Paperwork reduction pilot projects interim report

              fi ?t'2: f n
^^^»***j *^o*
BOB
THE PAPERWORK REDUCTION PILOT PROJECTS INTERIM REPORT
EXECUTIVE SUMMARY
The Child Nutrition and WIC Reauthorization Act of 1989 (P.L. 101-147), required the
Secretary of Agriculture to carry out pilot projects to test alternatives to annual application
and daily meal counting procedures in the National School Lunch Program (NSLP). The
intent of these projects is to test ways to reduce the administrative burden on schools with a
large percentage of students from low-income families, while maintaining program integrity.
The legislation specifically prescribes three pilot projects which test alternatives to the annual
application and meal counting requirements.
The Food and Nutrition Service is conducting a four-year study to determine the effects of
the pilot projects on paperwork burden and program integrity. This report presents
preliminary findings from the first two years of the study-school year 1990-91 when baseline
measures were taken and school year 1991-92, the first year of pilot operations. The final
report will include information from the final two years of pilot operations, school years
1992-93 and 1993-94, and will draw conclusions and make recommendations.
Study Objectives
The study was designed to answer research questions relating to the amount of time and cost
saved at the local level by reducing administrative requirements. In addition, pilot procedure
effects on student participation and Federal costs were also examined. Finally, the effect of
pilot procedures on program integrity is also explored.
This study collected data from two sources. Data were collected from pilot site schools and
School Food Authorities to measure pilot procedure effects on administrative burden,
participation, and Federal costs and accountability. In addition, data were collected through
an in-home survey of households at each pilot site to assess changes in family income and
eligibility for free or reduced price meals during the course of a lengthened application cycle.
Preliminary Findings
All findings included in this interim report are preliminary and based only one year of pilot
data. The final report will include data from the baseline year as well as three years of pilot
information. In addition, the final report will address issues not included in this repot, (see
Further Evaluation section below).
No-Fee Pilot Sites
In die no-fee pilot projects, schools serve meals to all children at no charge, regardless of the
income status of the child. Schools are not required to count meals by category (free,
reduced price and full price) because they receive Federal reimbursement based on set
claiming ratios. Schools count total meals served and then apply claiming ratios to determine
how many meals should be claimed for free reimbursement, how many for reduced price
reimbursement and how many for paid reimbursement. Claiming ratios are established using
historical participation, school enrollment, econometric models or other alternatives.
£
Four of the pilot project sites are operating no-fee programs: The School District of
Philadelphia, PA; Jersey City School District in Jersey City, NJ; Alisal Union School
District in Salinas, CA; and National City School District in National City, CA.
Time Savings
Each of the no-fee pilot projects was successful in reducing paperwork associated with the
school meals programs. By eliminating or reducing the frequency of the application process,
Philadelphia, Salinas, and National City saved between IS and 24 minutes per application in
processing time. Jersey City continues to process applications for all students every year.
The four sites saved between 41 minutes and 1 hour 33 minutes per school per day due to
simplification of the meal counting process. Instead of counting meals by category (e.g.,
free, reduced price and paid), the no-fee pilot sites count only total meals served. Time is
saved by not selling tickets at the beginning of the day, collecting tickets or money during
the meal service, and by not counting the tickets by category after the meal service is over.
Cost Savings
Through actual reductions in labor and direct costs coupled with reallocated labor hours--i.e.,
labor that was shifted from administration to meal preparation or other educational tasks~no-fee
school districts saved between $.09 and $.24 per meal.
Federal Costs
Federal costs in each of the no-fee sites rose as student participation increased. Average
daily meal rates in each site rose between 7 and 21 percent.
Alternative Application Pilot Sites
San Bernadino, California; Springfield, Oregon; and Lowell, Massachusetts each
implemented alternative application procedures. San Bernadino and Springfield are extended
eligibility sites where applications are processed once every two or three years rather than
every year. At the time of application, a child's eligibility is determined and remains the
same until the next time applications are collected. San Bernadino processes applications
every other year and conducts verification in the alternate year. Springfield petitioned to
process one third of their applications each year and verify 10 percent of those approved in
that year. Lowell developed a modified application to simplify their application process and
shift part of the procedure from the fall to the spring. All alternative application sites have
continued to count meals by category.
Two of the sites, Springfield and Lowell are no longer operating under pilot procedures.
Springfield did not fully implement their proposed pilot procedures. Lowell's simplified
application actually increased the time spent on paperwork so they made the decision to try
direct certification instead.
Time Savings
San Bernadino collects applications from students every other year and conducts verification
on 10 percent of all applications on file in the alternate year. In years when applications are
u
not processed, San Bernadino saves over 6,000 hours-about 20 minutes per application. In
years when verification is not conducted, San Bernadino saves about 510 hours-over 2 hours
per application
Cost Savings
San Bernadino saves about $.05 per meal in reallocated labor and direct costs during years
when applications are not processed.
Federal Costs
Increases in meal service cannot be directly attributed to the pilot.
Direct Certification Pilot Sites
Columbus, Ohio and the State of Maine are direct certification pilot projects. Direct
certification is a simplified method to determine eligibility for free meals under the National
School Lunch and School Breakfast programs. In place of applications, school officials may
contact food stamp or Aid to Families with Dependent Children (AFDC) offices for
documentation that children are members of a food stamp household or an AFDC assistance
unit. These children are automatically eligible for free meals. Besides simplifying the initial
application process, direct certification also makes the verification process less burdensome
by reducing the number of applications that must be verified.
When the pilot projects began in 1990, direct certification was newly approved by law (P.L.
101-147). The two pilot sites were chosen to test the usefulness of direct certification-
Columbus at the local level and Maine at the State level.
Time Savings
Columbus had an estimated 3,653 hour savings-about 13 minutes per application- in
application processing from the baseline to the first year of pilot operations. Data from
Maine were not available for this interim report.
Cost Savings
As a result of labor and direct savings (from postage, printing, etc.) Columbus saved about
$.01 per meal.
Federal Cost
Federal costs are affected by direct certification because program costs increase as the
number of free eligibles eating meals increases. In Maine, for example, the number of free
meals served increased by 812,000 between 1991 and 1992. At the same time, the numbers
of reduced price and paid meals declined indicating that children shifted from the reduced
price and paid categories into the free category-perhaps as a result of direct certification.
While participation effects resulting from direct certification cannot be separated from those
caused by shifts in the economy or other factors, there is some evidence that Federal costs in
the direct certification sites rose as a result of their pilot procedures. Both Columbus and the
State of Maine experienced much higher than average growth in the number of free eligibles
HI
between the baseline year (1991) and the first year of pilot operations (1992). Columbus
experienced a 23 percent increase in their number of free eligibles and Maine experienced a
35 percent increase while the national average growth in free eligibility was only 11 percent.
Provision 1 and 2 Pilot Sites
Atlantic City, NJ; Milford, MA; and Terrell County, GA were chosen to test alternatives to
Provisions 1 and 2-two alternative application counting and claiming processes that have
been allowed by law since 1977. As in the no-fee sites, meals were served to all children at
no charge and claimed using claiming percentages.
Atlantic City and Milford both ended their no-fee meal programs due to district costs coupled
with changes in personnel. Terrell County still runs a no-fee program but they claim meals
using current Provision 2 procedures rather pilot Provision 2 procedures~i.e., they use
participation percentages rather than enrollment percentages to prepare claims for
reimbursement.
In-Home Survey
The majority of the households, 91 percent, interviewed in 1991 and again in 1992 retained
their eligibility to receive free lunches. Households who were not eligible to receive free
lunches in 1991 did have notable increases, in benefits in 1992 and demonstrate a potential
impact on program integrity. However, these results should be interpreted carefully because:
1) the samples were small; and 2) the data were collected only once within each year, which
makes it is difficult to truly assess household income fluctuations.
Further Evaluation
Since the study's inception, a number of issues have arisen that should be addressed to
present a more complete picture of how the pilot procedures have affected district school
food service operations and program integrity.
In the no-fee sites, the final report will further examine the impact of no-fee procedures on
breakfast program finances and student participation. In addition, the final report will
present a more in-depth discussion of the impact of no-fee procedures on district finances.
Finally, the final report will present a comparison of enrollment-based versus participation-based
claiming percentages and will examine how well claiming percentage models match
actual meal service.
In the alternative application sites further evaluation will focus on two issues: 1) the impact
of direct certification on paperwork savings associated with extended applications; and 2)
whether applications for meal benefits accumulate over time resulting in service of free and
reduced price meals to children who may no longer be eligible.
Future evaluation of the direct certification pilot projects will focus on efficiencies gained
through experience with the direct certification process. For example, the question of
whether the incidence of duplicate applications is reduced from one year to the next will be
examined.
IV
INTRODUCTION
The Child Nutrition and WIC Reauthorization Act of 1989 (P.L. 101-147), required the
Secretary of Agriculture to carry out pilot projects to test alternatives to annual application
and daily meal counting procedures in the National School Lunch Program (NSLP). The
intent of these projects is to test ways to reduce the administrative burden on schools with a
large percentage of students from low-income families, while maintaining program integrity.
The legislation specifically prescribes three pilot projects which test alternatives to the annual
application and meal counting requirements.
The Food and Nutrition Service is conducting a four-year study to determine the effects of
the pilot projects on paperwork burden and program integrity. Twelve school districts were
selected to participate in the study. Of the twelve participating, three tested alternatives to
Special Assistance Provision 1 and 2 application and meal counting requirements and the
remainder tested other alternatives designed by the districts themselves. This report presents
findings from the first two years of the study-school year 1990-91 when baseline measures
were taken and school year 1991-92, the first year of pilot operations. The final report will
also include information from the final two years of pilot operations, school years 1992-93
and 1993-94.
This report is organized into seven chapters. Chapter 1 reviews background information on
the NSLP and presents the study objectives. Chapter 2 presents preliminary findings on the
no-fee pilot projects. Chapter 3 presents preliminary findings from the alternate application
pilot projects. Chapter 4 presents preliminary findings from the direct certification pilot
projects. Chapter 5 presents information from the pilot Provision 1 and 2 projects. Chapter
6 presents preliminary findings from the in-home survey portion of the study. Finally,
Chapter 7 summarizes the information presented in the preceding chapters.
CHAPTER 1
PROGRAM AND STUDY BACKGROUND
The Food and Nutrition Service (FNS), U.S. Department of Agriculture (USDA), is
responsible for the administration of five child nutrition programs, including the National
School Lunch Program (NSLP). The NSLP is the oldest and the largest of the child nutrition
programs. Since 1946, the NSLP has made it possible for schools to serve nutritious lunches
to students each school day. Annual Federal expenditures to States have increased from less
than $100 million in 1946 to approximately $4.7 billion in 1993. Approximately 42 million
children in 93,000 schools (grades K-12) located in more than 20,000 school districts across
the country participate in the NSLP. On an average day, about 25 million children
participate in the program.
NSLP Legislative History
The National School Lunch Act was first enacted in 1946 to "...safeguard the health and
well-being of the Nation's children and to encourage the domestic consumption of nutritious
agricultural commodities and other food, by assisting the States, through grants-in-aid and
other means, in providing an adequate supply of food and other facilities for the
establishment, maintenance, operation, and expansion of nonprofit school lunch programs."
The Act, as amended, authorizes cash payments to assist schools in making local purchases
of food and the distribution of commodities.
The Act directs local school authorities to serve lunches without cost or at a reduced price to
those children determined by school food authorities as being unable to pay the full cost.
Further, there is to be no physical segregation or discrimination against any child unable to
pay for his or her lunch. The schools are to utilize, insofar as practicable, surplus
commodities in their lunch programs.
Legislation during the 1970s focused on expanding the coverage of some programs,
improving the quality of the meals served and the dissemination of nutrition education
information. In addition, legislation was passed to reduce paperwork in low-income school
districts through the Special Assistance Certification and Reimbursement alternatives,
commonly known as Provision 1 and Provision 2. Later legislation, such as the Omnibus
Budget Reconciliation Act of 1981 (P.L. 97-35), was concerned with targeting benefits to
those most in need while improving program management and accountability.
