Food Stamp Program: Achieving Cost Neutrality in Minnesota's Family
Investment Program (Letter Report, 02/12/96, GAO/RCED-96-54).

GAO examined the methodologies that Minnesota used for estimating the
costs that would have been incurred if it had not implemented the
Minnesota Family Investment Program (MFIP), focusing on whether: (1)
Minnesota implemented the agreed-upon methodologies; (2) Minnesota
reasonably estimated fiscal year (FY) 1994 Food Stamp Program costs; and
(3) the Department of Agriculture (USDA) paid Minnesota the same amount
for food stamp benefits in FY 1994 than it would have paid in the
absence of MFIP.

GAO found that: (1) Minnesota agreed to use two methodologies for
estimating food stamp benefit costs that would have been incurred in the
absence of MFIP; (2) Minnesota randomly assigned applicant families to
either traditional welfare programs or MFIP and estimated the cost of
providing MFIP benefits and food stamp benefits in traditional welfare
programs; (3) Minnesota excluded all costs unique to MFIP which are paid
by the state to achieve cost neutrality; (4) Minnesota used the same
methodology to allocate allowable administrative costs among the welfare
programs; (5) Minnesota implemented the agreed-upon methodologies and
generated reasonable cost estimates for FY 1994, but a data processing
problem delayed results for 7 months; and (6) USDA overpaid Minnesota
$115,395 in excess food stamp benefits during the first 6 months of
MFIP, based on Minnesota's forecasts of what food stamp costs would have
been in the absence of MFIP.

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

 REPORTNUM:  RCED-96-54
     TITLE:  Food Stamp Program: Achieving Cost Neutrality in 
             Minnesota's Family Investment Program
      DATE:  02/12/96
   SUBJECT:  Public assistance programs
             State-administered programs
             Cost analysis
             Welfare benefits
             Overpayments
             Administrative costs
             Refunds to government
             Statistical methods
IDENTIFIER:  Minnesota Family Investment Program
             Minnesota Family General Assistance Program
             AFDC
             Aid to Families with Dependent Children Program
             Food Stamp Program
             Medicaid Program
             
******************************************************************
** 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 and the Secretaries of Agriculture
and Health and Human Services

February 1996

FOOD STAMP PROGRAM - ACHIEVING
COST NEUTRALITY IN MINNESOTA'S
FAMILY INVESTMENT PROGRAM

GAO/RCED-96-54

Cost of Minnesota's Family Investment Program

(150246)


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

  AFDC - Aid to Families with Dependent Children
  HHS - Department of Health and Human Services
  MFIP - Minnesota Family Investment Program
  USDA - U.S.  Department of Agriculture

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


B-266156

February 12, 1996

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

The Honorable Pat Roberts
Chairman
The Honorable E (Kika) de la Garza
Ranking Minority Member
Committee on Agriculture
House of Representatives

The Honorable Dan Glickman
The Secretary of Agriculture

The Honorable Donna Shalala
The Secretary of Health and Human Services

In April 1994, Minnesota began a 5-year, federally authorized welfare
reform demonstration project known as the Minnesota Family Investment
Program (MFIP).  Aimed at simplifying the welfare system, this
project consolidates the food assistance and cash benefits provided
by three programs--Aid to Families With Dependent Children (AFDC),
the Food Stamp Program, and Minnesota's Family General Assistance
Program--into a single monthly payment. 

The Food Stamp Act of 1977, as amended, requires that the federal
government spend no more for this project's food assistance component
in any fiscal year than it would have spent for the Food Stamp
Program.  That is, the demonstration project is to be cost neutral. 
As part of the process for ensuring cost neutrality, the act requires
the Secretary of Agriculture and the state of Minnesota to agree upon
methodologies for estimating what the food stamp benefit and
administrative costs would have been had there been no demonstration
project.  Furthermore, Minnesota has agreed to return to the U.S. 
Department of Agriculture (USDA) any funds it received in excess of
those that would have been paid under the traditional Food Stamp
Program.  The act also requires GAO to report periodically on whether
these cost-neutrality requirements have been met. 

