Supplement to Comptroller General's Report to the Congress, "The Food Stamp Program--Overissued Benefits Not Recovered and Fraud Not Punished" (CED-77-112 July 18, 1977)

Published by the Government Accountability Office on 1977-08-31.

Below is a raw (and likely hideous) rendition of the original report. (PDF)



Supplement To
Comptroller General’s Report To
The Congress, “The Food Stamp
Program--Overissued Benefits
Not Recovered And Fraud
Not Punished,”
(CED-77412 JULY 18, 1977)
This supplement        to an earlier GAO report
includes comments         from the Department     of
Agriculture     and GAO’s       response to those
comments.      Agriculture    did not dispute the
report’s    basic message that proper actions
are not being taken to recover overissuances
(estimated    at $590 million       a year) and to
punish recipient       fraud which     is a serious

Some of Agriculture’s     comments    tend to
obscure and minimize    the report’s message--
unjustifiably so in GAO’s opinion. GAO still
believes that the problems discussed in the

                                                           31, 1977
                      COMPTROLLER     GENERAL     OF THE UNITED   STATES
                                    WASHIN6TON.    DC.   20148


To the President of the Senate and the
Speaker of the House of Representatives
       We are issuing     this supplement      to provide  an evaluation
of the Department       of Agriculture's     comments on our July 18,
1977, report    to the Congress entitled          "The Food Stamp Program--
Overissued    Benefits    Not Recovered     and Fraud Not Punished"
(CED-77-112).      We did not receive       Agriculture's   comments in
time to include      them.in    the report.

        As is our normal policy         and practice,       we asked Agriculture
on May 25, 1977, to furnish            us its written       comments on a draft
of the report     within     30 days.      At Agriculture's        request,   this
deadline     was subsequently       extended    from June 24 to July 1, then
to July 8, and finally           to July 15.      We also met with top of-
ficials    of Agriculture's        Food and Nutrition        Service    and the Of-
fice of the Secretary          five times in late June and early            July to
discuss    our draft     report,    to obtain     oral comments from Agricul-
ture,    and to try to expedite         written    comments.
        Meanwhile,       the House of Representatives            had scheduled
floor    debate on a bill         to revise     the food stamp program's
authorizing       legislation.        In an effort      to provide      timely     in-
put to the legislative            process,    we briefed      the Committee        on
Agriculture       staff     on our findings       and recommendations.            At that
time the staff          said it would be helpful           if we could issue our
final     report    in time for the floor          debate,    and asked that we
make every effort           to do so.      In view of Agriculture's            repeated
delays      in furnishing      us with its comments, we decided                not to
wait any longer.            We issued the report        without     them on July 18,
      After   the report     was printed      and about the time it was
being issued,    we received     Agriculture's      written   comments in a
July 18, 1577. letter.         Agriculture      did not dispute   the report's
basic message, but raised        some issues and made some statements
which we believe     require    our response.

       We believe    that our report  shows serious     and widespread
problems     with the food stamp program and that       administrative
and legislative      actions are needed to correct      those problems.
        We are sending   copies of this   Supplement to the Director,
Office     of Management and Budget,    and to the Secretary of