Program Administration
FNS of the USDA implements the Program authorizing legislation and is generally
responsible for establishing regulations, policies and guidelines; monitoring program
performance; and providing program and administrative funds to the States. There are seven
FNS Regional Offices that primarily monitor and provide technical assistance to the State
agencies.
Within each State, responsibility for the administration of school nutrition programs usually
rests with the State education agency. State administering agencies enter into an agreement
with FNS that outlines the requirements for participation. The State agencies provide
technical assistance to local school districts and monitor program performance. State
administering agencies also collect and summarize data on the number of meals served by
each participating school food authority (SFA) and report the number of meals served, by
type, to the Food and Nutrition Service.
Local SFAs, typically school districts, enter into an agreement with the State agency that
outlines the requirements for program participation. At the local level, district staff
administer the program in the schools they supervise. Schools participating in the school
lunch and/or breakfast programs are required to prepare and serve meals according to USDA
established meal patterns. Schools are responsible for approving applications for free and
reduced price meals and verifying a sample of those applications. Schools count meals
served each day by eligibility status (free, reduced price or full price), SFA submits monthly
claims for reimbursement to the State agency.
Generally, any public school or nonprofit, private school of high school grade or under is
eligible to participate in the school nutrition programs. Public and licensed, nonprofit,
private residential child care institutions such as orphanages, homes for retarded children,
and temporary shelters for runaway children are also eligible.
Program Operations
Certification process for free and reduced price meals
While all children receive some Federal subsidies in cash and commodities, the USDA
establishes income eligibility criteria for needy children to receive free or reduced price
meals in the School Nutrition Programs. All income eligibility guidelines are derived from
the official Federal poverty guidelines and are updated annually. Family income and
household size jointly determine free and reduced price eligibility. Children from families
with incomes 130 percent of the official U.S. government poverty level or less ($18,655 for
a family of four in the 1994 school year) are eligible to receive free meals and those with
family incomes 185 percent of the poverty level or less ($26,548 for a family of four in the
1994 school year) are eligible to receive meals at a reduced price.
Generally, determination of benefits is made at the start of each .school year. On or about
the beginning of the school year, school food authorities must publicly notify the parents of
all children enrolled in school of the availability of free and reduced price lunch benefits.
Households apply for free or reduced price meals by completing an application provided by
the SFA. The application calls for households to list the names of all persons in the
household, the social security number of an adult household member, and the household's
income by source (i.e. earnings from work, welfare or unemployment payments, child
support or alimony, pensions, retirement, social security), unless an AFDC or food stamp
case number is provided in lieu of income information and the signature of an adult
household member.
In addition to an initial determination of benefits, SFAs are required to verify the eligibility
of a sample of students receiving free or reduced price lunches. If SFAs choose to verify a
random sample, they must verify at least 3 % of applications on file. SFAs may also choose
to verify a 1 % sample of those applications within $100 of the upper limit of free or reduced
price eligibility plus one half of one percent of households that provided a food stamp case
number in lieu of income information. Some smaller SFAs choose to verify all applications.
The verification process is required to be completed by December 15 of each year.
The total administrative responsibilities in the certification process at the school or SFA level
consist of: training for staff, advertisement/notification of benefits, eligibility determination,
roster creation, and verification of a sample of approved applications.
In FY 1993 nearly 55 percent of all lunches were served free or at a reduced price. Overall
average daily student participation rates vary with the costs to students with rates of 44
percent for students receiving full price meals, 71 percent for students receiving reduced
price meals, and 79 percent for students receiving free meals.1
Financial reimbursement
FNS pays States for each school lunch based on a national average payment factor multiplied
by the number of meals of each type served. All lunches earn a base level of
reimbursement. Free and reduced price meals earn an added benefit over and above the base
level. The payment factor is prescribed annually by the Secretary of Agriculture to cover
cost increases due to inflation. In the 1993-94 school year, the level of payment is $. 17 per
lunch for fully paid meals, $1.33 for reduced price meals and $1.73 for free meals. SFAS
that serve 60 percent or more of their meals free or at a reduced price receive an additional
$.02 per meal.
SFAs are offered an amount of commodities based on the number of lunches they serve, plus
as much of specified types of surplus commodities as they can use without waste. States
earn $.14 worth of entitlement commodities for each lunch served in the 1993-94 school
year.
All Federal meal payments are performance based. That is, the monthly claim for
reimbursement is limited to the number of free, reduced price and paid lunches actually
served to children eligible for such benefits for that month. Program regulations require
schools to take daily meal counts, at the point of service, which identify the number of free,
1 Burghardt, J., Gordon, A., Chapman, N., Gleason, P., Fraker, T. The School
Nutrition Dietary Assessment Study: School Food Service. Meals Offered and Dietary
Intakes (Mathematica Policy Research: Princeton, N.J., 1993). p. 133.
reduced price and paid reimbursable meals served. Meal counting procedures may not
overtly identify free and reduced price students.
School participation requirements
All public or private nonprofit schools are eligible to receive federal reimbursement for the
school nutrition programs if they:
Serve meals which meet the nutritional standards established by USDA;
Supply meals free or at a reduced price to all children determined to be eligible;
Do not discriminate against or overtly identify any child because of inability to pay
the full price of a meal;
Operate the food service on a nonprofit basis;
Operate the food service for all students without regard to race, color, national origin,
sex, age or handicap.
Special Assistance Certification and Reimbursement Alternatives
Public Law 96-166, enacted November 10, 1977, amended the National School Lunch Act to
authorize special assistance and certification procedures to reduce paperwork in the school
lunch and breakfast programs. These alternatives, commonly referred to as Provision 1 and
Provision 2, allow schools to reduce the annual certification and public notification
requirements of the school lunch program. Provision 2 also allows participating schools to
base claims for reimbursement on claiming percentages rather than a daily counting of meals
by eligibility category.
Under Provision 1, an SFA or a school having at least 80 percent of its enrolled children
determined eligible for free or reduced price meals may authorize the school to reduce annual
certification and public notification for those children eligible for free and reduced price
meals to once every two consecutive school years. Schools participating under Provision 1
must continue to record the number of free, reduced price and paid meals served daily as the
basis for calculating claims for Federal reimbursement. There were approximately 49
schools in 10 SFAs operating under Provision 1 in 1990 (the first year of this study).
Under Provision 2, an SFA or a school which serves free meals to all enrolled children in
that school may certify children for free and reduced price meals for up to three consecutive
school years. Meal counting procedures are also altered to reduce administrative burden.
During the first year, schools must take daily counts of the number of meals served by
eligibility category. These meal counts are converted into percentages of total meals served
each month (i.e. the number of free meals served in January is divided by the total number
of meals served that month). Reimbursement in the second and third years is calculated by
applying these monthly percentages to total meals counts for the corresponding months.
Thus, the daily recording of meals served by eligibility category is not required in the second
and third years, but schools must continue to count the total number of meals served each
month. Provision 2 schools must pay with funds other than Federal subsidies for free meals
served to reduced price and paid students. The school continues to receive reimbursement
for free, reduced price and paid students and must therefore base claims for reimbursement
on meal counts by eligibility category. There were approximately 358 schools in 148 SFAs
operating under Provision 2 in 1990.
Study Rationale
NSLP regulations impose requirements on school districts, food service, and local school
officials aimed at maintaining program integrity. Two areas are focused upon in this siudy:
the requirement to take annual applications to determine eligibility category (i.e., paperwork
burden) and the requirement to count lunches by category (i.e., program integrity). FNS is
testing ways to improve application and counting processes while maintaining program
integrity and imposing the least burden possible on the school district and school food service
workers and program participants.
Both the annual application requirements and the daily lunch counts are designed to ensure
that Federal funds are used for the purpose intended i.e., to provide children eligible for
free, reduced price, and paid meals a meal meeting the prescribed meal pattern. However,
these requirements for program integrity increase the administrative burden at the local level.
The need to balance paperwork burden and program integrity has long been recognized.
Special assistance Provisions 1 and 2 were enacted in 1977 to reduce paperwork in schools
with high application processing burden ~ those serving primarily low-income children. The
pilot projects authorized by P.L. 101-147 will allow FNS to assess these special assistance
provisions and several other methods to reduce paperwork in the areas of application
processing and meal counting.
Study Objectives
Much of the paperwork and administrative burden currently imposed on school officials is
necessary in order to provide reimbursement based on actual counts of meals served by
category. This study examines ways to reduce this burden and analyzes the effects such
reduction will have on program integrity.
The study has three main objectives. The first objective is to measure the reduction in
paperwork and administrative burden brought about by the pilot procedures. The second
objective is to profile changes to program operations in the school districts such as increases
in participation as a result of the pilot procedures. The third objective is to assess the impact
on local revenues, Federal costs, and the distribution of program benefits (program integrity)
in the pilot projects.
The need to administer Federal programs in an accountable manner necessitates certain
administrative procedures and recordkeeping, much of it at a local level. These procedures
result in more effective use of Federal funds and Federal tax dollars. As paperwork is
reduced, program integrity may decline and Federal expenditures increase, although local tax
dollars are saved. These demonstration projects should advance FNS' knowledge of the
trade-offs between local costs and administrative burdens and Federal program integrity
requirements.
The study is intended to answer the following questions:
• What is the reduction in administrative burden associated with alternate
application and certification procedures?
• What is the reduction in administrative burden associated with alternate meal
counting procedures?
• What is the reduction in administrative burden associated with alternate income
verification procedures caused by the pilot application procedures?
• In the pilot sites, do family size or income changes affect eligibility for free
and reduced price meals from year to year?
• What is the impact on local revenues?
• How are Federal program costs affected by alternate application and meal
counting procedures?
This study is being conducted over a four-year period (1990-1994) in 12 sites that
volunteered to participate in the pilot program. Abt Associates, a consulting firm located in
Massachusetts, conducted the first phase of data collection during the 1990-1991 school year.
The study was then awarded to Atlantic Resources Corporation to complete the data
collection and perform the analysis.
The study uses a pretest-posttest design in 8 of the 12 case studies: San Bernardino, CA;
Salinas, CA; Springfield, OR; Columbus, OH; Jersey City, NJ; Philadelphia, PA; National
City, CA and Lowell, MA. Data were collected in these sites in the first year of the study,
SY 1990-1991, to determine the level of baseline operations. The pilot procedures were
implemented in SY 1991-1992. Data were collected in SY 1991-1992 and again in SY 1992-
1993 to measure the effects of the alternate procedures. Data will be collected again in SY
1993-1994, the final year of the study, to further assess the effects of the alternate pilot
procedures. The data for all three years will be compared in order to assess changes that
occurred from year to year at each pilot site.
The remaining four pilot sites intended to begin pilot operations during the first year of this
study. These are Terrell County, GA; Atlantic City, NJ; Milford, ME, and the State of
Maine. For these sites, changes cannot be evaluated from baseline application and meal
counting procedures. Instead, comparisons will be made from year to year of the study to
evaluate how the pilot procedures have affected each site and to assess the impact of the
procedures that were implemented. During the first two years of this study, eight sites
successfully implemented pilot procedures ~ Jersey City, Salinas, Philadelphia, National
City, Columbus, and the State of Maine) Two sites, Springfield and Terrell County,
partially implemented pilot procedures. Atlantic City and Milford dropped out of the study.
This study collected data from two sources. Data were collected from pilot site schools and
SFAs to measure pilot procedure effects on administrative burden, participation, and Federal
costs and accountability. In addition, data were collected through an In-Home Survey of
households at each pilot site to assess changes in eligibility status during the course of a
lengthened application cycle.