In this report we (1) describe the methodologies that Minnesota
agreed to use for estimating the costs of the Food Stamp Program--for
both benefits and administration--that would have been incurred if
the demonstration project had not been implemented, (2) determine if
Minnesota implemented these methodologies, (3) assess the
reasonableness of these methodologies, as implemented, to estimate
the cost of the Food Stamp Program for fiscal year 1994, and (4)
compare the payments that would have been paid to Minnesota using the
agreed-upon methodologies with the actual payments in fiscal year
1994. 


   RESULTS IN BRIEF
------------------------------------------------------------ Letter :1

Minnesota agreed to use two different methodologies for estimating
the costs of the food stamp benefits and administration that would
have been incurred in the absence of the demonstration project.  For
food benefits, Minnesota agreed to a methodology that (1) randomly
assigns applicant families to either the traditional welfare programs
or the demonstration project and (2) uses the cost of the food stamp
benefits provided to those in the traditional welfare programs to
estimate the cost of providing similar benefits to those in the
demonstration project.  For administrative costs, Minnesota agreed to
a two-step methodology to achieve cost neutrality.  First, Minnesota
excludes all costs unique to the project, such as the project
director's salary, which are paid by the state.  Second, to allocate
project costs, Minnesota agreed to use the same methodology it
currently uses for allocating allowable administrative costs to
traditional welfare programs. 

Minnesota implemented the agreed-upon methodologies for estimating
and allocating costs for fiscal year 1994.  However, an unexpected
data-processing problem delayed by 7 months Minnesota's estimates of
the fiscal year 1994 food stamp benefit costs that would have been
incurred in the absence of the demonstration project.  Minnesota has
resolved its data-processing problem, and future estimates should be
more timely.  We believe that these methodologies as implemented
resulted in reasonable estimates for ensuring cost neutrality in
fiscal year 1994. 

USDA used Minnesota's forecasts of food stamp benefits that would
have been paid in the absence of the demonstration project as the
basis for making payments to Minnesota.  According to the
cost-neutrality estimates Minnesota developed for fiscal year 1994,
the state received $115,395 more from USDA for food stamp benefits in
the first 6 months of the demonstration project than it would have in
the absence of the project.  USDA billed Minnesota for this amount on
January 30, 1996. 


   BACKGROUND
------------------------------------------------------------ Letter :2

Minnesota's welfare reform demonstration project, authorized in
1989,\1 has multiple goals, including simplifying the welfare
system.\2 Because MFIP combines AFDC, the Food Stamp Program, and the
state's Family General Assistance Program into a single welfare
program, MFIP families receive financial assistance from only one
program, with a single set of rules and procedures. 

Program eligibility and benefit levels for MFIP differ from those of
the traditional AFDC and Food Stamp programs.  These differences can
result in families' qualifying for more benefits under MFIP than
under the traditional programs.  About 8,500 families\3 in seven
counties will receive MFIP benefits during the 5-year demonstration
project.  However, participation in MFIP is limited by the act to not
more than 6,000 families at the same time. 

The Department of Health and Human Services (HHS) is responsible for
AFDC, and USDA is responsible for the Food Stamp Program.  HHS is
taking the lead role for welfare reform demonstration projects,
including monitoring their cost neutrality. 

To conduct the demonstration project, Minnesota was granted certain
waivers to the requirements of AFDC and the Food Stamp Program by HHS
and USDA, respectively.  For example, under the demonstration
project, Minnesota can provide food stamp benefits in the form of
cash instead of coupons.  However, the act requires the state to give
MFIP families the option to receive food assistance in the form of
food stamp coupons.  Minnesota also agreed to an independent
evaluation of the demonstration project's impacts and implementation. 

Because of the cost-neutrality requirement for MFIP, Minnesota has to
pay any additional costs generated by the demonstration project. 
Cost neutrality is calculated separately for each program, and within
the food stamp portion of the demonstration project, the
cost-neutrality requirement must be met separately for benefit and
administrative costs.  Fiscal year 1994 is the first period for which
MFIP must be cost neutral.  Because MFIP began on April 1, 1994, the
cost-neutrality requirement for fiscal year 1994 covers a 6-month
period ending September 30, 1994. 


--------------------
\1 The food stamp and AFDC portions of the MFIP project are
authorized, respectively, by section 22 of the Food Stamp Act of
1977, as added by P.L.  101-202 (Dec.  6, 1989) and section 8015 of
the Omnibus Budget Reconciliation Act of 1989 (P.L.  101-239, Dec. 
19, 1989). 

