School Lunch Program: Cafeteria Managers' Views on Food Wasted by
Students (Letter Report, 07/18/96, GAO/RCED-96-191).

Pursuant to a congressional request, GAO provided information on food
waste from school lunches provided to school children under the National
School Lunch Program.

GAO found that: (1) school cafeteria managers had varying perceptions
about the degree to which food waste was a problem; (2) elementary
school cafeteria managers were more likely to perceive food waste as a
more serious problem; (3) the amount of food wasted varied by the type
of food, with cooked vegetables being wasted more often; (4) many
cafeteria managers believed that students' attention on recess or free
time, rather than lunch, contributed to waste; (5) many cafeteria
managers believed that allowing students to select what they wanted to
eat would reduce waste; and (6) most cafeteria managers were satisfied
with the federal commodities they received for use in the School Lunch
Program, but about 10 percent reported that they would rather not
receive at least half of the different types of commodities they
received under the program.

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

 REPORTNUM:  RCED-96-191
     TITLE:  School Lunch Program: Cafeteria Managers' Views on Food 
             Wasted by Students
      DATE:  07/18/96
   SUBJECT:  Food programs for children
             Children
             Elementary school students
             Secondary school students
             State-administered programs
             Nutrition research
             Surveys
             Public schools
             Statistical data
IDENTIFIER:  National School Lunch Program
             
******************************************************************
** This file contains an ASCII representation of the text of a  **
** GAO report.  Delineations within the text indicating chapter **
** titles, headings, and bullets are preserved.  Major          **
** divisions and subdivisions of the text, such as Chapters,    **
** Sections, and Appendixes, are identified by double and       **
** single lines.  The numbers on the right end of these lines   **
** indicate the position of each of the subsections in the      **
** document outline.  These numbers do NOT correspond with the  **
** page numbers of the printed product.                         **
**                                                              **
** No attempt has been made to display graphic images, although **
** figure captions are reproduced.  Tables are included, but    **
** may not resemble those in the printed version.               **
**                                                              **
** Please see the PDF (Portable Document Format) file, when     **
** available, for a complete electronic file of the printed     **
** document's contents.                                         **
**                                                              **
** A printed copy of this report may be obtained from the GAO   **
** Document Distribution Center.  For further details, please   **
** send an e-mail message to:                                   **
**                                                              **
**                                            **
**                                                              **
** with the message 'info' in the body.                         **
******************************************************************


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


Report to the Chairman, Committee on Economic and Educational
Opportunities, House of Representatives

July 1996

SCHOOL LUNCH PROGRAM - CAFETERIA
MANAGERS' VIEWS ON FOOD WASTED BY
STUDENTS

GAO/RCED-96-191

School Lunch Program

(150249)


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

  NSLP - National School Lunch Program
  SMSA - Standard Metropolitan Statistical Area
  OVS - Offer vs.  serve

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


B-271615

July 18, 1996

The Honorable William F.  Goodling
Chairman, Committee on Economic and
 Educational Opportunities
House of Representatives

Dear Mr.  Chairman: 

In fiscal year 1995, about 26 million students in about 94,000
schools (public and private schools and residential child care
institutions) nationwide were served lunches each day through the
National School Lunch Program (NSLP).  The program's total federal
costs for this period were over $5 billion--about $4.5 billion in
cash reimbursements and over $600 million in commodity foods, such as
beef patties, flour, and canned vegetables.  The schools
participating in the program had to offer lunches that included one
serving each of milk, meat or a meat alternate (such as peanut
butter), and bread or a bread alternate (such as pasta) and at least
two servings of vegetables and/or fruits.  Some concerns have been
raised about the amount of food provided in the school lunch program
that students throw away.  This discarded food is commonly referred
to as plate waste. 

To help address these concerns, you asked us to study plate waste in
the NSLP.  Specifically, we agreed to survey a random sample of
cafeteria managers in the public schools nationwide that participate
in the program to obtain the managers' perceptions on the (1) extent
to which plate waste is a problem, (2) amount of plate waste by type
of food, and (3) reasons for and ways to reduce plate waste.  We also
agreed to determine whether the perceptions of managers differed by
their school's level (elementary, middle, or high school), their
school's location (urban, suburban, or rural), and the proportion of
their school's lunches served free and at a reduced price.  In
addition, we agreed to ask cafeteria managers about their level of
satisfaction with the federal commodities used in the NSLP.\1

Our survey results represent the views of cafeteria managers in about
80 percent of the 81,911 public schools nationwide that participated
in the NSLP in the 1993-94 school year, the latest year for which a
comprehensive list of public schools was available.\2 All reported
differences between subgroups of respondents (e.g., cafeteria
managers in urban versus rural schools), unless otherwise stated, are
statistically significant.\3


--------------------
\1 In addition to our current work, we recently assessed the percent
of selected nutrients (calories, protein, saturated fat, and total
fat) wasted by students with various characteristics.  See Waste From
School Lunches (GAO/RCED-96-128R, May 8, 1996). 