Study Limitations
This interim study has a number of limitations that must be considered when interpreting the
findings. The study provides a preliminary analysis of 12 case studies that are not
representative of all schools or SFAs in the country. The information collected cannot be
used to make predictions or generalizations about schools throughout the Nation. In
addition, data were collected from a small number of sites, some sites could not provide all
the needed data, and there were some differences in how information was provided across
the sites.
Furthermore, it is not possible to isolate the effects of the implementation of the pilot
procedures from other changes that occurred during the demonstration period. Changes at
these pilot sites cannot be attributed solely to pilot procedures. For example, the economic
condition of the Nation must be considered when evaluating the results from this study. Any
increase in student participation in the NSLP pilot projects may be the result of the
worsening economic conditions of the country ~ increases in student participation cannot be
solely attributed to the pilot procedures. Finally, In-Home Survey data were not collected
from all twelve pilot sites and the sites where the In-Home Survey was conducted did not
have high response rates in the second year of data collection, SY 1991-1992.
Pilot Project Descriptions
Four types of pilot projects are included in the study: No-Fee, Direct Certification,
Alternative Application, and Pilot Provision 1 and 2. Table 1.1 presents a profile of
procedures by pilot site.
8
No-Fee
In a no-fee program, schools serve meals to all students without charge, regardless of the income
status of the child. Total meal counts are recorded daily, instead of meal counts by eligibility
category (i.e., free, reduced price, and paid). Claiming percentages are developed that reflect
the numbers of free, reduced price, and paid eligible children. These claiming percentages are
applied to total daily meal counts to calculate eligibility claims. Jersey City, NJ; Salinas, CA;
Philadelphia, PA; and National City, CA are operating no-fee programs.
Direct Certification
Direct Certification is a simplified method of determining eligibility for free meals under the
NSLP. In place of applications, school officials may contact food stamp or Aid to Families with
Dependent Children (AFDC) offices for documentation that children are members of a food
stamp household or AFDC assistance units. The children are automatically eligible for free
meals. The Direct Certification pilot projects include Columbus, OH and the State of Maine.
Alternative Application
Extended Eligibility
In an Extended Eligibility program, applications are processed once every two or three years.
At the time of application, a child's eligibility is determined and remains the same until the next
time applications are collected. The SFA continues to collect applications from new students
as they enroll in the school. The Extended Eligibility pilot projects are Springfield, OR, and
San Bernardino, CA.
Modified Application
Under the Modified Application pilot project, the SFA uses a simplified version of the original
application. The modified application is sent to students who were categorically eligible in the
previous school year. Lowell, MA, is using Modified Application.
Pilot Provision 1 and Pilot Provision 2
Provision 1
Pilot Provision 1 procedures extend application processing intervals to every third year (instead
of every two years), using enrollment-based claiming percentages, and counting total daily meal
counts. Enrollment-based claiming percentages are calculated by dividing the total number of
children approved for free and reduced price lunches by total enrollment. A No-Fee program
is optional. Only schools with more than 80 percent of students eligible for free or reduced
price meals may implement Provision 1 procedures. The Atlantic City, NJ, pilot site attempted
to implement Pilot Provision 1 procedures.
Provision 2
Provision 2 procedures require schools to implement a no-fee program. To calculate claiming
percentages applications are processed once every three years and used to compute
enrollment-based claiming percentages (as in Pilot Provision 1). In the second and third years,
the SFA may collect applications from new and previously non-approved students. Based on
these new applications, and on those updated from previously approved applications and changes
in enrollment, the SFA may adjust the claiming percentage annually. The test procedures differ
for Provision 2 because, for Pilot Provision 2, SFAs use claiming percentages in all three years.
Milford, ME, and Terrell County, GA, pilot sites attempted to implement Pilot Provision 2
procedures.
10
Table 1.1
SUMMARY OF PILOT PROJECT PROCEDURES3
No-Fee Projects
Jersey City,
New Jersey
Salinas, California:
Alisal Union
School District
The Jersey City pilot project includes providing a No-Fee meal program.
There has been no change in the application or verification processes. The
Jersey City School District develops enrollment-based claiming percentages
each year. Total meal counts are taken daily. Claiming percentages are
applied to total meal counts to calculate reimbursement claims.
The Alisal Union School District, located in East Salinas, is providing a
No-Fee meal program and is collecting and verifying applications once
eery three years. Enrollment-based claiming percentages are developed
be H.J on applications and total enrollment. Total meal counts are taken
daily.
Philadelphia,
Pennsylvania
The Philadelphia School District is providing a No-Fee meal program in 144
of its 272 schools. The school district combines direct certification of
children from food stamp, AFDC, and foster care rolls with data provided
from a socio-economic study to determine free and reduced price eligibility
and develop claiming percentages. Schools take total daily meal counts and|
multiply the total meal counts by the claiming percentages to determine
reimbursement claims. The application and verification processes have
been eliminated.
National City,
California:
National School
District
The National School District is providing a No-Fee meal program. Total
daily meal counts are taken daily. In the 1991-1992 school year claiming
percentages were developed based on the average number of free, reduced
price, and paid lunches served in the 1990-1991 school year. The district hai
developed a statistical model to calculate claiming percentages for the
1992-1993 and 1993-1994 school years. Claiming percentages are applied to
total meal counts to calculate claims. The application and verification
processes have been eliminated.
a.i!J IJ»: iui.
BregEiSEn
...«;r.^..:Iii.~::;:t.,.::..: .!-■:..
n-r"gil—4 '-■■ ■ - :,--:::- ---.■
Columbus, Ohio
The Columbus School District is performing district-wide Direct
Certification of children from families which receive food stamps or AFDC
benefits. There has been no change in meal counting procedures.
State of Maine
The Maine State Department of Education is conducting Direct Certification
from food stamp and AFDC rosters for students in the entire State of Maine.
Lists of free and reduced price students are compiled at the State office in
August of each year and sent to the appropriate School Food Authority. All
other students receive regular applications at the start of the school year in
September.
/OA
(Continued)
SUMMARY OF PILOT PROJECT PROCEDURES
Alternative Application Projects
Springfield,
Oregon
The original pilot procedures for the Springfield School District pilot project
included collecting applications from one third of the school population
annually. The application would be valid for three years. Each year
Springfield would verify 10 percent of all applications on file. Pilot
procedures have been partially implemented. The district verifies 10 percent
of all applications on file. There has been no change in meal counting
procedures.
San Bernardino,
California
The San Bernardino School District pilot procedures include collecting
applications once every two years and verifying applications over an
extended period (3 months) in alternate years when applications are not
processed. There has been no change in meal counting procedures.
Lowell,
Massachusetts
The Lowell School District is collecting a simplified NSLP application from
students who were categorically eligible in the previous year. Applications
are distributed and processed in the fall for all students not approved through
the simplified application process. There has been no change in verification
or meal counting procedures.
sajaafwg-iiiaaiirtiiiiSsajB
rovision
iiBiiaiii.i-i ii:i»ii.ifriiiiii'i ...-.' -.... ... r „^-.j-.l ir-_.ai —,_—^^
Atlantic City,
New Jersey
Pilot Provision 1 procedures include processing applications once every three
years, developing enrollment-based claiming percentages, and counting total
daily meal counts. Schools must have at least 80 percent of enrolled children
eligible for free or reduced price meals to participate. Offering a No-Fee meal
program is an option. However, the Atlantic City pilot site has dropped out
of the study.
Milford,
Maine
Pilot Provision 2 procedures include processing applications once every three
years, developing enrollment-based claiming percentages, and counting total
daily meal counts. A No-Fee meal program is required. The Milford pilot
site is no longer participating in this study as a Pilot Provision 2 project and is
no longer participating as a Provision 2 site. Milford is currently using typical
application, verification, and meal counting procedures.
Terrell County,
Georgia
Pilot Provision 2 procedures include processing applications once every three
years, developing enrollment-based claiming percentages, and counting total
daily meal counts. A No-Fee meal program is required. The Terrell County
pilot site did not implement the Pilot Provision 2 procedures.
Participation-based claiming percentages have been used to calculate claims
for reimbursement instead of the intended enrollment-based claiming
percentages. The Terrell County pilot site is currently operating under
Provision 2 procedures.
a) Some pilot sites have not implemented all or part of their pilot project procedures. These pilot sites include Springfield,
Atlantic City, Milford, and Terrell County. ... a
BLANK
PAGE
Table 1.2
PROFILE OF PILOT PROJECTS
(Figures Represent SY1990-1991 Unless Otherwise Noted)
Location of
Pilot Projects
Type of
Area
1990
Population*
School District
Number
of Schools
School
Enrollment
Pilot Projects
Number of
Schools
Implementing
Pilot Projects
Enrollment of
Schools in Pilot
Projects
Percent of Students in Pilot Projects
Eligible for:
Free Lunches
: I .■_:■; Jril'-t 1<MT I :: '>'-.. '■ WVP 1='" H:-. ■' "'■
Reduced Price
Lunches
Pilot School Site
School at Which
Data Were
Collected
Jersey City, NJ
Salinas, CA
Urban
Urban
228,537 37
108,777
28,000
5,281
15 10,534
5,281
.
■' '•' i1''—- '■■.'?—— kj—.
83.3%
77.0
5.7%
13.9
School #5
Elementary School
Virginia Rocca Barton
Elementary School
Philadelphia,
PA»>
Urban 1,585,577 272 196,451 144 108,525 88.3 4.7 George W. Childs
Elementary School
National City,
CA Suburban 54,249 10 6,410 10 6,410 63.5 14.5 New Horizons
Elementary School
iii-;.!L-;::i..'r:.i,-t . H -Xl.rtli .:'=&:}ixl.iirI'.*:.':. . . .'..- :.,.:■ ■ :::.-r:r.: .-»:::.•:,■ .;■■;■. •....:.■-..
;:|f.||..^V.ra^.ilpWjlil:,.Jj{.iif;iri. tf^ltfw •■l,-"-11'rl'»fi' ^r-^'r^'iiinii- n—f-'Vl^'--ir'-!;T'-:''- rp' 1-•!■" ^'-T' -H l';- f ■—r'l . ■■■-•}, r'l i ;r-r ' - Hi- I ]V ti ,.--.-■:-.—■ , ■■,'■ • i ' : —I" -i - ■-
Columbus, OH Urban 632,910 141 63,786 141 63,786 41.8 7.3 Everett Middle School
State of Maine State 1,227,928 693 215,149 693 215,149 19.4 6.1 N.A.
.ii:i;;,;:i;!:?Ui^:,i.!Ju;'i!.L.;;,:";; ,L!..^;:i:iiir;,i^;iiji^Lii^.;:.::;.;'^';:\L";::.i:rj;.:"!!:^;-;. ;:;;'■;: :?,:r.
i-xl-::JWI;«l.s.!i:;u;::ut.-;.:!J.;. •;;; ;. ts.ii --,iian4I^4fiH;,:»«,:Bi-»m1til-:f.i«,;:. -i ::■<■■■■■■:
sai-i
Springfield, OR Suburban 44,683 21 9^11 1 280 71.4 7.1 Brattain Elementary
School
San Bernardino,
CA
Urban 164,164 55 40,973 55 40,973 41.6 4.8 Jefferson Hunt
Elementary School
Lowell, MA? Urban 103,439 29 13,414 29 13,414 43.3 6.7 Washington
Elementary School
jipH(tW>.;ui.ii«..,.[.>r-»y,Bt;Hi,li..i.»a
■ ■•:lm:':'...1li-Ji.: inflni.ll!i«-nir. < Mis *.-.:iV- M. ■ i'r v...-:.-. ■: .
-»> .• -. bS Sal ! i: - , -. ... -.
JK-ir.l',i'.:"r:!"-tV;l' ;':^V\^V°,^:'uL!i-;n; f^!f^"^^..^;^- • r'.:;5l-'i::r';-''':,;;;-
,.: f .. .: : '-:i(.": : : ..