\2 The other goals of MFIP are (1) helping families increase their
income, (2) helping families move toward self-sufficiency, and (3)
reducing long-term dependency on welfare as the primary source of
family income. 

\3 In fiscal year 1994, Minnesota reported a monthly average of
133,006 households receiving food stamp benefits and a monthly
average of 63,754 cases receiving assistance under AFDC. 


   ESSENTIAL AGREEMENTS WERE
   REACHED ON COST-ESTIMATING
   METHODOLOGIES TO ENSURE COST
   NEUTRALITY
------------------------------------------------------------ Letter :3

Minnesota and the federal government agreed to two different
methodologies to estimate the food stamp benefit and administrative
costs that would have been incurred if the demonstration project had
not been implemented.  The formal agreements for the demonstration
project are between HHS and Minnesota and are spelled out in a
document known as Special Terms and Conditions, which was effective
on March 31, 1994.  According to HHS, MFIP's requirements to ensure
cost neutrality are more restrictive than those of other, more recent
welfare demonstration projects. 


      DIFFERENT METHODOLOGIES WERE
      AGREED UPON FOR ESTIMATING
      BENEFIT AND ADMINISTRATIVE
      COSTS
---------------------------------------------------------- Letter :3.1

To estimate benefit costs, Minnesota agreed to use a research design
that (1) randomly assigns applicant families to a control group
eligible for traditional welfare program benefits, including food
stamps, or to a group eligible for MFIP benefits and (2) uses the
cost of food stamp benefits provided to families assigned to the
control group to estimate what the cost would have been to provide
traditional food stamp benefits to those assigned to MFIP.  To help
ensure comparability of the two groups, Minnesota is to assign
families to each group in accordance with an agreed-upon random
sampling plan.  Minnesota agreed to estimate food stamp benefit costs
by applying specific formulas to data on the cost of providing food
stamp benefits to the control group.\4 These estimates are to be
calculated annually by March 31.  Therefore, the estimate for fiscal
year 1994 was due by March 31, 1995. 

Minnesota and the federal government have not agreed upon certain
other methodologies relating to food stamp benefits that are required
by section 22 of the Food Stamp Act.  These methodologies were
intended to calculate excess costs associated with those MFIP
families who received food benefits in the form of food stamp
coupons.  While Minnesota has proposed methodologies to comply with
this requirement, in a January 3, 1994, letter, Minnesota requested
an exemption from these requirements because they provide no
additional assurance of overall cost neutrality or any additional
information needed to reconcile costs.  Minnesota officials told us
that this is because any additional costs associated with coupons are
accounted for in the methodology for estimating benefits.  As of
January 2, 1996, Minnesota had not received a written reply to its
request from either USDA or HHS.\5

For administrative costs, Minnesota agreed to use a cost allocation
plan to ensure administrative cost neutrality.  This plan involves
two steps.  First, Minnesota will identify the direct and indirect
administrative costs that are unique to the demonstration project. 
These costs, such as the MFIP director's salary, are to be excluded
from federal payments and are to be fully paid for by the state. 
Second, to allocate administrative costs that are eligible for
federal reimbursement for MFIP, Minnesota agreed to use the
methodology it currently uses for allocating allowable administrative
costs among the Food Stamp and other federal and state programs. 
This methodology uses data from the state's time study survey, which
measures the amount of time county employees spend on various federal
and state welfare programs.  The results of this survey are used to
allocate administrative costs across 14 federal and state programs. 


--------------------
\4 In fiscal year 1995, Minnesota made some adjustments to the
sampling plan that could require changes in the formulas used to
calculate the benefit estimate in future years. 

\5 In addition, section 22(d)(3) of the Food Stamp Act contemplated
that the food assistance benefits and administrative costs paid by
the federal government would include certain adjustments estimated
pursuant to an agreed-upon methodology.  According to Minnesota
officials, these adjustments were related to an earlier research
design to estimate benefit costs and are no longer applicable. 


      FORMAL AGREEMENTS ON
      METHODOLOGIES ARE WITH HHS
---------------------------------------------------------- Letter :3.2

The formal agreements on the cost-neutrality methodologies for food
stamps are between Minnesota and HHS, not between Minnesota and USDA. 
However, according to USDA officials, USDA participated in the
negotiations regarding the cost-neutrality methodologies and agreed
to them.  The formal agreements are between Minnesota and HHS because
HHS has assumed the lead role for welfare reform demonstration
projects that include its programs--AFDC and Medicaid--and USDA's
Food Stamp Program. 

