2

The School Meals Programs

This chapter describes criteria for eligibility and the process for application, certification, verification, participation, meal counting, and reimbursement in the National School Lunch Program (NSLP) and the School Breakfast Program (SBP), as well as the limitations of the current administrative process. It is essential to understand all elements of the school meals programs before considering alternative procedures that could reduce administrative burden and make it possible to provide nutritious meals to a greater number of the nation’s schoolchildren.1

ADMINISTRATIVE PROCESS OF THE SCHOOL MEALS PROGRAMS

This section describes the overall flow of the administration of the school meals programs and then provides detail on eligibility; certification; verification; participation; and counting, claiming, and reimbursement.

Process Flow

Figure 2-1 illustrates the flow of the school meals process, from determining the eligibility of students to serving them nutritionally qualified meals, noting that the distributions of students and meals

____________

1 This chapter draws heavily on Chapter 2 of the panel’s interim report (National Research Council, 2010).



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 16
2 The School Meals Programs T his chapter describes criteria for eligibility and the process for application, certification, verification, participation, meal counting, and reimbursement in the National School Lunch Program (NSLP) and the School Breakfast Program (SBP), as well as the limitations of the current administrative process. It is essential to understand all elements of the school meals programs before considering alternative procedures that could reduce administrative burden and make it possible to provide nutritious meals to a greater number of the nation's schoolchildren. 1 ADMINISTRATIVE PROCESS OF THE SCHOOL MEALS PROGRAMS This section describes the overall flow of the administration of the school meals programs and then provides detail on eligibility; certification; verification; participation; and counting, claiming, and reimbursement. Process Flow Figure 2-1 illustrates the flow of the school meals process, from determining the eligibility of students to serving them nutritionally qualified meals, noting that the distributions of students and meals 1This chapter draws heavily on Chapter 2 of the panel's interim report (National Research Council, 2010). 16

OCR for page 16
THE SCHOOL MEALS PROGRAMS 17 ET: All Students--True Eligibility (Unobserved) Certification process CT: Approved Students--Truth (Unobserved) Certification errors CO: Approved Students--Observed Participation MO: Meals Served--Observed Under Traditional Approach Participation Response to Universal Free Meals MU: Meals Served--Universal Free Meals (Unobserved) FIGURE 2-1 School meals process and distributions of enrolled students and FIG2-1.eps meals served across free, reduced-price, and full-price categories: Traditional approach and universal free meals. SOURCE: Prepared by the panel. served across the free, reduced-price, and full-price meal categories at each point differ. The first two boxes and the first oval in the figure reflect distributions based on all enrolled students; the second oval and last box relate to average daily meals served. For simplicity, we have assumed that the process depicted in the figure occurs instantaneously

OCR for page 16
18 USING ACS DATA TO EXPAND ACCESS TO THE SCHOOL MEALS PROGRAMS and have ignored how the distributions and the relationships among them change over time.2 The top box in the figure, labeled "ET: All Students--True Eligibility (Unobserved)," represents the distribution of all enrolled students by their true eligibility status, including those who are eligible for free meals using program rules described below; those who are eligible for reduced-price meals using those rules; and all other students, who are eligible only for full-price meals. As noted, this distribution is not observed. The process by which students are identified and approved as being eligible for free or reduced-price meals is known as certification.3 Students who are found to be eligible through the certification process become approved students. The second box in the figure, labeled "CT: Approved Students--Truth (Unobserved)," represents the distribution of all enrolled students accord- ing to a certification process with no errors. Some students who are eligible for free or reduced-price meals decline to participate in the certification process. All students who either do not apply or are not directly certified for free meals are eligible only for full-price meals, as are those students who apply but are found to be ineligible for free or reduced-price meals. The distribution, CT, is not observed. The number of students in the free category of CT will be less than or equal to the number in the free category of ET by the number of students who were not directly certified and who did not apply for benefits. Enrolled students who were not directly certi- fied and did not apply for benefits will be in the full-price category of CT. Likewise, students in the reduced-price category of ET who did not apply will be in the full-price category of CT. The first oval (and third item) in the figure, labeled "CO: Approved Students--Observed," represents the distribution of enrolled students into categories of approved for free meals, approved for reduced-price meals, and eligible for full-price meals, in which the categories of approved for free or reduced-price meals are as determined by the actual operation of the certification process and maintained in school records. The certifica- tion process is described below. The difference between CT and CO is due to misclassification of students (errors) during the certification process. For example, some students who are eligible for free meals may have been approved for reduced-price meals. 2As discussed below, a student paying full price at the beginning of the school year can be approved for free meals later in the year if, for example, the family's income falls. Once approved, the student can continue to receive free meals for the remainder of the year (and up to 30 days into the next year until a new eligibility determination is made), even if the family's income rises above the eligibility threshold for free meals. 3The certification process encompasses both direct certification by comparison of student enrollment lists with state and local lists of participants in several means-tested programs and the solicitation, submission, and review of applications.