Atlantic City,
NJ Urban 37,986 11 6,320 1 395
Milford, ME Rural 2,278 955 255
62.0b
39.2
10.4"
15.7
Chelsea Junior
High School
Helen S. Dunn
Elementary School
Terrell County,
GA Rural 10,653 1,880 460 78.9 12.0 Lilly Cooper Primary
School
a) Source: 1990 Census of Population
b) SY 1991-1992 Data
c) SY 1989-1990 Data
/c!Cs
CHAPTER 2
NO FEE PILOT PROJECTS
In the no-fee pilot projects, schools serve meals to all children at no charge, regardless of the
income status of the child. Schools are not required to count meals by category (free, reduced
price and full price) because they receive Federal reimbursement based on set claiming ratios.
Schools count total meals served and then apply claiming ratios to determine how many meals
should be claimed for free reimbursement, how many for reduced price reimbursement and how
many for paid reimbursement. Claiming ratios are established using historical participation,
school enrollment, econometric models or other alternatives.
Because a claiming ratio does not allow schools to identify which children should be charged for
a reduced or full price meal, no money is collected and school districts must somehow absorb
the loss of student payments. This restriction makes the no-fee option most viable in schools
serving a high proportion of low-income students.
Four of the pilot project sites are operating no-fee programs: The School District of
Philadelphia, PA; Jersey City School District in Jersey City, NJ; Alisal Union School District
in Salinas, CA; and National City School District in National City, CA. Each of these districts
serves meals in some or all of their schools at no charge to all students, takes total daily meal
counts and develops their claim for Federal reimbursement based on set claiming ratios. The
pilot procedures in each site differ principally in how their claiming percentages have been
developed.
TABLE 2.1
NO-FEE PILOT PROJECT SITES
Characteristics of the School Districts in the 1991-92 School Year
No-fee Pilot Site Total Schools Number of
Schools in the
Pilot
Number of
Children in the
Pilot
Percent Free
Eligible in Pilot
Percent
Reduced Price
Eligible in Pilot
Philadelphia, PA 272 144 108,525 88% 5%
Jersey City, NJ 37 15 10,534 83% 6%
Salinas, CA 7 7 5,281 77% 14%
National City, CA 10 10 6,410 64% 15%
Three of the pilot sites (Philadelphia, Salinas, and Jersey City) are operating using claiming
percentages based on enrollment in school and one is operating using claiming percentages based
on actual participation. Enrollment percentages are calculated by determining the eligibility
status of all children in a school and converting the numbers into percentages. If a school of
100 students has 75 students eligible for free meals, then 75 percent of each day's meal count
11
will be claimed as free. An enrollment-based claiming percentage assumes that children in all
income categories eat school lunches at about the same rates because of the zero price charged.
Two of the sites (Salinas and Jersey City) using enrollment-based claiming percentages collect
applications to determine the number of children in school eligible for free and reduced price
meals. Jersey City continues to collect, review, approve, and verify applications annually in
every school. Schools that certify over 80 percent of students for free or reduced price lunches
are included as pilot schools. In the 1991-92 school year, 15 of 31 K-8 schools participated in
the no-fee program. In 1992-92, 8 elementary schools were added for a total of 23
participating. None of the senior high schools have participated in the pilot because they could
not collect enough applications to meet the 80 percent threshold.
The Alisal Union School District (Salinas, CA) collects, approves and verifies applications once
every three years. The district consists of seven elementary schools and pilot procedures have
been implemented in all schools. Based on applications collected in 1990-91, approximately 77
percent of children attending school in Alisal are eligible for free meals and 14 percent are
eligible to receive reduced price meals.
The third site using enrollment-based claiming percentages (Philadelphia) designed a socio-economic
study to determine claiming percentages rather than using applications. The study,
conducted by researchers from Temple University, combined information on food stamp and Aid
to Families with Dependent Children (AFDC) households with in-home interviews. The district
used direct certification to determine how many children in each school would be eligible for
free meals based on receipt of food stamps or AFDC. Researchers then interviewed 2,500
households not eligible for food stamps/AFDC to determine what percentage of those households
would be eligible for free or reduced price meals. The information from the two sources was
combined to create enrollment-based claiming percentages for each school in the district.
Philadelphia implemented pilot procedures in all schools in the district with 70 percent or more
students eligible for free meals.
The fourth no-fee site (National City) is using participation-based claiming percentages to
calculate claims for Federal reimbursement. Participation-based claiming percentages are
developed based on the actual number of meals served to free, reduced price and full price
eligible children. Throughout the first year of the pilot (1991-92), National City based their
claiming percentages on the actual number of meals served to free, reduced price and full price
eligible children in the prior year. During that year, however, the district hired a statistician
from the University of San Diego to create a linear regression model that estimates and annually
updates participation percentages. The model is based on ten years of historic claiming patterns
and includes economic and demographic variables such as regional unemployment rates, gross
regional product and race/ethnicity to make annual updates. The district began using the model
to calculate claims in October, 1992.
12
Paperwork Reduction
Time Savings
No-fee programs have the potential to significantly reduce program paperwork by eliminating
all or part of the application and meal counting and claiming processes. Three of the four pilot
projects reduced their program paperwork by changing the application process. Philadelphia and
National City eliminated applications and Salinas reduced application collection to once every
three years.
TABLE 2.2
APPLICATION PROCESSING AND VERIFICATION TIME SAVED
IN NO-FEE PILOT SITES FROM BASELINE TO YEAR 1
No-Fee Pilot Site
Philadelphia, PA
Total Hours Saved in
Application Processing
and Verification
20,480
Number of Free and
Reduced Price
Applications on File
83,641
Time Savings per
Application Not
Processed
15 min
Jersey City, NJ 9,468
Salinas, CA 1,906 4,799 24 min
National City, CA 1,131 4,977 14 min
Philadelphia, Salinas, and National City saved between 14 and 24 minutes per application in
processing time. This includes the time it takes to distribute, collect, transport, review, enter
information into a computer file and file the application. It also includes the time it takes to
verify a sample of applications on file. The 20,480 hours saved in Philadelphia represent over
10 staff years - about 1 staff year for every 14 schools in the pilot program.
Salinas had the highest processing time per application and so achieves the greatest savings by
eliminating the application process. The Alisal School District in Salinas has a very high
population of Spanish-speaking migrant farm workers. The school food service director annually
hired bilingual translators to help parents read, understand, and complete the application for meal
benefits. In addition, the application and verification processes in Salinas are complicated by
the nature of migrant work-families move often and may be difficult to locate.
Jersey City shows no savings in the application and verification processes because they continue
to collect applications annually.
In addition to simplifying the application process, no-fee pilot procedures eliminate many staff
duties related to meal counting. Under a no-fee system only the total number of meals is
counted on a daily basis. Under standard procedures, food service workers must use a ticket,
roster or automated system to record meals served by free, reduced price and full price
_
13
categories. Generally food service workers must then spend time after the meal service tallying
categorical counts.
TABLE 2.3
MEAL COUNTING TIME SAVED IN NO-FEE PILOT SITES
- "
No-Fee Pilot Site Total Hours Saved by
Eliminating Meal
Counting by Category
Number of Pilot Schools Time Savings per
School per Day
Philadelphia, PA 39,144 144 1 hour 30 min
Jersey City, NJ 3,588 15 1 hour 18 min
Salinas, CA 1,000 8 41 min
National City, CA 2,821 10 1 hour 33 min
Prior to pilot implementation, three sites used manual procedures to count meals by category.
Philadelphia and Jersey City used a ticket system requiring tickets to be distributed by school
administrators and collected/counted by school food service staff at the end of the meal service.
National City used a roster system organized by classroom. Each of the three sites using manual
counting systems saved between 78 and 93 minutes per school per day by implementing no-fee
meal counting procedures.
Salinas saved about 41 minutes per day per school by counting only total meals. Prior to pilot
implementation, Salinas used an Aecutab system. The Aecu tab system uses automated tab
tickets. Tickets are coded (by free, reduced price and full price categories) and sectioned so that
the food service worker taking meal counts can feed them through an automated terminal which
will read them and remove a section. The automated terminal then tallies the meals by eligibility
category. Because Salinas had been using a more automated and less labor-intensive system
prior to pilot implementation, they saved less time per day strictly due to pilot procedures.
Cost Savings
Cost savings at the pilot sites resulted from a combination of factors including: 1) elimination
of direct costs; 2) shifts in labor from administration to meal preparation; 3) reductions in use
of off-budget staff (including school administrators or teachers); 4) actual staff reductions.
By eliminating the application and meal count by category processes, a number of direct cost
categories can be cut back or eliminated such as application printing, postage, meal tickets,
ticket display boards, meal counting equipment (e.g., Accutab machines), armored car contracts
and money orders, etc. Labor costs can be reduced through not hiring temporary workers to
process applications or through eliminating staff positions. By reducing costs in these areas,
pilot sites saved between one half cent and nine cents per meal.
14
TABLE 2.4
ACTUAL SAVINGS IN ADMINISTRATIVE COSTS IN NO-FEE SITES
Baseline to Year 1 Comparison
1 Pilot Site Reductions in
Labor Costs
Reductions in
Direct Costs
Additional
Admin Costs**
Meals Served
(Baseline Year)
r^—: 1 Admin Savings per
Meal
Philadelphia, PA $ • 64,939 $- 22,970 $ + 42,000 9,619,061 $.005
Jersey City, NJ $ 0 $ 0 $ o 1,017,001 $.00
1 Salinas, CA $-24,110 $- 18,000 $ 0 667,333 $.06
I National Cky, CA $-79,325 $- 3,942 $+ 1,487 900,143 $.09
** Additional Administrative costs are the result of the socio-economic survey conducted in Philadelphia and statistical model development in
National City.
Labor Costs: Three sites had actual reductions in labor costs. By eliminating the application
process, Philadelphia and Salinas did not have to hire temporary help to review and process
applications. Philadelphia was able to save over $30,000 in temporary labor costs and Salinas
saved over $13,000. In addition, both Philadelphia and Salinas were able to reduce staff
required to handle money and distribute tickets. As a result, Philadelphia saved almost $35,000
and Salinas saved about $10,000. Finally, Salinas was able to save $1,000 in overtime charges.
National City also had savings in actual labor costs. Prior to pilot implementation, the food
service office paid money into a general fund to reimburse the cost of using school staff (e.g.,
school secretaries, teachers, etc.) to process and verify applications. In 1991-92 the amount was
reduced by over $26,000 because schools no longer had to process and verify applications. In
addition, the central office saved about $1,500 by cutting 60 hours of office overtime. Finally,
the school district cut school food service staff by 14.5 hours per day (in 10 district schools) as
a result of simplifying the meal counting process. This reduction was achieved through attrition.
The district originally achieved a labor savings of 25.5 hours per day, however, 11 of the hours
had to be added back in to accommodate increases in meals prepared and served.
Direct Costs: Philadelphia, Salinas and National City also had reductions in direct costs as a
result of changes in the application and meal counting processes. Philadelphia saved a total of
$22,970 by not printing applications ($5,250), not printing meal tickets ($12,000) and not
handling money-purchasing money orders and hiring armored vehicles~($5,720).
Salinas saved about $18,000 in direct costs by not printing and mailing applications ($3,300),
not printing tickets or purchasing ticket display boards ($10,500), and not having to maintain
computers and Accutab machines ($4,200).
In contrast to the three sites which eliminated or reduced the incidence of application processing,
Jersey City did not achieve actual cost reductions because they continued to process applications
each year. However, Jersey City did achieve marked savings by reallocating labor away from
the meal counting process.
15
Reallocated Labor: By simplifying the meal count process, Jersey City saved about $.24 per
meal. All of the labor hours saved in Jersey City were reallocated to other activities within the
school food service account or to school administration. The total estimated annual value of
labor savings is $240,991. Of this, $106,809 (44.3 percent) is related to ticket distribution and
counting, $109,620 (45.5 percent) is related to money collection, and $24,562 (10.2 percent)
is related to SFA billing for moneys collected. Of the total estimated savings, $134,772 (55.9
percent) is school administrative staff time, and $106,219 (44.1 percent) is school food service
staff time.
