A sum
several .~~~ ,~~~IIM" ,B f
making
ments in recovering
claims and
reducing certification
errors in the
Food Stamp Program.
A report on
efforts to reduce
· fraud, waste, and
abuse in the Food
Stamp Program
ASER~
United States
Department of
Agriculture
Food and
Nutrition
Service
FNS-246
May 1984
.. ,
Introduction
We are pleased to present the
Food and Nutrition Service's
fourth report on strategies to
reduce waste, fraud, and abuse
in the Food Stamp Program. This
Reducing Errors
in Certification
For many States, the big news
in certification is computers.
Computers are not only capable
of quickly cross-checking client
information, but also are helping
caseworkers reduce errors through
"machine-assisted" interviews.
On other fronts, demonstration
projects in Pennsylvania, New
Mexico, and North Carolina now
have results to share on using
second party reviews, verification
specialists, and different
verification techniques to reduce
certification errors.
Computers Speed
Certification In
Alaska, Illinois
During the 1970's, computer
technology helped to modernize
many welfare offices. New
machines could accurately compute
benefits and screen for
duplication of services, freeing
2
issue relays more information on
error reduction strategies, in
keeping with this year's Operation
Awareness emphasis on error
prevention. There are several articles
that may be helpful to you
in your error prevention activities,
including information on
three demonstration projects
that we know will be of interest
to you.
We have recently reviewed the
results of a full year of operation
of the State Exchange Project.
Topics attracting attention last
year included fraud control, investigative
and -prosecution
techniques, fraud and fair hearing
procedures, and claims tracking
and collection strategies. In
many cases, States have been
able to successfully incorporate
these ideas into their own systems.
I would like to personally
thank each of the States that
hosted a State Exchange meeting.
We hope that in its second
year of operation, the State Exchange
Project wi II provide for
even greater information sharing,
especially in the field of
error prevention.
caseworkers from arduous math
worksheets and card file
searches.
Now a new age of computers
promises to make certification
light-years more efficient through
machine-assisted eligibility interviews
and verification of many
kinds of data while the client is
in the office.
States like Alaska and Illinois
are experimenting with online
verification, a method of electronically
cross-checking unemployment
compensation, residency,
or other data at the point of
application. This kind of checking
can prevent errors before
they get into the system, saving
money, time, and energy from
being spent later in claims collection
and fraud investigations.
States can also use the systems
to manage their programs
better. The computers can improve
employee accountability
and workload monitoring. They
can automatically send out
monthly reports and issuance
We would like to take this opportunity
to share with you some
of the results of our joint efforts
to combat waste, fraud, and
abuse in the Food Stamp Program
over the past 2 years.
From 1981 to 1983, the number
of fraud hearings rose 42 percent,
completed fraud investigations
jumped 195 percent, and
fraud prosecutions increased by
93 percent. In addition, the
dollar amount of recipient
claims collected increased by
106 percent. We want to encourage
you to keep up the good
work. The challenge of making a
substantial dent in our error rate
deserves the same commitment
of attention and resources. We
are convinced that an all-out effort
on prevention of errors holds
great promise for making a real
impact.
Virgil L. Conrad
Deputy Administrator
for Family Nutrition Programs
Food and Nutrition Service
notices, and can check for errors
of logic in casefiles.
Alaska
A pioneer in making the most
of the new computer technology
is Alaska. Its online Eligibility
and Issuance System (EIS) for
food stamps and Aid to Families
with Dependent Children (AFDC)
is currently being used in much
of the State and will be fully implemented
next month. Alaska
estimates the system will save
about $1.4 million annually by
preventing food stamp errors
that result from unreported income
and changes in household
size, as well as mathematical or
policy mistakes.
The system will als·o cut down
on errors resulting from mandatory
monthly reports. It can generate
the monthly reports and
keep track of all the information
that comes in on them. An eligibility
worker can then compare
it online with information from
the following month, and make
changes online. In fact, the computer
will not allow a benefit to
be issued unless an eligibility
technician authorizes it, based
on a monthly report.
Other errors can be prevented
by the "front end" screening of
information at the time a client
applies for food stamps. When
an applicant comes to the welfare
office, a clerk can enter into
the computer basic information
the client provides on the first
two pages of the application.