\2 Approximately 94 percent of all public schools participated in the
NSLP in the 1993-94 school year. 

\3 A statistically significant difference means that the difference
between subgroups is too large to be attributed to chance. 


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

Cafeteria managers varied in the extent to which they perceived plate
waste as a problem in their school.  Although the majority perceived
plate waste as little or no problem, almost one in four reported that
it was at least a moderate problem.  By school level, managers in
elementary schools were more likely to perceive waste as a problem
than managers in middle or high schools.  By school location and by
schools serving different proportions of free and reduced-price
lunches, the extent to which managers viewed waste as a problem did
not differ. 

The amount of waste varied by the type of food included in the school
lunch, according to cafeteria managers.  For example, the average
amount of waste for cooked vegetables was 42 percent, compared with
11 percent for milk. 

Cafeteria managers strongly agreed on some of the reasons for and
ways to reduce plate waste.  For example, 78 percent cited students'
attention being on recess, free time, or socializing rather than
eating as a reason for waste.  Almost 80 percent believed that
allowing students to select only what they want to eat would reduce
plate waste.  We found few variations in the responses of cafeteria
managers concerning the reasons for and ways to reduce plate waste by
their school's level, location, and proportion of free and
reduced-price lunches served. 

Most cafeteria managers reported satisfaction with the federal
commodities they received for use in the school lunch program. 
However, about 10 percent reported that they would rather not receive
at least half of the different types of commodities provided. 


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

The NSLP is designed to provide school children with nutritionally
balanced and affordable lunches to safeguard their health and
well-being.  The program, administered by the U.S.  Department of
Agriculture's Food and Consumer Service, is available in all 50
states, the District of Columbia, and the U.S.  territories. 

The schools participating in the NSLP receive a cash reimbursement
for each lunch served.  In turn, the schools must serve lunches that
meet federal nutritional requirements and offer lunches free or at a
reduced price to children from families whose income falls at or
below certain levels.  For school year 1995-96, the schools were
reimbursed $1.795 for each free lunch, $1.395 for each reduced-price
lunch, and $0.1725 for each full-price lunch. 

Furthermore, for each lunch served, the schools receive commodity
foods--14.25 cents' worth in school year 1995-96.  The Department
provides a billion pounds of commodity foods annually to states for
use in the NSLP.  States select commodity foods from a list of more
than 60 different kinds of food, including fresh, canned, and frozen
fruits and vegetables; meats; fruit juices; vegetable shortening and
oil; and flour and other grain products.  The variety of commodities
depends on the quantities available and market prices.  According to
the Department, federal commodities account for about 20 percent of
the food in the school lunch program. 

Through school year 1995-96, the schools were required to offer
lunches that met a "meal pattern" established by the Department.  The
meal pattern specified that a lunch must include five items--a
serving of meat or meat alternate; two or more servings of vegetables
and/or fruits; a serving of bread or bread alternate; and a serving
of milk.  The meal pattern was designed to provide nutrients
sufficient to approximate one-third of the National Academy of
Sciences' Recommended Dietary Allowances. 

Effective school year 1996-97, the schools participating in the
program will be required to offer lunches that meet the Dietary
Guidelines for Americans.  Among other things, these guidelines,
which represent the official nutritional policy of the U.S. 
government, recommend diets that are low in fat, saturated fat, and
cholesterol.  In meeting these guidelines, the schools may use any
reasonable approach, within guidelines established by the Secretary
of Agriculture, including using the school meal pattern that was in
effect for the 1994-95 school year. 

All students attending the schools that participate in the NSLP are
eligible to receive an NSLP lunch.  In fiscal year 1995, about 58
percent of the eligible students participated in the program.  About
49 percent of the participating students received free lunches, 7
percent received reduced-price lunches, and 44 percent received
full-price lunches.  The students who do not participate in the
program include those who bring lunch from home, eat off-campus, buy
lunch a la carte at school or from a school canteen or vending
machine, or do not eat at all. 

Concerns about plate waste prompted the introduction into the NSLP of
the offer versus serve (OVS) option more than a decade ago.  Under
this option, a school must offer all five food items in the NSLP meal
pattern, but a student may decline one or two of them.  In a school
that does not use this option, a student must take all five items. 
All high schools must use the OVS option, and middle and elementary
schools may offer it at the discretion of local officials.  According
to a 1993 Department report, 71 percent of the elementary schools and
90 percent of the middle schools use the OVS option. 


   EXTENT TO WHICH CAFETERIA
   MANAGERS PERCEIVED PLATE WASTE
   AS A PROBLEM
------------------------------------------------------------ Letter :3

Cafeteria managers varied in the extent to which they perceived plate
waste as a problem in their school during the 1995-96 school year. 
Ninety percent of the managers provided an opinion on plate waste. 
The majority of those with an opinion did not perceive it as a
problem.  However, 23 percent of those with an opinion reported that
it was at least a moderate problem.\4 Figure 1 presents cafeteria
managers' perceptions of the extent to which plate waste was a
problem in their school. 

