Food Assistance: Efforts To Control Fraud and Abuse in the WIC Program
Can Be Strengthened (Chapter Report, 08/30/1999, GAO/RCED-99-224).

GAO examined fraud and abuse in the Department of Agriculture's (USDA)
Special Supplemental Nutrition Program for Women, Infants, and Children
(WIC), focusing on: (1) what is known about the amount of fraud and
abuse in the program, including the levels of fraud and abuse being
detected by the state and local agencies responsible for administering
WIC; and (2) the efforts taken to prevent and detect fraud and abuse and
barriers to these efforts.

GAO noted that: (1) USDA has no recent estimates on the overall levels
of vendor, participant, or employee fraud and abuse occurring in the WIC
program; (2) the Department has some information on that portion of
fraud and abuse detected by state agencies; (3) in estimating the extent
of WIC fraud and abuse, USDA has relied on two studies, completed in
1988 and 1991, that provide information on some types of vendor and
participant fraud and abuse; (4) USDA is updating these two studies and
expects to have this information late in 2000 for vendors and late in
1999 for participants; (5) regarding fraud and abuse that has been
detected, USDA collects information on an annual basis from state WIC
agencies on vendors but not on participants or employees; (6) while this
information is limited, it is valuable because it helps USDA monitor
states' efforts to detect vendor fraud and abuse; (7) at the state and
local levels, according to GAO's survey, agencies reported detecting
higher levels of vendor fraud than of participant and employee fraud;
(8) the number of vendors identified as committing fraud or abuse in a
2-year period--October 1, 1996, through September 30, 1998--represents
about 9 percent of all vendors in the WIC program as of September 30,
1998; (9) in comparison, the number of participants identified as
committing fraud or abuse of a serious nature during this 2-year period
represented an estimated .14 percent of the average monthly number of
participants in fiscal year 1998; (10) very little employee fraud or
abuse was reported; (11) WIC agencies use a variety of strategies to
prevent and detect vendor, participant, and employee fraud and abuse;
(12) regarding vendors, states differ in the management procedures they
use to control fraud and abuse, including the methods they use to limit
the number of vendors they authorize to participate in WIC and the
amount and type of vendor monitoring they perform; (13) with respect to
participants, state and local agencies use a variety of strategies,
ranging from educating participants on program rules and regulations to
using a computer system to verify that a participant is eligible on the
basis of participation in another program; (14) agency officials
identified several barriers to preventing and detecting fraud and abuse;
and (15) these barriers include a lack of: (a) federal criteria to
evaluate whether a state has authorized an appropriate number of vendors
given the resources available to manage vendors; (b) information on
participant fraud and abuse being detected; and (c) policies or
procedures regarding potential employee conflict-of-interest situations.

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

 REPORTNUM:  RCED-99-224
     TITLE:  Food Assistance: Efforts To Control Fraud and Abuse in the
	     WIC Program Can Be Strengthened
      DATE:  08/30/1999
   SUBJECT:  State-administered programs
	     Food programs for children
	     Reporting requirements
	     Internal controls
	     Infants
	     Children
	     Women
	     Program abuses
	     Fraud
	     Federal/state relations
IDENTIFIER:  Special Supplemental Food Program for Women, Infants, and
	     Children
	     WIC

******************************************************************
** This file contains an ASCII representation of the text of a  **
** GAO report.  Delineations within the text indicating chapter **
** titles, headings, and bullets are preserved.  Major          **
** divisions and subdivisions of the text, such as Chapters,    **
** Sections, and Appendixes, are identified by double and       **
** single lines.  The numbers on the right end of these lines   **
** indicate the position of each of the subsections in the      **
** document outline.  These numbers do NOT correspond with the  **
** page numbers of the printed product.                         **
**                                                              **
** 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.                                         **
**                                                              **
** A printed copy of this report may be obtained from the GAO   **
** Document Distribution Center.  For further details, please   **
** send an e-mail message to:                                   **
**                                                              **
**                                            **
**                                                              **
** with the message 'info' in the body.                         **
******************************************************************

Cover
================================================================ COVER

Report to Congressional Committees

August 1999

FOOD ASSISTANCE - EFFORTS TO
CONTROL FRAUD AND ABUSE IN THE WIC
PROGRAM CAN BE STRENGTHENED

GAO/RCED-99-224

WIC Fraud and Abuse

(150290)

Abbreviations
=============================================================== ABBREV

  EBT - electronic benefits transfer
  FNS - Food and Nutrition Service
  GAO - General Accounting Office
  ITO - Indian tribal organizations
  NSA - nutrition services and administration
  USDA - U.S.  Department of Agriculture
  WIC - Special Supplemental Nutrition Program for Women, Infants and
     Children

Letter
=============================================================== LETTER

B-282910

August 30, 1999

Congressional Committees

This report examines fraud and abuse in the U.S.  Department of
Agriculture's Special Supplemental Nutrition Program for Women,
Infants and Children (WIC).  As part of our review, we collected
information, through surveys and interviews, from USDA's Food and
Nutrition Service and state and local WIC agencies on the extent of
fraud and abuse in the program.  Specifically, we (1) describe what
is known about the amount of fraud and abuse in the program,
including the levels of fraud and abuse being detected by the state
and local agencies responsible for administering WIC, and (2) examine
the efforts taken to prevent and detect fraud and abuse and barriers
to these efforts.  We are providing this report because of
congressional concerns about the potential for fraud and abuse in the
WIC program and the lack of reliable information on fraud and abuse
in the program. 

We are sending copies of this report to Senator Thad Cochran,
Chairman, and Senator Herbert Kohl, Ranking Minority Member,
Subcommittee on Agriculture, Rural Development, and Related Agencies,
Senate Committee on Appropriations; Representative Joe Skeen,
Chairman, and Representative Marcy Kaptur, Ranking Minority Member,
Subcommittee on Agriculture, Rural Development, Food and Drug
Administration, and Related Agencies, House Committee on
Appropriations.  We are also sending copies of this report to the
Honorable Daniel R.  Glickman, Secretary of Agriculture and the
Honorable Jacob J.  Lew, Director, Office of Management and Budget. 
We will also make copies available to others upon request. 

If you have any questions about this report, please contact me at
(202) 512-5138.  Major contributors to the report are listed in
appendix II. 

Lawrence J.  Dyckman
Director, Food and
Agriculture Issues

List of Recipients

The Honorable Richard G.  Lugar
Chairman
The Honorable Tom Harkin
Ranking Minority Member
Committee on Agriculture, Nutrition, and Forestry
United States Senate

The Honorable William F.  Goodling
Chairman
The Honorable William (Bill) Clay
Ranking Minority Member
Committee on Education and the Workforce
House of Representatives

The Honorable John R.  Kasich
Chairman
The Honorable John M.  Spratt, Jr.
Ranking Minority Member
Committee on the Budget
House of Representatives

The Honorable Pete Domenici
Chairman
The Honorable Frank Lautenberg
Ranking Minority Member
Committee on the Budget
United States Senate

EXECUTIVE SUMMARY
============================================================ Chapter 0

   PURPOSE
---------------------------------------------------------- Chapter 0:1

In recent years, the Congress has expressed increased concern about
the vulnerability to fraud and abuse of the U.S.  Department of
Agriculture's (USDA) Special Supplemental Nutrition Program for
Women, Infants and Children (WIC).  The WIC program provides
nutritious supplemental foods and nutrition education and assistance
to lower-income pregnant, breast-feeding, and postpartum women;
infants; and children to the age of 5, who are at nutritional risk. 
WIC is a discretionary grant program that serves as many individuals
as the available funding permits.  Therefore, actions taken to reduce
losses resulting from fraudulent activities will make resources
available to serve more eligible people.  In 1999, the Congress
appropriated $3.9 billion for WIC, which is expected to serve about
7.4 million participants in an average month. 

Because of congressional concerns about the potential for fraud and
abuse in the WIC program and because of the lack of reliable
information about fraud and abuse, GAO (1) described what is known at
the federal, state, and local levels about the amount of fraud and
abuse in the program, including information on detected fraud and
abuse, and (2) examined the efforts taken to prevent and detect fraud
and abuse and barriers to these efforts. 

This report is based on the results of a nationwide survey of state
WIC agencies; a representative random sample of local WIC agencies;
site visits conducted in five states accounting for about 42 percent
of the total average monthly participation for fiscal year 1998; and
discussions with agency officials.  See appendix I for a more
detailed discussion of the methodology used to conduct this work. 

   BACKGROUND
---------------------------------------------------------- Chapter 0:2

WIC is a federally funded nutrition assistance program administered
by USDA's Food and Nutrition Service.  Food benefits are typically
provided to participants in the form of vouchers that can be used to
obtain approved foods at authorized retail outlets and food stores,
commonly referred to as vendors.  Vendors, participants, and
employees can engage in a variety of fraudulent or abusive
activities.  For example, vendors could charge the WIC program more
for a food item than the item's shelf price.  Participants could have
misrepresented facts affecting their eligibility, such as income, in
order to receive program benefits.  Finally, employees could obtain
benefits for friends or family who are not eligible for the program. 

WIC operates in the 50 states and the District of Columbia, through
33 Indian tribal organizations, and in Guam, the U.S.  Virgin
Islands, American Samoa, and the Commonwealth of Puerto Rico.  These
88 government entities administer the program through about 1,846
local WIC agencies.  These local agencies are public or private
nonprofit health or human services agencies and serve participants
through the clinics located in their service area.  Although USDA
sets the regulations for the WIC program and provides guidance, the
states have flexibility in the policies and procedures they use. 
USDA's Office of the Inspector General has completed several reviews
of WIC program operations at specific locations within the past 5
fiscal years; however it has not completed a comprehensive nationwide
review of the program since 1988, when it examined the food benefit
delivery system and vendor monitoring. 

   RESULTS IN BRIEF
---------------------------------------------------------- Chapter 0:3

USDA has no recent estimates on the overall levels of vendor,
participant or employee fraud and abuse occurring in the WIC program. 
The Department has some current information on that portion of fraud
and abuse detected by state agencies.  In estimating the extent of
WIC fraud and abuse, USDA has relied on two studiesï¿½completed in 1988
and 1991ï¿½that provide information on some types of vendor and
participant fraud and abuse.  USDA is updating these two studies and
expects to have this information late in 2000 for vendors and late in
1999 for participants.  Regarding fraud and abuse that has been
detected, USDA collects information on an annual basis from state WIC
agencies on vendors but not on participants or employees.  While this
information is limited, it is valuable because it helps USDA monitor
states' efforts to detect vendor fraud and abuse.  At the state and
local levels, according to GAO's survey, agencies reported detecting
higher levels of vendor fraud than of participant and employee fraud. 
The number of vendors identified as committing fraud or abuse in a
2-year periodï¿½October 1, 1996 through September 30, 1998ï¿½represents
about 9 percent of all vendors in the WIC program as of September 30,
1998.  In comparison, the number of participants identified as
committing fraud or abuse of a serious nature during this 2-year
period represented an estimated .14 percent of the average monthly
number of participants in fiscal year 1998; the number reported as
committing less serious offenses represented an estimated 1.64
percent of the average monthly number of participants.  Very little
employee fraud or abuse was reported.  Information provided by state
and local agencies on the amount of detected fraud and abuse does not
present a complete picture of fraud and abuse in the program, partly
because the level of detection efforts differ among both state and
local agencies.  In addition, regardless of detection efforts, some
violations go undetected. 

WIC agencies use a variety of strategies to prevent and detect
vendor, participant, and employee fraud and abuse.  Regarding
vendors, states differ in the management procedures they use to
control fraud and abuse, including the methods they use to limit the
number of vendors they authorize to participate in WIC and the amount
and type of vendor monitoring they perform.  With respect to
participants, state and local agencies use a variety of strategies,
ranging from educating participants on program rules and regulations
to using a computer system to verify that a participant is eligible
on the basis of participation in another program, such as Medicaid. 
Concerning employees, strategies such as supervisory review and
protection of food vouchers from theft are used by most local
agencies; however only about half of the local agencies reported
having a conflict-of-interest policy for WIC employees who are also
WIC participants.  Agency officials identified several barriers to
preventing and detecting fraud and abuse.  These barriers include a
lack of (1) federal criteria to evaluate whether a state has
authorized an appropriate number of vendors given the resources
available to manage vendors, (2) information on participant fraud and
abuse being detected, and (3) policies or procedures regarding
potential employee conflict-of-interest situations.  GAO is making
several recommendations designed to overcome these barriers. 

   PRINCIPAL FINDINGS
---------------------------------------------------------- Chapter 0:4

      USDA DOES NOT HAVE OVERALL
      ESTIMATES OF FRAUD AND ABUSE
-------------------------------------------------------- Chapter 0:4.1

USDA is updating its two studies that provide information on vendor
and participant fraud and abuse.  It expects to have this updated
information late in 2000 for vendors and late in 1999 for
participants.  USDA will use this information to estimate the overall
levels of potential fraud and abuse in the program.  While these two
studies do not address all types of vendor and participant fraud and
abuse, such as authorized vendors who redeem vouchers obtained from
unauthorized vendors and individuals who participate but are
ineligible on the basis of facts other than income, USDA will use
this information to estimate the levels of potential fraud and abuse
in the program associated with vendors overcharging the program and
participation by individuals whose income level makes them ineligible
for the program. 