HHS has taken the lead role because the main focus of these welfare
reform demonstration projects is changing AFDC.  In this role, HHS
formally negotiates the terms and conditions for those projects,
including cost-neutrality provisions, and is responsible for ensuring
that these cost-neutrality requirements are carried out.  As of
November 6, 1995, HHS had the lead responsibility for overseeing 36
approved welfare reform demonstration projects, including MFIP, that
involve USDA's Food Stamp Program.  USDA and HHS have no documented
procedures on how they are to work together on MFIP or on these other
welfare reform demonstration projects. 


      AGREEMENTS WITH MINNESOTA
      ARE MORE RESTRICTIVE THAN
      THOSE WITH OTHER STATES
---------------------------------------------------------- Letter :3.3

The cost-neutrality requirements for MFIP are more restrictive than
those for more recently approved welfare reform demonstration
projects.  According to an HHS official, while MFIP must be cost
neutral separately for benefit and administrative costs for the Food
Stamp Program and for AFDC, demonstration projects are now typically
only required to be cost neutral to the federal government overall. 
Under these latter circumstances, costs for the Food Stamp Program
and AFDC would be considered in total in determining overall federal
cost neutrality.  In addition, in September 1994, HHS formally
announced more flexible cost-neutrality principles and procedures to
be used for welfare demonstration projects.  In accordance with this
announcement, HHS generally determines the cost neutrality over the
life of a demonstration project, not annually as is required for
MFIP.  Because the cost-neutrality requirements for MFIP are mandated
by the Food Stamp Act, the recent changes in cost-neutrality policy
and practice cannot be applied to MFIP. 


   MINNESOTA IMPLEMENTED
   AGREED-UPON METHODOLOGIES AND
   GENERATED REQUIRED ESTIMATES
   FOR FISCAL YEAR 1994
------------------------------------------------------------ Letter :4

For fiscal year 1994, Minnesota implemented the agreed-upon
cost-neutrality methodologies and generated the required estimates of
food stamp benefit and administrative costs that would have been
incurred in the absence of the demonstration project. 

To estimate the cost of the food stamp benefits, Minnesota
implemented the agreed-upon approach.  This approach, which was
implemented with the assistance of a contractor experienced in the
design and implementation of such research projects,\6 included the
use of computerized procedures to randomly assign families to the
control or the MFIP group. 

To ensure proper implementation, according to Minnesota officials,
county employees involved in MFIP received training on the rationale
for, and the methods that would be used to make, the random
assignment.  The evaluation contractor and state MFIP officials
monitored the implementation of the sampling procedures to ensure
that established procedures were followed and to resolve any
unanticipated problems. 

The random assignment of families was implemented without major
problems, according to Minnesota officials.  Between April 1, 1994,
and September 30, 1994, 3,687 families were randomly assigned to the
MFIP group, and 2,796 were randomly assigned to the control group.\7

While implementing this methodology, Minnesota encountered a
data-processing problem that prevented it from obtaining needed data
on the cost of the food stamp benefits.  This unanticipated problem
occurred because existing reports generated by the state's computer
system\8 were not identifying all of the food benefit costs
associated with the control and MFIP cases.  As a result, Minnesota
could not generate these estimates by the agreed-upon date of March
31, 1995.  By October 31, 1995, Minnesota had resolved this problem
and provided the initial estimates to HHS.  The final estimates were
provided to HHS and USDA on November 29, 1995.  Minnesota computed
the estimates on the basis of the agreed-upon formulas with a minor
refinement proposed by HHS.\9 According to Minnesota officials,
future estimates are expected to be more timely. 

Minnesota also implemented the agreed-upon cost allocation
methodology to determine what the administrative costs would have
been in the absence of the MFIP demonstration project. 


--------------------
\6 Manpower Demonstration Research Corporation. 

\7 For evaluation purposes, the sample size for the MFIP group is
greater than for the control group.  This difference is taken into
account in the formulas used to calculate the estimate of benefit
costs for cost-neutrality purposes. 

\8 MAXIS is Minnesota's integrated eligibility and benefits
determination system. 

\9 An adjustment was made for a small number of individuals from
control group families who joined MFIP group families. 