TABLE 2.5
TOTAL SAVINGS IN ADMINISTRATIVE COSTS IN NO FEE SITES
Baseline to Year 1 Comparison
1 Pilot Site Actual
Reductions in
Labor
Reallocated
K/ibor
Direct Savings Additional
Admin Costs
Total Admin
Savings per
Meal
Philadelphia, PA $ - 64,939 $-1,768,069 $ - 22,970 $ + 42,000 $.19
Jersey Cky, NJ $ 0 $- 240,991 $ o $ 0 $.24
1 Salinas, CA $-24,110 $- 17,640 $ - 18,000 $ o $.09
\ National City, CA $-79,325 $- 40,590 $- 3,942 $ + 1,487 $.14
Philadelphia achieved $. 18 per meal in paperwork savings-for a total savings of $. 19 per meal-by
reallocating labor away from the application/meal counting processes. Philadelphia saved
about 23,460 hours (163 hours per school) or $1.1 million in school vice-principal and teacher
time by not requiring them to review applications and maintain rosters. The hours saved were
reallocated to other educational activities. The district also saved over 5,000 hours worth
$219,000 of administrative aide time by not distributing meal tickets on a daily basis.
Philadelphia also saved almost 27,000 hours of the district's food service administration time
(worth over $438,000) by eliminating the verification process and simplifying the accounting
process.
Salinas was able to reallocate about $17,700 away from application processing and meal
counting. The time saved was used to provide labor at two school sites that became multi-track
schools-schools that operate year round. Without the pilot project, Salinas would have had to
make additional hires to provide labor at the year round schools which now operate without a
summer vacation period for school staff.
Salinas had an additional savings associated with the pilot project which is not shown on the
above table. Implementing the pilot procedures resulted in monetary savings and labor hour
savings which enabled the district to establish breakfast programs at seven schools. Without the
pilot program, the breakfast program would have cost the district an additional $20,000. The
$20,000 represents the combined approximate cost in labor at all schools in the district
implementing a breakfast program. Without a no-fee program, two workers would have been
16
required at each site-one to serve meals and one to count by category. As a result of the pilot,
only one worker is needed to administer the breakfast program at each site saving the district
approximately $20,000.
As previously discussed, National City had an actual reduction of labor hours due to pilot
implementation—approximately 14.5 labor hours per day were cut from 10 district schools. In
addition, about 11 hours of staff time per day was reallocated from the meal counting process
to meal preparation. The reallocation was necessary to accommodate increases in the number
of meals served.
Effects on Participation
No-fee programs have considerable potential for increasing student participation, particularly in
high schools or other schools where current participation rates are lower than average. The
absence of price is a significant incentive to participate. And because meal counting is
simplified-that is, there is no meal ticket system-a no-fee system can eliminate any stigma
associated with accepting a frse meal.
Three of the four no-fee pilot sites had lower-than-average free and full price participation rates
prior to beginning the pilot program. Philadelphia and Jersey City also had lower-than-average
reduced price participation rates. The potential for increases in participation rates were highest
in these areas.
TABLE 2.6
BASELINE AVERAGE DAILY MEAL RATES IN PILOT SCHOOLS
School Year 1990-91
H Category National
Average
Rates
Philadelphia
PA
Jersey City, NJ Salinas, CA National City,
CA
[ Free 80% 65% 68% 77% 83%
1 Reduced Price
74% 26% 40% 82% 84%
1 Full Price 40% 3% 8% 32% 61%
| TOTAL 54% 53% 60% 73% 78%
From the baseline year to the first year of pilot procedures, participation rates increased in all
of the no-fee sites. District-wide participation increased the most in Jersey City where the
average daily meal rate rose from 60 percent of children selecting a school lunch to 73 percent-a
21 percent increase in participation. As a result of increased student participation, the number
of meals served in Jersey City rose by 30 percent from the baseline to the first year of the pilot.
In each of the pilot sites, the percentage increase in total meals is higher than the percentage
increase in participation because of higher enrollments. In other words, the 30 percent increase
17
in the number of meals served in Jersey City is a result of a 21 percent increase in participation
coupled with more children in school.
Salinas experienced an 18 percent increase in the average daily meal rate and a 25 percent
increase in total meals served. Again, the 25 percent increase in total meals served is a result
of the 18 percent increase in participation coupled with ai increased number of children in
school.
TABLE 2.7
AVERAGE DAILY MEAL RATES IN NO-FEE PILOT SITES
A Baseline to Year 1 Comparison
| No-Fee Pilot Site Avg Daily
Meal Rate
Baseline
Avg Daily
Meal Rate
Year 1
Percent
Increase in
ADM rate
Actual Increase in
Meals Served
Philadelphia, PA 53% 63% 18% 2,725,394
Jersey City, NJ 60% 73% 21% 302,206
Salinas, CA 73% 87% 18% 163,941
National City, CA 78% 84% 7% 86 372
National City experienced only a 7 percent increase in its average daily meal rate an
percent increase in the number of meals served. National City is an elementary-only school
district that already had very high rates of student participation prior to pilot implementation.
In fact, National City was the only no-fee pilot site to begin the demonstrate n with higher than
average participation rates in all meal categories. Consequently, their potential for increases in
participation was limited.
In Philadelphia, the only site to include high-schools in the pilot, the overall average daily meal
rate rose by 18 percent and total meals served increased by 28 percent. The average daily meal
rate in elementary schools rose by only 4 percent but in high schools it rose by 186 percent.
As in National City, elementary schools already had high rates of participation before pilot
implementation so potential for increase in participation rates was limited.
TABLE 2.8
AVERAGE DAILY MEAL RATES IN PHILADELPHIA
A Baseline to Year 1 Comparison p- Avg Daily
Meal Rate
Baseline
Avg Daily
Meal Rate
Year 1
Percent
Increase in
ADM rate
Actual Increase in
Meals Served
Elementary Schools 72% 75% 4% 1,022,030
Middle Schools 35% 50% 41% 861,460
High Schools 13% 38% 186% 841,904
TOTAL 53% 63% 18% 2,725,394
18
However, while Table 2.8 demonstrates that participation rates rose the most in high schools
where potential for increase was greatest, it also shows that 38 percent of the new meals were
served in elementary schools since elementary school students represent over half of the children
in Philadelphia's pilot program.
Federaj Costs
Federal costs have risen in each of the no-fee pilot sites as a result of increases in meals served.
Higher Federal reimbursement rates contributed slightly to increased Federal costs in Year 1 of
pilot operations-i.e., about 3 percent of additional Federal dollars spent in each of the pilot sites
is attributable to increased Federal reimbursement rates and 97 percent is attributable to
additional meals served.
TABLE 2.9
INCREASES IN FEDERAL DOLLARS TO NO-FEE PILOT SITES
A BASELINE TO YEAR 1 COMPARISON
Pilot Site Federal Dollars •
Baseline
Federal Dollars •
Yearl
Increase in Federal
Dollars - Baseline
to Year 1
Additional Meals
served in Year 1
Philadelphia $ 16,587,494 $ 20,095,038 $ 3,507,544 2,725,394
Jersey City $ 1,737,543 $ 2,181,868 $ 444,325 302,206
Salinas $ 1,084,670 $ 1,338,494 $ 253,825 163,941
National City $ 1,290,680 $ 1,465,854 $ 175,174 86,372
Changes in District Revenues
The National School Lunch Program is supported by revenues from several sources including
Federal and State per meal reimbursements, local payments and student payments. In a no-fee
program student payments are eliminated which lowers the average per meal reimbursement--
i.e., the actual amount of money that is available to produce each meal. However, because
some administrative duties are eliminated, districts may achieve actual labor and direct cost
savings which reduces their cost to produce a meal. Philadelphia saved one half cent per meal,
Salinas saved $.06 per meal and National City saved $.09 per meal as a result of pilot
implementation.
Nationally, schools received an average of $1.74 per meal in Federal reimbursements, State
contributions and student payments in 1992.
19
NATIONAL AVERAGE PER MEAL REVENUES
Amount
Source in millions
Federal Cash Payments $3,810
Federal Commodity
Payments $ 574
State Contributions $ 446
Student Payments $2,304
Total $7,134
Total Lunches Served 4.101
Average Revenue per Meal $ 1.74
Because no-fee school districts must make up the difference between the Federal reimbursement
received for full price and reduced price meals and the cost to produce a meal, this type of
program is o.ily feasible in districts that do not rely heavily on student fees to support their meal
programs. Nationally, during the 1992 school year, students paid about $2.3 billion for school
lunches-about S3 percent of Federal cash and commodity lunch reimbursements for that year.
In comparison, prior to pilot implementation, National City had student payments equal to 12
percent of Federal reimbursements, Jersey City and Salinas had student payments equal to 2
percent of Federal reimbursements in pilot schools, and Philadelphia had student payments equal
to 1 percent of Federal reimbursements in pilot schools.
During the baseline year, Philadelphia had total lunch revenues (in the 144 pilot schools) of
$17.6 million including $16.6 in Federal reimbursements, about $770,000 in State revenues and
$202,000 in student payments. Total revenues divided by total meals served show that
Philadelphia earned about $1.83 per meal prior to pilot implementation. During the first year
of pilot operations, total revenues to Philadelphia increased by $3.5 million due to the 28 percent
increase in meals served. Both Federal revenues and State revenues increased but student
revenues were eliminated. As a result, Philadelphia earned about $1.71 per meal in the first
year of pilot operations—a $. 12 decrease from the prior year.
As previously discussed, due to reductions in labor and direct costs, Philadelphia had actual
savings of about one half cent per meal as a result of the pilot procedures (p. 15). Therefore,
in order to operate a no-fee program, Philadelphia had to find ways to cut their cost of meal
production by more than $.11 per meal. Philadelphia showed a total administrative savings of
about $. 19 per meal including reallocated labor. However, most of the total administrative
savings accrued to school administrators such as principals, secretaries, etc. rather thai: to the
food service account.
20
TABLE 2.10
PHILADELPHIA CHANGES IN REVENUE
Baseline to Year 1 Comparison
Federal
Revenues
State Revenues Student
Payments
Total Revenues Per Meal
Revenue
I Baseline Year $16,587,495 $ 769,525 $ 202,499 $17,559,519 $ 1.83
1 Year 1 $20,095,039 $ 987,556 $ 0 $21,082,595 $ 1.71
1 Absolute Change $ 3,507,544 $ 218,031 $ -202,499 $ 3,523,076 $-.12
1 Percent Change + 21% + 28% -100% + 20% -7%
Prior to no-fee implementation, Jersey City earned about $1.83 per meal in pilot schools. As
a result of an increase in the number of meals served during the first year of the no-fee program
coupled with increased Federal and State per meal reimbursement rates, total revenues to Jersey
City increased by about 25 percent. However, the number of meals served in Jersey City grew
by 30 percent resulting in a net per meal revenue decrease. Jersey City's per meal revenue
decreased by about $.07 from $1.83 to $1.76~which is still $.02 above the national average.
TABLE 2.11
JERSEY CITY, CHANGES IN REVENUE
Baseline to Year 1 Comparison
Federal
Revenues
State
Revenues
Student
Payments
Total Revenues Per Meal
Revenue
BaiaBoi Year $ 1,737,543 $ 84,876 $ 35,671 $ 1,858,090 $ 1.83
1 Year 1 $2,181,869 $ 138,168 $ o $ 2,320,037 $ 1.76
| Absolute Change $ 444,326 $ 53,292 $ -35,671 $ 461,947 $-.07
| Percent Change + 26% + 63% - 100% + 25% -4%
Salinas had revenues of about $1.80 per meal prior to no-fee implementation. During the first
year of pilot operations, per meal revenues fell to about $1.73~a $.07 decrease. However,
Salinas also achieved an actual per meal labor savings of $.06 through reductions in paperwork.
Therefore, Salinas actually forfeited about $.01 per meal due to the no-fee program.