   Figure 1:  Extent to Which
   Cafeteria Managers Perceived
   Plate Waste From School Lunches
   as a Problem in Their School,
   1995-96 School Year

   (See figure in printed
   edition.)

Note:  This figure is based on the responses of the 90 percent of the
cafeteria managers who had an opinion on the extent to which plate
waste from school lunches was a problem in their school.  The
remaining 10 percent did not know whether plate waste was a problem. 

By school level, we found some variation in cafeteria managers'
perceptions of plate waste.  As figure 2 shows, managers at
elementary schools were more likely than those at middle or high
schools to report that plate waste from school lunches was at least a
moderate problem during the 1995-96 school year. 

   Figure 2:  Percent of Cafeteria
   Managers Reporting That Plate
   Waste From School Lunches Was
   at Least a Moderate Problem, by
   School Level, 1995-96 School
   Year

   (See figure in printed
   edition.)

Note:  The difference is not statistically significant between middle
and high schools. 

By school location and by schools serving different proportions of
free and reduced-price lunches, we found no statistically significant
differences in cafeteria managers' perceptions of plate waste. 

We also considered the extent to which cafeteria managers perceived
plate waste as a problem by asking them to compare the amount of
waste from school lunches with the amount of waste from packed
lunches from home.  Sixty-three percent of the managers were able to
make this comparison.  Of these, 79 percent believed that the amount
from school lunches was less than or the same as the amount from
packed lunches.  (See fig.  3.)

   Figure 3:  Cafeteria Managers'
   Comparison of the Amount of
   Plate Waste From School Lunches
   With the Amount of Waste From
   Packed Lunches From Home

   (See figure in printed
   edition.)

Note:  This figure is based on the responses of the 63 percent of the
cafeteria managers who could differentiate between the amounts of
plate waste from school lunches and from packed lunches from home. 
For the remaining 37 percent, 21 percent reported few, if any, packed
lunches at their school, while the other 16 percent could not tell
the difference between packed and school lunch waste.  The percents
in this figure do not add to 100 because of rounding. 


--------------------
\4 The sampling error for percents presented in this report is plus
or minus no more than 5 percentage points, unless otherwise indicated
in app.  I. 


   AMOUNT OF PLATE WASTE BY FOOD
   TYPE
------------------------------------------------------------ Letter :4

Cafeteria managers reported large variations in the amount of waste
from eight different types of food that may be included as part of
the school lunch.  For each food type, managers reported how much of
the portions served, on average, was wasted.  On the basis of the
managers' responses, we estimate that the average amount wasted
ranged from a high of 42 percent for cooked vegetables to a low of 11
percent for milk.\5 Figure 4 shows our estimate of the average
percent of waste for each of the eight food types. 

   Figure 4:  Amount of Food
   Portion Wasted, by Food Type

   (See figure in printed
   edition.)

Source:  GAO's analysis of survey data. 

By school level, the amount of waste varied for all food types except
canned or processed fruits.  In general, the waste reported for each
food type was highest in the elementary schools and lowest in the
high schools.  (See fig.  5.)

   Figure 5:  Variation in Amount
   of Waste, by School Level, for
   Seven Food Types

   (See figure in printed
   edition.)

Note:  Differences in the amount of waste are not statistically
significant between elementary and middle schools for fresh fruits
and between middle and high schools for meat alternates, meats, and
milk. 

Source:  GAO's analysis of survey data. 

By school location, the amount of waste varied for three food
types--cooked vegetables, raw vegetables/salads, and milk.  For
example, for each of these food types, the urban schools reported
more waste than the rural schools.  (See fig.  6.)

   Figure 6:  Variation in Amount
   of Waste, by School Location,
   for Three Food Types

   (See figure in printed
   edition.)

Note:  Differences in the amount of waste are not statistically
significant between urban and suburban schools for cooked vegetables
and between suburban and rural schools for raw vegetables/salad and
milk. 

Source:  GAO's analysis of survey data. 

By schools serving different proportions of free and reduced-price
lunches, the average amount of waste varied for four food types--raw
vegetables/salads, fresh fruits, canned or processed fruits, and
milk.  (See fig.  7.)

   Figure 7:  Variation in Amount
   of Waste, by Schools Serving
   Different Proportions of Free
   and Reduced-Price Lunches, for
   Four Food Types

   (See figure in printed
   edition.)

Note:  Differences in the amount of waste are not statistically
significant between schools serving under 30 percent free and
reduced-price lunches and schools serving 30 to 70 percent free and
reduced-price lunches for fresh fruits, canned or processed fruits,
and milk.  In addition, differences in the amount of waste are not
statistically significant between schools serving 30 to 70 percent
free and reduced-price lunches and schools serving over 70 percent
free and reduced-price lunches for fresh fruits and canned or
processed fruits. 

Source:  GAO's analysis of survey data. 


--------------------
\5 To estimate the percent of waste for each food type, we
substituted percents for the descriptions of the amount of food
students throw away or waste that were provided in the survey (i.e.,
"hardly any or none" is 0 percent, "less than half" is 25 percent,
"about half" is 50 percent, "more than half" is 75 percent, and "all
or almost all" is 100 percent).  We then averaged these percents for
all of the cafeteria managers responding to our survey. 