                                   of the United      States

             FRAUD NOT PUNISHED"        (CED-77-112,      JULY 18,      1977)
        In commenting       on a draft     of our report        entitled       "The
Food Stamp Program-- Overissued             Benefits       Not Recovered        and
Fraud Not Punished"          (CED-77-112,      July 18, 1977),           the
Department     of Agriculture       said that it is committed                to
more vigorous      action     and is determined          to see improvement
in the fraud and overissuance              area.     We believe
Agriculture's      commitment      and determination          are encouraging,
but they are long overdue.             According       to its comments,
Agriculture's      thinking     is generally        in line with all but
one of the legislative          recommendations          in the report.
Regarding     administrative       changes,      Agriculture       said that
some of our recommendations            are already        being implemented
and that others       will    be carefully       considered.
        Agriculture        does not dispute         the basic message of our
report--     that proper        actions    are not being taken (1) to
recover      overissuances         which,    according    to Agriculture's
latest     data,     constitute       almost 12 percent       of the total
cost of program benefits,                or about $590 million          a year, and
 (2) to punish fraud which is a serious                   problem     in the
program.        Its comments, however,            raise   some issues and
include      some statements          and objections      which tend to
obscure and minimize             the report's       message.      Some of the
objections        are embellished         with assertions       that seem to
have little        or no hard data or other clear               evidence    to
support     them.
        Agriculture's   comments are discussed               in more detail
below    and are included   in this Supplement               as appendix    I.
Discussion      of legislative       recommendations
        Agriculture        stated  that the Administration's         proposal
for overhauling          food stamp legislation       included     most of
the legislative          changes recommended in our report.             It was
referring       to our     recommendations     that administrative
adjudication        of   most recipient     fraud be authorized        and
that the Federal           share of the States'     costs for investigating
and prosecuting          suspected    fraud be increased       from 50 to 75
        Our report    includes     a section    on congressional       actions
to increase     overissuance       recovery,    which sets out the
major legislative        proposals     considered     by this and the
previous    Congress on this         subject.     The section    discusses
various    approaches      the Congress has been considering            to
improve overall       program integrity,         especially   those
affecting     overissuances.         These include
       --suspending        individuals       from   program     participation,
       --requiring       States     to pay a part       of    food   stamp       bonus
       --clarifying      the Secretary     of Agriculture’s       authority
           to exercise     flexibility   in handling       claims
           against   program recipients,
       --increasing      from 50 to 75 percent  the Federal                    share
           of the cost of investigating     and recovering                   food
           stamp benefit     overissuances, and
       --allowing       States  to retain    50 percent    of all            funds
           recovered     as a result    of such investigation                and
We state     in the report        that Agriculture’s    legislative
proposal     for this year        included   many of the same provisions.
        Regarding      our recommendation            that States      be authorized
to retain      some part of the money recovered                   from recipients
of overissuances,          Agriculture         agreed that it may be useful
to authorize        States    to retain        a portion       of fradulent     over-
issuances      recovered      but strongly         disagreed      that States
should retain         a portion      of Federal        dollars    overissued      due
to the State’s         own errors.         It said that this would operate
as a disincentive          for States        to make improvements            to prevent
overissuances         in the first       place.        It also said that food
stamp officials         could arrange          for overissuances         with quick
recoveries,       thereby     resulting        in the retained        portion     being
diverted     from the Federal           Treasury       to the States        or, in
some places,        to the counties.
         We believe   that,      generally,      the State and local
officials      and workers       who are carrying        out the day-to-day
operations       of a multi-billion         dollar    Federal      program are
worthy of greater        trust      than is indicated         by Agriculture’s
statement      and that Agriculture’s            blanket    indictment      of the
integrity      of these people         is unfounded.        During our reviews
of the food stamp program,               we noted various        problems     in

program administration        at the State and local          level,   but
nowhere did we observe anything             even closely    approximating
the lack of honesty       Agriculture       alludes    to.  Such dishonesty
would require     the systematic       collusion      of many people
in food stamp offices       without     any individual      realizing
personal  gain because the retained              funds would go into
State or county treasuries.            Such dishonesty,       if discovered,
should be dealt      with severely,       but we believe      such instances,
although  theoretically       possible      under our recommendation,
would happen very rarely          if at all.
        As discussed        in our report     (see p. 40), we believe
Agriculture      would need to closely           monitor   the results
of its quality         control   program and investigate          and
reverse      any increasing       trends   in program errors        that
result      in overissuances.          It has authority      to withhold
program administrative            funds from States      found not to be
administering        the program efficiently          and effectively.
         Agriculture      also proposed       that the legislation       be
revised       to make it easier       for it to collect       from States
the value of food stamp benefits                 overissued   because of
State negligence.            Under existing       law, States     can be
required        to repay Agriculture        the value or overissued
benefits        only if there was gross negligence            or fraud
by the State.          Under Agriculture’s          proposal,   such
repayment        could be required       if overissuances       resulted
from simple negligence,              This proposal       may be helpful
 in recovering        overissuances,       but it has significant
        Such a provision       could only be applied             if enough
cases were brought         to Agriculture’s         attention        to show
negl igence.       There is some doubt,         however,        whether    it
would be feasible        for Agriculture        to monitor         the States
closely     enough for it to identify           a significant
proportion       of all overissuances       that occur,          so that
it could determine         whether    the States       were negligent
and the amount they should repay.                 It may not be
reasonable       to expect the States       to put forth           the extra
effort     and money necessary        to effectively          identify     and
report     overissuances     to Agriculture         if they will        be
required     to repay the value of the overissuances.
        As discussed       in chapter  4 of our report,     we believe
a built-in      incentive     is needed to encourage     the States
to identify       and recover     food stamp overissuances     on
a day-to-day       basis regardless      of the cause of the
overissuance--      honest errors     by food stamp personnel       or
recipients,       negligence    by food stamp personnel,      or fraud