      USDA IS IMPROVING
      INFORMATION COLLECTED ON
      DETECTED VENDOR FRAUD AND
      ABUSE BUT COLLECTS NO
      INFORMATION ON DETECTED
      PARTICIPANT OR EMPLOYEE
      FRAUD AND ABUSE
-------------------------------------------------------- Chapter 0:4.2

In the past, USDA collected aggregate information from state agencies
on the amount of detected vendor fraud and abuse.  Recently, it has
developed a system for collecting profile data on all vendors
participating in the WIC program, including information on vendors
detected as having committed fraud or abuse.  These data should
enable USDA to develop basic information on the number and
characteristics of individual vendors identified by state and local
agencies as having committed fraud and abuse.  USDA does not collect
information from state or local agencies on participants or employees
identified as having committed fraud and abuse.  Furthermore, at the
state agency level, 21 of the 51 agencies (the 50 states and the
District of Columbia) reported that they do not maintain data on
participant fraud and abuse. 

      STATE AND LOCAL AGENCIES
      REPORT DETECTING RELATIVELY
      MORE VENDOR THAN PARTICIPANT
      OR EMPLOYEE FRAUD AND ABUSE
-------------------------------------------------------- Chapter 0:4.3

For the 2-year period GAO examined, relatively more vendor than
participant or employee fraud and abuse was identified by state and
local agencies. 

  -- Vendors.  State agencies reported identifying about 3,771
     vendors as having committed fraud or abuse.  This number
     represents about 9 percent of all vendors in the program as of
     September 30, 1998.  There was substantial variation in the
     number of vendors detected as having committed fraud or abuse
     among the states, ranging from 15 states reporting no detection
     to 6 states reporting 25 percent or more of their vendors
     detected.  The levels detected may reflect, in part, the level
     of effort to detect fraud and abuse by the state agencies as
     well as the types of strategies the states use to monitor the
     program. 

  -- Participants.  Local agencies identified an estimated 7,074
     participants as having committed one or more of seven types of
     serious fraud and abuse, such as exchanging food vouchers for
     cash.  This number represents an estimated .14 percent of the
     average monthly participation in fiscal year 1998.  In addition,
     local agencies identified an estimated 79,271 participants as
     having committed one or more of three types of less serious
     fraud and abuse, such as redeeming food vouchers outside of the
     authorized dates.  This number represents about 1.64 percent of
     the average monthly participation in fiscal year 1998. 
     Furthermore, an estimated 58 percent of the local agencies
     detected no incidences of serious participant fraud or abuse and
     an estimated 28 percent detected no incidences of less serious
     fraud or abuse.  An estimated 22 percent of local agencies
     detected no fraud or abuse in either category, and an estimated
     4 percent did not respond to the question.  At the state level,
     21 of the 51 state agencies reported that they do not maintain
     data on participant fraud and abuse.  Moreover, local agencies
     in the 21 states that reported not maintaining data on
     participant fraud and abuse were more likely to identify no
     participant fraud or abuse in the serious category. 

  -- Employees.  In GAO's survey, only 4 percent of local agencies
     identified any documented cases of employee fraud or abuseï¿½an
     estimated 48 individuals nationwide for the 2-year period. 

      AGENCIES VARY IN THEIR
      EFFORTS TO PREVENT AND
      DETECT FRAUD AND ABUSE AND
      ENCOUNTER SOME BARRIERS
-------------------------------------------------------- Chapter 0:4.4

State and local agencies use a variety of strategies to prevent and
detect fraud and abuse in the WIC program.  In an effort to
effectively manage a review of vendors, 42 of the 51 state agencies
reported making some effort to limit the number of authorized
vendors; 8 states reported not limiting the number of vendors; and
one state, Mississippi, does not use the vendor distribution system. 
While all the states met the minimum federal requirement to monitor
at least 10 percent of their vendors, they varied substantially in
whether they conducted monitoring that was of a routine nature or was
a more detailed investigation of vendor activities, such as
compliance buysï¿½when an undercover investigator poses as a
participant and attempts to exchange WIC vouchers inappropriately,
such as for nonapproved food or nonfood items.  For example, while
eight states reported that none of their vendors were subject to
compliance buys, New York State reported that about 40 percent of its
vendors were subject to compliance buys, and the District of Columbia
reported that 94 percent of its vendors were subject to compliance
buys.  States also vary in the procedures they use to identify
vendors they consider more likely to abuse the programï¿½high-risk
vendors. 

To prevent and detect participant and employee fraud and abuse, state
and local agencies use a variety of strategies.  Almost all local WIC
agencies take measures to educate participants on program rules and
responsibilities and to protect food vouchers from theft or
fraudulent alteration.  The methods to implement strategies also
vary.  For example, while almost all states reported using computer
systems to maintain participant data and identify participants in
more than one WIC program, some states reported using these systems
to verify an applicant's income eligibility for the WIC program on
the basis of participation in another income-based program. 
Regarding employee fraud and abuse, many agencies require supervisory
review of employee activities; however, not all agencies have
policies to deal with a potential conflict-of-interest for employees
who may also receive WIC benefits or who certify and issue benefits
to the same individual. 

Several barriers hamper agencies' efforts to prevent and detect fraud
and abuse.  Regarding vendors, USDA lacks clear and specific criteria
to assess whether the number of vendors authorized by a state meet
the regulatory requirements for both participant access and
management.  Criteria describing more specific considerations to be
used in selecting vendors for participation in the program would
assist state agencies in managing vendors.  Without such criteria,
some states may authorize more vendors than they can effectively
manage with their available resources.  For participants, USDA lacks
reliable data because it does not require state and local agencies to
report detected participant fraud and abuseï¿½who is committing the
fraud, how often, and what type--as well as the amount of program
funds that are lost.  This information is valuable for evaluating the
effectiveness of agency efforts aimed at preventing and detecting
participant fraud and abuse.  Without these data, USDA and state
agencies are hindered in their management of the program, including
identifying the need for changes necessary to improving program
integrity.  Furthermore, it is possible that not collecting such
information may send a message to some agency officials that
preventing and detecting participant fraud and abuse is a low
priority and thus damage the public's trust in the program. 
Regarding employees, USDA does not require agencies to have a policy
on potential employee conflict-of-interest situationsï¿½employees who
also participate in the program and may issue their own benefits and
employees who certify an individual as eligible to receive benefits
and issue benefits to the same individual.  The lack of a policy may
hinder the prevention and detection of employee fraud and abuse. 
Finally, state and local agencies reported that their staffing and
funding resources are inadequate to prevent and detect fraud, and
USDA officials reported that resource constraints at the federal
level have limited their fraud prevention and detection efforts. 

   RECOMMENDATIONS
---------------------------------------------------------- Chapter 0:5

To enhance USDA's ability to improve program integrity and to
encourage better fraud and abuse prevention and detection efforts
among state and local WIC agencies, GAO recommends that the Secretary
of Agriculture direct the Administrator of the Food and Nutrition
Service to take the following actions: 

  -- Amend the regulations on vendor management to ensure that the
     states limit their authorized vendors to a number that they can
     effectively manage and issue guidance to the states on the
     specific criteria the Food and Nutrition Service will use to
     assess the states' compliance with the regulations and the
     actions a state would need to take if the Service determines
     that the state has authorized more vendors than it can
     effectively manage;

  -- Work with the state WIC agencies and the National Association of
     WIC Directors to develop and implement cost-effective strategies
     for the states to use in collecting and maintaining information
     on incidences of participant fraud and abuse, which would be
     periodically reported to the Food and Nutrition Service.  Such
     information should include the nature of the fraud detected and
     the associated dollar losses; and

  -- Require state agencies to have a policy and procedures for
     addressing employee conflicts of interest. 

   AGENCY COMMENTS AND GAO'S
   EVALUATION
---------------------------------------------------------- Chapter 0:6

GAO provided USDA with a draft of this report for review and comment. 
USDA generally agreed with the information and recommendations
contained in the draft report.  However, USDA raised some questions
about specific steps that would be necessary to implement two of the
recommendations.  Regarding the recommendation relating to
strengthening vendor management, USDA noted that the intent of the
recommendation could be achieved through a combination of regulatory
change and the issuance of program guidance to the states. 
Concerning the recommendation to determine the costs and benefits of
developing a national database on participant fraud and abuse, USDA
noted that it could develop cost-effective strategies for collecting
and maintaining such data by working with state WIC agencies and the
National Association of WIC Directors.  USDA believes this would
achieve the desired result without a lengthy and costly formal
cost-benefit study.  GAO agrees that these modifications to the two
recommendations would address the problems that GAO found and has
revised the draft recommendations to reflect USDA's suggestions. 
USDA also provided GAO with a number of technical comments and
clarifications that were incorporated into the report as appropriate. 

INTRODUCTION
============================================================ Chapter 1

The Special Supplemental Nutrition Program for Women, Infants and
Children (WIC) is a federally funded nutrition assistance program
that is designed to improve the health of lower-income pregnant,
breast-feeding and postpartum women; infants; and children up to age
5, who are at nutritional risk.  WIC provides participants with
nutritious supplemental foods, nutrition education, and referrals to
health care services.  It was established in 1972 by Public Law
92-433 as an amendment to the Child Nutrition Act of 1966.  USDA's
Food and Nutrition Service (FNS) administers WIC through federal
grants to state agencies.  In 1999, the Congress appropriated $3.9
billion for the WIC program, which is expected to serve about 7.4
million participants in an average month. 

In recent years, the Congress has expressed increased concern about
the WIC program's vulnerability to fraud and abuse.  In response to
some of these concerns, the William F.  Goodling Child Nutrition
Reauthorization Act of 1998 (P.L.  105-336) (the Goodling Act), which
reauthorized WIC and other child nutrition programs, contained
provisions specifically aimed at improving the integrity of the WIC
program.  Because WIC is a discretionary grant program that serves as
many individuals as the available funding permits, any actions that
FNS, state, and local WIC agencies take to reduce losses from
fraudulent activities make resources available to serve more eligible
people. 

   WIC ADMINISTRATION AND FUNDING
---------------------------------------------------------- Chapter 1:1

FNS' headquarters and seven regional offices administer the WIC
program through a federal/state partnership in which FNS makes funds
available in the form of grants to state agencies that, in turn,
provide program benefits to participants through local WIC agencies. 
Each state agency is responsible for developing guidelines to ensure
that WIC benefits are effectively delivered to eligible participants. 
WIC operates in all 50 states and the District of Columbia, through
33 Indian tribal organizations, and in the U.S.  Virgin Islands,
American Samoa, Guam, and the Commonwealth of Puerto Rico.  These 88
government entities administer the program through 1,846 local WIC
agencies.  These local agencies are public or private nonprofit
health or human service agencies; they can also be an Indian health
service unit, a tribe, or an intertribal council.  Local agencies
serve participants directly or through one or more clinics located in
their service area.  Staff at local WIC agencies and clinics approve
applicants for participation, distribute food benefits, and provide
nutrition education to eligible individuals.  Local WIC agencies vary
in average monthly participation, serving from about 307,000 at the
largest local agency to fewer than 35. 

Grants to the states are divided into food grants and nutrition
services and administration (NSA) grants.  Food grants cover the
costs of supplemental foods and are allocated to the states through a
formula that is based on the number of individuals in each state who
are eligible for WIC benefits.  WIC foods are designed to supplement
the participant's diet and are high in protein, calcium, iron, and
vitamins A and C.  These foods include milk, juice, eggs, cereal,
peanut butter, dried beans, tuna fish, carrots, and infant formula. 
Specially prescribed infant formula is also available to meet unusual
dietary or health-related conditions within parameters established in
federal regulations.  Each state designates the types and amounts of
foods that local WIC agencies can prescribe to meet the nutritional
needs of each participant. 

The NSA grants are allocated to the states through a formula that is
based on factors such as a state's number of projected program
participants and a salary differential for state and local government
employees by industry.  NSA grants cover the costs of program
administration, including start-up, monitoring, auditing,
accountability for food delivery systems, nutrition education,
breast-feeding promotion and support, outreach, certifying
applicants, developing and printing food vouchers, and activities
associated with detecting and preventing fraud and abuse. 

In fiscal year 1998, total expenditures under food grants were about
$2.8 billion, and NSA expenditures were about $1.1 billion. 
Nationally, in fiscal year 1998, expenditures for food benefits
averaged about $31.75 per participant per month, and expenditures for
activities covered under the NSA grant averaged about $12 per
participant per month.  Federal WIC appropriations totaled $3.9
billion in fiscal year 1998, up from $3.73 billion annually in fiscal
years 1996 and 1997.  The program is primarily funded by federal
appropriations; some states supplement the federal grant with their
own funds. 