21
TABLE 2.12
SALINAS, CA CHANGES IN REVENUE
Baseline to Year 1 Comparison 1 Federal
Revenues
State Revenues Student
Payments
Total Revenues Per Meal
Revenue
1 Baseline Year $ 1,084,670 $ 89,706 $ 23,594 $ 1,197,970 $ 1.80
I Year 1 $ 1,338,495 $ 98,232 $ 0 $ 1,436,727 $ 1.73
1 Absolute Change $ 253,825 $ 8,526 $ -23,594 $ 238,757 $-.07
1 Percent Change + 23% + 10% -100% + 20% -4% |
National City had the largest decrease in per meal revenues because they were the site with the
lowest proportion of free and reduced price eligible students and the most dependent on student
payments. Prior to no-fee implementation, National City earned about $1.73 per meal in
Federal and State reimbursements plus student payments. During the first year of pilot
operations, per meal revenues fell by $.14 to $1.59. However, National City also achieved the
highest rate of actual labor savings--$.09 per meal. Therefore, actual per meal loss to National
City is about $.05.
TABLE 2.13
NATIONAL CITY, CA CHANGES IN REVENUE
Baseline to Year 1 Comparison
Federal
Revenues
State Revenues Student
Payments
Total Revenues Per Meal
Revenue
Baseline Year $ 1,290,681 $ 104,436 $ 158,422 $ 1,553,539 $ 1.73
Yearl $ 1,465,853 $ 106,716 $ 0 $ 1,572,569 $ 1.59
| Absolute Change $ 175,172 $ 2,280 $ -158,422 $ 19,030 $-.14
1 Percent Change + 14% + 2% - 100% + 1% -8%
Further Evaluation
During the final year of the pilot projects a number of additional areas affecting no-fee systems
will be examined. These include the impact of no-fee procedures on the breakfast program, a
comparison of participation-based versus enrollment-based claiming percentages, and a
discussion of the impact of no-fee procedures on district finances.
22
CHAPTER 3
ALTERNATIVE APPLICATION PILOT PROJECTS
San Bernardino, California; Springfield, Oregon; and Lowell, Massachusetts each implemented
alternative application procedures. San Bernardino and Springfield are extended eligibility sites
where applications are processed once every two or three years rather than every year. At the
time of application, a child's eligibility is determined and remains the same until the next time
applications are collected. San Bernardino processes applications every other year and conducts
verification in the alternate year. Springfield petitioned to process one third of their applications
each year and verify 10 percent of those approved in that year. Lowell developed a modified
application to simplify their application process and shift part of the procedure from the fall to
the spring. All alternative application sites have continued to count meals by category.
TABLE 3.1
ALTERNATIVE APPLICATION PROJECT SITES
Characteristics of the School Districts
Pilot Site Total
Schools
Number of
Schools in Pilot
Number of
Children in Pilot
Percent Free
Eligible in Pilot
Percent Reduced
Price Eligible in
Pilot
San Bernardino, CA 52 52 40,973 42% 1 5%
Springfield, OR 21 1 280 71% 7%
Lowell, MA 29 29 14,000 41% 6%
Time Savings
San Bernardino collects applications from students in all 52 district schools every other year.
Verification of a 10 percent sample of all applications on file is conducted during the alternate
year when applications are not processed. Although the distrk* is using B two-year cycle for
applications, it still sends applications to all households during the alternate year to capture new
students and those who may be newly eligible for free or reduced price meals.
TABLE 3.2
APPLICATION PROCESSING AND VERDTICATION
TIME SAVED IN SAN BERNARDINO
Alternative
Application Pilot
Site
San Bernardino
Total Hours Saved
in App Processing
6,012
Total Hours
Saved in
Verification
510 20 minutes/
application
•Note: The district actually spent aboutTTiourspervenncauoncompTeted. However, in order to standardize thenumbers^enumber of hours
spent to verify 239 applications was divided by all 18,318 applications on file to arrive at 2 minutes per application on file.
Number of Free
and Reduced
Apps on File
18,318
Time Saved per
Application Not
Processed
Time Savings per
Verification Not
Completed
2 minutes/ application
23
In years when applications are not processed, San Bernardino saves over 6,000 hours of
administrative staff time by not collecting and processing applications for all students in their
school district. Most of those hours are attributable to school clerks. The data suggest that a
clerk at each school works almost full time for about 3 weeks to collect, review, process, and
prepare applications for data entry. Data are then entered at the district level where data
processors spend about 18 hours per year preparing the data base. In addition, the district's
food service senior secretary spends over 14 hours annually drafting the public notice/parent
letter and training staff.
San Bernardino saves about 510 hours of administrative work in years when verification is not
conducted. In the 1991-92 school year, San Bernardino selected a focused sample and verified
239 applications at a rate of over 2 hours per verification. When the amount of time for
verification activities is averaged across all applications on file, San Bernardino spent about 2
minutes per application. Therefore, in the 1992-93 school year when applications were
processed in San Bernardino but no verification activities were performed, the district saved 510
hours of paperwork.
Cost Savings
Cost savings from alternate application procedures accrue in two ways: 1) labor that would have
been used to approve applications can be redirected to other activities; and 2) the same volume
of applications do not have to be printed and distributed.
The San Bernardino SFA annually pays $290,000 to a general fund to repay the district for
NSLP paperwork—including application processing, verification, meal counting, etc.~performed
by school administrators. This amount has not been reduced as a result of the pilot. Most of
the labor savings associated with pilot procedures (99 percent) accrue to the district's
administrative staff, allowing them to reallocate labor from NSLP paperwork to other activities.
TABLE 3.3
TOTAL SAVINGS IN ADMINISTRATIVE COSTS REALIZED
IN YEARS WHEN APPLICATIONS ARE NOT PROCESSED
Pilot Site
San Bernardino
Actual Reductions
in Labor
Reallocated Labor
68,107
Direct Savings
045
Additional
Administrative Costs
Total Admin
Savings per Meal
.05
While San Bernardino saves about $.05 per meal-or $1,368 per school-in reallocated labor and
direct costs during years when applications are not processed, they save very little during years
when verification is not conducted. Only about $6,197 in labor costs is associated with district-wide
verification activities which translates into about $119 per school and less than one cent
per meal.
24
TABLE 3.4
TOTAL SAVINGS IN ADMINISTRATIVE COSTS REALIZED
IN YEARS WHEN VERD7ICATION NOT PERFORMED
Pilot Site
San Bernardino
Actual Reductions
in Labor
Reallocated Labor
6,197
Direct Savings Additional
Administrative Costs
Total Admin
Savings per Meal
.004
The $6,197 represents 406 labor hours at the school level-about 8 hours per school per year-plus
100 hours of district senior secretary time plus about 3 hours of data processing time.
Direct Savings: In 1991-92 when applications were not processed, San Bernardino had total
direct savings of $3,045 which represents $1,465 saved by not having to print applications plus
$1,580 saved by not doing a Spanish language translation of applications, letters, and scan sheets
used for data entry. No direct cost savings are associated with not performing verification every
other year.
Effects on Participation
Increases in participation in San Bernardino cannot be directly attributed to the pilot since
increases were occurring across the nation. However, San Bernardino did have increases in
enrollment and participation in excess of what occurred in California as a whole. San
Bernardino experienced a 17 percent increase in enrollment between the baseline and first year
of pilot operations~from 19,756 to 23,156 students. In addition, they experienced a 6 percent
increase in average daily participation-from 42.8 percent participating on an average day to 45.3
percent participating. This is a much higher increase than occurred in the State. During that
time period, California's school enrollment and average daily participation each rose less than
one percent.
Further Evaluation
The San Bernardino school district implemented Direct Certification in the 1992-93 school year.
Continuing evaluation of this pilot project will examine the effect of direct certification on
paperwork savings associated with extended applications. In addition, future evaluation will
focus on determining whether applications for free and reduced price meal benefits accumulate
over time, i.e., whether applications stay on file for children who are no longer eligible for
benefits.
The two other alternate application sites-Springfield, Oregon and Lowell, Massachusetts had
difficulties implementing their pilot procedures. Lowell found that their proposed pilot actually
increased district level paperwork slightly. As a result, they have elected to discontinue the
pilot. Springfield only partially implemented their proposed procedures.
Springfield intended to collect applications from one-third of the school population on an annual
basis and certify applications for a three year period. In addition, Springfield planned to verify
10 percent of all applications on file each year. After selection, the district chose to have only
one school implement pilot procedures. The school was unable to completely implement the
25
proposed pilot procedures. While the district did verify 10 percent of all applications on file
from the pilot school, they continued to collect and review applications from all students every
year. As a result, paperwork actually increased. The pilot school had an enrollment of 220
children eligible for free and reduced price meals. Using current verification procedures, a 3
percent sample-about 7 applications-would be verified annually. However, under the pilot, a
10 percent sample-22 applications-are being verified annually increasing district paperwork by
approximately IS hours.
Lowell developed a modified application used for students who were categorically eligible—that
is, were members of households receiving food stamps or Aid to Families with Dependent
Children. The application contained only the child's name, food stamp or AFDC case number
and an adult signature block. Children who were categorically eligible in the prior year were
given the simplified application in June-near the end of the school year. The simplified
applications were processed during the summer months using permanent school food service
staff. The list of children potentially eligible for free lunches based on receipt of food stamps
or AFDC was then sent to the local welfare office for confirmation. Applications for all other
students were approved in the fall. Lowell developed this process in order to make more
efficient use of permanent staff and avoid paying overtime costs at the beginning of the school
year to process applications. Lowell experienced problems implementing their proposed
procedures and has elected to participate in Massachusetts' direct certification program instead
of continuing the pilot.
Initially Lowell experienced difficulties eliciting the cooperation of the local welfare office.
Prior to the pilot project, the local welfare office participated in the district's verification—that
is, the Lowell SFA sent a list of children to the welfare office and the welfare office matched
the list with computer files of food stamp and AFDC recipients. However, while the agency
was willing to assist with verification of 100-200 applications, they were less willing to do a
computer match for 3,000 applications-the total number of children categorically eligible for
free meals in the Lowell school district.
Lowell also had problems with duplicate applications. In order to ensure that all children had
the opportunity to apply for free meals and to avoid overt identification, Lowell distributed
applications to every child in the district whether or not the child had previously completed a
simplified application. While the letter accompanying the application stated that those previously
approved did not have to complete the new application, Lowell found that many of the families
completing simplified applications in the spring also filled out full applications in the fall. As
a result, paperwork in Lowell actually increased slightly and Lowell continued to pay about the
same amount of overtime to employees to complete applications during the fall.
Using Massachusetts state-wide direct certification, the State sends vouchers to families receiving
food stamp or AFDC benefits. Lowell will be charged $.35 for each student certified by the
State in their district. As a result, Lowell will pay around $1,050 to use the state-wide system.
During the final year of data collection, the issue of duplicate applications under the
Massachusetts direct certification system will be examined. It will be interesting to explore
26
whether the incidence of children returning both a State voucher and a completed application
declines under the new system.
27
28
CHAPTER 4
DIRECT CERTOTCATTON PILOT PROJECTS
Columbus, Ohio and the State of Maine are direct certification pilot projects. Direct certification
is a simplified method to determine eligibility for free meals under the National School Lunch
and School Breakfast programs. In place of applications, school officials may contact food
stamp or Aid to Families with Dependent Children (AFDC) offices for documentation that
children are members of a food stamp household or an AFDC assistance unit. These children
are automatically eligible for free meals. Besides simplifying the initial application process,
direct certification also makes the verification process less burdensome by reducing the number
of applications that must be verified.
While the SFA does not have to seek, approve, or keep on file applications for these direct-certified,
categorically-eligible students, there are other requirements that the SFA must meet.
The SFA must maintain the certified child's name on a list generated by the AFDC or food
stamps office and notify households that: 1) their children are eligible for free meals, 2) no
further application is required, 3) they should notify the school if they do not want their child
to receive free benefits, and 4) they must notify the school when they are no longer certified to
receive food stamps or AFDC benefits.