Background on current or prior
contacts with the welfare system,
and a special alert if the
person has been disqualified for
fraud, will appear on the screen.
At the eligibility interview, the
caseworker can enter the remaining
information from the application
into the computer terminal
on his or her desk. All of the elements
of identification can then
be screened. If the social security
number has been used by
anyone else in the State, the
worker will know it instantly. If
the employment the client reports
is different from what the
State department of labor is reporting,
an alert will go to the
The Illinois Department of Public Aid is
testing a combined application for persons
applying for food stamps, AFDC,
and SSI. Questions asked of applicants
worker online. Once the information
is entered, the computer
can calculate the benefits for
both food stamps and AFDC
accurately.
• For more information contact:
Eric Hanson, EIS Project Director,
Alaska Department of Health and
Social Services, (907) 465-3355.
Illinois
Other States are testing the
new technology as well. Illinois,
for example, has a computerized
certification system up and running
on a pilot basis in three
welfare offices located in
Chicago and Springfield. The
system uses a combined application
for food stamps,
AFDC, Assistance for the Aged,
Blind, and Disabled, and
Medical Assistance.
At the Illinois test sites, applicants
don't even have to fill out
application forms. The receptionist
enters their name and reason
for visiting the office, and determines
if an acting or pending
case exists. Then a "screener"
asks the client about 30 questions
and enters the answers into
the terminal. A printer prints
come up on a computer screen at an
eligibility worker's desk. The worker then
types in the answers and a high-speed
printer types out a summary.
out what's known as a "name
file search" at this point. This
shows whether a client has ever
gotten assistance in the State
before. A second search matches
the data supplied by the client
with information from the State's
department of labor on prior
work history and receipt of
unemployment insurance.
The new system saves time,
according to Springfield employees.
All Illinois counties
must run these same clearances,
but most have someone sit at a
keyboard, plug in the information,
and ask for clearances.
With the new system, clearances
come back automatically, and
right while the client is at the
interview.
In the last phase, the client
meets with the eligibility worker,
who asks questions as they are
displayed on a computer screen.
The worker types in the answers
and the documentation provided.
The computer even tells whether
the documentation is sufficient.
In this way, the new system
makes the handling of cases
more consistent, because it
eliminates some of the subjective
decisions that a casew·orker
must make in determining eligibility.
After this interview, a highspeed
printer types out a summary
of the questions and
answers. The applicant signs
this form and it becomes a part
of his or her file.
• For more information contact:
Tim Grace, Food Stamp Director,
Illinois Department of Public
Aid, (217) 782-1355.
3
supervisor. Since Brunswick
County is a rural area with a
small caseload, only one caseworker
was assigned the role of
verification specialist.
All verification specialists
handled a much smaller caseload
than other caseworkers. Their interviews
with applicants were
longer and involved more verification
than the norm.
Since households identified
as error prone via statistical
analysis may be required to furnish
additional verification, no
waivers to existing regulations
were required.
The staff in Mecklenburg
County developed a worksheet
that centered on all monthly
household expenses rather than
allowable food stamp deductions.
Much of the intensive
interview and verification activity
focused on the households'
management of expenses. The
clerk assigned to the Mecklenburg
verification unit routinely
checked property tax records,
motor vehicle. records, and
school reccrrds. Home visits
were rarely used.
The cost of this special casehandling
was about double the
cost of processing a normal
case. But the results were impressive.
In Mecklenburg County,
the error rate of error-prone
cases declined by 31.5 percent.
RTI projected that if all counties
achieved a reduction comparable
to Mecklenburg's, over $3.3 million
in allotment costs would be
saved statewide.
North Carolina is now using
verification specialists throughout
the State to interview and
certify all error-prone households,
including any household
consisting of five or more members.
IFor more information contact:
Larry Goolsby, North Carolina
Department of Human
Resources, (919) 733-4570.
Hawaiian Caseworkers
Reduce Errors
One problem many States
have in common is how to avoid
making the same certificati o n
errors over and over. Since June
1982, Hawaii has been workiBg
to solve this problem by using a
6
• •
new corrective action technique
that has helped reduce the
State's cumulative allotment
error rate from 9 percent to 6.3
percent in just 6 months. The
technique itself is very simple,
and uses elements of the Japanese
idea of a "quality circle."