   WIC ELIGIBILITY AND
   CERTIFICATION
---------------------------------------------------------- Chapter 1:2

State WIC agencies establish program eligibility criteria that are
based on federal guidelines.  To qualify for the program, WIC
applicants must show evidence of health or nutritional risk, such as
anemia, that is medically verified by a health professional.\1 In
addition, participants must have incomes at or below 185 percent of
the poverty guidelines that are established annually by the
Department of Health and Human Services.\2

For fiscal year 1998, a family of four, for example, could qualify
for the WIC program with an income of $30,433 or less.  State
agencies are required to automatically accept as income-eligible
those individuals who document their or a certain family member's
participation in the Food Stamp Program, Medicaid, or Temporary
Assistance for Needy Families Program.\3

To be certified, applicants must reside within the jurisdiction of
the state agency, and meet both the program's income and nutritional
risk criteria.  The Goodling Act increased certification requirements
to include the physical presence of each applicant at the time of the
certification in WIC and required that applicants provide
documentation of household income and residency.\4 Once applicants
are certified for participation, they return to the WIC agency
periodically (ranging from 1 to 3 months, depending upon nutritional
risk or other factors) to receive their food benefits and nutrition
education. 

Over the past 10 years, the number participating in the WIC program
in an average month more than doubled, from 3.4 million in 1988 to
7.4 million in 1998.  During this same period, the number of infants
enrolled in the WIC program grew from about 27 percent of all infants
born in the United States to about 47 percent. 

--------------------
\1 Nutritional risk means detrimental or abnormal nutritional
conditions detectable by biochemical or anthropometric measurements. 
Other nutritionally related conditions include dietary deficiencies
that impair or endanger health. 

\2 Poverty guidelines are established separately for Alaska and
Hawaii. 

\3 Participants may simultaneously participate in WIC and in the Food
Stamp Program because WIC is a supplemental nutrition program and has
separate income guidelines. 

\4 The Goodling Act provides a waiver in specific situations for
infants and children if the agency determines that the physical
presence requirement would present an unreasonable hardship.  In
addition, in establishing income eligibility on the basis of
participation in means-tested programs, the Goodling Act requires
participants to present documentation of their current enrollment. 

   FOOD DELIVERY SYSTEMS
---------------------------------------------------------- Chapter 1:3

The WIC food benefit, referred to as a food package, can be provided
through retail purchase, home delivery, direct distribution, or any
combination of these.  Generally, participants receive their food
benefits in the form of a check or a voucher (hereafter referred to
as a voucher) that they use to purchase specific foods at authorized
retail grocery stores, referred to as vendors.  Only the vendors
authorized by state agencies may redeem food vouchers in exchange for
providing supplemental foods to participants.  Vendors are authorized
to participate in the program for a specified period of
timeï¿½generally 1 to 2 years--and are subject to renewal by the state
agency.  By regulation, each state agency is permitted to authorize
an appropriate number of vendors in order to ensure adequate
participant access and effective management of the vendors.  State
agencies are permitted to establish criteria to limit the number of
vendors that they authorize and are encouraged to consider the impact
of authorization decisions on small businesses.\5 In providing
supplemental foods to participants, vendors must accept vouchers only
within the time period specified on the voucher and are authorized to
provide only the foods specified on the voucher that is presented by
the participant at the time of purchase.  In addition, vendor prices
for WIC foods cannot exceed current prices charged to other
customers. 

--------------------
\5 The Goodling Act requires that all states, in selecting a retail
store for participation in the program, take into consideration the
prices that the store charges for WIC foods compared with the prices
that other stores charge for the foods, and establish procedures to
ensure that vendors selected for participation do not subsequently
raise prices to levels that would otherwise make the store
ineligible. 

   VENDOR, PARTICIPANT, AND
   EMPLOYEE FRAUD AND ABUSE
---------------------------------------------------------- Chapter 1:4

Vendor fraud or abuse is an intentional or deliberate action taken to
violate program regulations, policies, or procedures.  Actions
include, but are not limited to, accepting food vouchers for cash,
which is known as trafficking, or providing credit toward the
purchase of unauthorized items; giving cash or credit for returned
food items that were purchased with food vouchers; altering food
vouchers or accepting expired vouchers; charging more than the shelf
price or exceeding the maximum price allowed by WIC; or charging for
food that the participant does not receive. 

Participant fraud or abuse is an activity or action by WIC
participants taken to obtain benefits to which they are not entitled
and/or to misuse the benefits they receive.  Actions include
misrepresenting facts that are used to determine eligibility (such as
income, the age of children, or the existence of children);
exchanging food vouchers for nonapproved items (such as cash, alcohol
or tobacco products, or sundries); selling or giving away food
obtained with vouchers; participating at more than one local WIC
agency simultaneously, thereby receiving multiple benefits, also
called dual participation; or verbally abusing WIC vendors and/or WIC
employees. 

Employee fraud or abuse is an intentional and deliberate action that
violates program regulations, policies, or procedures.  Actions
include, but are not limited to, misappropriating food vouchers;
altering food vouchers; entering false or misleading information in
case records; or creating fictitious or nonexistent participant
files. 

   PROGRAM REQUIREMENTS RELATED TO
   THE PREVENTION AND DETECTION OF
   FRAUD AND ABUSE
---------------------------------------------------------- Chapter 1:5

Program regulations contain several provisions relating to the
prevention and detection of vendor fraud and abuse.  The regulations
require that state agencies conduct on-site monitoring visits of
their vendors to, among other things, survey the types and levels of
abuse and errors, if any, among participating food vendors and to
take corrective action, as appropriate.  States are required to
monitor a representative sample of at least 10 percent of their
authorized food vendors annually.  On-site monitoring visits may
include an examination of cashier check out procedures and a review
to determine whether supplemental program foods are available and
whether WIC-approved food prices are clearly marked and do not exceed
agreed-upon limits.  More in-depth investigations--compliance buys,
trafficking buys, and inventory audits--are also used.  Compliance
buys occur when an undercover investigator poses as a participant and
attempts to exchange WIC vouchers inappropriately, such as for
nonapproved food or nonfood items.  Trafficking buys occur when an
undercover investigator attempts a more flagrant misuse of vouchers
usually attempting to exchange them for cash but also for firearms,
ammunition, explosives, or controlled substances (drugs).  Inventory
audits are sometimes performed to check that the store's inventory is
adequate and supports the quantities of WIC foods reported as sold. 
Regulations also require states to have a system in place to identify
problematic or high-risk vendors and to conduct on-site monitoring
and further investigationï¿½such as compliance buys--as appropriate. 

FNS regulations specify the types of sanctions that agencies can
impose on vendors identified as committing fraud or abuse.\6
Sanctions include warning letters, monetary fines or penalties, and
disqualification from the program for up to 3 years.  Recent
regulatory changes have strengthened vendor sanctions.  The Goodling
Act requires state agencies to permanently disqualify vendors
convicted of trafficking in food vouchers or selling firearms,
ammunition, explosives, or controlled substances in exchange for food
vouchers.  An exception to the rule can be made when disqualification
would cause hardship to participants or when the owner had, at the
time of the violation, an effective policy and program in effect to
prevent these violations and was not aware of or did not approve of
the violation.  When a state agency permits a vendor to continue to
participate instead of disqualifying the vendor, the state agency
must impose a civil monetary penalty.  In addition, FNS has recently
issued regulations that mandate uniform sanctions for the most
serious offenses, increase the maximum time for vendor
disqualification, other than those permanently disqualified from the
program, to 6 years, and provide for use as program income the funds
obtained from civil monetary penalties and fines.  All violations
require a pattern of incidences to warrant a mandatory sanction,
except the violations for vendors convicted of trafficking or the
illegal sale of alcoholic beverages or tobacco products, which only
require one incidence to warrant a mandatory sanction. 

Regarding participant fraud and abuse, regulations require the states
to ensure that participants meet program eligibility requirements for
medical or nutritional risk, residency, and income, and are in a
category served by the WIC program (pregnant, postpartum,
breast-feeding women, infants, and children to 5 years).  Regarding
income eligibility, the Goodling Act added the requirement that all
participants provide documentation of household income or
participation in certain public assistance programs, including Food
Stamps, Medicaid, and Temporary Assistance for Needy Families.\7
States, however, are not required to verify the documentation
presented.  The Goodling Act also requires that all applicants be
physically present at each certification determination but waives
that requirement for certain situations in which compliance would
pose a barrier to participation. 

FNS regulations specify the sanctions that can be imposed on
participants identified as committing fraud or abuse.  Sanctions
include oral warnings, written letters of warning, and
disqualification from WIC for up to 3 months. 

Regarding employee fraud, FNS regulations require state agencies to
establish an on-going management evaluation system that includes at
least reviewing local agencies' financial and participation reports,
developing corrective action plans to resolve program deficiencies,
monitoring the implementation of the corrective action plans,
performing on-site visits, and monitoring local agency operations. 
Monitoring of local agencies includes evaluating management,
certification, nutrition education, accountability, financial
management systems, and food delivery systems.  The monitoring
reviews must be conducted at least once every 2 years and include
on-site reviews of a minimum of 20 percent of the local agency's
clinics. 

State agencies are also responsible for controlling and accounting
for supplemental foods and food vouchers.  State agencies must
reconcile food vouchersï¿½redeemed, lost, stolen, expired, and voided. 
The only conflict-of-interest requirement in the regulations is that
state agencies must ensure that no conflict of interest exists
between any local agency and the food vendor or vendors within the
local agency's jurisdiction.\8

We have recently issued reports and presented testimony on several
aspects of FNS' WIC program, however, this is the first report
dealing with the prevention and detection of vendor, participant, and
employee fraud and abuse in the program.\9 USDA's Office of Inspector
General has completed reviews of prevention and detection of fraud
and abuse in the WIC program at specific locations within the past 5
fiscal years, however, it has not completed a comprehensive review of
the program since 1988, when it examined the food benefit delivery
system and vendor monitoring. 

--------------------
\6 State statutes may impose sanctions for program violations in
addition to those mandated by federal regulations.  In addition,
federal regulations require that state agencies disqualify a WIC
vendor who has been disqualified from the Food Stamp Program. 

\7 This income documentation requirement can be waived for applicants
for whom documentation is not available and for those (such as
homeless persons) for whom it would present a barrier to
participation. 

\8 There is a general provision in the WIC regulations identifying
penalties for individuals identified as committing fraud.  This
provision would include WIC employees. 

\9 Food Assistance:  Information on WIC Sole-Source Rebates and
Infant Formula Prices (GAO/RCED-98-146, May 11, 1998); Food
Assistance:  WIC Program Issues (GAO/T-RCED-98-125, Mar.  17, 1998);
Food Assistance:  Information on Selected Aspects of WIC
(GAO/T-RCED-98-128, Mar.  17, 1998); Food Assistance:  Working
Women's Access to WIC Benefits (GAO/RCED-98-19, Oct.  16, 1997); Food
Assistance:  A Variety of Practices May Lower the Costs of WIC
(GAO/RCED-97-225, Sept.  17, 1997); and WIC:  States Had a Variety of
Reasons for Not Spending Program Funds (GAO/RCED-97-166, June 12,
1997). 

   OBJECTIVES, SCOPE, AND
   METHODOLOGY
---------------------------------------------------------- Chapter 1:6

Because of congressional concerns about the potential for fraud and
abuse in the WIC program and the lack of reliable information on this
issue, we (1) described what is known about the amount of fraud and
abuse in the program, including the levels of fraud and abuse being
detected by the state and local agencies responsible for
administering WIC, and (2) examined the efforts taken to prevent and
detect fraud and abuse and barriers to these efforts. 

To obtain information on the extent of fraud and abuse in the
program, agency efforts to prevent and detect fraud and abuse, and
barriers to implementing strategies to prevent and detect fraud and
abuse, we conducted a mail survey of all state WIC agencies,
including Indian tribal organizations and U.S.  Territories.  Using a
similar survey, we surveyed a random sample of 500 local WIC agencies
from a nationwide list of about 1,846 local agencies provided to us
by FNS.  Our survey asked the directors of state and local agencies
to provide information on (1) the number and types of vendor,
participant, and employee fraud and abuse; (2) the number and types
of sanctions given to offenders, and (3) strategies used to prevent
and detect fraud and abuse.  We also asked directors to provide
information on agency activities during the application process and
on factors that limit them from implementing additional fraud
prevention and detection strategies. 

For state agencies, we received survey responses from all 50 states
and the District of Columbia, and 29 of the 37 Indian tribal
organizations and U.S.  Territories.  For local agencies, we received
survey responses from 458 of the 500 local agencies in our sample;
this gave us a response rate of 91.6 percent.  We used the responses
to the survey of local agencies to develop overall results that are
representative of those that would be obtained from the 1,691 local
agencies that would have responded had we mailed the survey to all
local agencies.  As with all sample surveys, our statistical
estimates based on data obtained from our local agency survey contain
a sampling errorï¿½the potential error that arises from not collecting
data from all local agencies.  Unless otherwise indicated, all
reported estimates based on our local survey have a sampling error of
no more than 5 percentage points if the estimate is represented in
terms of a percent.  Sampling errors are reported in appendix I. 