When the pilot projects began in 1990, direct certification was newly approv by law (P.L.
101-147). The two pilot sites were chosen to test the usefulness of direct certification-
Columbus at the local level and Maine at the State level. Since 1990, however, many States and
individual school districts have implemented direct certification. Currently, at least sixteen
States have made statewide direct certification available.2
TABLE 4.1
DIRECT CERTIFICATION PILOT SITES
Characteristics of the State or District in the 1990-91 School Year
Direct Certification Pilot Site Total Sckoote Number of
ScbookntPilot
Number of
Children in Pilot
Percent Free
Eligible in Pilot
Percent Reduced
Price Eligible
Colon.bus, OH 141 141 63,786 42* 7*
Maine State 712 712 205,457 17* 5%
Time Savings
Columbus had an estimated 3,653 hour savings in application processing from the baseline to
the first year of pilot operations. This represents 3,760 hours of time saved at the school level
2 States that provide direct certification on a statewide basis include: Alaska,
Illinois, Indiana, Maryland, Massachusetts, Michigan, Minnesota, Missouri, North
Dakota, Ohio, Oklahoma, Pennsylvania, Rhode Island, Texas, Utah, Washington.
29
from not having to receive and process over 17,000 applications minus 107 hours of additional
time spent on direct certification at the district level. At the district level about 107 more hours
were spent during the first year of direct certification in developing the public release notice and
the letter to parents explaining the new process as well as implementing the new procedure.
This may be a one-time investment of effort-data collection in future school years will indicate
whether direct certification implementation continues to be more time consuming at the district
level.
During the first year of direct certification, Columbus did not have higher time savings due to
inexact computer tape matches—i.e., many children's names on the food stamp and AFDC tapes
could not be electronically matched with names on school registers. Names that could not be
electronically matched were subject to manual matching using addresses and parent names to
resolve questions about student identity. Again, future data collection will examine whether this
continues to be a problem.
School level data collection indicated about 27 hours of time saved using the direct certification
process. Extrapolated to 141 schools, 3,760 hours were saved. This total may grow in future
years. During the first year of direct certification, about SO percent of directly certified students
also submitted applications. School staff spent considerable time matching applications to the
direct certification list sent from the Department of Human Services. Data from years 2 and 3
may show that there is a direct certification learning curve—i.e., as the process becomes more
familiar to staff and students, fewer duplicate applications will be filed and more paperwork
savings will be achieved. In fact, preliminary data from the second year of pilot operations
shows that the number of duplicate applications submitted was reduced from 50 percent to about
22 percent.
In addition to time savings from reductions in applications processed, Columbus achieved
savings by reducing their verification sample size. Districts are not required to verify
certifications made through the direct certification process. Columbus draws a 3 percent sample
of all applications on file for verification. Because over 17,000 applications were directly
certified, the number of verifications required was reduced by 521. The data for time savings
due to changes in verification in Columbus are not sufficient for direct estimation, however, one
can estimate time savings based on national average time to complete a verification. The Study
of Income Verification in the National School Lunch Program indicated that SFAs spend about
1 hour per verification completed to select a sample, notify the household, collect the
information, make a determination and make any changes to a child's eligibility status.3 Based
on this, Columbus saved about 521 hours by not having to verify that number of applications.
Study of income verification, p. xxiii.
30
TABLE 4.2
APPLICATION PROCESSING AND VERIFICATION TIME SAVED IN DKECT
CERTIFICATION PILOT SITES FROM BASELINE TO YEAR 1
Direct Certification Pilot Site Total Hours Saved in
Aspkcrtiou rYoceanag and
Verification
Number of Applications
Directly Certified
Time Savings per
Application not proccased
Columbus, OH 3,653 17,362 13 min
Maine State not available 20,287 not available
During the 1992 school year, the State of Maine reduced the number of applications processed
by about a third-20,287 were directly certified out of a total 59,212 approvals. Actual time
saved in Maine due to direct certification is not available. Maine began to directly certify
applications in the baseline year of data collection, therefore, no comparison of time spent in
application processing and verification activities can be made between baseline and pilot
procedures. As in Columbus, schools in Maine received a high rate of duplicate applications
during the first year of direct certification-about 33 percent of students directly certified also
submitted applications. Future data collection will focus on time savings achieved after direct
certification has been in place for two or more years.
Cost Savings
Cost savings from direct certification can result in two ways: 1) direct savings from reduced
postage, fewer envelopes or applications needed; and 2) labor savings from reduced hours spent
in application processing and verification activities.
TABLE 4.3
TOTAL SAVINGS IN ADMINISTRATIVE COSTS IN
DIRECT CERTIFICATION SITES
Pilot Site Actual
Labor
Reallocated
Labor
Direct Savings Additional Admin
urn
Total Aomin
Savings per Meal
Columbus, OH $ 0 $105,825 $ 0 $ 5,035 $.01
State of Maine $2,354 not available $ 0 $ 10,000 not available
Columbus had no actual reductions in labor costs-that is, no workers had hours actually reduced
from their work schedules due to this procedure. However, Columbus saved about $100,2%
in reallocated labor due to changes in the application process and about $5,529 due to decreases
in the number of verifications required. These hours and dollars were rechanneled to other
administrative or educational activities within the school district. Again, as the direct
certification process becomes more familiar, it is expected that Columbus will experience higher
cost savings.
31
Columbus experienced no direct savings, in fact, they had increased direct costs due to increased
postage. Columbus provides a package of important school and services-related information to
each student on the first day of school each year. In each package is an application for free or
reduced price school meals plus a month's worth of free or reduced price tickets for students
who were qualified in the previous year. Columbus had hoped to be able to include a parental
notification in packages going to the homes of children who were directly certified. However,
as of the first year of data collection, Columbus was not able to complete the direct certification
process in time to include notices in the first day packages. As a result, Columbus mailed letters
to directly certified households informing them that their children had already been approved and
exhorting them not to complete the application. Since over 17,000 students were directly
certified in the first year, postage for letters sent to parents cost about $5,035.
The State of Maine experienced a $2,354 reduction in the cost of conducting income verification
due to direct certification. Income verification for the entire State is conducted at the State
level. The State uses a focused verification sample—1 percent of applications listing income
within $50 of the eligibility guidelines are sampled plus 1/2 percent of those categorically
eligible (because their families receive food stamps or AFDC) are verified. Mai.ie estimated that
101 verifications were avoided since over 20,000 students were directly certified. Since the
State hires a consultant to conduct verifications annually, the verification change reduced actua'
labor hours and costs associated with the process.
Maine had an increase in direct costs of $10,000. This amount represents postage, envelopes,
paper plus a 29 minute computer run. Tlr.se costs are paid for by tie Maine Department of
Human Services (not the Department of Education) which actually conducts the direct
certification procedure.
Effects on Participation
Participation in direct certification sites increases if the number of children certified for free
meals increases due to the new procedure. Both Columbus and the State of Maine experienced
much higher than average growth in the number of free eligibles between the baseline year
(1991) and the first year of pilot operations (1992).
TABLE 4.4
INCREASE IN NUMBERS OF FREE ELIGIBLE STUDENTS AT DIRECT
CERTIFICATION PILOT SITES
Free FMgiabi, 1991 FrccEagiUcf. 1992 Pcfccat Iarreate
Col—PM, OH 26,667 32,732 23%
MM* State ^4,671 46,810 35 X
U.S. Total 11,630,300 12,856358 11%
32
During the 1990-1992 time period, the nation as a whole experienced an 11 percent growth in
the number of free eligibles, so not all the growth in Columbus and Maine can be attributed to
direct certification. However, Columbus experienced growth equal to more than 200 percent
and Maine experienced growth equal to more than 300 percent of the national average.
Columbus' growth in free eligibility also outstripped Ohio's growth. During that time period,
Ohio experienced only a 10 percent increase in the number of free eligibles in the State.
Federal costs are affected by direct certification because program costs increase as the number
of free eligibles eating meals increases. In Maine, for example, the number of free meals served
increased by 812,000 between 1991 and 1992. At the same time, the numbers of reduced price
and paid meals declined indicating that children shifted from the reduced price and paid
categories into the free category-perhaps as a result of direct certification.
TABLE 4.5
CHANGES IN MEALS SERVED IN THE STATE OF MAINE
School Year FrecMeak
Served
Reduced
Meals Served
Paid Meals
Served
Total Mods
Served
1991 5,691,302 1,471,663 9,624,802 16,787,767
1992 6,503,295 1,470,554 9,005,726 16,979,575
Gunge 811,993 - 1,109 - 619,076 191,808
As a result in categorical shifts in meals served (plus a 3.4 percent increase in reimbursement
rates), total Federal dollars reimbursed to Maine increased by almost $1.6 million or 13 percent.
Further Evaluation
Future evaluation of the direct certification pilot projects will focus on efficiencies gained
through experience with the direct certification process. For example, the question of whether
the incidence of duplicate applications is reduced from one year to the next will be examined.
33
34
CHAPTER5
PROVISIONS 1 AND 2
In 1977, Congress amended the National School Lunch Act to authorize special assistance and
certification procedures to reduce paperwork in the school lunch and breakfast programs. P.L.
9S-166 provided for alternative approaches, commonly referred to as Provision 1 and Provision
2, to standard application and meal counting and claiming procedures. Provision 1 and 2 have
not been widely used. At the beginning of the pilot project period in 1990, only 346 schools
in the country operated under either Provision 1 or 2.
Under Provision 1 (as permitted by current law), schools with at least 80 percent of students
eligible for free or reduced price meals may take applications once every two years instead of
annually. Schools participating under Provision 1 must continue to record the number of free,
reduced price, and paid meals served daily as the basis for calculating claims for Federal
reimbursement. A no-fee program is an option. There were 59 schools in 11 SFAs operating
under Provision 1 procedures in the first year of the pilot program, SY 1990-1991.
Under Provision 2 (as permitted by current law), an SFA or a school which serves meals at no
charge to all children may take applications once every three years. Meal counting procedures
are also altered to reduce administrative burden. During the first year, schools must take daily
counts of the number of meals served by free, reduced price and paid eligibility categories.
These meal counts are converted into percentages of total meals served each month, e.g., the
number of free meals divided by the total number of meals served that month equals the free
claiming percentage. Reimbursement in the second and third years is calculated by applying
these monthly percentages to total meal counts for the corresponding months. Daily meal counts
by eligibility category are not required in the second and third years, but schools must continue
to count the total number of meals served each month. There were 287 schools in 110 SFAs
operating under Provision 2 procedures during the first year of the pilot program, SY 1990-
1991.
The Child Nutrition and WIC Reauthorization Act of 1989 (P.L. 101-147) specifically prescribed
alternatives to Provisions 1 and 2 to be tested through the pilot projects. The procedures for
Pilot Provision 1 include: processing applications once every three years, taking total daily meal
counts, and using enrollment-based claiming percentages. The procedures for Pilot Provision
2 include: processing applications once every three years, taking total daily meal counts, and
using enrollment-based claiming percentages. A No-Fee program is required for both Provisions
1 and 2. Table 5.1 summarizes current Provision 1 and 2 procedures and Pilot Provision 1 and
2 procedures.
Pilot Provision 1 and 2 procedures were to be tested in sites that already participated in current
law Provision 1 or 2. The new procedures were to be compared to old Provision 1 or 2
procedures for paperwork reduction and overall efficiency. Letters were sent to all schools
known to be operating as Provision 1 or 2 schools at the beginning of the project (1990) to
request their participation as pilot sites. Very few schools responded. Each of the three pilot
35
BLANK
PAGE
Table 5.1
COMPARISON OF PROVISION 1 AND 2
AND PILOT PROVISION 1 AND 2 PROCEDURES
ORIGINAL PROVISIONS PILOT PROVISIONS
PROVISION 1
School must have 80% of
students eligible for free or
reduced price.
