2002, unless USDA waives the
requirement. Collectively, EBT systems now supply about 70 percent of all
food stamp benefits.

Of the 29 states with statewide electronic benefit systems, only 4--Florida,
Missouri, South Carolina, and Texas--independently and proactively analyze
their electronic databases to identify suspect recipients. Apparently, these
states viewed this activity as essential to their efforts to improve the
integrity of the Food Stamp Program. Additionally, since 1994, USDA's Office
of Inspector General has identified about 34,000 suspected traffickers in
Maryland and provided this information to that state. All five of these
states invested the resources necessary to investigate suspect recipients
and disqualify those engaged in trafficking. For example, for fiscal years
1998 and 1999, these five states were responsible for disqualifying about 99
percent of all the individuals nationwide who were removed from the Food
Stamp Program for trafficking benefits. Although not as aggressive or
proactive as these 5 states, 9 other of the 29 states investigated suspect
recipients--identified by other sources, such as FNS through its efforts in
disqualifying storeowners--and disqualified those who engaged in
trafficking. The remaining 15 states did not disqualify any recipient for
trafficking during the 2-year period.

Table 1 shows the actions taken by the 29 states and FNS against suspect
recipients during fiscal years 1998 and 1999.

Table 1: Number of Trafficking Storeowners and Recipients Removed From the
Food Stamp Program in the 29 States With Statewide EBT Programs, Fiscal
Years 1998-99

Continued

              Does the state   Does FNS   Does the state                Number of
               independently  send a list use FNS' list                trafficking
               use EBT data  of suspected    or other      Number of   storeowners
    State       to identify   trafficking   sources to    recipients   disqualified
                 suspected    recipients   investigate   disqualified      by
                 recipient      to the    and disqualify
               traffickers?    state?a      recipient                      FNS
                                          traffickers?b
 Alabama      No             No           No             0            10
 Alaska       No             c            No             0            0
 Arkansas     No             No           Yes            10           3
 Colorado     No             Yes          No             0            6
 Connecticut  No             Yes          No             0            10
 District of
 Columbia     No             Yes          Yes            5            35
 Florida      Yes            Yes          Yes            83           114
 Georgia      No             Yes          No             0            44
 Hawaii       No             No           No             0            4
 Idaho        No             c            No             0            0
 Illinois     No             Yes          Yes            3            31
 Kansas       No             c            Yes            6            0
 Louisiana    No             No           Yes            10           29
 Maryland     No             Yes          Yes            3,000        35
 MassachusettsNo             Yes          No             0            2
 Minnesota    No             c            Yes            4            0
 Missouri     Yes            Yes          Yes            335          2
 New HampshireNo             c            No             0            0
 New Mexico   No             No           Yes            20           4
 North Dakota No             c            No             0            0
 Oklahoma     No             No           Yes            25           5
 Oregon       No             c            No             0            0
 Pennsylvania No             Yes          No             0            59
 Rhode Island No             c            No             0            0
 South
 Carolina     Yes            Yes          Yes            393          54
 South Dakota No             c            No             0            0
 Texas        Yes            No           Yes            2,966        75
 Utah         No             Yes          Yes            13           2
 Vermont      No             c            No             0            0
 Total                                                   6,873        524

Note: These states implemented statewide EBT systems at different dates, but
all had statewide EBT systems in effect by Apr. 1, 1999. See app. II for the
date each state implemented EBT.

a FNS currently does not consistently provide a list of suspected
trafficking recipients to the states.

b Other sources include USDA's Office of Inspector General and hotline
calls.

c FNS had not disqualified any storeowners in this state.

Source: Food and Nutrition Service.