by recipients.        The Department     of Justice     concurs whole-
heartedly      that the program must provide         some financial
incentive      to the States      for making aggressive       efforts
to recover       food stamp overissuances,       rather    than the
current    financial     disincentive.
        Agriculture        said that its proposal             for legislation
requiring       that recipients         who are administratively
determined        to have committed          fraud be suspended from
participation          in the program calls           for automatic
suspension--        no warnings       would be permitted.            Although
we would generally             expect   that most recipients
committing        fraud would be suspended              from the program
under a system of administrative                   sanctions,       our
recommendation           envisioned     providing       administrators        some
discretion        and enough flexibility            to make exceptions
and adjustments           in unusual      cases--to       issue warnings
of suspension           in special     cases where this might seem more
appropriate         than suspension.
      'Agriculture's         proposal       included      another     provision
that the Federal           penalty     for food stamp fraud             involving
less than $100 (misdemeanor                 fraud)     be reduced so that
such cases could be prosecuted                   before    Federal     magistrates
instead     of in U.S. District             Courts.       Our report        discusses
this proposal          (see p. 35) and notes that it might not
have the major beneficial                impact contemplated           because most
cases of suspected            food stamp recipient            fraud are not being
seriously       considered        for Federal       prosecution.          It light    of
the burdens already             facing     Federal     prosecutors        in handling
more serious         crime cases,        the outlook       for large numbers
of Federal       prosecutions         of food stamp recipient               fraud
because of a shift            to the jurisdictional             level     of Federal
magistrates        courts     is somewhat uncertain.
        The Administration's          legislative   proposal    also would
require      the Secretary       of Agriculture   to set standards        for
State administration            of the program,   including     staffing
standards.       Agriculture       would be authorized       to withhold
administrative        funds for any State not complying            with
these standards         or the State plan of operation.            The
Department      stated     that this would provide        it with a new
tool for addressing           the problem of recovering        overissuances.
        Staffing      standards for food stamp program operations
have not been established           for the States,     but they have been
authorized        since July 1974 by section       15(b) of the Food
Stamp Act of 1964 (7 U.S.C.            2024).   This section     also
authorizes        the Secretary   to withhold    administrative       funds
from any State not providing            an adequate    number of qualified
personnel       to administer   the program efficiently         and

effectively,         In light     of this,     the problems      in recovering
overissued      food stamp benefits           might not be significantly
affected     by simply      enacting    another     legislative     provision.
We believe,       however,     that establishing         and vigorously
enforcing     staffing      standards      for States      could contribute
to improvement.
        In its comments Agriculture              also mentioned       an
Administration         proposal     to require      each recipient       to
be given a form, at the time the recipient                      is approved
to receive       food stamps, which would describe                the changes
the recipient        is required        to report--such       as changes in
income-- and provide           a means for reporting          them.    The
form would also describe              the penalties       for fraud and
misrepresentation.             Use of such a form, whether            prescribed
administratively          or legislatively,         might be helpful        in
preventing       food stamp overissuances             in some cases,
especially       those due to honest errors;              but it would not
seem to directly          affect    the recovery        of overissuances
that occur.
Discussion       of administrative            recommendations
         Agriculture         said that it is committed                to more
vigorous       action      in the fraud and overissuance                   area, that
some of our recommendations                   are already        being implemented,
and that the others              will    be carefully        considered.           It
said that program administration                      in regard to fraud and
overissuance          has shown marked improvement                  since completion
of our review,           as evidenced        by substantial           increases        in
the establishment             of claims      and amounts recovered.
According        to Agriculture's          supporting        figures,       the dollar
amounts of claims             established        increased      from $6.3 million
in fiscal        year 1975 to $13.8 million                in 1976 and to
$15 million         in the three quarters              ended June 1977.              However,
as discussed          in our report         (see p. 18), Agriculture                 did
not have accurate             or complete        information        on the claims
the States        had established.             Thus, the reported             increase
in claims        established        could represent          improved       overissuance
identification           and claims       establishment,          or it could
simply      represent       better     reporting       and recordkeeping.
       Recoveries    of overissued         benefits    for the 9 months
ended March 31, 1977 (the most recent                information    available),
were about $5.6 million           compared with about $4.2 million
for the previous       9 months.       while     the percentage   increase
(33 percent)      is substantial,        the base ($4.2 million)         used
in computing      it is small;      relatively      small dollar    amount
increases    can produce large percentage              increases.

        The claims established      and the amounts recovered
are still     infinitesimal    compared with the estimated    $590
million    in overissuances     each year.
        A further       factor      that may have contributed          to the
increase      in reported         claims     is the increase      in program
growth.      As previously            noted,   the reported    value of
established       claims was $6.3 million             for fiscal     year
1975 and $15 million              for the three quarters         ended
March 1977--$20           million      on an annualized      basis.      Pro-
gram benefits         increased        from $4.4 billion      for fiscal
year 1975 to $5.1 billion                 for the year ended June 1977--
an increase       of $700 million.
Discussion       of other       report    matters
        Recipient      eligibility
        Agriculture       noted that 95 percent             of the households
receiving       food stamp benefits            are eligible       for program
participation        and that only 5 percent              are ineligible.
These figures        are accurate        but misleading         in the overall
context      of our report.           Over issued benefits         do not
relate      only to ineligible          participants.          Agriculture’s
quality      control     data  clearly       shows that nearly          two-thirds
of the overissued           benefits     are received         by households        which
are basically        eligible       for program participation               but which,
for one reason or another,               are getting        too much in food
stamp benefits.
        Locations      reviewed
       Agriculture       questioned         whether     the locations           included
 in our review       represented         a reasonable        cross section           of
the more than 3,000 local                food stamp projects            in the pro-
gram.    It said that over one-third                  of the food stamp case-
load is in the South,             including      substantial         participation
in areas that are predominantly                  rural      or small-town,           and
that we had not included               any southern         or rural      projects
in our review.          Agriculture         said that it had filed                or is
planning     to file     gross negligence           claims      against       three
of the areas included             in our review,          indicating        poor over-
all administration           in these projects.              It also said that
the areas we reviewed            were not a representative                  sample
because six of the eight               projects     were big cities             where
crime rates are above the national                    average.
       We made our review at (1) the Food and Nutrition
Service’s    headquarters       in Washington,     D.C.,    (2) three              of
its regional     offices--Chicago,       Illinois;     Princeton,
New Jersey    (since     moved to Robbinsville,        New Jersey);                and