To better understand activities, problems, and limitations affecting
agency efforts to identify fraud and abuse, we interviewed staff at
the agencies for five states--California, Florida, Illinois, New
York, and Texas.  We judgmentally selected these agencies to include
states with high levels of participation and to provide geographic
diversity.  These five states account for about 42 percent of the
total average monthly participation in WIC for fiscal year 1998.  We
also interviewed officials at 14 local WIC agencies in the following
locations:  Irwindale, Sacramento, San Diego, and San Marcos, in
California; Dania and Miami, in Florida; Chicago and Springfield, in
Illinois; Albany and Saratoga Springs, in New York; and Austin,
Dallas, and San Antonio, in Texas.  We interviewed officials in FNS
headquarters as well as in regional offices in Atlanta, Chicago,
Dallas, and San Francisco to obtain information on overall program
operations, policies, and guidance.  For additional perspective, we
interviewed an assistant attorney general in Illinois, the state
director in Pennsylvania, and the executive directors at the Center
on Budget and Policy Priorities and the National Association of WIC
Directors.\10

We conducted our work from April 1998 through July 1999 in accordance
with generally accepted government auditing standards.  GAO contacts
and staff acknowledgements are listed in appendix II. 

--------------------
\10 The Center on Budget and Policy Priorities is an independent
nonprofit research and analysis organization located in Washington,
D.C.  The National Association of WIC Directors is a voluntary
membership organization of state and local WIC directors, WIC
nutrition coordinators, and members of corporate organizations that
provide leadership to the WIC community. 

WIC FRAUD AND ABUSE
============================================================ Chapter 2

FNS has no recent estimates on the overall level of fraud and abuse
occurring in the WIC program but does have some information on vendor
fraud and abuse detected by state agencies.  For estimating the
extent of fraud and abuse, FNS has relied on information from two
studies, completed in 1988 and 1991.\11 These studies, which are
currently being updated, provide information on the extent to which
vendors overcharge the program and the extent to which individuals
are participating in the program who are not eligible for it on the
basis of their income.  Regarding fraud and abuse that has been
detected, FNS collects data on vendor fraud and abuse identified by
state agencies but does not collect data on participant or employee
fraud and abuse detected by state and local agencies.  FNS has
collected aggregate information from states on vendor fraud and abuse
in the past and has recently taken steps to build a database of all
WIC vendors that will include information on fraud and abuse. 

State agencies maintain information on detected vendor fraud and
abuse, but only 30 of the 51 state agencies reported that they
maintain data on detected participant fraud and abuse.\12 Local
agency directors, however, were able to provide information on
participant fraud and abuse detected at their agencies.  State and
local agencies provided information on employee fraud and abuse. 
Regarding the fraud and abuse being detected by state and local
agencies, the level of vendor fraud and abuse reported by state
agencies was higher than the levels of participant and employee fraud
and abuse identified by local agencies.  Specifically, state agencies
reported that a total of 3,771 vendors were identified as having
committed various types of fraud or abuse during the 2-year period
covered by our surveyï¿½fiscal year 1997 and fiscal year 1998.\13 This
number represents about 9 percent of the WIC authorized vendors
nationwide as of September 30, 1998.\14

According to our survey of local agencies, during the 2-year period,
an estimated 7,074\15 participants were identified as having
committed fraud or abuse in one of seven categories of serious fraud
and abuse, such as exchanging food vouchers for cash.  This number
represents about .14 percent of the average monthly number of
participants at local agencies in fiscal year 1998.\16 For three
types of less serious fraud and abuse, such as redeeming food
vouchers outside of the authorized dates on the vouchers, our survey
found the number of participants identified as having committed such
offenses to be an estimated 79,271, which represents about 1.64
percent of the average monthly number of participants at local
agencies in fiscal year 1998.  Concerning employee fraud and abuse,
about 4 percent of local agencies suspected fraud and abuse, but an
estimated 48 employees nationwide were identified as having committed
fraud or abuse. 

--------------------
\11 WIC Income Verification Study, 1988, Final Report; WIC Vendor
Issues Study, 1991, Final Report. 

\12 In this report, we are presenting data reported by the 51
geographic state agencies (50 states and the District of Columbia)
separately from the Indian tribal organizations and U.S.  Territories
(ITO/Territories).  Seventeen of the 29 ITO/Territories that
responded to our survey reported that they maintain data on detected
participant fraud and abuse. 

\13 We are not reporting vendor data for ITO/Territories because our
survey did not separate out ITO/Territories that use other state
agencies' vendors. 

\14 As of Sept.  30, 1998, about 43,000 vendors were authorized to
participate in the WIC program in the 51 states, as reported by state
agencies. 

\15 See app.  I for confidence intervals of our estimates. 

\16 The estimates of .14 percent for serious offenses and 1.64
percent for less serious offenses are provided to put the estimated
number of participants identified as having committed fraud or abuse
into perspective.  These estimates use the number of participants
reported as having committed fraud or abuse during the 2-year period
(fiscal years 1997 and 1998) as a numerator and the average monthly
participation for fiscal year 1998 as a denominator.  Our estimates
would have been lower if we had collected information on the number
of participants reported as committing fraud or abuse in only one
fiscal year.  The confidence intervals for these two estimates are
reported in app.  I.  ITO/Territories reported less fraud and abuse
in both categories. 

   FNS DOES NOT HAVE CURRENT
   ESTIMATES OF PROGRAM FRAUD AND
   ABUSE AND COLLECTS DATA FROM
   WIC AGENCIES ON ONE TYPE OF
   DETECTED FRAUD AND ABUSE
---------------------------------------------------------- Chapter 2:1

FNS does not have systems in place to generate annual or biennial
estimates on the levels of vendor, participant or employee fraud and
abuse; however, it does collect data from state agencies on vendor
fraud and abuse.  For its estimates of fraud and abuse occurring in
the program, FNS has relied on information provided by two studies, a
1991 vendor study and a 1988 participant study.  FNS is currently
updating these studies. 

According to the 1991 vendor study, overcharging by an estimated 22
percent of vendors nationwide resulted in overcharges that amounted
to an estimated 1.9 percent of the total dollar value of WIC food
vouchers redeemed in 1991.  This represented about $39.5 million of
an estimated $2.09 billion in retail redemptions that year.\17 This
study also found that .6 percent of the vendors undercharged for a
1991 redemption value of $11.9 million.  The participant study found
that an estimated 5.7 percent of WIC participants nationwide were not
eligible for the benefits they received on the basis of their income. 
Food benefits provided to these income-ineligible participants were
estimated to represent 5.8 percent or $84 million out of the total of
$1.5 billion in food benefits provided in 1988.  While these studies
provide valuable information about vendor and participant fraud and
abuse in the program, they did not determine the extent to which
vendor overcharges and participation by income-ineligible individuals
were due to fraud and abuse or to inadvertent errors.  Moreover,
neither study provided complete estimates of all the fraud and abuse
occurring in their respective areas.  For example, the vendor study
did not address trafficking, and the participant study did not
address areas of ineligibility other than income. 

For vendors, FNS is conducting a study designed to, among other
things, estimate the levels of vendor fraud and abuse.  As in the
1991 study, the estimates will be based on investigations, in this
case involving more than 4,500 compliance buysï¿½when undercover agents
attempt to exchange WIC vouchers for nonapproved or nonfood
itemsï¿½from a statistically representative sample of approximately
1,500 vendors throughout the contiguous United States.  Like the
previous study, this new effort will provide FNS with valuable
information about fraud and abuse in the program, but will not
provide a complete picture of fraud and abuse because it excludes,
for example, information on authorized vendors who redeem vouchers
obtained from unauthorized vendors.  FNS expects to issue the study
late in 2000. 

FNS is also updating its 1988 participant study to estimate, among
other things, the number of ineligible participants.  This study is
designed to (1) develop a national estimate of the number of
participants who are ineligible to receive WIC benefits because of
their income and (2) identify the cost of the food benefits provided
to the individuals who are not income-eligible for the program.  Like
the previous study, this new effort will not provide a complete
picture of fraud and abuse because it excludes, for example,
information on the misuse of WIC benefits or those who receive
multiple benefits.  In addition, this new effort will not determine
whether benefits issued incorrectly to individual participants
occurred as a result of fraud and abuse or inadvertent error.  FNS
anticipates that this study will be issued late in 1999. 

Regarding detected fraud and abuse, FNS recognizes the value of
maintaining some information on vendor fraud and abuse in order to
monitor states' detection efforts and, therefore, collects such
information from the state agencies.  In the past, FNS collected only
aggregate information on vendor fraud and abuse from the states.  FNS
has recently begun to create a new database of information on
individual WIC vendors from state agency data, including information
on detected fraud and abuse.  This database should enable FNS to
develop information on the number and characteristics of individual
vendors identified by state and local agencies involved in fraud and
abuse.  FNS plans to issue its first report with information
contained in the new database in the fall of 1999.  FNS does not
collect information on incidences of detected participant and
employee fraud and abuse from state agencies. 

--------------------
\17 This amount was annualized on the basis of redemptions in June
and July of 1991. 

   AGENCIES DETECTED VARYING
   LEVELS OF VENDOR, PARTICIPANT,
   AND EMPLOYEE FRAUD AND ABUSE
---------------------------------------------------------- Chapter 2:2

Nationwide, the relative level of detected vendor fraud and abuse was
substantially higher than the level of detected participant or
employee fraud and abuse, according to our analysis of state and
local WIC agencies' responses to our surveys.  Regarding vendor fraud
and abuse, the amount reported by the states varied considerably. 
For participants, local agencies also identified varying levels of
fraud and abuse.  We relied on local agency data to identify
participant fraud and abuse because 21 of the 51 state agencies
reported that they do not maintain data on participant fraud and
abuse.  Regarding employee fraud and abuse, local agencies identified
few confirmed incidences.  Information provided by state and local
agencies on the amount of detected vendor, participant, and employee
fraud and abuse does not present a complete picture of fraud and
abuse in the program, partly because the level of detection efforts
differ among both state and local agencies.  In addition, regardless
of detection efforts, some violations go undetected. 

      DETECTED LEVELS OF VENDOR
      FRAUD AND ABUSE
-------------------------------------------------------- Chapter 2:2.1

Nationwide, state WIC directors identified 3,771 vendors as having
committed fraud or abuse during the 2-year period we surveyedï¿½fiscal
year 1997 through fiscal year 1998.  This number represents about 9
percent of all vendors participating in the program as of September
30, 1998.  The variation in the number of vendors detected among
states was substantial, ranging from 15 states reporting no vendors
detected to a high of six states reporting 25 percent or more vendors
detected as having committed fraud and abuse.  Moreover, the District
of Columbia, with 31 vendors, reported that 23 vendors committed
fraud or abuse.  Figure 2.1 shows the variation by state in the
number of vendors reported as having committed fraud or abuse during
the 2-year period, expressed as a percentage of vendors participating
in the program as of September 30, 1998. 

   Figure 2.1:  Levels of Detected
   Vendor Fraud and Abuse by
   State, October 1996 Through
   September 1998

   (See figure in printed
   edition.)

Note:  For the 15 states reporting no vendor fraud or abuse detected,
eight statesï¿½Alaska, Iowa, Idaho, Maine, North Dakota, Nebraska, New
Mexico, and South Carolinaï¿½did not suspect or become aware of any
vendor fraud or abuse, five statesï¿½Hawaii, Missouri, Montana, New
Hampshire, and Rhode Islandï¿½suspected vendor fraud or abuse but did
not confirm any incidences, Georgia suspected vendor fraud or abuse
but was unable to confirm the fraud or abuse at the time of our
survey, and Mississippi does not use the vendor distribution system. 

Source:  GAO's analysis of survey data from state agencies. 

The types of vendor fraud and abuse can vary dramatically, ranging
from relatively minor violations such as allowing product
substitutionsï¿½for example, large eggs for medium eggs or 2-percent
milk for whole milkï¿½to major violations, such as exchanging vouchers
for cash.  Therefore, comparing the level of detected vendor fraud
and abuse among states may not accurately reflect the nature of the
violation or the extent to which program integrity is being impaired. 
Some serious vendor fraud and abuse violations reported by agency
officials included the following: 

  -- At a Los Angeles convenience store, undercover investigators
     purchased nonfood items, including diapers and cigarettes, using
     WIC vouchers.  The vendor also purchased food vouchers for cash
     from the investigators. 

  -- During compliance buys in Chicago, vendors paid investigators as
     little as 10 cents on the dollar for WIC vouchers, while other
     investigators were allowed to buy prohibited items, including
     alcoholic beverages, with their vouchers. 

  -- In another Chicago case, undercover agents found that a family,
     through their retail stores, bought WIC vouchers from
     participants for cash and set up false corporations to redeem
     the vouchers.  The redeemed value of the WIC vouchers involved
     totaled more than $580,000 over a 7-month period. 

  -- In Texas, employees of a large grocery chain were identified as
     altering already-redeemed food vouchers by increasing their
     redeemed value.  The resulting overcharges to the program
     totaled over $6,800.  State employees uncovered this situation
     when they were entering data from the vouchers into the computer
     system for redemption and reconciliation. 

      DETECTED LEVELS OF
      PARTICIPANT FRAUD AND ABUSE
-------------------------------------------------------- Chapter 2:2.2

According to our survey of local agencies, the estimated number of
participants detected as having committed serious types of fraud and
abuse was substantially lower than the number identified as having
committed less serious types of fraud and abuse.  In addition, an
estimated 58 percent of the local agencies detected no incidences of
serious fraud or abuse and an estimated 28 percent detected no
incidences of less serious fraud or abuse.  An estimated 22 percent
of local agencies detected no fraud or abuse in any category and 4
percent did not respond to the question.  We relied on our survey of
local agencies for information on participant fraud and abuse because
21 of the 51 state WIC agencies reported that they did not maintain
data on participant fraud and abuse.  In analyzing the participant
fraud and abuse identified by local agencies, we found variation
among local agencies in the amount detected.  Local agencies in the
21 states that reported not maintaining data on participant fraud and
abuse were more likely to identify no fraud or abuse in the serious
categories than in the 30 states that reported maintaining such data. 