   REASONS FOR AND WAYS TO REDUCE
   PLATE WASTE
------------------------------------------------------------ Letter :5

When responding to a list of possible reasons for plate waste at
their school, the cafeteria managers most frequently selected a
nonfood reason--"student attention is more on recess, free time or
socializing than eating." When responding to a list of possible ways
to reduce plate waste, the managers most often viewed actions that
would involve students, such as letting students select only what
they want, as more likely to reduce plate waste than other actions. 


      REASONS FOR PLATE WASTE
---------------------------------------------------------- Letter :5.1

Seventy-eight percent of the cafeteria managers cited a nonfood
reason--students' attention on recess, free time, or
socializing--when asked why students at their school did not eat all
of their school lunch.  Figure 8 shows the percent of managers who
identified each of the nine reasons listed in our survey as either a
minor, moderate, or major reason for plate waste in their school. 

   Figure 8:  Reasons for Plate
   Waste Cited by Cafeteria
   Managers

   (See figure in printed
   edition.)

By school level, the percent of managers selecting a reason for plate
waste varied for four of the reasons provided in our survey.  (See
fig.  9.) For example, elementary school managers were much more
likely than middle or high school managers to report "amount served
is too much for age or gender" as a reason for plate waste. 

   Figure 9:  Variation in Reasons
   for Plate Waste Cited by
   Cafeteria Managers, by School
   Level

   (See figure in printed
   edition.)

Note:  The difference is not statistically significant between
elementary and middle schools for "attention on recess, free time,
socializing" and between middle and high schools for "attention on
recess, free time, socializing," "do not like that food," and "amount
served is too much for age or gender."

By school location, the percent of cafeteria managers selecting a
reason for plate waste varied for four of the reasons provided in our
survey.  (See fig.  10.) For example, managers at urban schools were
more likely than those at suburban and rural schools to report that
students "do not like that food" as a reason for plate waste. 

   Figure 10:  Variation in
   Reasons for Plate Waste Cited
   by Cafeteria Managers, by
   School Location

   (See figure in printed
   edition.)

Note:  The difference is not statistically significant between urban
and suburban schools for "take more than they can eat"; between urban
and rural schools for "attention on recess, free time, socializing,"
"not hungry," and "take more than they can eat"; and between suburban
and rural schools for "attention on recess, free time, socializing"
and "do not like that food."

By schools serving different proportions of free and reduced-price
lunches, cafeteria managers' perceptions differed somewhat for three
reasons.  For example, managers in schools serving under 30 percent
free and reduced-price lunches were more likely than managers in
schools serving over 70 percent free and reduced-price lunches to
cite "take more than they can eat" as a reason for plate waste.  (See
fig.  11.)

   Figure 11:  Variation in
   Reasons for Plate Waste Cited
   by Cafeteria Managers, by
   Schools Serving Different
   Proportions of Free and
   Reduced-Price Lunches

   (See figure in printed
   edition.)

Note:  The difference is not statistically significant between
schools serving under 30 percent and schools serving 30 to 70 percent
free and reduced-price lunches for all three reasons; between schools
serving under 30 percent and schools serving over 70 percent free and
reduced-price lunches for "attention on recess, free time,
socializing"; and between schools serving 30 to 70 percent and
schools serving over 70 percent free and reduced-price lunches for
"not enough time to finish eating."

In addition to asking cafeteria managers to respond to a list of
possible reasons for plate waste, we asked them to identify the
effect on plate waste of the NSLP's requirements for types of food
and serving sizes that were in effect at the time of our survey.  The
managers believed that, overall, the minimum federal serving sizes
provided about the right amount of food for the students at their
school.  (See fig.  12.)

   Figure 12:  Cafeteria Managers'
   Opinions on the Minimum Serving
   Sizes Required by the NSLP

   (See figure in printed
   edition.)

Note:  Percents do not add to 100 because of rounding. 

Furthermore, for each of four minimum serving size requirements that
were in effect at the time of our survey, most cafeteria managers
reported that each requirement did not result in more plate waste at
their school.  However, two requirements--serving at least
three-fourths of a cup of fruits/vegetables daily and serving at
least eight servings of breads/grains weekly--were viewed as
resulting in more plate waste by about one-third and one-quarter of
the managers, respectively.  Figure 13 shows the percent of cafeteria
managers who reported that the minimum serving sizes for the four
requirements resulted in more waste. 

   Figure 13:  Percent of
   Cafeteria Managers Reporting
   Increased Plate Waste at Their
   School Because of Minimum
   Serving Size Requirements

   (See figure in printed
   edition.)

In addition, we asked cafeteria managers about the potential effect
on plate waste of increasing the minimum serving sizes for
fruits/vegetables and breads/grains.\6 For fruits/vegetables, 62
percent of the middle and high school managers said that increasing
the amount from three-fourths of a cup to one cup daily would cause
more waste.  For breads/grains, 53 percent of the middle and high
school managers said that increasing the number of weekly servings
from 8 to 15 would increase plate waste; and 69 percent of the
elementary school managers reported that increasing the number of
servings of breads/grains from 8 to 12 weekly would cause more plate
waste. 