For every error found in a
quality control review, the worker
responsible must analyze the
cause of the error-be it procedural,
such as filling out an input
document Incorrectly, or
substantial, such as misinterpreting
a policy-and submit a
written report to the unit supervisor.
That analysis must go
beyond simply correcting the individual
case error. The worker
must determine the cause of the
error, and this often involves
contacting the client to discuss
the error.
But it doesn't stop there. At
monthly unit meetings, the caseworker
must make a presentation
to the entire unit on the case,
the type of error, its cause, and
how that error can be prevented.
What Hawaii has found is that
when one worker makes an error,
the cause of the error is often
present throughout the unit. For
instance, if a worker incorrectly
interprets a policy and, as a
result, commits an error, often
others in the unit are suffering
from the same misunderstanding.
The cause of that policy misinterpretation
might be lack of train~
ng in a specific area or because
the supervisor misunderstands a
policy and is passing it down to
workers incorrectly. Whatever
the reason, the unit-wide presentation
has the effect of eliminating
not one error but many errors
of the same type.
In another instance, one group
of caseworkers realized that
many procedural errors were occurring
near the computer deadline,
as they rushed to complete
input documents. Armed with
this information, the unit developed
a plan that staggered the
workload, doing recertifications
earlier in the month and avoiding
the end-of-the-month crush .
At first, the new error analysis
system took some getting used
to on the caseworkers' part. But
Hawaii's management has tried
to maintain a nonjudgmental atmosphere
in the unit meetings
so that caseworkers don't feel
picked on. And, by sharing their
experiences and analyzing their
own errors, caseworkers have
found that they now are much
more aware of the part they play
in the system. They are also seeing
that what they're learning is
not only helping them reduce
their own errors but, surprisingly
enough, is also cutting the incidence
of "client errors."
This new error analysis system
is in use in all local offices
throughout Hawaii. The system
is a small part of Hawaii's overall
corrective action strategy but it
seems to be having a considerable
effect on their error rate.
•For more information contact:
Christina Lam, Hawaii Department
of Social Services and
Housing, (808) 548-0660.
Claims-Getting Back
What Recipients Owe
Keeping track of claims can
be one of the more difficult administrative
tasks in operating
the Food Stamp Program. Usually
there are so many different departments
and divisions involved
in processing claims that the
simple act of knowing where a
specific case is at a given moment
is almost impossible.
Some States, such as New
Jersey, have found placing
responsibility for recovering
claims in one unit to be a great
help. Other States, like Arizona,
are finding that minicomputers
can be an inexpensive, yet effective,
aid in organizing and tracking
recoveries.
Tried and True
Methods Boost
New Jersey Collections
A concentration of efforts on
the part of New Jersey's State
claim unit and county welfare
agency personnel has made
New Jersey one of the Nation's
leaders in the collection of
recipient claims. New Jersey
currently ranks eighth nationally
in dollars collected for recipient
claims.
The State's commitment, combined
with new food stamp
regulations on fraud and claims
collection, has dramatically increased
New Jersey's collections.
During 1980, the State collected
$149,000 in food stamp
claims. By the end of 1983, this
annual amount increased to
more than $1.2 million.
Burlington County, located in
southern New Jersey, has made
a significant contribution to the
State's overall success. Although
the Burlington County welfare
office handles only a mediumsized
food stamp caseload, it
ranks second in the State in
total monies recovered from
food stamp claims. The County
also ranks first in the State in
the collection of "agency error"
overissuances.
Burlington County's success
can be attributed to its decision
to locate all responsibility for
Maureen McGlashan, of Burlington
County's collections unit, works with a
Victor Waller, supervisor in the Burling·
ton County welfare office, points to an
impressive increase in collections activ·
ity. Burlington County ranks first in New
claim actions in one unit. From
the point that a potential claim
is identified, total responsibility
for all actions belongs to the
collections unit. These actions
include: determining the amount
of loss, preparing the claim,
referring claims to legal agencies,
meeting with and arranging
payment schedules with clients,
maintaining fiscal records, following
up on existing claims,
and reporting to the State
claims unit.
County welfare personnel be-client
to set up a payment schedule .
Jersey in the collection of "agency error
overissuances," and is second in the
total monies recovered from food stamp
claims.
lieve that the collections unit
has been especially successful
due to its rapport with the legal,
fiscal, and income maintenance
units. Burlington County also
tries to run its collections unit
as much like a private collection
agency as possible.