The number of participants identified by local agencies as having
committed fraud or abuse in one or more of seven serious categories,
such as misrepresenting their income, during the 2-year period
covered by our survey was an estimated 7,074.  This number represents
about .14 percent of the average monthly number of participants in
fiscal year 1998. 

Working from a list of seven serious types of fraud and abuse
categories contained in our local agency survey, participants
receiving multiple benefitsï¿½obtaining two sets of food vouchers for
the same time periodï¿½was the most frequently identified offense. 
Figure 2.2 shows the number of participants identified as having
committed serious fraud and abuse, by category, during the 2-year
period. 

   Figure 2.2:  Estimated Number
   of Participants Having
   Committed Various Types of
   Serious Fraud and Abuse,
   October 1996 Through September
   1998

   (See figure in printed
   edition.)

Note:  Some participants were identified as having committed more
than one type of serious fraud or abuse.  The levels of detected
fraud for the multiple benefits category and the misrepresented
income categories were significantly higher than those of the three
least frequent categories.  All other differences among categories
were not statistically significant.  See app.  I for confidence
intervals for these estimates. 

Source:  GAO's analysis of survey data from the local agencies. 

During our site visits, agency officials further explained these
types of serious fraud and abuse and provided the following examples: 

Multiple benefits.  Participants who receive multiple benefits obtain
two sets of vouchers for the same time period.  For example, one
local agency official in Texas told us that a participant who was
enrolled in and receiving benefits from two clinics within the state
was discovered by a WIC employee when the participant presented her
identification card issued at one clinic to officials at the other
clinic.  The participant was subsequently counseled about program
rules and removed from the program at one local agency but continued
to receive benefits at the other agency. 

Exchanging vouchers for nonapproved or nonfood items.  Participants
commit fraud when they purchase nonapproved or nonfood items with
vouchers.  For example, a participant attempted to exchange vouchers
for alcohol, according to an agency director in Maryland.  When the
director became aware of this attempt, she sent the participant a
warning letter. 

Misrepresentation of income.  Participants commit fraud by
understating their income so that it falls within WIC guidelines,
thus meeting the WIC income eligibility requirement.  This happens,
for example, when an applicant reports only her income and does not
report the income of her spouse or other adults considered part of
the household unit. 

Misrepresentation of eligibility status.  Participants commit fraud
by misrepresenting factors affecting their eligibility status, such
as their residency, pregnancy status, or exact number of qualifying
dependents. 

Exchanging vouchers for cash.  Participants commit fraud by
exchanging WIC vouchers for cash.  For example, according to an
agency director in New York State, a participant completed the
voucher by filling in an amount greater than the amount of the
purchases and asked for and received the difference in cash. 

Giving away food vouchers or WIC food.  In some instances,
participants have committed program fraud by giving away food
vouchers or the WIC food.  For example, a local agency director in
California told of a participant who exchanged her WIC vouchers for
car rides. 

Claiming fictitious dependents.  In a few instances, participants
have been known to create fictitious dependents in order to be
eligible for the WIC program or to obtain more WIC benefits than they
would otherwise be entitled to.  In one instance, a couple in the Los
Angeles area created 60 sets of twins in order to obtain infant
formula, the highest-priced WIC food item.  The Goodling Act
addresses this by requiring the physical presence of each individual
seeking certification for participation in the program. 

The amount of fraud and abuse identified by local agencies varied. 
An estimated 58 percent of local agencies identified no participants
committing serious fraud and abuse, while for an estimated 4 percent
of the agencies, the number of participants identified represented
over 1 percent of their average monthly number of participants in
fiscal year 1998. 

According to our analysis of participant data, local agencies in the
21 state agencies that reported not maintaining data were more likely
to detect no serious participant fraud and abuse.  Figure 2.3 shows
the percentages of local agencies that identified no serious
incidences of fraud and abuse in the 21 states that reported that
they do not maintain data and in the 30 states that reported they do
maintain data. 

   Figure 2.3:  Percent of Local
   Agencies Identifying No Serious
   Participant Fraud and Abuse in
   the States That Do Not and Do
   Maintain Data on Participant
   Fraud and Abuse, October 1996
   Through September 1998

   (See figure in printed
   edition.)

Source:  GAO's analysis of survey data from state and local agencies. 

We also asked local agencies to identify less serious types of
participant fraud and abuse.  The number of participants identified
as having committed fraud or abuse in one or more of three less
serious categories was an estimated 79,271.  This number represents
about 1.64 percent of the average monthly number of participants in
fiscal year 1998.  The type of less serious offenses most commonly
identified by local agencies was redeeming food vouchers outside the
authorized dates.  Figure 2.4 shows the number of participants
identified as having committed each type of less serious fraud and
abuse listed in our survey in the 2-year period. 

   Figure 2.4:  Estimated Number
   of Participants Having
   Committed Less Serious Fraud
   and Abuse by Type, October 1996
   Through September 1998

   (See figure in printed
   edition.)

Note:  Some participants were identified as having committed more
than one type of less serious fraud or abuse.  The level of detected
abuse for redeeming vouchers outside the authorized dates was
significantly higher than for the other two categories, which were
not significantly different from one another.  See app.  I for
confidence intervals for these estimates. 

Source:  GAO's analysis of survey data from local agencies. 

During our site visits, agency officials further explained these
types of less serious participant fraud and abuse and provided the
following examples: 

Redeemed vouchers outside authorized dates.  This type of abuse
involves participants who use vouchers either before or after the
valid ï¿½use byï¿½ dates printed on the vouchers.  Vouchers are generally
dated for a 30-day period and valid use requires redemption within
that period.  At a local agency in California, the director told us
of a case in which a participant altered the expiration date on an
expired voucher so that she could purchase items after the voucher
had expired.  At another agency in Kansas, a director reported 25
incidences of participants who cashed their vouchers outside of the
authorized dates. 

Selected incorrect brands or incorrect quantities.  There are a
variety of ways in which this abuse can occur.  For example, a
participant may choose a larger size of juice than permitted, such as
a 64-ounce bottle instead of a 46-ounce bottle. 

Committed verbal abuse.  Agency directors also reported that some
participants were verbally abusive to either WIC vendors or local
agency staffï¿½a program abuse. 

A larger percentage of local agencies identified incidences of fraud
and abuse in the less serious category than in the serious
categoryï¿½67 percent and 40 percent respectively.  An estimated 74
percent of the local agencies identified some type of participant
fraud or abuse.  As with the serious category, there was variation
among agencies in the amounts of less serious fraud and abuse
identified.  An estimated 28 percent of local agencies identified no
participants as having committed less serious fraud and abuse, while
for an estimated 10 percent of the agencies, the number of
participants identified represented more than 5 percent of their
average monthly number of participants in fiscal year 1998. 

Agency officials also identified about 2,902 participants who
committed other types of fraud and abuse that were not specifically
mentioned in the lists provided in our survey.  Some of these types
could be considered serious fraud and abuse, while others could be
considered less serious.  One of the serious types specified by some
local agencies we visited was selling or returning infant formula for
cash. 

There are relatively few documented cases nationwide of selling or
returning infant formula to stores for cash, but local WIC directors
told us that this activity was of concern because it suggests a
serious health risk to infants and because the high cost of formula
can make the activity attractive to some participants.  A Texas
director told us of a participant who advertised a case of infant
formula for sale in a local newspaper.  After the director contacted
the participant to advise her that selling WIC formula was against
program rules, the participant returned the unneeded formula to the
clinic.  According to FNS officials, a one-month supply of WIC
formula can have a retail value of almost $100. 

      DETECTED LEVEL OF EMPLOYEE
      FRAUD AND ABUSE
-------------------------------------------------------- Chapter 2:2.3

The number of employees identified by local agencies as suspected of
having committed and actually having committed fraud or abuse during
the 2-year period our survey covered were very small, about 78 and 48
respectively.  The low number of actual cases of identified employee
fraud or abuse can be explained, in part, because on some occasions
when an employee is suspected of fraud, he or she quits, and agency
staff are unable to document the incidence.  We did not collect
survey information on the types of employee fraud and abuse
committed.  However, in our discussions with agency officials, the
lack of separation of duties was cited as a problem.  For example, at
one local agency in Illinois, a clinic supervisor had access to the
computer system and participant records as well as the printer system
to issue or print vouchers.  The supervisor created hundreds of
fictitious participant records and printed food vouchers for each
participant on a day when she was not scheduled to work.  She then
sold these food vouchers to local vendors for a portion of their
stated value. 

AGENCIES VARY IN THEIR EFFORTS TO
PREVENT AND DETECT FRAUD AND ABUSE
AND ENCOUNTER SOME BARRIERS
============================================================ Chapter 3

State and local WIC agencies vary in their efforts to control vendor,
participant, and employee fraud and abuse, as the flexibility
provided by WIC regulations allows.  Regarding vendors, states differ
in their management procedures, including the methods they use to
limit the number of vendors authorized to participate in the WIC
program and the amount and type of monitoring they perform.  With
respect to participants, state and local WIC agencies vary in their
efforts to ensure that only eligible individuals receive benefits and
that the benefits are used as intended.  For employees, WIC agencies
differ in the extent of conflict-of-interest controls they have in
place to prevent and detect employee fraud and abuse. 

While federal, state, and local WIC agencies have undertaken a number
of activities to prevent and detect the different types of fraud and
abuse, several barriers that can inhibit their success were
identified.  These barriers are a lack of (1) criteria to assess
whether the states have authorized an appropriate number of vendors;
(2) information on the number and characteristics of participants who
have committed fraud; (3) policies or procedures regarding potential
employee conflict-of-interest situations; and (4) resources to devote
to fraud control efforts. 

   STATES DIFFER IN EFFORTS TO
   MANAGE VENDORS
---------------------------------------------------------- Chapter 3:1

The management of vendorsï¿½the authorization, types and extent of
monitoring, identification of high-risk vendors, and sanctions
imposed on violatorsï¿½varied among the states.  As allowed by federal
regulations, state agencies can choose the methods they use to ensure
that they are effectively managing WIC vendors.  The practices they
choose to use and their levels of effort vary considerably.  Recent
legislation imposes new vendor monitoring requirements in some of
these areas and could produce greater uniformity.  In addition,
recent changes in FNS regulations should increase uniformity in
sanctioning vendors by mandating maximum uniform sanctions for the
most serious offenses and increasing the maximum time for
disqualifying vendors from the program. 

      STATES VARY IN METHODS USED
      TO LIMIT THE NUMBER OF
      AUTHORIZED VENDORS
-------------------------------------------------------- Chapter 3:1.1

FNS regulations require that states relying on the retail
distribution of WIC food benefits authorize an appropriate number of
vendors that can be effectively managed while ensuring adequate
access for participants.  Of the 51 state WIC agencies, 42 reported
making some effort to limit or contain the number of authorized
vendors.\18 Eight states reported that they do not limit the number
of vendors, and one state, Mississippi, distributes WIC products
through state-operated centers.  Table 3.1 shows the various methods
states used to limit the number of vendors. 

                               Table 3.1
                
                Methods States Used to Limit the Number
                               of Vendors

                                                      Number of states
Method used                                               using method
--------------------------------------------------  ------------------
Vendor food prices cannot exceed limits set by the                  26
 state\a
Ratio of the number of participants to vendors\b                    12
Competitive bidding for vendor slots\c                               5
Absolute or fixed number of vendors                                  4
Other vendor ratios                                                  3
Other methods\d                                                     19
----------------------------------------------------------------------
Notes:  Some states reported using more than one method. 

\a State sets maximum allowable food prices that vendors can charge
the WIC program. 

\b The number of WIC participants per vendor in an area predetermined
by a state agency. 

\c States determine the number and appropriate distribution
(location) of WIC vendors.  Vendors compete for the authorized slots. 

\d Other methods reported include having vendors (1) meet minimum
stock requirements, (2) keep prices charged the program within 10
percent of the lowest priced store in the community, (3) not sell gas
or alcohol, (4) be conveniently located, and (5) maintain their
stores in good condition.  Requiring a demonstrated need for a store
in an area was also reported. 

Source:  GAO's analysis of survey data from state agencies. 

States' efforts to limit the number of vendors can vary, depending in
part on the method used and, in some cases, the strictness of the
selection criteria imposed within the method.  For example, when
competitive bidding is used, a predetermined number of vendor slots
are available, and vendors submit their prices for WIC foods to bid
on those slotsï¿½the lowest bidders become the authorized vendors.  By
limiting vendor food prices to an amount pre-set by the state agency,
the strictness of the selection criteria may limit the number of
vendors who participate.  For example, if the state restricts the
food prices to an amount that is below the general market rate, fewer
vendors may be willing to participate than if it restricts the food
prices to an amount nearer the general market rate. 