--------------------
\6 Regulations published in the Federal Register on June 13, 1995,
modified the meal pattern requirements by increasing the portion
sizes for fruits/vegetables and breads/grains according to grade
level.  The Healthy Meals for Children Act (P.L.  104-149, May 29,
1996) modified the National School Lunch Act to allow school food
authorities to use the meal pattern in effect for the 1994-95 school
year.  The use of this meal pattern will allow the schools to
continue to use serving sizes for fruits/vegetables and breads/grains
that were in effect prior to the June 13, 1995, regulations. 


      WAYS TO REDUCE PLATE WASTE
---------------------------------------------------------- Letter :5.2

Of 11 possible actions listed in the survey to reduce plate waste,
cafeteria managers viewed actions involving students in the choice of
food, such as letting students select only what they want and seeking
students' opinions regularly about menus, as more likely to reduce
plate waste than other actions.  (See fig.  14.)

   Figure 14:  Cafeteria Managers'
   Views on Ways to Reduce Plate
   Waste

   (See figure in printed
   edition.)

By school level, there was some variation in the views of cafeteria
managers for two of the actions to reduce plate waste listed in our
survey.  (See fig.  15.) For example, elementary school managers were
more likely than high school managers to identify "reduce federally
required portion sizes" as an action that would cause a little or a
lot less plate waste. 

   Figure 15:  Variation in
   Cafeteria Managers' Views on
   Ways to Reduce Plate Waste, by
   School Level

   (See figure in printed
   edition.)

Note:  The difference is not statistically significant between middle
and high schools for "reduce federally required portion sizes" and
"replace federal commodities with cash." For "replace federal
commodities with cash," the difference is statistically significant
between elementary and high schools but not between middle and high
schools.  Because there are more elementary schools than middle or
high schools in the sample, the statistical estimates for elementary
schools are more precise. 

By school location, there was some variation in the views of
cafeteria managers for four of the actions listed in our survey.  For
example, managers in urban schools were more likely than managers in
rural schools to cite "seek student opinions regularly about menus"
as an action that would cause less plate waste.  (See fig.  16.)

   Figure 16:  Variation in Ways
   to Reduce Plate Waste Cited by
   Cafeteria Managers, by School
   Location

   (See figure in printed
   edition.)

Note:  The difference is not statistically significant between urban
and suburban schools except for "start first lunch period later,"
between urban and rural schools for "start first lunch period later,"
and between suburban and rural schools for "provide nutrition
education for students" and "replace federal commodities with cash."

By schools serving different proportions of free and reduced-price
lunches, there was no variation in cafeteria managers' views on ways
to reduce plate waste.  Managers in each group--schools serving under
30 percent free and reduced-price lunches, schools serving between 30
and 70 percent free and reduced-price lunches, and schools serving
over 70 percent free and reduced-price lunches--had similar opinions
about the general level of effectiveness for the 11 potential actions
to reduce waste that were listed in the survey. 

In addition, most managers reported that two approaches already in
place in most schools result in less plate waste.  Eighty percent of
the managers said that the OVS option results in less waste, and 55
percent said that offering more than one main dish or entree daily
results in less waste. 


   CAFETERIA MANAGERS' LEVEL OF
   SATISFACTION WITH FEDERAL
   COMMODITIES
------------------------------------------------------------ Letter :6

Most cafeteria managers reported satisfaction with various aspects of
the federal commodities received at their school for use in school
lunches.  The managers' level of satisfaction was highest for the
taste and packaging of the commodities and lowest for the variety of
foods available and the quantity of individual commodities.  Figure
17 shows the percent of cafeteria managers who were satisfied, and
the percent who were dissatisfied, with the federal commodities
provided for school lunches. 

   Figure 17:  Cafeteria Managers'
   Satisfaction and
   Dissatisfaction With Federal
   Commodities Used for School
   Lunches

   (See figure in printed
   edition.)

Over 70 percent of the managers reported that they wanted all or
almost all of the different commodities received.  However, about 10
percent reported that they would prefer not to receive about half or
more of the different commodities they were sent.  (See fig.  18.)

   Figure 18:  Cafeteria Managers'
   Views on Receiving the
   Different Federal Commodities

   (See figure in printed
   edition.)


   AGENCY COMMENTS
------------------------------------------------------------ Letter :7

We provided copies of a draft of this report to the Department's Food
and Consumer Service for its review and comment.  We met with agency
officials, including the Deputy Administrator, Special Nutrition
Programs. 