San Francisco,    California,             and (3) eight                 food    stamp
projects   in five States              as shown below.
       --Alameda        County      (Oakland),           California
       --Cook      County      (Chicago),         Illinois
       --Cuyahoga        County      (Cleveland),                Ohio
       --Luzerne        County      (Wilkes-Barre),                Pennsylvania
       --Peoria       County      (Peoria),           Illinois
       --Philadelphia            County       (Philadelphia),             Pennsylvania
       --San      Francisco       County       (San Francisco),                California
       --Wayne      County       (Detroit),           Michigan
         In selecting     locations       for our review,     it was not
feasible,      within   the limits        of our available      resources,
to obtain      a mix of locations           that would provide
scientific      assurance     of being representative           of the
Nation’s     3,000 project        areas.       Our principal    objective
was to identify       weaknesses        in the program’s      system of
procedures      and their      implementation,         and to recommend
corrective      measures that would apply to the overall
program,     not just     to the project         areas we reviewed.
       We chose primarily        large     cities      because that is
where the bulk of the food stamp benefits                    are dispensed.
The five States       included     in our review dispensed              about
25 percent     of the total      nationwide         food stamp benefits
in fiscal     year 1976.       Even in the South,           over half of
the benefits      were distributed         in metropolitan       areas.
We included     two nonurban       areas--Peoria         County,   Illinois,
and Luzerne County,         Pennsylvania--        to see if the activities
we were reviewing        were carried        out significantly          different
in areas having a lot of small towns.                    We did not set
out to review only locations            that we knew in advance were
administering      these activities          poorly.
         As discussed      in our report    (see p. 18), Agriculture
did not have accurate          overall   indicators       of which projects
or States were doing a good job of identifying                      and
recovering      overissuances      and punishing       recipient      fraud.  The
implication       that the southern      States,     including       southern
metropolitan        areas,   were doing a much better            job than the
areas we visited         does not seem to be based on solid
statistical       data or systematic       analyses.        In any event,
Agriculture       agrees that the problems          we reported       are

serious   and widespread   and warrant              intensive      and
extensive   corrective   measures.
          From a program integrity               standpoint,       the only
reliable        data Agriculture          had on the relative           quality
of States'          administration         of the food stamp program
was quality           control     data.     This data,       however,     concerns
mainly       the quality         of the evaluation          of applicants'
eligibility           for benefits       rather     than the quality         of
activities          related      to recovering        overissuances.        The
latest       quality      control     data we could get from
Agriculture,            showing error        rates and related        data by
State       (quality      control     data is not available           or
reliable        by project),         showed that the percentage              of
program benefits              overissued       in the States which we
reviewed         (15.5 percent)         was near the national           average
 (14.1 percent).              (The national        average     for the subsequent
quality        control      period    was about 12 percent,           but
Agriculture           said it was not yet ready to give us the
data by State.)
      Agriculture's       reference    to high crime rates for the
major metropolitan        areas included       in our review seems
to say that      crime rates     are indicative      of the tendency
of recipients       in various    areas toward food stamp fraud.
We do not have, or know of, any hard facts                that show
this  to be so.
        We note, however,       that Federal    Bureau of Investigation
crime statistics       for 1975 (the most recent         available)    show
that the rates for property          crime (as opposed to violent
crime)     per 100,000 population       for the six l/ metropolitan
areas we visited       were about the same (5,56x)         as the
national      average for all metropolitan         areas (5,530),
including      those in the South.        The average property      crime
rate per 100,000 population          for the entire      Nation was
4,800.      We do not view these statistics           as showing that
recipients       in the locations    we visited     are far more
likely     to commit food stamp fraud than would be true for
the Nation       as a whole.
       Agriculture      also stated  that there is solid        evidence
of far more vigorous         action against     fraud   in areas of
the country       we did not visit,    especially     in the South.
Although     it did not have data on most of the State