   METHODS USED PROVIDED
   REASONABLE COST ESTIMATES
------------------------------------------------------------ Letter :5

The methodologies for estimating the costs of the Food Stamp Program
in the absence of the MFIP demonstration project, and the manner in
which they were implemented, resulted in reasonable estimates of
costs for fiscal year 1994.  We believe that the estimates are
reasonable for both the food benefit and administrative costs. 


      FISCAL YEAR 1994 ESTIMATE OF
      FOOD BENEFIT COST WAS BASED
      ON SOUND METHODOLOGY
---------------------------------------------------------- Letter :5.1

The estimate used to ensure the cost neutrality of the food stamp
benefits provided under MFIP was based on a sound research
methodology.  The strength of this methodology hinges on its use of a
control group selected at random to estimate the cost for the MFIP
group.  Because the random assignment was, by definition, made by
chance--as with a lottery or the toss of a coin--the two groups are
expected to be similar except for the program by which they receive
benefits.  In fact, data provided by the evaluation contractor showed
that the random assignment process was resulting in MFIP and control
groups that were similar.\10

Because the groups were similar, the cost of providing food stamp
benefits to the control group provided the best estimate of what the
cost would have been to provide food stamp benefits to the group
receiving MFIP benefits. 

Furthermore, Minnesota took reasonable steps to ensure the integrity
of the random assignment of cases to the MFIP and control groups. 
For example, the random assignment procedures were designed to
prevent "gaming" of who is assigned to the MFIP or control group. 
This was done by giving responsibility for making the random
assignment to the MFIP evaluation contractor, located in New York
City, not the county worker dealing with the potential MFIP
recipient.  In addition, the contractor used a computer program that
had been used in other evaluation studies to make such random
assignments.  Minnesota and the contractor also monitored the
integrity of the procedures by periodically comparing the
characteristics of the control and MFIP groups to make sure they were
similar. 


--------------------
\10 A comparison of the families assigned to the MFIP and control
groups found no statistically significant differences in
characteristics such as employment status, education level, previous
welfare experience, and ethnicity. 


      FISCAL YEAR 1994
      ADMINISTRATIVE COST ESTIMATE
      WAS REASONABLE
---------------------------------------------------------- Letter :5.2

The fiscal year 1994 administrative cost estimate was reasonable even
though the cost allocation approach used to calculate the estimate
had some limitations.  A cost allocation approach may not be the most
reliable method for ensuring cost neutrality because this approach
depends on the accuracy of the state's identification of
administrative costs unique to the demonstration project. 

Despite this limitation, the cost allocation procedures Minnesota
used produced a reasonable administrative cost estimate for fiscal
year 1994.  We found that the ratio of administrative to benefit
costs for MFIP was comparable to the ratio of administrative to
benefit costs for the Food Stamp Program in Minnesota in fiscal year
1994.  Table 1 shows the ratio of benefit to administrative costs for
fiscal year 1994. 



                                Table 1
                
                 Ratio of Benefits to Federal Share of
                  Administrative Costs for Food Stamp
                       Program, Fiscal Year 1994

                          Food Stamp Program      Administrative  Rati
Program                             benefits               costs     o
------------------------  ------------------  ------------------  ----
Minnesota, excluding            $236,977,553         $22,289,005    9%
 MFIP
MFIP                            $1,612,634\a            $111,610    7%
----------------------------------------------------------------------
\a Estimate of total food stamp benefits that would have been paid to
MFIP participants in the absence of the demonstration project. 


   RECONCILIATION OF COST
   ESTIMATES TO PAYMENTS HAS
   OCCURRED FOR FISCAL YEAR 1994
------------------------------------------------------------ Letter :6

From April 1, 1994, to September 30, 1994, the first 6 months of
MFIP's operations, Minnesota received $1,728,029 in payments from
USDA for MFIP food benefits, which was $115,395 more than it would
have received in the absence of the demonstration project.  Because
there was no way to accurately calculate in advance what USDA's
payment for MFIP families should be, Minnesota requested and received
payments from USDA that were based on the state's forecasts of what
the food stamp costs would have been in the absence of the
demonstration project.  Table 2 shows the difference between the
payments USDA made for food benefits to MFIP participants and
Minnesota's subsequent estimate of the comparable costs for food
benefits under the traditional Food Stamp Program. 