Agency officials questioned why our survey results generalize to 80
percent, rather than 100 percent, of all the public schools that
participated in the NSLP in the 1993-94 school year.  Relatedly,
agency officials asked if we had analyzed the characteristics of
nonrespondents.  We generalized our results to 80 percent of the
public schools because we used a conservative statistical approach
that required us to generalize our results only to the overall level
reflected by our response rate, in this case 80 percent.\7 We did not
analyze the characteristics of nonrespondents because we believe that
such an analysis alone would not allow us to generalize our survey
results to 100 percent of the public schools that participated in the
NSLP in the 1993-94 school year.  To generalize to 100 percent of the
public schools, we believe it would also be necessary to analyze
information about perceptions of plate waste from a subsample of
cafeteria managers who did not respond to our survey.  This analysis
would allow us to assess whether the opinions of these managers
differed significantly from those of the managers who completed and
returned a survey. 

Further, the Department commented that our survey's list of possible
reasons for plate waste did not permit cafeteria managers to select
other possible reasons, including meal quality and palatability.  We
agree that these reasons may affect plate waste.  However, we
included two related reasons for plate waste--"they [students] do not
like that food" and "they [students] do not like the way the food
looks or tastes." We believe these two reasons address, in part, meal
quality and palatability.  In addition, respondents had the
opportunity to identify other reasons contributing to plate waste. 
Less than 5 percent of the respondents specified other reasons that
they considered to be at least a minor reason for plate waste. 

The Department also commented that we did not solicit the views of
children or their parents/caretakers.  We agree that the views of
cafeteria managers present only one perspective on the extent of, and
reasons for, plate waste and that valuable information could be
obtained from a comprehensive, nationwide study of the views of
children and their parents/caretakers.  The time and resources
associated with such a study could be substantial. 

In addition, the Department commented that our study did not address
whether there was more or less plate waste in the NSLP than in other
lunch settings--such as at home or in restaurants.  While identifying
the amount of waste in different lunch settings was not an objective
of our study, our survey asked cafeteria managers if they perceived
the amount of waste from school lunches as more, less, or about the
same as the amount of waste from lunches brought from home.  Our
survey results found that, of those cafeteria managers who were able
to assess differences in the amount of plate waste, 79 percent
believed that the amount from school lunches was less than or the
same as the amount from lunches brought from home. 

Finally, agency officials provided some technical and clarifying
comments that we incorporated into the report as appropriate. 


--------------------
\7 Our approach is consistent with that of W.G.  Cochran, Sampling
Techniques, 2nd ed.  (New York:  John Wiley & Sons, Inc., 1967), pp. 
355-357. 


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

To develop the questions used in our survey of cafeteria managers, we
reviewed the NSLP's regulations and research addressing the issue of
waste in the program.  Furthermore, we spoke with representatives
from school food authorities,\8 the American School Food Service
Association, and the Department's Food and Consumer Service.  We
refined our questions by pretesting our survey with the cafeteria
managers of 18 schools in Illinois, Pennsylvania, South Carolina,
Texas, Virginia, West Virginia, and the District of Columbia. 

We mailed our survey to a random sample of 2,450 cafeteria managers
in public schools in the 50 states and the District of Columbia.  We
selected our sample from the 87,100 schools listed in the National
Center for Education Statistics' Common Core of Data Public School
Universe, 1993-94, the latest year for which a comprehensive list of
public schools was available.  This document did not identify whether
a school participated in the NSLP.  Eighty percent (1,967) of those
surveyed returned a survey.  Of these, about 4 percent (80) reported
that their school did not participate in the NSLP, while the
remainder (1,887) reported that their school participated in the
program.  Our survey results generalize to 65,743 of the 81,911
public schools nationwide that participated in the NSLP in the
1993-94 school year.  This number may vary for individual questions,
depending on the response rate to the question. 

As with all sample surveys, our results contain sampling
error--potential error that arises from not collecting data from the
cafeteria managers at all schools.  Unless otherwise indicated in
appendix I, the sampling error for the survey results presented in
this report is plus or minus no more than 5 percentage points. 

Sampling error must be considered when interpreting differences
between subgroups, such as urban and rural schools.  All differences
we report are statistically significant unless otherwise noted. 
Statistical significance means that the difference we observed
between subgroups is too large to be attributed to chance. 

We conducted our review from July 1995 through June 1996 in
accordance with generally accepted government auditing standards.  We
did not, however, independently verify the accuracy of the cafeteria
managers' responses to our survey. 

Appendix II contains a more detailed description of our survey
methodology.  Appendix III contains a copy of our survey and
summarizes the responses. 


--------------------
\8 School food authorities are responsible for the management of
school food services at one or more schools and have the legal
authority to operate the NSLP. 


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

As agreed with your office, unless you publicly announce its contents
earlier, we plan no further distribution of this report until 7 days
from the date of this letter.  At that time, we will send copies of
this report to the appropriate congressional committees, interested
Members of Congress, the Secretary of Agriculture, and other
interested parties.  We will also make copies available to others on
request. 