L/Peoria   was classified      as a metropolitan                area and the
   San Francisco   metropolitan     area included                Alameda

activities        associated     with the punishment      of recipient
fraud,      available      data on State prosecutions       shows that,
for the 9 months ended March 1977, three States--
California,         Texas, and Ohio-- accounted      for 82 percent
of all food stamp fraud prosecutions.                These three
States      dispensed      17 percent   of the total    food stamp
bonus.        Two of these States--California          and Ohio--were
included       in our review.
        Suspected      recipient       fraud
         Agriculture         expressed        concern     that a chart        in the
report      (see p. 26), indicating                that over half of the
claims     established           in the locations           we reviewed       were
considered         to involve         suspected      recipient       fraud,     could
be misleading.             It said that the claims                established       may
not be representative                 of all overissuances             at those
locations        regarding         the proportion         involving       suspected
recipient        fraud,      because claims          are not established            for
many of the overissuances                   that occur.         Agriculture       also
said that it was skeptical                    of the accuracy          of food stamp
offices'       classifications            of claims       because no hearing
or adjudication            had occurred         at that time and because
caseworkers          often    fill     out the claim forms which are the
basis for the classifications.
        Agriculture's         position      is that caseworkers             have
a vested      interest      in erroneously          classifying        claims
as suspected          fraud to make their           own performance          appear
better     and to blame recipients              for their         own mistakes.
Accoraing       to Agriculture,          the caseworker's           best
protection        is often     to indicate        that a recipient           mis-
represented         the facts,       because this        is the one way in
which the caseworker             is completely         "off     the hook".
Agriculture         agreed that fraud         is a serious          problem,
but believes          that its incidence         may be substantially
overstated        in the chart        in question.
       Over the years Agriculture       has been repeatedly      asked
for data on the extent     of food stamp fraud,      but none has
been provided    because it was said not to be available.                                 ,
We also asked for any data it had on suspected           recipient
fraud that might be better       than the data we obtained         at
the projects   reviewed.    Agriculture     told us that none was
        In our report       we do not contend      that the data we
collected     from the projects       is scientifically        representative
of the entire       Nation    to the point    where we could confidently
say that over half of the estimated               $590 million     overissued
annually    results      from suspected    fraud or misrepresentation.

Insofar     as we know, however,            Agriculture   has no
scientific        evidence      showing,    on a national    basis,  how
much of the overissued              food stamp benefits      result  from
suspected       fraud or misrepresentation.             Moreover,
Agriculture         officials     gave us no facts      to support   their
contention        that the claims        data we collected      from the
projects      is not reasonably          representative    of the over-
issuance     situation        in those projects.
       As explained        in our report      (see p. 27), the claims
data we obtained         at the projects       neither    indicated    that
overissuances        due to project     office      error   were under-
reported,     nor that overissuances           due to recipient-
furnished      information     were overreported.           On the contrary,
we found that about 38 percent              of the overissuance        claims
at four projects         for which we had data on this matter
were classified         as resulting     from project       office  error--
hardly    evidence      of a coverup     syndrome.
         Further , there were indications                  that in cases of
overissuance           claims      attributable      to problems      with recipient-
furnished         information,           the project   offices     gave recognition
to the possibilities                 of recipient     misunderstandings       or
other circumstances                that seemed to negate suspicions             of
fraud or deliberate               misrepresentations.            For example,     every
case of income or asset understatement                        by a recipient     was
not arbitrarily             classified        as suspected     fraud.     Many were
classified          as simple misunderstandings               or omissions.       It
may be that some misclassifications                      occurred;     however,
we have no reason to suspect                    that the records       at the
project       offices       were grossly        in error     or that workers      in
project       offices       lack integrity.
       Recovering       over issuances
       Agriculture       also stated     that our        report      does not
adequately       recognize    the difficulty     in        recovering     non-
fraudulent      overissuances      from indigent           families,     noting
that a 1975 Department           of Agriculture          study showed that
84 percent      of the food stamp households                 had liquid    . assets
of less than $100.
         In discussing       the recovery       of overissuances,         our
point     was that food stamp offices              were not systematically
evaluating         each overissuance       case to determine         whether
it was appropriate           to attempt      recovery.      The report      points
out (see p. 17) that the Food and Nutrition                      Service’s
instructions         require    project    offices     to consider      a house-
hold’s     ability      to repay an overissuance           before    deciding
what to do about recovery.                We found,      however,    that entire
categories         of overissuances     --such     as those due to agency

error --were      being regarded    as uncollectible      without
considering       each household’s     ability    to repay the
benefits     they erroneously      received.      Our report
concludes      (see p. 24) that each claim should be
individually       evaluated   as to the appropriateness          of
pursuing     its collection.
       Administrative          disqualifications
        Agriculture        noted that administrative                disqualifica-
tion of recipients             obtaining      benefits       fraudulently        was
used more frequently              in locations        other than the ones
we visited,        particularly          in the South.          Our report
states      (see p. 23) that Agriculture                 officials        told us
that such disqualifications                 are being made in other
locations.         However I we have been unable to obtain
specific       data from Agriculture            as to the prevalence              of the
disqualification           procedure.         We recognize         in the report       that
this procedure          could be a viable           alternative         to criminal
prosecution         in many cases.
       Quality      control     reviews
       Finally,      Agriculture      stated    that our report     does
not adequately         explain   that the information         on overall
overissuances        is based on a small statistical             sample
from its quality         control    program;      that time-consuming,
in-depth      reviews    are conducted       of the cases selected
for such samples:          and that this      type of review would
not be practicable           when large     numbers of cases are
       Our report   states    quite   clearly    (see p. 7) that
the scope of quality        control   reviews    is beyond what can
normally    be expected     of caseworkers      on a day-to-day
basis.     This same statement       was contained      in the draft
we gave Agriculture       for comment.        The point    of the
report   is not that the numbers of claims            established     or
the amounts recovered        should match the numbers or
amounts shown by the statistical            sample.     Rather,   the
report   says that a number of measures are available                by
which the identification          and recovery     of overissuances
could be made more effective.