                                Table 2
                
                MFIP Food Benefit Payments Compared With
                Food Benefit Estimates, Fiscal Year 1994

Actual and estimated payments                        Amount of payment
----------------------------------------  ----------------------------
USDA's total food stamp payments--cash                      $1,728,029
 and coupons--to Minnesota for MFIP
 cases
Estimate of total food stamp benefits                        1,612,634
 that would have been paid in the
 absence of MFIP
Amount Minnesota owes to USDA                                 $115,395
----------------------------------------------------------------------
The difference of $115,395 is an excess cost that Minnesota must pay
back to USDA to meet the cost-neutrality requirements.  USDA billed
Minnesota for this amount on January 30, 1996.  Minnesota can choose
between paying USDA within 30 days or having its current year funding
reduced by this amount. 

The $115,395 in excess costs does not represent the total additional
food benefits provided to MFIP families in fiscal year 1994. 
Minnesota calculated that it provided $2,208,026, or $595,392 more in
food benefits for MFIP in fiscal year 1994 than it would have
provided in Food Stamp Program benefits in the absence of the
demonstration project.  However, because of the cost-neutrality
requirement, USDA's payment liability was limited to $1,612,634. 

Because Minnesota used a cost allocation method, rather than
estimating costs on the basis of a control group, it does not need to
further reconcile administrative costs for fiscal year 1994.  (For
MFIP administrative costs, Minnesota received a total of $111,610 for
the 6-month period.  Of this total, $103,202 was for indirect
administrative costs incurred by counties and $8,408 was for indirect
administrative costs incurred by the state.)


   COMMENTS FROM FEDERAL AGENCIES
   AND THE STATE OF MINNESOTA
------------------------------------------------------------ Letter :7

We provided copies of a draft of this report to the Secretaries of
Agriculture and Health and Human Services and to the Commissioner of
Minnesota's Department of Human Services for review and comment. 
USDA and Minnesota agreed with the report's findings.  Because of the
federal government furlough, HHS did not provide comments on the
draft.  However, we had previously met with the HHS official
responsible for the cost neutrality of welfare reform demonstration
projects, who agreed with the facts as presented in the report.  We
incorporated his suggestions for minor technical corrections and or
clarification as appropriate.  USDA's and Minnesota's comments appear
in appendixes I and II, respectively. 


   SCOPE AND METHODOLOGY
------------------------------------------------------------ Letter :8

In developing information for this report, we spoke with and obtained
documents from officials with USDA, HHS, and Minnesota (state and
county) and with the evaluation contractor regarding the
cost-neutrality requirements, agreements on the methodologies,
implementation of those methodologies and formulas, and repayment of
excess costs.  We also observed the implementation of MFIP procedures
in one urban and one rural county in Minnesota.  To assess the
reasonableness of the methodologies and resulting estimates, we
conducted a qualitative analysis of the methods used for estimating
the food benefits.  To assess the reasonableness of the
administrative cost estimates, we compared the ratios of
administrative costs to benefits provided for both the demonstration
project and the Food Stamp Program in the state.  We also obtained
and compared USDA's fiscal year 1994 payments to Minnesota with the
estimates of what those payments would have been in the absence of
the demonstration project. 

We conducted our work between June 1995 and January 1996 in
accordance with generally accepted government auditing standards. 
However, we did not verify the accuracy and reliability of data from
Minnesota's and USDA's financial and other management information
systems. 


---------------------------------------------------------- Letter :8.1

We are sending copies of this report to appropriate congressional
committees, interested Members of Congress, and other interested
parties.  We will also make copies available on request. 

Please contact me at (202) 512-5138 if you or your staff have any
questions.  Major contributors to this report are listed in appendix
III. 

Robert A.  Robinson
Director, Food and
 Agriculture Issues




(See figure in printed edition.)Appendix I
COMMENTS FROM THE U.S.  DEPARTMENT
OF AGRICULTURE
============================================================== Letter 




(See figure in printed edition.)Appendix II
COMMENTS FROM THE STATE OF
MINNESOTA
============================================================== Letter 


MAJOR CONTRIBUTORS TO THIS REPORT
========================================================= Appendix III

Thomas E.  Slomba, Assistant Director
Andrea Wamstad Brown, Project Leader
John F.  Mitchell
Carol Herrnstadt Shulman


*** End of document. ***