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

Sincerely yours,

Robert A.  Robinson
Director, Food and
 Agriculture Issues


ESTIMATES WITH SAMPLING ERRORS
THAT EXCEED 5 PERCENT FOR PERCENTS
PRESENTED IN THIS REPORT
=========================================================== Appendix I

                (Sampling errors in percent)

                                        Estimate
                                               d  Sampling
Description                              percent     error
--------------------------------------  --------  --------
Middle school cafeteria managers            69.4       6.0
 reporting students "do not like that
 food" as reason for plate waste at
 their school (fig. 9)
High school cafeteria managers              71.4       6.2
 reporting students "do not like that
 food" as reason for plate waste at
 their school (fig. 9)
Middle school cafeteria managers            43.9       6.6
 reporting students "take more than
 they can eat" as reason for plate
 waste at their school (fig. 9)
High school cafeteria managers              60.5       6.9
 reporting students "take more than
 they can eat" as reason for plate
 waste at their school (fig. 9)
Middle school cafeteria managers            18.1       5.1
 reporting that "amount served is too
 much for age or gender" as reason for
 plate waste at their school (fig. 9)
High school cafeteria managers              23.2       5.9
 reporting "amount served is too much
 for age or gender" as reason for
 plate waste at their school (fig. 9)
Urban school cafeteria managers             58.1       5.6
 reporting "not hungry" as reason for
 plate waste at their school (fig. 10)
Suburban school cafeteria managers          48.4       5.3
 reporting "not hungry" as reason for
 plate waste at their school (fig. 10)
Urban school cafeteria managers             50.5       5.4
 reporting "take more than they can
 eat" as reason for plate waste at
 their school (fig. 10)
Cafeteria managers at schools serving       45.9       5.2
 over 70 percent free and reduced-
 price lunches reporting students
 "take more than they can eat" as
 reason for plate waste at their
 school (fig. 11)
Middle school cafeteria managers            41.4       6.2
 reporting "reduce federally required
 portion sizes" as a way to reduce
 plate waste (fig. 15)
High school cafeteria managers              43.4       6.7
 reporting "reduce federally required
 portion sizes" as a way to reduce
 plate waste (fig. 15)
Middle school cafeteria managers            47.9       7.7
 reporting "replace federal
 commodities with cash" as a way to
 reduce plate waste (fig. 15)
High school cafeteria managers              57.9       7.3
 reporting "replace federal
 commodities with cash" as a way to
 reduce plate waste (fig. 15)
Urban school cafeteria managers             43.9       6.2
 reporting "replace federal
 commodities with cash" as a way to
 reduce plate waste (fig. 16)
Suburban school cafeteria managers          47.5       5.7
 reporting "replace federal
 commodities with cash" as a way to
 reduce plate waste (fig. 16)
----------------------------------------------------------

SCOPE AND METHODOLOGY
========================================================== Appendix II

The Chairman of the House Committee on Economic and Educational
Opportunities asked us to study plate waste in the National School
Lunch Program (NSLP).  Specifically, we agreed to survey cafeteria
managers in public schools nationwide that participate in the NSLP to
obtain their perceptions on the (1) extent to which plate waste is a
problem, (2) amount of plate waste by type of food, and (3) reasons
for and ways to reduce plate waste.  We agreed to determine whether
the perceptions of managers differed by their school's level
(elementary, middle, or high school), their school's location (urban,
suburban, or rural), and the proportion of their school's lunches
served free and at a reduced price (under 30 percent free and reduced
price, 30 to 70 percent free and reduced price, or over 70 percent
free and reduced price).  In addition, we agreed to ask cafeteria
managers about their level of satisfaction with federal commodities
used in the program. 

To develop the questions used in our survey of cafeteria managers, we
reviewed the NSLP's regulations and research addressing the issue of
waste in the program.  Furthermore, we spoke with representatives
from school food authorities,\9 the American School Food Service
Association, and the U.S.  Department of Agriculture's Food and
Consumer Service.  We refined our questions by pretesting our survey
with the cafeteria managers of 18 schools in Illinois, Pennsylvania,
South Carolina, Texas, Virginia, West Virginia, and the District of
Columbia.  Generally, the questions on our survey concerned the
1995-96 school year. 

We mailed our survey to a random sample of 2,450 cafeteria managers
in public schools in the 50 states and the District of Columbia.  We
selected our sample from the 87,100 schools listed in the National
Center for Education Statistics' Common Core of Data Public School
Universe, 1993-94, the latest year for which a comprehensive list of
public schools was available from the National Center for Education
Statistics.  This document did not identify whether a school
participated in the NSLP.  We sent as many as two followup mailings
to each cafeteria manager to encourage response.  Eighty percent
(1,967) of those surveyed returned a survey.  Of these, about 4
percent (80) reported that their school did not participate in the
NSLP, while the remainder (1,887) reported that their school
participated in the program.  Our survey results generalize to 65,743
of the 81,911 public schools nationwide that participated in the NSLP
in the 1993-94 school year.  This number may be lower for individual
questions, depending on the response rate for the question. 

The results of our survey of cafeteria managers cannot be generalized
to schools that opened after school year 1993-94; to private schools;
to most residential child care institutions; to schools in the U.S. 
territories; and to schools represented by the survey nonrespondents. 