The methods a state chooses to deliver foods can also account for
variation in the number of vendors authorized.  For example, Vermont
and parts of Ohio use dairies to deliver WIC foods directly to
participants' homes and have few authorized vendors.  In Chicago,
many small vendors selling primarily alcohol and tobacco have been
replaced by 15 food centers operated through a partnership between
the state WIC agency and Catholic Charities of the Archdiocese of
Chicago.  These food centers distribute WIC products in exchange for
special WIC food vouchers that have no monetary value and that cannot
be used at other WIC vendors.  Other factors can affect states'
efforts to limit the number of vendors as well, such as population
densityï¿½lower density areas may require fewer participants per
vendor. 

These factors, as well as the methods used, can result in
considerable variation in the number of participants per authorized
WIC vendor.  Our survey found the ratio of participants to vendors
varied among the states.  For example, North Dakota reported 57
participants per vendor, the District of Columbia reported 535
participants per vendor, and Vermont reported 709 participants per
vendor. 

--------------------
\18 In order to avoid double-counting vendors, we are not reporting
vendor data for ITO/Territories because our survey did not separate
those ITO/Territories that use the vendors of other state agencies. 

      STATES VARY IN VENDOR
      MONITORING BUT MEET THE
      10-PERCENT REQUIREMENT
-------------------------------------------------------- Chapter 3:1.2

To manage and control vendors, FNS requires the states to conduct
on-site monitoring of a sample of at least 10 percent of their
vendors each year.  These routine monitoring visits may include such
activities as reviewing cashier checkout procedures, examining the
availability and prices of supplemental program foods, and insuring
that WIC products are clearly identified, as well as providing vendor
education and training.  The states use a variety of types of vendor
visits to satisfy FNS' required 10-percent sample of vendors
monitored.  For example, one state uses the initial visit to a retail
storeï¿½made to approve the store as a WIC vendor.  Another state
counts technical assistance visits towards satisfying this
requirement.  When further investigation is indicated, monitoring
visits may also include compliance buys.  According to our analyses
of fiscal year 1998 monitoring activity reported by the states, all
the states conducted on-site monitoring for at least 10 percent of
their vendors.  On a nationwide basis, the states reported that at
least 40 percent of all vendors were subjected to monitoring visits
and about 12 percent were subjected to compliance buys during that
year.  Some vendors were also subject to trafficking buys. 

The type of vendor monitoring activity varies by state.  Some states
reported that most of their monitoring consisted of routine visits
while others noted that most monitoring consisted of compliance buys. 
The percentage of vendors subjected to compliance buys in fiscal year
1998 varied substantially among the states.  For example, while eight
states reported that none of their vendors were subjected to
compliance buys, New York State reported that about 40 percent of its
vendors were subjected to them, and the District of Columbia reported
that 94 percent were subjected to them.  The differences we
identified may not mean that the states conducting more intensive
monitoring have more problematic vendors; it could mean that these
states devote more resources to this type of investigation. 

The states we visited varied in the number of compliance buys
conducted and in the way they were conducted.  In 1998, for example,
Florida reported relying on volunteers from the local agencies' staff
to conduct compliance buys at 15 of its about 2,000 authorized
vendors.  Florida put a moratorium on compliance buys early in 1998,
thereby limiting the number of vendors subject to such buys, because
of concerns about the safety of the volunteers.  In Illinois, which
contracts with a private investigative firm to conduct compliance
buys, 76 of almost 1,800 vendors were subject to such buys.  New York
used state agency employees to conduct compliance buys at about 1,700
of its over 4,200 vendors in 1998. 

      MOST STATES USE SYSTEMATIC
      METHODS FOR IDENTIFYING
      HIGH-RISK VENDORS
-------------------------------------------------------- Chapter 3:1.3

As part of a state's vendor monitoring, FNS regulations require the
state agency to design and implement a system to identify high-risk
vendors.  The regulations state that the criteria for identifying
high-risk vendors may include such considerations as the level and
severity of suspected vendor overcharges to the program and
participant complaints.  Forty-eight of the 50 states that use
vendors reported that they use certain types of indicators to
identify high-risk vendors, and 43 of these states reported using a
statistical method.\19 While our survey did not ask the states to
identify the high-risk criteria that they were using, we discussed
the criteria with agency officials interviewed during our site
visits.  The types of indicators they reported using ranged from
objective indicators, such as the volume of WIC foods sold, to
subjective indicators, such as association with another store known
to abuse the program.  Because of this variation, a vendor considered
high-risk in one state may not be identified as high-risk in another
state. 

Monitoring high-risk vendors is a prudent method for detecting fraud
and abuse.  At the time of our review, FNS regulations required that
the states investigate high-risk vendors as appropriate.  However,
they were silent on defining the appropriate number of identified
high-risk vendors that were to be monitored.  Thus, a state could
have been in compliance without monitoring any high-risk vendors if
it determined that no monitoring was appropriate.  Furthermore, had a
state determined that it was appropriate to monitor a high-risk
vendor, the regulations did not specify the type of monitoring
required.  The Goodling Act should reduce the flexibility in the
regulations on high-risk vendors by requiring the states to conduct
compliance-buy investigations of all vendors so identified.  On June
16, 1999, FNS issued proposed rules that would require state agencies
to conduct annual compliance buys or inventory audits on all
high-risk vendors, up to the 10 percent minimum of all authorized
vendors.  If less than 10 percent of the vendor population is
identified as high-risk, state agencies must make up the difference
by conducting compliance buys or inventory audits on vendors selected
at random.  This would replace the current requirement to monitor a
representative sample of 10 percent of all vendors. 

FNS' proposed rules, issued on June 16, 1999, are intended to make
vendor management more consistent across states.  The proposal
acknowledges that there is significant variation in agencies'
practices for managing vendors and that differences in state systems
have resulted in the inconsistent treatment of vendors across and
within state agencies.  It also acknowledges that this inconsistency
has resulted in unacceptable levels of vendor fraud and program
noncompliance.  If implemented, these rules would reduce some of the
variation we found among the states in terms of the number of vendors
authorized, vendor monitoring, and the identification of high-risk
vendors.  The proposed rules clearly emphasize the need to limit the
number of vendors to ensure effective program oversight and would
require state agencies to establish criteria to limit the number of
vendors authorized.  Although the proposal does not specify the
criteria that state agencies should use when limiting vendor numbers,
it does require that when developing such criteria, state agencies
consider the participant-to-vendor ratios.  The proposal does not
identify the criteria FNS would use to assess whether the states have
authorized more vendors than they can effectively manage. 

--------------------
\19 Hawaii and Idaho reported not using such indicators during the
period covered by our survey because they were in the process of
redesigning high-risk criteria.  Hawaii's redesign was part of its
installation of a statewide computer system, and Idaho's redesign was
part of an update of its computer system. 

      STATES IMPOSE A VARIETY OF
      SANCTIONS ON ABUSIVE VENDORS
-------------------------------------------------------- Chapter 3:1.4

FNS regulations provide the states with latitude in choosing the type
of sanctions imposed on vendors detected as committing fraud and
abuse.  The states reported imposing sanctions that ranged from
warning letters for less serious offenses to disqualification for the
most serious offenses.  More than half of the vendors committing
fraud and abuse during the 2-year period our survey covered received
a warning letter only.  Table 3.2 shows the percent of vendors who
violated program rules and the sanctions they received. 

                               Table 3.2
                
                Type of Sanctions the States Imposed and
                    Percentage of Identified Vendors
                          Receiving Sanctions

                                                    Percent of vendors
Type of sanction                                    receiving sanction
--------------------------------------------------  ------------------
Warning letter only                                                 58
Temporary disqualification or suspension from the                   25
 WIC program
Monetary fine or penalty                                            21
Sanction points\a                                                    9
Other                                                                4
No warning letter or other sanction                                  2
----------------------------------------------------------------------
Note:  The total does not add to 100 percent because some vendors
could have received more than one sanction during the 2-year period. 

\a Vendors are assessed sanction points for offenses such as allowing
the purchase of a similar but not WIC-approved food or accepting the
return of food purchased with WIC vouchers for cash or credit towards
another purchase.  Accumulation of a specified number of points can
adversely affect the renewal of the vendor's contract and, in extreme
situations, can subject the vendor to a 1-year disqualification. 

Source:  GAO's analysis of survey data from state agencies. 

The Goodling Act strengthened WIC sanctions by requiring that vendors
convicted of trafficking and other serious violations be permanently
disqualified.  FNS issued regulations in March 1999 that mandate
uniform sanctions across state agencies for the most serious vendor
violations and increase the maximum period for disqualification,
other than those permanently disqualified, to 6 years.  The
implementation of these mandatory sanctions is intended to curb
vendor fraud and abuse in the WIC program.  In instances where vendor
disqualification would impose undue hardship on participants, states
must assess a civil monetary penalty in lieu of disqualification. 
FNS' March 1999 regulations further specify that all civil monetary
penalties and fines must be used as program income by the state
agency. 

   AGENCIES VARY IN THE STRATEGIES
   USED TO PREVENT AND DETECT
   PARTICIPANT FRAUD AND ABUSE
---------------------------------------------------------- Chapter 3:2

State and local agencies vary in the strategies they use to ensure
that only eligible individuals receive WIC benefits and that they use
the benefits appropriately.  Table 3.3 shows the strategies included
in our survey and the number of state agencies and the estimated
percent of local agencies using each strategy.  Some strategies, such
as educating participants on program rules and regulations, were used
by many state and local agencies, while others, such as using
computer systems to verify eligibility on the basis of participation
in other income-eligible programs, were used by fewer state and local
agencies. 

                                        Table 3.3
                         
                             Strategies to Prevent and Detect
                           Participant Fraud and Abuse and the
                          Number of State Agencies and Estimated
                           Percent of Local Agencies Using the
                                        Strategies

                                                     Number of states   Estimated percent
Strategy                                                   (51 total)   of local agencies
-------------------------------------------------  ------------------  ------------------
Educate participants on program rules and fraud                    51                  95
 and abuse
Use a computer system to maintain participant                      50                  89
 data
Use a computer system to identify dual                             49                  73
 participation within state
Use a computer system to print food vouchers                       47                  85
Use complaint forms for vendors and/or others to                   44                  79
 report participant fraud or abuse
Share and receive fraud and abuse prevention                       36                  67
 information and strategies with other WIC state
 agencies, other government agencies, and/or
 interest groups
Provide fraud and abuse awareness training for                     36                  44
 WIC staff
Require documentation of income beyond self-                       35                  70
 declaration by applicant
Require documentation of applicant's residence                     26                  45
Maintain an 800 number hot-line or other                           23                  \a
 publicized phone number for reporting fraud or
 abuse
Perform more local agency evaluations than                         18                  \a
 required
Use a computer system to verify adjunct                            16                  50
 eligibility\b
Contact other states when participants move into                   12                  \a
 the state
-----------------------------------------------------------------------------------------
\a Local agencies were not asked to provide information on these
strategies. 

\b Eligibility based on applicant's participation in another,
income-qualifying program. 

Source:  GAO's analysis of survey data from state and local agencies. 

Every state agency and an estimated 95 percent of the local agencies
reported using the strategy of educating participants on program
rules and fraud and abuse.  However, they differed in the methods
used to educate participants.  During our visits, officials told us
that such methods include having the participant read and sign a
program statement and giving each new participant a pamphlet
explaining participant responsibilities. 

Regarding the strategy of requiring income documentation to establish
an applicant's income eligibility for WIC, 14 states and an estimated
29 percent of local agencies did not require applicants to provide
proof of income.  The Goodling Act mandates that all state and local
agencies require applicants to present documentation of income; FNS
implemented this requirement in February 1999.  Nevertheless, the
manner in which this requirement is implemented may continue to
differ.  Agencies that have historically required income
documentation have varied in the procedures they follow when
documentation is not provided at the time of application.  For
example, in Texas, agency officials told us that they require
evidence of income without exception.  If an applicant is homeless or
has no income, a statement to this effect is required from someone,
such as a member of the clergy or a social worker.  In contrast, in
Illinois, if the individual does not have documentation at the time
of initial application, the state will allow a 30-day waiver; thus an
applicant can be certified and receive benefits for the first month
and be required to bring in evidence for the second month of
participation.  Such variation is likely to continue. 

Differences in implementation occur in the use of other strategies as
well.  While almost all states use a computer system to determine
whether an applicant is already participating in the WIC program at
another location, thereby receiving multiple benefits for the same
time period, a practice known as dual participation, they varied in
how they used their systems.  For example, California's computer
system provides real-time information on participation at the time of
application, thereby immediately preventing duplicate enrollment.  In
contrast, in Texas, after an applicant is certified and issued
benefits, the system checks participant files, and, if there is
duplication, the local agency is notified the next day.  When dual
participation is detected, any corrective action or sanction would
typically occur the next time that the participant came in to pick up
benefits.  Texas officials explained that there were too few
instances of dual participation to incur the costs associated with
improving the timeliness of the information. 

In addition, we identified some states that check for dual
participation beyond their borders.  In our survey, 16 of the 51
state agencies reported that they have written agreements or working
arrangements with other state agencies to check for dual
participation between the states. 