As the table shows, a total of 6,873 recipients nationwide had been
disqualified during fiscal years 1998 and 1999. About 99 percent of the
disqualifications occurred in five states--Maryland, 44 percent; Texas, 43
percent; South Carolina, 6 percent; Missouri, 5 percent; and Florida, 1
percent.4 In these states, officials principally relied upon the results of
a detailed analysis of EBT databases to identify suspect recipients and
invested the necessary resources to investigate them and to disqualify those
found to engage in trafficking. These states coordinated their work with FNS
and USDA's Office of Inspector General to ensure their investigations did
not overlap, which could have alerted trafficking storeowners that
recipients with whom they had engaged in trafficking were being
investigated.

Two of the states--Florida and Texas--generally use the strategy that FNS
follows in identifying trafficking storeowners. That is, they analyze their
EBT data first to identify stores likely to be engaged in trafficking and
then they identify likely trafficking recipients using those stores.
However, the states' lists of suspected traffickers are more comprehensive
than FNS'. For example, Texas often identifies hundreds of recipients
suspected of trafficking at each store identified as likely to be engaged in
trafficking. On the other hand, FNS generally limits its identification of
recipients to the few cases that it needs to support its actions against a
storeowner. After recipients are identified as suspected traffickers, the
states investigate to confirm whether trafficking actually occurred before
disqualifying those found to be engaging in trafficking.

Other states take a different approach. For example, Missouri identifies
suspect recipient traffickers by profiling all recipients in the EBT
database without regard to specific stores. Between January and August 1999,
Missouri identified about 500 recipients for possible investigation.
Maryland uses information provided by USDA's Office of Inspector General
through a special arrangement to investigate and take action against
recipients.5 Since 1994, when the Inspector General began to provide this
list to Maryland, the state's food stamp trafficking unit has disqualified
about 7,700 recipients out of about 34,000 referrals, including 3,000 during
fiscal years 1998 and 1999. These referrals were associated with eight
trafficking storeowners.

Of the 15 states that had not taken any action against trafficking
recipients, 5 had received the identity of suspected recipients from FNS.
According to officials in these states, they did not investigate suspect
recipients because the investigations are time-consuming and costly and it
was not cost-effective to do so. The officials in the five states that
disqualified about 99 percent of all those removed from the program
nationwide for trafficking agreed that acting against suspect traffickers
was not cost-effective. However, these officials and FNS officials agree
that identifying suspect recipients and disqualifying those who traffic is
an essential activity for maintaining the integrity of the Food Stamp
Program. They maintain that their efforts act as a deterrent by discouraging
other recipients from engaging in trafficking. They also recognize that they
have a fiduciary responsibility to operate the program effectively and
efficiently. In this regard, FNS has established improving the integrity of
the program as a major goal in complying with principles of the Government
Performance and Results Act of 1993 (Results Act).

Extent of RecipientTrafficking, as Prescribed by the Results Act

Although EBT data have been available since 1993 to analyze food stamp
transactions for trafficking, FNS has only recently taken steps to encourage
the states to target recipients engaged in trafficking.6 In July 1999, FNS
instructed its regional offices to work with the states to reach agreements
on how best to use the EBT data now available to identify, investigate, and
disqualify trafficking recipients. It also proposed a regulation to permit
the states to retain 35 percent of the funds they collect from the penalties
assessed against recipients disqualified from the program for trafficking.
However, the Service will not be able to determine how effective these
actions will be in reducing the overall level of trafficking because it
lacks a current, reliable estimate of the extent of trafficking. Such an
estimate would also allow FNS to better adhere to the principles of the
Results Act, which state that goals and strategies be quantifiable and
measurable. EBT data could be used to develop a current, reliable estimate
of the extent of trafficking, to establish goals and strategies for reducing
trafficking, and to measure the effectiveness of these strategies in
reducing trafficking.