We matched the 1,887 survey responses to information about each
school in the Common Core of Data.  We used the Common Core of Data
to identify school location and to validate survey responses on
student enrollment and school level.  From this validation, we
determined that a number of the surveys were completed for the
surveyed school's district rather than for the individual school.  In
those cases, we used information from the Common Core of Data to
determine the surveyed school's level (e.g., elementary) and student
enrollment.  We assumed that the school served the same proportion of
free and reduced-price lunches as the district.  Unless otherwise
stated in the survey response, we also assumed that districtwide
opinions about plate waste applied to the surveyed school. 

Table II.1 shows the number of cafeteria managers responding to our
survey, by school level. 



                               Table II.1
                
                Number of Cafeteria Managers Responding,
                            by School Level

School level                     Number responding    Percent of total
------------------------------  ------------------  ------------------
Elementary                                   1,181                62.6
Middle                                         277                14.7
High                                           256                13.6
Other                                          173                 9.1
----------------------------------------------------------------------
Note:  We defined an elementary school as any school serving children
in grade 6 and under or any school serving children through grade 8
provided that the school also serves children in grade 3 or under. 
We defined a middle school as any school with a minimum grade level
of 4 through 8 and a maximum grade level of 7 through 9.  We defined
a high school as any school serving children only in grades 9 and up. 
Some schools, such as those serving children in kindergarten through
grade 12, did not meet these definitions, and we did not include them
in our analysis of differences by school level. 

Table II.2 shows the number of cafeteria managers responding, by
school location.



                               Table II.2
                
                Number of Cafeteria Managers Responding,
                           by School Location

School location                  Number responding    Percent of total
------------------------------  ------------------  ------------------
Urban                                          426                22.6
Suburban                                       472                25.0
Rural                                          989                52.4
----------------------------------------------------------------------
Note:  We defined urban as large and mid-size central cities of
standard metropolitan statistical areas (SMSAs).  We defined suburban
as the urban fringe of large and mid-size cities in SMSAs and large
towns not in SMSAs with populations of 25,000 or more.  We defined
rural as areas with populations of less than 2,500 as well as small
towns not in SMSAs. 

Table II.3 shows the number of cafeteria managers responding, by
schools serving different proportions of free and reduced-price
lunches. 



                               Table II.3
                
                Number of Cafeteria Managers Responding,
                by Schools Serving Different Proportions
                   of Free and Reduced-Price Lunches

Schools serving different
proportions of free and
reduced-price lunches            Number responding    Percent of total
------------------------------  ------------------  ------------------
Schools serving under 30                       451                23.9
 percent
Schools serving 30 to 70                       811                43.0
 percent
Schools serving over 70                        445                23.6
 percent
Undetermined                                   180                 9.5
----------------------------------------------------------------------
As with all sample surveys, our results contain sampling
error--potential error that arises from not collecting data from
cafeteria managers at all schools.  We calculated the sampling error
for each statistical estimate at the 95-percent confidence level. 
This means, for example, that if we repeatedly sampled schools from
the same universe (i.e., Common Core of Data) and performed our
analyses again, 95 percent of the samples would yield results within
the ranges specified by our statistical estimates, plus or minus the
sampling errors.  In calculating the sampling errors, we used a
conservative formula that did not correct for sampling from a finite
population.  The sampling error for most of the survey results
presented in this report is plus or minus no more than 5 percentage
points. 

Sampling error must be considered when interpreting differences
between subgroups, such as urban and rural schools.  For each
comparison of subgroups that we report, we calculated the statistical
significance of any observed differences.  Statistical significance
means that the difference we observed between two subgroups is larger
than would be expected from the sampling error.  When this occurs,
some phenomenon other than chance is likely to have caused the
difference.  Statistical significance is absent when an observed
difference between two subgroups, plus or minus the sampling error,
results in an interval that contains zero.  The absence of a
statistically significant difference does not mean that a difference
does not exist.  The sample size or the number of respondents to a
question may not have been sufficient to allow us to detect a
difference.  We used the chi square test of association to test the
significance of differences in percentages between two subgroups and
the t-test for differences in means. 

We conducted our review from July 1995 through June 1996 in
accordance with generally accepted government auditing standards.  We
did not, however, independently verify the accuracy of the cafeteria
managers' responses to our survey. 



(See figure in printed edition.)Appendix III

--------------------
\9 School food authorities are responsible for the management of
school food services at one or more schools and have the legal
authority to operate the NSLP. 


RESULTS OF SURVEY OF SCHOOL
CAFETERIA MANAGERS
========================================================== Appendix II



(See figure in printed edition.)



(See figure in printed edition.)



(See figure in printed edition.)



(See figure in printed edition.)



(See figure in printed edition.)



(See figure in printed edition.)



(See figure in printed edition.)



(See figure in printed edition.)



(See figure in printed edition.)



(See figure in printed edition.)


MAJOR CONTRIBUTORS TO THIS REPORT
========================================================== Appendix IV

Thomas Slomba, Assistant Director
Rosellen McCarthy, Project Leader
Sonja Bensen
Carolyn Boyce
Jay Scott
Carol Herrnstadt Shulman


*** End of document. ***