We also identified differences in implementation when an applicant
presents documentation of participation in one of the means-tested
programsï¿½Food Stamps, Medicaid, or Temporary Assistance for Needy
Families.  By participating in one of these programs, the applicant
automatically meets the WIC income eligibility requirements. 
Nationwide, for the agencies that provided estimates, about 58
percent of their participants, on average, were income-eligible on
the basis of their participation in another means-tested program. 

While FNS regulations do not require WIC agencies to verify an
applicant's documentation of participation in another
income-qualifying program, our survey found that local agencies
verify about 31 percent of this information, on average.  In
verifying this information, local agencies take such steps as
checking computer systems or telephoning service providers to verify
program participation.  In some states, verification can also be done
by an automated check of Medicaid participation.  For example, in
California, agency staff use the statewide WIC computer system, which
accesses the state's automated Medicaid database, to verify an
applicant's status in Medicaid; in New York State, this verification
is accomplished by scanning the applicant's Medicaid card.  On the
other hand, Florida officials told us that the card scan system of
verification was used sparingly because each time the card is scanned
there is a 28-cent cost to the local agency.  As shown in table 3.4,
the percentage of such verification in fiscal year 1998 varied among
agencies, with an estimated 45 percent conducting no verification. 

                               Table 3.4
                
                  Estimated Percent of Local Agencies
                      Conducting Various Levels of
                      Verification of Applicants'
                     Participation in Other Income-
                          Qualifying Programs

                                                     Estimated percent
Level of verification                                of local agencies
--------------------------------------------------  ------------------
None verified                                                       45
Less than 10 percent verified                                        6
10 to 50 percent verified                                           13
More than 50 but less than 100 percent verified                     12
100 percent verified                                                14
Did not respond to question                                          9
----------------------------------------------------------------------
Note:  The total does not add to 100 percent due to rounding of
numbers. 

Source:  GAO's analysis of survey data from local agencies. 

Once individuals are certified as eligible and are participating in
the WIC program, local agencies can become aware of suspected fraud
from several sources, including vendors, other participants, or the
general public.  Agency officials told us that, generally, they first
discuss the suspected situation with the participant.  Depending upon
the participant's response or the severity of the situation, other
methods may be used.  In responding to a survey question about how
they handle these suspected cases, an estimated 30 percent of local
agencies said they deal with the situations themselves without
contacting the state agency, while other local agencies refer some
situations to the state agency or handle the situations jointly. 
Some of this variation may be explained by the fact that some states,
such as California and New York, have a separate unit for dealing
with suspected participant fraud and abuse. 

WIC agencies can impose various sanctions when a participant is
identified as having committed fraud or abuse.  Of the approximately
7,074 participants identified as committing serious fraud and abuse
in the 2-year period we reviewed, about 1,787 received suspensions or
temporary disqualification from the program.  Serious types of fraud
and abuse include dual participation, exchanging vouchers for
nonapproved or nonfood items, or misrepresenting income.  The
estimated 79,271 participants committing less serious violations were
most frequently given an oral warning or counseling about program
rules.  The less serious types of fraud include redeeming vouchers
outside of the authorized dates and selecting incorrect brands or
quantities.  Table 3.5 shows the estimated number of participants
committing serious and less serious violations and the sanctions they
received. 

                               Table 3.5
                
                    Estimated Number of Participants
                  Committing Serious and Less Serious
                   Violations and the Sanctions They
                                Received

                                  Estimated number    Estimated number
                                   of participants     of participants
                                committing serious     committing less
Sanction                                violations  serious violations
------------------------------  ------------------  ------------------
Oral warning or counseling                   5,121              57,547
 about program rules
Written letter of warning                    2,282              19,379
Suspension or temporary                      1,787                 336
 disqualification for 1 to 3
 months
Other                                          428                 694
No sanction                                     33                  \a
----------------------------------------------------------------------
Note:  The total number of sanctions administered exceeds the total
number of participants who committed fraud and abuse, in part because
participants may receive more than one sanction. 

\a We were unable to develop a reliable estimate. 

Source:  GAO's analysis of survey data from local agencies. 

   AGENCIES USE SEVERAL STRATEGIES
   TO CONTROL EMPLOYEE FRAUD AND
   ABUSE
---------------------------------------------------------- Chapter 3:3

State and local WIC agencies employ over 18,000 individuals to manage
and deliver WIC benefits.  FNS regulations that aid in preventing and
detecting employee fraud require the security and accountability of
vouchers, including providing physical security during transporting,
receiving, and issuing them.  State agencies are also responsible for
reconciling the vouchers and ensuring that there is no
conflict-of-interest between local agencies and vendors.  In addition
to these activities, agency officials use a variety of strategies to
prevent and detect employee fraud and abuse, according to their
responses to a list we provided in our survey.  Table 3.6 shows the
strategies agencies use and the percent of local agencies using them. 

                               Table 3.6
                
                 Strategies Used to Prevent and Detect
                    Employee Fraud and Abuse and the
                  Estimated Percent of Local Agencies
                          Using the Strategies

                                                     Estimated percent
                                                     of local agencies
Strategy                                                using strategy
--------------------------------------------------  ------------------
Take measures to protect food vouchers and/or                       96
 blank food voucher stock from theft or fraudulent
 alteration
Supervisor's review of employee activities at                       93
 local agencies/clinics
Require that local agency/clinic staff report                       69
 suspected employee fraud or abuse to the state
Separate duties within the certification process                    68
 so that the employee certifying does not issue
 food vouchers
Share and receive information about strategies                      60
 with other agencies, such as other local or state
 WIC agencies, other government agencies, and/or
 interest groups
Have a policy for conflict-of-interest for WIC                      52
 participants who are also employed by the WIC
 program
Make unannounced visits to clinics\a                                47
Provide fraud and abuse awareness training for                      45
 local agency/clinic staff
Use a statistical method to detect employee fraud                    7
----------------------------------------------------------------------
\a Unannounced visits by supervisor or other agency management. 

Source:  GAO's analysis of survey data from local agencies. 

As shown in the table, almost all local agencies had policies in
place to protect food vouchers.  In addition, at some agencies we
visited the agency's management reviews included physical security
and program integrity.\20 Most local agencies also used supervisor's
review of employees' activities.  However, other policies that would
enhance prevention and detection are not as widely used.  Two of
these policies relate to conflict-of-interestï¿½when employees who
participate in the WIC program might certify their own eligibility
and issue their own benefits, and when there is no separation of
duties so that an employee can certify and issue benefits to the same
individual.  For example, an estimated 45 percent of the local
agencies do not have conflict-of-interest policies for employees who
also receive WIC benefits, which could result in a situation in which
employees certify themselves as eligible to participate and issue
their own WIC benefits.  Furthermore, an estimated 30 percent of the
local agencies do not separate duties within the certification
process.  In this case, employees could certify and issue WIC
benefits to relatives and friends.  While these situations in and of
themselves may not compromise program integrity, they present the
appearance of impropriety.  WIC agency officials told us that duties
are not always separated because, for example, some clinics have too
few employees to do so.  However, even in these situations, prudent
precautions can be taken.  For instance, one agency uses a separate
agency number for issuing WIC benefits to employee participants and
another agency requires a supervisor's signature when, because
staffing levels are low, an employee is going to both certify and
issue benefits to the same individual.  FNS regulations do not
require WIC agencies to have policies on conflict-of-interest for
employee-participants or for separation of duties. 

In addition to the strategies we identified in our survey, we asked
WIC agency officials about other activities they were using.  One
activity reported by FNS' Southwest Region was a federal/state
project in which WIC officials worked together to design a guide for
preventing and detecting employee abuse.  This guide outlines a
series of suggested actions aimed at the prevention and detection of
employee abuse and has been adopted by several states in the region. 
For example, it sets a minimum expectation that agencies establish a
policy relating to conflict-of-interest.  It also outlines practices
related to separation of duties, including practices to be followed
in clinics with few employees. 

--------------------
\20 Local agencies not taking measures to protect vouchers from theft
would include those in Mississippi, where the voucher system is not
used, and those in Vermont, where voucher stock is housed at a
separate state agency and not accessible to local WIC/Health
Department staff. 

   AGENCIES FACE BARRIERS TO
   PREVENTING AND DETECTING FRAUD
   AND ABUSE
---------------------------------------------------------- Chapter 3:4

Agency officials identified several barriers that can inhibit their
efforts to prevent and detect fraud and abuse.  These barriers are a
lack of (1) federal criteria to evaluate whether a state has
authorized an appropriate number of vendors, (2) information on
detected participant fraud and abuse, (3) a policy or procedures on
potential employee conflict-of-interest situations, and, (4)
insufficient resources to devote to fraud detection and control
efforts. 

      REGULATIONS LACK SPECIFIC
      CRITERIA TO ASSESS WHETHER
      STATES HAVE AUTHORIZED AN
      APPROPRIATE NUMBER OF
      VENDORS
-------------------------------------------------------- Chapter 3:4.1

FNS regulations require that the states authorize an appropriate
number and distribution of vendors to ensure adequate participant
access and effective review of its authorized vendors.  However, FNS
regulations currently lack criteria for evaluating whether states
have authorized an appropriate number of vendors.  The regulations
also do not provide the states with clear and specific criteria to
use in determining whether the number of vendors is at a manageable
level and appropriate for the resources states have available for
vendor management.  In vendor regulations proposed in 1990, FNS
recognized the need to establish criteria to ensure that the states
effectively and efficiently use program resources.  In addition, in
its proposed regulations, FNS recognized that if a state authorized
more vendors than necessary its resources might not be sufficient to
ensure the effective oversight of vendors, thus increasing the
opportunities for fraud and abuse or forcing curtailment of other
critical state and local agency activities.  These regulations were
never made final.  FNS' recently proposed rules reemphasize the need
to limit the number of vendors, and while calling for the states to
establish criteria to use in limiting the number of vendors
authorized, the proposal does not specify any criteria.  According to
some state officials, the lack of criteria for determining what
constitutes an appropriate number of vendors has hindered efforts to
place limits on vendors in the face of political pressures arising,
in part from the retail community. 

Monitoring is a critical aspect of effective vendor management and is
resource-intensive because of the numerous tasks that must be
performed.  These tasks include making on-site visits to vendors,
analyzing the data obtained as well as information on vendors'
redemptions, and determining the need for and conducting more
intensive compliance-buy investigations.  In our survey, state agency
officials were asked how easy or difficult it is for them to monitor
their vendors.  Regarding the required routine on-site monitoring of
at least 10 percent of authorized vendors each year, four state
agencies responded that it was somewhat difficult to conduct this
monitoring.  However, with respect to conducting the more
resource-intensive compliance buys, 22 state agencies responded that
it was somewhat or very difficult.  Moreover, 24 state agencies
reported that it was somewhat or very difficult to design and
implement a system to identify high-risk vendorsï¿½which may require
compliance-buy investigations.  State agencies will face further
challenges in allocating resources to manage vendors under the
Goodling Act's mandate that compliance buys be conducted at all
high-risk vendors. 

      INFORMATION IS LACKING ON
      PARTICIPANT FRAUD
-------------------------------------------------------- Chapter 3:4.2

FNS does not collect information on the number and characteristics of
participants who engage in fraud and abuse.  Regarding state
agencies, 21 of the 51 states reported that they do not maintain
these data.  Lacking such information, FNS is not able to assess the
extent of participant fraud and abuse, evaluate state and local
agencies' efforts to control it, or identify the changes needed to
improve program integrity. 

A number of agency officials reported that they lack adequate
information on how to identify situations that are at high-risk for
participant fraud and abuse.  Such information could be useful to WIC
agencies in identifying high-risk situations and in improving their
knowledge of fraud detection strategies.  In our survey, 20 of the 51
state agencies and an estimated 43 percent of local agency directors
reported that information on identifying these situations was less
than adequate.  During a site visit, one agency director told us that
information on identifying high-risk situations would help her and
her staff know what to watch for and what to do when a potential
fraud situation arises.  Agency officials also provided information
on their overall knowledge of strategies for detecting participant
fraud, with officials in 21 of the 51 state agencies and officials in
an estimated 47 percent of local agencies identifying a
less-than-adequate level of overall knowledge. 

      REGULATIONS DO NOT ADDRESS
      POTENTIAL CONFLICTS OF
      INTEREST FOR EMPLOYEES
-------------------------------------------------------- Chapter 3:4.3

Agency officials cited a lack of a policy or procedures for potential
conflicts-of-interest situations for employees that may hinder the
prevention and detection of employee fraud and abuse.  Regarding
employee-participants, 12 of the 51 state agencies surveyed reported
that they do not have a conflict-of-interest policy for
employee-participants to ensure that employee-participants do not
certify or issue benefits to themselves; an estimated 45 percent of
the local agencies lacked such a policy.  Regarding separation of
duties, eight of the state agencies surveyed reported that they do
not separate duties within the certification process so that the
employee certifying eligibility does not issue vouchers; an estimated
30 percent of the local agencies similarly do not separate duties. 

      INADEQUATE RESOURCES
-------------------------------------------------------- Chapter 3:4.4

Many state and some local agency officials identified staffing and
funding resources as hindering additional efforts to prevent and
detect fraud and abuse in the program.  For example, as shown in
table 3.7, inadequate staffing and funding resources were identified
by almost half of the states and about one-quarter of local agencies
as adversely affecting their efforts. 