       In summary, we believe     that our report   oojectively
points    out the seriousness    and extensiveness   of the
overissuance     problem   in the food stamp program--including

overissuances     attributable      to suspected   recipient      fraud
and misrepresentation.         Major improvements       along   the
lines   discussed     in our report    are needed.


APPENbIX    I                                                                                APPENDIX I

                              DEPARTMENT              OF AGRICULTURE
                                   OFFICE        OF   THE    SECRETARY

                                   WASHINGTON.              D.   C.   20250     s

                                                                              July     18,     1977

     Mr. Henry Eschwege, Director
     Community and Economic Development
     U.S. General Accounting   Office
     Washington,  D.C.  20548

     Dear Mr.   Eschwege:

     In responding  to your report   on food stamp overissuance                      and fraud,
     the Department would like    to cover three basic areas:

       1.   The Administration's         recommendations     to Congress for
            legislative      overhaul    of the Food Stamp Program include            a
            number of provisions         for dealing   with fraud and overissuance.
            These provisions,        which have now been largely        adopted by
            the Senate and by the House Agriculture              Committee,     include
            most of the legislative          recommendations     in your report
            plus several      additional     measures.    In the key area of
            administrative       sanctions    against  recipient    fraud,    our
            recommendation       is tougher than yours.         The Administration        is
            committed      to dealing    with the fraud and overissuance           issue.

       2.   The Administration        plans to undertake     more vigorous     action
            administratively       to improve the States'       performance    in this
            area.     The GAO report      does not give sufficient       emphasis to
            the fact that the Department's          actions   in this area, and the
            results     in the field,     have improved substantially       since the
            GAO audit was done.          There is still    much more that should
            be done, however, and we will         move on these matters.

       3.   There are serious      problems with the GAO report,            stemming
            largely   from the use of a sample in which 6 out of 8 of the
            survey areas were major metropolitan               centers with high crime
            rates and with--in      several    instances--unusually         poor adminis-
            tration   of food stamp operations.            There is solid evidence
            of far more vigorous        action   against     fraud and overissuance
            in other areas of the country           and especially      in the South,
            but no Southern area or highly           rural     area is included    in the
            GAO survey.      The situation     with regard to fraud and overissuance
            in the country      in general is simply not as bleak as is pictured
            in much of this big city-oriented             report.     In addition,   the GAO
            report  gives the impression         that there is more fraud than the
            Department    believes    to be the case.

     We should also point out that over 95 percent of all households        receiving
     stamps nationwide  do meet all. basir eligibility   requirementsand    that less
     than 5 percent of those on the rolls   are actually    ineligible   and should
     not be there.

APPENDIX I                                                               APPENDIX I

   Mr. Henry Eschwege                                                2

   I.   Legislative   recommendations
   The Administration      agrees with the GAO recommendation for administrative
   sanctions against recipients       who commit fraud, so that such recipients
   can be punished even if no court action takes place.        Consequently the
   Administration's     food stamp proposal submitted on April 15 recommended
   a flat disqualification       of such recipients.

   We note that the GAO report would call for disqualification      or probation
   of recipients   guilty of fraud.   We disagree with the probation idea.
   Fraud should result in disqualification,     pure and simple.   Provisions
   requiring   such disqualification  for varying periods of time appear in
   both the Senate and the House Agriculture     Conrni;tee Dill.

   The Administration  also agrees with the GAOrecommendation that the
   federal share of the costs of state fraud investigations   and prose-
   cutions be raised from 50 percent to 75 percent.    This provision also
   appeared in the Administration's   food stamp proposal, and has also
   been adopted by the Senate and by the House Agriculture   Committee,

   The Administration     has also recommended that the federal penalty
   for misdemeanor recipient      fraud be reduced so that such cases may
   be tried in Federal Magistrates Court instead of U.S. District       Courts.
   It is very difficult      to persuade U.S. Attorneys' Offices to take
   recipient    fraud cases involving small amounts of money before a U.S.
   District    Court.   The Department believes that more prosecutions and
   convictions     will result if such cases need only go before a Magistrate.

  Another key Administration   recommendation involves procedures for
  reporting  changes of income by recipients  during a certification  period.
  Many such changes do not get reported by recipients,   or are reported by
  recipients  but do not result in any adjustment in benefits by case-
  workers.   This is an important area of error and overissuance.