Although some states began using EBT statewide in the mid-1990s, FNS only
began in 1999 to develop regional plans with the states for a joint effort
to reduce recipient trafficking using EBT data. The seven draft plans we
reviewed were all different, but they generally included such activities as
defining the federal and state roles for identifying recipients suspected of
trafficking and developing processes for routinely sharing information. All
these plans provide that FNS would submit to the states the names of suspect
recipients associated with storeowners disqualified from the program.
However, all states could target a more extensive list of suspect recipients
for investigation. FNS' plans do not set goals for the number of recipients
to be investigated and/or disqualified. Furthermore, none of the plans
described how they contribute to FNS' overall effort to reduce trafficking.
As of December 1999, all of these plans were still in the preliminary
stages.

In a separate action, recognizing the high cost to the states of
investigating and disqualifying trafficking recipients, FNS recently
proposed an amendment to the food stamp regulations to allow the states to
establish claims to recover the benefits trafficked by those disqualified
from the program and to keep 35 percent of the funds collected. This portion
of the collected funds that the states would retain is in addition to the 50
percent in administration costs that FNS already provides the states for
managing the program. Under this proposed regulation, the states could keep
some of the funds collected from a trafficker--the disqualified household
member; more probably, however, the family's benefit would be reduced by up
to 20 percent each month until the amount trafficked is repaid.

To illustrate, assume that a family of four had received $400 in monthly
food stamp benefits and that one household member was disqualified for
trafficking a total of $600. Three family members would still be entitled to
receive monthly benefits totaling $300. The state could collect the money
directly from the member disqualified or more likely recoup the benefits
trafficked by offsetting benefits provided to other family members. The
state could reduce future monthly benefits to the family by $60 (20 percent
of $300) for 10 months to recover the $600 of benefits trafficked. The state
would retain 35 percent of the amount recouped ($210).

The six states we visited were all aware of the terms of the proposed
regulatory change and how it provided a financial incentive to take action
against recipient traffickers. However, they differed on whether the
potential amounts to be recovered would be adequate for funding the expenses
of independently identifying, investigating, and penalizing recipients who
traffic benefits.

the Extent of Recipient Trafficking

FNS is not able to measure the effectiveness of its or the states' efforts
in reducing the overall level of trafficking because it lacks current,
reliable information on the extent of trafficking. The last estimate,
developed in 1995, used 1993 data and did not rely on EBT data. With the
introduction of EBT systems in 1993, FNS has an important tool for
developing current estimates of the extent of trafficking at the local,
state, and national levels. Using these estimates, FNS could establish goals
for reducing trafficking on the basis, for example, of the value of benefits
trafficked each year. FNS could then develop strategies to efficiently and
effectively reduce trafficking and use EBT data to measure the extent to
which it was achieving its goals. This approach to addressing the issue of
recipient trafficking would be based upon a more reliable estimate of the
extent of trafficking and therefore would be more consistent with the
provisions of the Results Act, which prescribe that valid data be used to
measure performance.

FNS' actions to help the states reduce recipient trafficking are not being
guided by the best available estimate of the extent of recipient
trafficking, which would enable it to better set appropriate goals and
strategies, as prescribed by the Results Act. Instead, as currently proposed
in the draft regional plans, FNS would work with the states only to
investigate the number of suspect recipients identified as being involved
with trafficking storeowners. In its fiscal year 1999 and 2000 performance
plans, FNS' goal is to disqualify 1,201 stores annually. FNS could realize
this goal but not substantially reduce the overall level of trafficking
because the stores disqualified may be those stores with relatively low
levels of trafficking. FNS has not set priorities for targeting the
trafficking stores--for example, the volume of transactions and/or the value
of the benefits trafficked. If FNS set such priorities and identified these
storeowners, additional states might have an incentive to examine more
suspect recipients purchasing at these stores because the likelihood of
recipient trafficking would be greater.

The use of EBT data to identify suspect recipients has proven effective for
developing a pool of recipients likely to be trafficking as a first step in
conducting investigations and disqualifying those found to have engaged in
trafficking. The five states extensively using EBT data have disqualified
about 99 percent of the 6,873 traffickers disqualified nationwide from the
program. FNS and these states recognize that identifying and disqualifying
traffickers is a costly undertaking. However, they also rightly recognize
that their efforts can serve as a deterrent to future trafficking and that
they have a fiduciary responsibility to maintain the integrity of the Food
Stamp Program. The aggressive and specific techniques used by these states
in conducting their efforts to minimize trafficking could provide useful
models to the other states.