                               Table 3.7
                
                  State and Local Agencies Identifying
                Less Than Adequate Staffing and Funding
                  Levels for Preventing and Detecting
                Vendor, Participant, and Employee Fraud
                               and Abuse

                                                     Estimated percent
                                Number of 51 state   of local agencies
                                agencies reporting         identifying
                                        inadequate          inadequate
                                         resources           resources
------------------------------  ------------------  ------------------
Resources for vendor
----------------------------------------------------------------------
Number of staff                                 22                  \a
Funding                                         24                  \a

Resources for participant fraud and abuse prevention and detection
----------------------------------------------------------------------
Number of staff                                 24                  27
Funding                                         25                  38

Resources for employee fraud and abuse prevention and detection
----------------------------------------------------------------------
Number of staff                                 24                  16
Funding                                         22                  27
----------------------------------------------------------------------
\a Local agencies were not asked to identify resources to prevent or
detect vendor fraud and abuse. 

Source:  GAO's analysis of survey data from state and local agencies. 

During our site visits, some local agency officials told us that
workload demandsï¿½certifying applicants; delivering client services,
such as health care referrals and nutrition education; and meeting
other program requirements, such as voter registrationï¿½often left
little or no time to follow up on suspected cases of fraud and abuse. 
For example, one official said that taking time to check on
individuals suspected of providing incomplete documentation of income
would detract from time spent providing services to other clients. 
Some state agency directors also shared this view. 

Similarly, according to FNS officials, the management of the WIC
program, which includes fraud and abuse prevention and detection, is
hindered by a lack of resources.  FNS WIC officials pointed out that
FNS' staffing resources for managing the program have decreased,
while participation has increased.  In recent years, because of
resource constraints, the results of FNS management evaluations of
state and local agencies, which can address various fraud prevention
issues, have not been summarized and shared among FNS regions.  In
addition, resource constraints are adversely affecting the agency's
ability to implement the electronic benefits transfer (EBT) system,
according to FNS officials.\21 The use of such a system could improve
efforts to prevent and detect vendor and participant fraud and abuse. 
Finally, in developing and implementing a WIC vendor database, to
include data on vendor fraud and abuse activities, FNS has been
hindered by lack of resources.  This system was developed and
implemented by FNS without additional resources. 

--------------------
\21 EBT is a system that allows participants to electronically
transfer their benefits to a vendor account to pay for WIC foods. 

   CONCLUSIONS
---------------------------------------------------------- Chapter 3:5

Although FNS regulations require that state agencies authorize an
appropriate number and distribution of vendors in order to ensure
that participants have adequate access and that state agencies can
effectively review vendors, they do not provide criteria for making
such an assessment.  If states approve more vendors than they can
effectively manage, the integrity of the vendor delivery system could
be compromised, and fraud and abuse may go undetected.  Regulations
and guidance describing more specific considerations to be used in
limiting the number of vendors participating in the program would
assist state agencies in this effort.  Such considerations could also
be used to assist FNS in its responsibility to evaluate whether the
states are effectively and efficiently managing reviews of their
vendors.  FNS' June 16, 1999 proposed rules would require that state
agencies establish criteria to limit the number of vendors
authorized, but they do not specify the criteria to be used.  The
proposal does not identify criteria FNS would use to assess if states
authorize more vendors than they can effectively manage. 

State and local WIC agencies vary considerably in the strategies they
use to prevent and detect program fraud and abuse, which increases
the complexities involved in managing the program and improving
program integrity.  Given these conditions, reliable data on detected
participant fraud and abuseï¿½who is committing the fraud and how
often, what types of fraud are being committed, and how much program
funding is lostï¿½are important in evaluating the effectiveness of
agency efforts aimed at preventing and detecting participant fraud
and abuse.  FNS does not require state and local WIC agencies to
report information on participant fraud and abuse.  While there would
be costs associated with collecting such information, the absence of
these data hinders FNS' and state agencies' ability to manage the
program, including identifying the changes needed to improving
program integrity.  Furthermore, it is possible that not collecting
such information may send a message to some agency officials that
preventing and detecting participant fraud and abuse is a low
priority and thus damage the public's trust in the program. 

Finally, regarding employee fraud and abuse, neither FNS' regulations
nor policies address conflicts of interest that develop when an
employee is responsible for both certifying eligibility and issuing
benefits or when an employee is also a WIC participant.  Internal
controls in these areas are necessary to minimize employee fraud and
abuse and to emphasize program integrity. 

   RECOMMENDATIONS
---------------------------------------------------------- Chapter 3:6

To enhance USDA's ability to improve program integrity and encourage
better efforts to prevent and detect fraud and abuse by state and
local WIC agencies, we recommend that the Secretary of Agriculture
direct the Administrator of the Food and Nutrition Service to take
the following actions: 

  -- Amend the regulations on vendor management to ensure that the
     states limit their authorized vendors to a number they can
     effectively manage and issue guidance to the states on the
     specific criteria FNS will use to assess their compliance with
     the regulations and the actions they would need to take if FNS
     determines that they have authorized more vendors than they can
     effectively manage;

  -- Work with the state WIC agencies and the National Association of
     WIC Directors to develop and implement cost-effective strategies
     for the states to use in collecting and maintaining information
     on cases of participant fraud and abuse, which would be
     periodically reported to FNS.  Such information should include
     the nature of the fraud detected and the associated dollar
     losses; and,

  -- Require state agencies to have a policy and procedures for
     addressing potential employee conflicts of interest. 

   AGENCY COMMENTS AND OUR
   EVALUATION
---------------------------------------------------------- Chapter 3:7

We provided the Food and Nutrition Service with copies of a draft of
this report for review and comment.  We met with agency officials,
including the Associate Administrator for FNS and the Associate
Deputy Administrator, Special Nutrition Programs.  FNS generally
agreed with the report's findings and recommendations.  However, the
officials raised some questions about the specific steps that would
be necessary to implement two of our proposed recommendations. 
Concerning the recommendation that FNS strengthen its vendor
management regulations, FNS officials commented that the intent of
the recommendation could be achieved through a combination of
regulatory change and the issuance of program guidance to the states. 
We agreed and revised our recommendation accordingly.  With respect
to the recommendation that FNS determine the costs and benefits of
developing a national database of information on participant fraud,
FNS officials commented that by working with state WIC agencies and
the National Association of WIC Directors they could develop
cost-effective strategies to collect and maintain data on participant
fraud and abuse at the state and national levels without carrying
out, what they believed would be, a lengthy and costly formal
cost-benefit study.  It was not our intent that the agency carry out
a lengthy cost-benefit study.  We revised our recommendation to have
FNS work with the states to implement cost-effective strategies. 

The agency officials also stated that they believed the information
contained in the report would help to highlight the fundamental
responsibility that state and local WIC agencies have in preventing
and detecting fraud and abuse on a day-to-day basis. 

METHODOLOGY AND ANALYSIS USED IN
THE MAIL SURVEY
=========================================================== Appendix I

In developing the questionnaires for our mail survey, we conducted
pretests of our state survey at five state WIC agencies managing the
Special Supplemental Program for Women, Infants and Children (WIC)
and one Indian tribal organization, and of our local survey with
directors of local WIC agencies in five states and the District of
Columbia.  GAO staff visited local WIC agencies to conduct each
pretest consisted of a visit to a local WIC agency by GAO staff. 
During these visits, we attempted to simulate the actual survey
experience by asking agency directors and staff to fill out the
survey.  We interviewed the director and staff to ensure that (1) the
questions were readable and clear, (2) the terms were precise, (3)
the survey did not place an undue burden on local agency directors,
and (4) the survey appeared to be independent and unbiased in its
point of view.  We also obtained a review of our surveys from
managers at FNS. 

In order to maximize the response to our surveys, we mailed a
pre-notification letter to all of the agencies in the survey about 1
week before we mailed the survey.  We also sent a reminder letter to
nonrespondents about 4 weeks after the initial mailing survey and a
replacement survey for those who had not responded after about 8
weeks.  After reviewing all of the survey responses, we contacted
agencies by telephone to clarify answers for selected questions. 

For local agencies, since we used a sample (called a probability
sample) of 500 of the 1,846 local WIC agencies to develop our
estimates--each estimate has a measurable precision, or sampling
error, which may be expressed as a plus/minus figure.  A sampling
error indicates how closely we can reproduce from a sample the
results that we would obtain if we were to take a complete count of
the universe using the same measurement methods.  By adding the
sampling error to and subtracting it from the estimate, we can
develop upper and lower bounds for each estimate.  This range is
called a confidence interval.  Sampling errors and confidence
intervals are stated at a certain confidence levelï¿½in this case, 95
percent.  For example, a confidence interval at the 95-percent
confidence level means that in 95 out of 100 instances, the sampling
procedure we used would produce a confidence interval containing the
universe value we are estimating.  Table I.1 lists the sampling
errors and confidence intervals for selected information about fraud
and abuse categories from the local agency survey. 

                                        Table I.1
                         
                         Sampling Errors and Confidence Intervals
                           of Estimates of Fraud and Abuse from
                         Information in the Local Agency Surveys

                                                                           Confidence
                                                                            interval
                                                                       ------------------
                                                             Sampling
                                                   Estimate     error      From        To
-------------------------------------------------  --------  --------  --------  --------
Chapter 2
Participants reported as committing fraud in one     0.141%     0.067    0.074%    0.208%
 or more of seven serious categories represented
 as a percentage of the average monthly
 participation in fiscal year 1998\a (page 23)
Number of participants reported as committing         7,074     3,231     3,843    10,305
 fraud in one or more of seven categories of
 serious fraud (page 23)
Number of participants reported as receiving          3,566     2,494     1,072     6,060
 multiple benefits ï¿½ dual participation (page 29)
Number of participants reported exchanging            2,049     1,888       161     3,937
 vouchers for nonapproved or nonfood items (page
 29)
Number of participants reported as                      886       420       466     1,306
 misrepresenting their income (page 29)
Number of participants reported as                      561       433       128       994
 misrepresenting facts affecting eligibility,
 other than income (page 29)
Number of participants reported exchanging              233       120       113       353
 vouchers for cash (page 29)
Number of participants reported giving away             218       163        55       381
 vouchers or food (page 29)
Number of participants reported claiming phantom        122       112        10       234
 or nonexistent dependents (page 29)
Participants reported as committing fraud in one     1.643%    0.7644   0.8786%   2.4074%
 or more of three less serious categories
 represented as a percentage of the average
 monthly participation in fiscal year 1998\a
 (page 23)
Number of participants reported as committing        79,271    36,840    42,431   116,111
 fraud in one or more of three categories of less
 serious fraud (page 23)
Number of participants reported redeeming food       59,810    26,965    32,845    86,775
 vouchers outside the authorized dates on the
 vouchers (page 33)
Number of participants reported selecting            25,209    25,074       135    50,283
 incorrect brands or quantities of food (page 33)
Number of participants reported as being verbally     8,625     4,283     4,342    12,908
 abusive to WIC vendors and/or WIC employees
 (page 33)
Number of participants reported as committing         2,902     1,644     1,258     4,546
 other types of fraud or abuse (page 34)
Number of employees who work in the WIC program          78        14        64        92
 who were suspected of fraud or abuse (page 35)
Number of employees who work in the WIC program          48        21        27        69
 who actually committed fraud or abuse (page 35)
Chapter 3
Number of participants reported as receiving an       5,121     3,125     1,996     8,246
 oral warning or counseling about program rules
 for serious violations (page 45)
Number of participants reported as receiving a        2,282     1,726       556     4,008
 written letter of warning for serious violations
 (page 45)
Number of participants reported as receiving a        1,787     1,455       332     3,242
 suspension or temporary disqualification from
 WIC for serious violations (page 45)
Number of participants reported as receiving some       428       310       118       738
 other sanction for serious violations (page 45)
Number of participants reported as receiving no          33        28         5        61
 sanction for serious violations (page 45)
Number of participants reported as receiving an      57,547    32,673    24,874    90,220
 oral warning or counseling about program rules
 for less serious violations (page 45)
Number of participants reported as receiving a       19,379     5,827    13,552    25,206
 written letter of warning for less serious
 violations (page 45)
Number of participants reported as receiving a          336       298        38       634
 suspension or temporary disqualification from
 WIC for less serious violations (page 45)
Number of participants reported as receiving some       694       601        93      1295
 other sanction for less serious violations (page
 45)
Number of participants reported as receiving no          \b        \b        \b        \b
 sanction for less serious violations (page 45)
-----------------------------------------------------------------------------------------
\a The sampling error may be too low because of high variability in
the numerator of the estimated percentages. 

\b We were unable to develop a reliable estimate. 

GAO CONTACTS AND STAFF
ACKNOWLEDGMENTS
========================================================== Appendix II

GAO CONTACTS

Lawrence J.  Dyckman, (202) 512-5138
Thomas E.  Slomba, (202) 512-9910

ACKNOWLEDGMENTS

In addition to those named above, Carolyn M.  Boyce, Clifford J. 
Diehl, D.  Patrick Dunphy, Judy K.  Hoovler, Barbara L.  Scharl, and
Carol Herrnstadt Shulman, made key contributions to this report. 

*** End of document. ***