A No-Fee meal program is an option.
Applications collected once In 2 years.
Meals counted by eligibility category.
Schools Participating = 59
School Food Authorities Participating = 11
Applications collected once in 3 years.
Claiming percentages based on annual
enrollment.
Study Sample = 1
PROVISION 2
School need not have 80% of
students eligible for free or
reduced price.
A No-Fee meal program is required.
Applications taken once In 3 years.
Claiming percentages based on monthly
participation in base year.
Schools Participating = 287
School Food Authorities Participating =110
H
Applications taken once in 3 years;
applications taken every year from new and
previously uncertified students.
Claiming percentages based on annual
enrollment and adjusted based on school
turnover.
Study Sample = 2
sites that joined the projects was actively recruited by FNS staff. Atlantic City, NJ agree to
participate as a Provision 1 site and Terrell County, GA and Milford, ME as Provision 2 sites.
TABLE 5.2
PROVISION 1 AND 2 PILOT SITES
Characteristics of the School Districts in the 1990-91 School Year
ProvBion 1 or 2 Pilot Site Total Scfcoob Number of
Scnook in Pilot
Number of
Children in Pilot
Percent Free
Eligible in Pilot
Percent Reduced
Price Eligible in
Punt
Atlantic City, NJ 11 1 395 62% 10*
Milford, ME 4 4 955 25* 12*
Terrell County, GA 3 3 460 79* 12*
Provision 1
FNS recruited Atlantic City to participate in the study as a Provision 1 pilot site. The pilot
procedures were intended to be implemented at only one school, Chelsea Junior High School.
The Atlantic City pilot site discontinued their participation in the study after the first year of data
collection. There was a change in administration and the new Director of Food Services decided
not to implement pilot procedures.
Provision 2
FNS recruited Milford to participate in the study as a Provision 2 pilot site. During the 1990-
1991 school year (the first year of the intended pilot), the district changed administrations and
withdrew from the pilot program (new Provision 2), and the Provision 2 program altogether.
Their decision was based on the fact that only about 37 percent of students were certified as
eligible for free lunches and local moneys were used to subsidize the lunch program each year.
Because the town was facing the possibility of a 100 percent increase in property taxes for
schools, the School Committee decided to discontinue providing lunches to all children at no
charge, and dropped from Provision 2 status altogether. Milford is currently using typical
application, verification, and meal counting procedures.
FNS also recruited Lilly Cooper Primary School in Terrell County to participate in the study as
a Pilot Provision 2 school. The Terrell County pilot site did not implement the Pilot Provision
2 pilot procedures because enrollment-based claiming percentages were never used to calculate
reimbursement claims. The State of Georgia computes all participation-based claiming
percentages for its Provision 2 schools and continued to apply participation-based percentages
to total counts supplied by Terrell County. The decision on whether to compute and use
enrollment-based percentages for the remaining period of the study is underway.
The final report will present both enrollment- and participation-based claiming percentages for
Terrell County as well as estimates of time spent counting meals to show which system is more
efficient for the district.
37
In summary, although the legislation prescribed that pilot procedures should be tested in
Provision 1 and 2 schools, each of the three sites recruited experienced problems in
implementing the pilot procedures.
38
CHAPTER 6
IN-HOME SURVEY: PRELIMINARY FINDINGS
An important objective of this study is to examine if Federal costs differ under the pilot
procedures. Pilot procedures may affect Federal costs in several ways. First, reimbursement
claims based on claiming percentages will differ from claims based on actual meal counts
because the use of average claiming percentages does not capture day-to-day fluctuations based
on factors such as menu or meal quality. Family size or income may change during the time
lunch applications are retained by the school, i.e., up to three years in Provision 1 and Provision
2 pilot projects. Any such household change, which should change a child's eligibility status
but is not captured under the pilot procedures, will lead to an incorrect payment or program
benefits.
Research suggests that family size and income do vary significantly over time. The NSLP
regulations state that families must report changes in income over $50 to their school because
school lunch eligibility may be affected. A recent study of income verification in the NSLP
found that an estimated 62 percent of all households verified had a change in monthly income
of $50 or more between application processing and verification, i.e., three to four months. An
estimated IS percent of households correctly verified in the fall had income or household size
changes by the spring and were sufficiently large to alter their meal benefit status. An additional
25 percent of households reviewed in the study, but not verified, had benefits in error by the
spring.4 However, it is important to note that these results were based on a national sample of
households; the In-Home Survey conducted for the current study is based on a sample of low-income
households.
Purpose of the In-Home Survey
The In-Home Survey (IHS) is designed to collect information from households once during each
school year to determine whether family size or income changes affect eligibility status for free
and reduced price meals from year to year. The information is helpful in determining the
likelihood that eligibility status changes between lengthened application periods based on income
and household size changes. IHS for Provision 1 and Provision 2 sites is conducted in years one
and three, and surveys for alternative sites are conducted in years two and four.
Because changes in family status during the year are not monitored in normal program
operations, the current study focuses on changes in income and family size between years.
Between-year changes in family status should be captured at the time of application. Unless
households self-report changes, they will go unobserved longer under extended eligibility
procedures. If a significant number of children change eligibility status, e.g., lose eligibility for
free meals through increased household income, it is possible that the alternate system does not
*R.G. St Pierre and M. MOM, Study of Income Verification in the National School Lunch Program:
jve Income Verification Procedures. Cambridge, MA, Abt Associates, Inc. 1989.
39
ensure adequate program integrity. That is, income changes may signify the need for a yearly
application process.
Administration of the In-Home Survey
Personal interviews were to be conducted with 60 households for SY 1990-91 and SY 1991-92
in each of the 11 pilot sites, for a total sample of 660 households. Each SFA provided a list of
all students who had been approved for free or reduced price meals to serve as the sample.
However, preliminary data are available for only 214 households. The following factors reduced
the number of respondents for the two-year period:
• Atlantic City discontinued participation in the study.
• Data were not collected in SY 1990-91 in the San Bernardino and Springfield
projects.
• Interviews encountered problems in gaining access to respondents in Jersey City
and Salinas, and data were not collected in these pilot projects.
• Eight of the 222 households interviewed in SY 1991-92 were dropped because of
incomplete data.
TABLE 6-1
IN-HOME SURVEY HOUSEHOLD RESPONDENTS IN SY 1990-91 AND 1991-92
Pilot Project Number of Households
Interviewed in 1990-91
Number of Households
Interviewed in 1991-92
Columbus 59 54
Lowell 61 47
Milford 60 31
National City 60 27
Philadelphia 60 20
Terrell County 60 43
TOTALS 360 222
Note: The actual number of households included in the IHS preliminary analyiii is 214; S cases were dropped due to
incomplete data.
40
Table 6-1 shows the actual six pilot projects where the IHS was administered in both school
years. These pilot projects include:
• Terrell County and Milford: Provision 2 sites.
• Lowell: Simplified applications from food stamp and AFDC household in the
spring of each year.
• National City: No-Fee.
• Philadelphia: No-Fee with Direct Certification of children from food stamp,
AFDC, and foster care rolls.
• Columbus: District-wide Direct Certification of children from families which
receive food stamp or AFDC benefits.
As Table 6-1 illustrates, the number of respondents decreased from SY 1990-91 to SY 1991-92.
Many of the respondents who were interviewed in SY 1990-91 had moved or could not be
located the second year. The overall attrition rate is 39.4 percent. In four individual cases,
some or all of the information was unknown and eligibility could not be computed.
Limitations of the Data
The IHS data have a number of limitations that require caution in drawing statistical conclusions.
First, documentation of income varied considerably; it was generally greater than required for
applications, but less than necessary for verification. Secondly, the attrition rate in the sample
poses possible response bias which cannot be explored in this study. Thirdly, the sample sizes
of households ineligible for free lunches, i.e., eligible for reduced price lunches or ineligible for
benefits, are very small. Nevertheless, the IHS data provide some useful findings and insights.
Table 6-2 summarizes the eligibility status of all 214 households from SY 1990-91 to SY 1991-
92. From SY 1990-91 to SY 1991-92,18 percent of the households changed from one eligibility
status to another, i.e., 176 (82 percent) retained their eligibility status between the two years.
There was a 9 percent decrease in benefits for households who were eligible for free or reduced
price lunches in SY 1990-91. Moreover, there was a 9 percent increase in the number of
households who were eligible for reduced price lunches or ineligible for benefits in SY 1990-91.
It is important to note that 91 percent of the households retained eligibility for free lunch from
SY 1990-91 to SY 1991-92.
41
TABLE 6-2
CHANGES IN ELIGIBILITY STATUS FROM SY 1990-91 TO SY 1991-92
SY 1990-91
Eligibility Status
Free/Categorical
(N=137)
SY 1991-92 Eligibility Status
Free/
Categorical
<N=136)
Free/Income
(N=50)
16
Reduced
Price
(N=21)
Paid
(N=7)
0
Free/Income
(N=52) 15 24 10
Reduced Price
(N=13)
te?
Paid
(N=12)
The shaded numbers indicate households that retained the same eligibility status between years.
Free/Categorical and Free/Income households are eligible to receive free lunches.
Table 6-3 summarizes the eligibility status of all 214 households in SY 1990-91 and SY 1991-92
by pilot project. It is important to note the small sample sizes in each site. Only a small
number of households experienced changes in benefits between the two years. However,
Milford households experienced a 42 percent change in benefits. Of the 42 percent, 23 percent
lost eligibility for free lunches and 19 percent gained eligibility to receive free lunch. Overall,
the total number of households eligible for free meals remained quite stable from year to year.
Moreover, increases in eligibility appear to have offset reductions and losses of benefits.
Impacts on Program Integrity
The majority of the households, 91 percent, interviewed in SY 1990-91 and again in SY 1991-92
retained their eligibility to receive free lunches. Households who were not eligible to receive
free lunches in SY 1990-91 did have notable increases in benefits in SY 1991-92 and
demonstrate the potential impact on program integrity. However, these results should be
interpreted carefully because 1) the samples were small and 2) the data were collected only once
within each year, which makes it is difficult to truly assess household income fluctuations.
Moreover, in considering Extended Eligibility, it would be important to undertake a more
intensive data collection effort to assess household income fluctuations within and between
years.
42
Table 6.3
HOUSEHOLD CHANGES IN ELIGIBILITY STATUS BY SITE
FROM SY 1990-1991 TO SY 1991-1992
ST 19901991
ST 1991-1992 Eligibility Status
Eligibility Statua rree/Categorical Free/Income Reduced Price
s
Paid
Columbus (N=54)
Free/Categorical 34 2 0
Free/Income 8
Reduced Price 0
Paid 0
Lowell (N=47)
Free/Categorical 28 4
T-*r :•
Free/Income i±. 6 1
Reduced Price 0 0 6
Paid 0 1 jip.;*Zir±\ J .
Milford (N=31)
Free/Categorical 0
Free/Income
Reduced Price 0
Paid 0 n
National City (N=22)
Free/Categorical •7 'i
Free/Income .....:.-„... .. r^
i . ** «...
Reduced Price
M-.T!'"-'-
* r. . i«S*
I>.IJII.I>. ...
Paid 0 1 0
Philadelphia (N=18)
Free/Categorical
■ — 1 •-. • f •■l^iJiC ,■ ■ m, -.,, if. ■ ■■■„ ,„. Free/Income i
Reduced Price 0 0 ■ 0 • ;!*y§:
rrr
Paid 0
-—i—---.—•,; ,ar,—i
Free/Categorical
Free/Income
Terrell County (N-42)
iu
Tia -'■ ■ : ■-• a-.'-'i-'iratat.:? ;i : "A
Reduced Price 0 pi*-::*:*;.' 2
-iri..
Paid 0 1 1 0 '■:
Note: The gray boxes indicate households that retained the same eligibility status between years. The Free/Categorical and Free/Income
households are eligible to receive free lunches h