FNS and state efforts to identify, investigate, and disqualify trafficking
recipients are being conducted in the absence of a reliable estimate of the
extent of trafficking. Without such an estimate, neither FNS nor the states
can measure their performance and determine the most effective deterrents to
trafficking. With the increasing use of EBT data, more reliable estimates
can be developed. If FNS and the states are to work together to deter
trafficking, developing such estimates is an important step in setting
appropriate goals and strategies and measuring the effectiveness of those
strategies, as prescribed by the Results Act.

To improve the integrity of the Food Stamp Program, we recommend that the
Secretary of Agriculture direct the Administrator of the Food and Nutrition
Service to (1) work with the five states currently using EBT data to
determine the best techniques for using these data to identify suspected
recipient traffickers and work with the other states with statewide EBT
systems to implement the best techniques, as appropriate and (2) use EBT
data to periodically develop reliable estimates of the extent of trafficking
and use these estimates to develop goals and appropriate strategies for
reducing trafficking.

We provided a draft of this report to the U. S. Department of Agriculture
for its review and comment. We met with Food and Nutrition Service
officials, including the Directors of the Program Accountability Division
(Food Stamp Program) and the Grants Management Division. These officials
generally agreed with the report's findings, conclusions, and
recommendations. They stated that the Service's ability to implement the
recommendation on using EBT data to maintain reliable estimates of the
extent of trafficking would be sharply limited because the Service has
neither the specialized software to analyze large data sets nor the
expertise needed to use it properly. They also stated that the Congress had
recently prohibited the Service from using program funds to conduct such
studies or evaluations. However, while the Congress has restricted the
Service's use of certain research funds, it has provided millions of dollars
to the Service for fighting fraud and improving the integrity of the Food
Stamp Program. For example, the Congress provided the Service $5 million in
fiscal year 2000 to use in preventing, identifying, and prosecuting fraud,
among other purposes, and further directed the Service to spend at least $3
million more to improve the integrity of the Food Stamp and Child Nutrition
programs. Furthermore, the President's budget for fiscal year 2001 proposes
$13 million for these purposes. The FNS officials also stated that the
Service had recently developed, using old research funding, an estimate of
the extent of trafficking nationwide as of 1998, and plans to release the
estimate by the end of March 2000. These officials also provided a number of
editorial and technical comments, which we incorporated into the report,
where appropriate.

We conducted our work from May 1999 through January 2000 in accordance with
generally accepted government auditing standards. Appendix I discusses our
scope and methodology.

As arranged with your office, unless you announce its contents earlier, we
plan no further distribution of this report until 30 days after the date of
this report. At that time, copies of this report will be sent to
congressional committees responsible for the Food Stamp Program; the
Honorable Dan Glickman, Secretary of Agriculture; the Honorable Shirley
Watkins, Under Secretary of Food, Nutrition, and Consumer Services, USDA;
the Honorable Samuel Chambers, Administrator, Food and Nutrition Services,
USDA; and the Honorable Roger Viadero, Inspector General, USDA. We will also
provide copies to others on request.

If you or your staff have any question about this report, please contact me
at (202) 512-5138. Key contributors to this report are listed in appendix
III.