  At present, Department regulations    require that changes in income of
  more than $25 a month be reported by recipients    within 10 days of their
  occurrence.    However, there is no standard form or procedure for
  reporting   such changes in most areas. Recipients are simply advised
  of this requirement when first   applying or first  certified.

  The Administration's    proposal would require that every recipient
  receive, at the time of certification,     a reporting form designed or
  approved by the Secretary.      The form would spell out the requirements
  for reporting    such changes on the form.    In addition,    the House
  Agriculture   Committee added the provision that civil and criminal
  penalties for fraud and misrepresentation      be contained on the reporting
  form. We believe that this reporting procedure will significantly
  increase recipient    reporting  of changes in income and state agency
  actions in response to such reports.      A definite    decrease in over-
  issuance should result.
APPENDIX I                                                                  APPENDIX I

 Mr. Henry Eschwege                                                     3

 Finally,    the Administration    proposal directs the Secretary to establish
 standards for efficient       and effective  administration   of the program,
 including staffing     standards.     States failing without good cause to
 meet   these standards, or to carry out the approved state plan of operation
 would be subject to a loss of up to 100 percent of the federal share of
 their administrative      funds,   It is the Administration's    intention  to
 require states to include their actions against fraud and overissuance
 in their state plans, so that failure        to comply with such a plan would
 subject a state to a loss in administrative         funds.
 Moreover, the GAO report observes in several places that insufficient
 staff was given as a reason for inaction on fraud or overissuance by
 a local project area. The provision calling for the Secretary to
 establish staffing standards and authorizing  him to withhold administra-
 tive funds if they are not met provides a new tool for addressing this

 The Administration  proposal also authorizes the Secretary to seek a court
 injunction  against a state that fails to comply with federal regulations
 or with its own state plan.
 A final set of legislative  changes recommended by the Administration would
 authorize the Secretary to bill states for the payment of food stamps
 overissued due to the negligence of the state.   At present, the Department
 can seek recovery only in cases of gross negligence by the state.
  The sole legislative   recommendation in the GAOreport which is not in
  the Administration   proposal is the recommendation that states retain a
  portion of any overissuances collected.
 We can agree that it may be useful to authorize states to retain a portion
 of fraudulent overissuances recovered.   We stronplv disagree. however,
 with  the suggestion that states should retain a portion of federal dollars
 overissued due to the state's own nealizence.

                                be a s                 It    could operate as a
 disincentive   for states to implement improvements         needed to prevent
 administrative    overissuances in the first  place.        Moveover, there are
 dangerous potentials      for abuse here. Food stamp       officials could arrange
 for overissuances with quick recoveries,     thereby       resulting in 50 percent of
 the overissuances being diverted from the Federal           Treasury into the state
 (or in some areas,     the county food stamp office).

APPENDIX I                                                                                APPENDIX I

   Mr. Henry     Eschwege                                                         4

   In addition,     the GAO recommendation           would run contrary      to the provisions
   of the Administration's           proposal    for punishing,     not rewarding     states
   that overissue      stamps due to their          own negligence.       We wish to bill
   states   for such overissuances,           not reward them by letting         them keep a
   portion   of the federal        funds they improperly        distributed.       For example,
   we recently    billed    California       for overissuances       in San Francisco      (one
   of the project      areas in the GAO report),           and are planning      to bill
   Detroit    and Cleveland      for overissuances        due to duplicate      ATP redemptions,
   another issue referred          to in the report.

   II.    Administrative          Actions

   The Administration    is committed    to more vigorous   action   in the fraud
   and overissuance   area and is determined       to see improvement   in this
   area.   Some GAO recommendations      for administrative    changes are already
   underway; others will    be carefully     considered.

   Although     it receives      little  attention   in the GAO report,   the
   administration       of the Food Stamp Program in regard to fraud and over-
   issuance has shown marked improvement             since the GAO audit was completed.
   The dollar      value of claims being established         is now several   hundred
   percent greater        than it was in fiscal       1975.  The dollars  being recovered
   are also on the increase.            Activity   is now particularly   strong in a
   number of southern         states.

   III.    Problems        with   the GAO report

   The GAO should have made at least some effort                    to audit localities        that
   represented     a reasonable       cross-section       of food stamp project        areas.
   Over one-third      of the food stamp caseload             is in the South, and there
   is substantial      participation        in a number of counties         that are predomi-
   nately    rural  or small town.          Not a single      southern or heavily        rural
   project    area was included         in the GAO audit.         Six of the eight project
   areas were big cities--Chicago,              Detroit,     Cleveland,    Philadelphia,
   Oakland, and San Francisco.              Although     gross negligence      claims are not
   filed   lightly   by this Department          and have been limited         in number, one
   of these six cities         already has a gross negligence            claim filed       against
   it and two others are likely             to be hit with gross negligence             claims
   in the near future.          To suggest or imply in any way that these areas
   (which all also have crime rates above the national                     average) present
   any sort of representative             sample of what is happening with food stamp
   fraud or overissuance          nationwide      is false.