The County's collections unit
also believes that the February
1983 food stamp regulations
allowing recoupment of allotments
and waiving of administrative
fraud hearings have increased
the amounts collected.
The agency has also benefited
from New Jersey's Set-Off Individual
Liability (SOIL) program.
The SOIL program allows the
collections unit to offset food
stamp claims against State income
tax refunds and homestead
rebates.
Burlington County's future
plans include a data processing
system that will automate many
of its manually completed processing
procedures.
IFor more information contact:
Pat DeMarzo, New Jersey
Department of Human Services,
(609) 633-6050.
Automation Helps Arizona
Keep Tabs on Claims
Many States find claims recovery
to be an overwhelming
administrative task. For Arizona,
help arrived in 1983, in the form
7
of a Basic IV minicomputer.
Since implementing its automated
tracking system, Arizona
has increased its recoveries
22-fold-from $450 a month to
$10,000 a month.
Under its old system, there
was no way for the State's
recovery unit to tell where a
given case was. If a quality control
review, audit, or management
evaluation found an overissuance,
a report was sent to a
local office to investigate and
establish a claim. But there was
no tracking system to follow up
with the local office if the report
was not returned within 30 days.
Similarly, several different documents
or duplicate documents
on a claim might be sent to different
branches of the State government
simultaneously. This
muddled the chain of decision
on the claim and obscured the
exact status of the case.
Under the new system, a minicomputer
makes Arizona's Restoration
and Recovery Unit (RRU)
the hub of the entire process.
Only one claim file is created,
and its location is logged onto
the computer by a clerk before it
is sent anywhere. When a file is
sent to a local office for followup,
that information is recorded.
Likewise, when the file returns
to the RRU from the local office,
Members of the FNS training section
tape "Let's Get Interviewed," an inhouse
production for eligibility workers
on interviewing skills. The videotape includes
a segment structured like a tele-
8
that information is logged onto
the computer even before the
file is forwarded to an RRU
worker for action.
Even though this might seem
to be a more c'ircuitous and
time-consuming route for paperwork
to travel, the system has
proved to be almost foolproof in
accurately locating case files.
The Arizona system is very
simple to set up. The computer
software used to structu're the
system is part of a standard
sales/accounts receivable package
that's easy to adapt to
monitor food stamp claims. The
system contains 23 primary data
fields which can be used to
record the case name, address,'
Social Security number, the
local office that carries the
case, the status and dollar value
of the claim, and, most importantly,
who is currently holding
the claim. Arizona uses a secondary
field to record historical
data on each case.
The usefulness of the minicomputer
is more than as a
glorified recordkeeper, however.
The system has the flexibility to
generate a number of reports,
flag delinquent cases, pinpoint
local offices that are not taking
action on claims, calculate productivity
reports for workers and
offices, and generate accurate
vision game show, with contestants,
game board, and bell. For more information
contact: Joseph H. Pinto, FNS National
Office, (703) 756-3471 .
Lydia Sermeno, of Arizona's Restoration
and Recovery Unit, uses the Basic IV
word processor to enter data on a claim.
Arizona has been able to increase the
number of claims recoveries 22-fold,
since introducing a new automated
system.
data on the value of claims at
different stages of processing.
Arizona's new system has
also been enhanced by the addition
of new staff to establish and
monitor claims. In several of the
large local offices, Arizona has
designated one full-time staff
member to process claims. This
means one staff member, well
versed in claims and program
rules, is now available for calculating
the claim amount.
For Arizona, adding extra staff
to process food stamp claims
was made possible because
those positions in the central office
were eligible for 75 percent
funding from FNS. The computer
system provides the support so
the increased staff is used effi ciently.
Arizona has developed a handbook
that describes their procedures
in more detail.
IFor a copy of the handbook,
or more information contact:
Diane Ross, Arizona Department
of Economic Security,
(602) 255-5137.
The Food Stamp Program is an equal
opportunity program. If you believe you
have been discriminated against because
of race, color, national origin, sex, age,
handicap, religious creed, or political
beliefs, write immediately to the Secretary
of Agriculture, Washington, D.C. 20250.
Home Economics and Household Collection