Sincerely yours,
Lawrence J. Dyckman
Director, Food and Agriculture Issues

Scope and Methodology

To determine the extent to which states with statewide electronic benefit
transfer (EBT) systems were identifying and disqualifying recipients who
engage in food stamp trafficking, we visited six states--Florida, Georgia,
Illinois, Louisiana, Maryland, and Texas--with cities that our analysis of
EBT data showed had high rates of trafficking. In each of those states, we
reviewed available policies, procedures, and practices, and interviewed
officials responsible for investigating food stamp trafficking. We also
obtained data from the Food and Nutrition Service's (FNS) regional offices
on the 29 states that had statewide EBT systems in effect before April 1999,
as well as from FNS' management information systems. We limited our work to
the states with statewide EBT systems because these systems provide a
detailed database of all transactions between recipients and stores. We did
not review the activities of the remaining states that distribute benefits
through coupons because these states can identify suspected recipients
engaged in trafficking only by conducting extensive undercover
investigations.

To determine the actions FNS has taken to encourage states to identify and
disqualify recipients engaged in trafficking, we interviewed FNS officials
at headquarters and four FNS regional offices--the Southeast Region in
Atlanta, Georgia; the Southwest Region in Dallas, Texas; the Midwest Region
in Chicago, Illinois; and the Mid-Atlantic Region in Robbinsville, New
Jersey. We also reviewed FNS' strategic and annual performance plans and
draft plans recently developed by each FNS regional office for working with
the states within their jurisdiction to pursue recipients engaged in
trafficking. We conducted follow-up telephone contacts, as needed, to
discuss the details of these regional plans and to obtain information on
actions taken by the states. We also analyzed relevant legislation,
implementing regulations, and program guidance. In addition, we met with
officials in the U.S. Department of Agriculture's Office of the Inspector
General and reviewed their reports on this program relating to fraud and
abuse.

We conducted our work from May 1999 through January 2000 in accordance with
generally accepted government auditing standards.

Twenty-Nine States With Statewide EBT Data, as of April, 1999

Source: Food and Nutrition Services.

 State                Date EBT system was implemented
 Alabama              November 1997
 Alaska               June 1998
 Arkansas             April 1998
 Colorado             February 1998
 Connecticut          October 1997
 District of Columbia October 1998
 Florida              October 1998
 Georgia              November 1998
 Hawaii               August 1998
 Idaho                February 1998
 Illinois             November 1997
 Kansas               March 1997
 Louisiana            December 1997
 Maryland             April 1993
 Massachusetts        October 1997
 Minnesota            October 1998
 Missouri             June 1998
 New Hampshire        January 1999
 New Mexico           August 1995
 North Dakota         March 1997
 Oklahoma             January 1998
 Oregon               May 1998
 Pennsylvania         September 1998
 Rhode Island         October 1998
 South Carolina       December 1995
 South Dakota         March 1997
 Texas                November 1995
 Utah                 April 1996
 Vermont              October 1998

Key Contacts and Staff Acknowledgments

Lawrence J. Dyckman, (202) 512-5138
Ron E. Wood, (202) 512-5138

In addition to those named above, John K. Boyle, Paul J. Pansini, Carol
Herrnstadt Shulman, and Michael Tropauer made key contributions to this
report.

(150139)

Table 1: Number of Trafficking Storeowners and Recipients Removed From the
Food Stamp Program in the 29 States With Statewide EBT Programs, Fiscal
Years 1998-99 7
  

1. For this review, we included the 29 states (including the District of
Columbia) that had statewide EBT systems as of Apr. 1, 1999, and had
sufficient time to use EBT data for analyzing recipients' purchasing
patterns.

2. Food Stamp Program: Information on Trafficking Food Stamp Benefits
(GAO/RCED-98-77, Mar. 26, 1998).

3. In some states, trafficking can also be prosecuted in state court.

4. Florida's statewide EBT system became operational in Oct. 1998,
Maryland's in Apr. 1993, Missouri's in June 1998, South Carolina's in Dec.
1995, and Texas' in Nov. 1995.

5. The Office of the Inspector General analyzes EBT data to generate a list
of all suspect recipients associated with storeowners who have been
convicted of trafficking. This list is more complete than the information
FNS would provide in connection with its investigation of trafficking
storeowners.

6. Only five states had adopted statewide EBT systems before 1997.
*** End of document. ***