   An example regards disqualification     of households     for fraud.     While the
   GAO report  found little  use of disqualification      in the areas it surveyed,
   use of disqualification   is much heavier      in some areas, particularly     in
   the South.

APPENDIX I                                                                    APPENDIX I

   Mr. Henry Eschwege                                                     5

   We are also concerned that the report may lead to a false impression
   concerning the incidence of fraud.   A chart in the report indicates                  '
   that local personnel in five of the cities surveyed believed that fraud
   might be involved in about 30 percent of the overissuance cases where
   claims were established in their areas, and that these cases involved
   about half of the dollar value of all overissuances for which claims
   were established in these areas during the period studied.

   We believe GAO should have explained more fully what these statistics
   mean. As the report explains elsewhere, claims are now established
   for only a small percentage of overissuances.   It may well be that
   those selected cases for which claims are established are more likely
   to be fraud cases than are overissuances in general.   It cannot be
   assumed that the cases for which claims are established are representa-
   tive of all overissuances.

   Even more troubling    is the manner in which the cases for which claims are
   established are identified      as suspected fraud.      The person filling     out
   the claims form simply puts down on the form whether the overissuance
   resulted from agency error, recipient       misunderstanding,      or fraud or mis-
   representation.    At the time this claim is filled        out, no administrative
   hearing or court adjudication      has occurred.    Moreover, the person filling
   out the form is often the caseworker involved in the particular              case.
   The caseworker has a vested interest       in indicating     that the overissuance
   did not result from his or her own error.         Similarly,     the caseworker may
   have an interest    in indicating   the case was not due to recipient         misunder-
   standing, since this may mean the caseworker did an inadequate job of
   explaining the requirements,       The caseworker's best protection        is often
   to indicate the recipient misrepresented the facts, since this is the one
   way in which the caseworker is completely "off the hook."

   For these reasons,     we are skeptical of the accuracy of the statistics    in
   the chart, and we    should like to emphasize that those statistics     cannot
   be applied to all    overissuances.    We believe that fraud is a serious
   problem, but that    its incidence may be substantially    overstated in the
   chart in question.

   In other areas, we believe the GAO report does not afford sufficient
   recognition  to the difficulty of recovering overissuances that were non-
   fradulent when those overissuances were made to families who remain
   indigent.   The Department's September 1975 survey of food stamp house-
   holds found that 84 percent of the households have liquid assets of less
   than $100.

   The report also does not adequately explain that the large majority of
   overissuance cases are not known to states and localities.          Our figures
   on overissuances come from quality control samples of a very small
   proportion of the caseload, and are then projected statewide and nation-
   wide. Quality control reviewers spend an average of 12 hours on each
   case, whereas regular certification        workers spend less than 1 hour on
   an average.   If the initial   certification      interview were 12 hours long,

APPENDIX I                                                                      APPENDIX I

     Mr.     Henry   Eschwege                                                     6

     the number of errors        would be vastly     reduced--but      the administrative
     costs     of the program.would     mushroom to extraordinary         levels.      The
     Administration's        goal is to simplify     the certification       process so
     that caseworkers        can do a more effective       job of avoiding      errors    in
     the time that is available.          We also expect that some of the administrative
     and man-hour savings that will         derive     from elfmination      of the food stamp
     purchase requirement        will  be applied    to doing a more effective          job of
     certifying      applicants   and of dealing    with overissuances.


     Assistant Secretary        for
                                       \ -ILcc
       Food and Consumer        Services
Copies of GAO reports are available to the general
public at a cost of $1.00 a copy. There is no charge
for reports     furnished      to Members       of Congress and
congressional     committee        staff members.      Officials   of
Federal,    State, and local governments              may receive
up to IO copies free of charge.                 Members       of the
press; college      libraries,    faculty     members,     and stu-
dents; and non-profit           organizations    may receive up
to 2 copies free of charge. Requests for larger quan-
tities should be accompanied              by payment.

Requesters entitled   to reports        without    charge should
address therr requests to:

             U.S. General Accounting     Office
             Distribution Section,   Room 4522
             441 G Street, NW.
             Washington,  D.C. 20548

Requesters   who    are required     to pay for reports
should   send their   requests   with checks or money
orders to:

             U.S. General Accounting          Office
             Distribution Section
             P.O. Box 1020
             Washington,  D.C. 20013

Checks or money orders should be made payable                     to
the U.S.      General   Accounting   Office. Stamps               or
Superintendent      of Documents   coupons   will not             be
accepted.   Please do not send cash.

To expedite   filling your order, use the              report num-
ber in the lower left corner and the                   date in the
lower right corner of the front cover.

GAO reports      are now available       on microfiche.     If such

       UNITED STATES                                POSTAGE   AND    FEES   PAID


   WASHINGTON, D.C. 20548

                                                                       THIRD CLASS