OCR for page 16
THE SCHOOL MEALS PROGRAMS 19 On any given day, a student may bring a meal from home or pur- chase a meal that does not qualify for reimbursement because it does not satisfy the nutritional requirements of the school meals programs. Hence, schools must count the total number of reimbursable meals served each day and note whether each student taking a meal is approved for a free or reduced-price meal or must pay full price. The last two distributions in Figure 2-1 reflect the distribution of average daily reimbursable meals served across the three categories. The second oval (and fourth item) in the figure, "MO: Meals Served-- Observed Under Traditional Approach," represents the distribution of meals served across the free, reduced-price, and full-price categories in a school that uses the traditional procedures for certifying students and claiming reimbursement. While some students never participate (take meals) or participate on only some days, others participate every day. When students line up in the cafeteria with their trays, a cashier determines whether each meal served qualifies as reimbursable under the school meals programs according to such criteria as food group composition and serving size. The cashier determines whether the stu- dent is approved for a free or reduced-price meal in a way that ensures there will be no overt identification of the student's eligibility category. 4 This process provides the meal counts maintained in school records that are used to determine federal reimbursements in the school meals programs. The third box (and fifth item) in the figure, "MU: Meals Served-- Universal Free Meals (Unobserved)," represents the participation distri- bution when meals are provided free to all students. The distribution is unobserved because meals are not counted by category when they are provided free under a special operating provision such as Provisions 2 or 3 or the American Community Survey (ACS) Eligibility Option (AEO). The available evidence suggests that if meals are provided at no cost, more students participate. This distribution is important in assessing the costs and benefits of a new provision, and a primary objective of the panel was to determine whether there is a reliable and operationally feasible method for estimating this distribution for a school, group of schools, or school district using available data. 4The Richard B. Russell National School Lunch Act (Section 9B(10):3-22) states (10) No physical segregation of or other discrimination against any child eligible for a free lunch or a reduced-price lunch under this subsection shall be made by the school nor shall there be any overt identification of any child by special tokens or tickets, announced or published list of names, or by other means. Available: http://www.fns.usda.gov/cnd/governance/Legislation/NSLA_12-13-10.pdf.

OCR for page 16
20 USING ACS DATA TO EXPAND ACCESS TO THE SCHOOL MEALS PROGRAMS Eligibility Students are eligible for free school meals if their family's "current" income is at or below 130 percent of the poverty guideline for their family size. Current income requested on the application form "may be for the current month, the amount projected for the first month the application is made for, or for the month prior to application" (U.S. Department of Agriculture/Food and Nutrition Service, 2011b). Students are "cat- egorically eligible" for free meals if someone in the family participates in certain other means-tested public assistance programs targeting the low-income population. Specifically, students are categorically eligible for free meals if their families receive assistance from the Supplemental Nutrition Assistance Program (SNAP, formerly the Food Stamp Program), Temporary Assistance to Needy Families (TANF), or the Food Distribu- tion Program on Indian Reservations (FDPIR). A student also is categori- cally eligible if he/she is enrolled in a Head Start or Even Start program or is (1) a homeless child as determined by the school district's homeless liaison or by the director of a homeless shelter, (2) a migrant child as determined by the state or local Migrant Education Program coordinator, or (3) a runaway child who is receiving assistance from a program under the Runaway and Homeless Youth Act and is identified by the local edu- cational liaison. With the passage of the Healthy, Hunger-Free Kids Act of 2010, foster children also are categorically eligible for free meals. Students who are not eligible for free meals are eligible for reduced- price meals if their family's "current" income is greater than 130 percent of the poverty guideline and at or below 185 percent of the poverty guideline. All other students are eligible only for full-price meals (U.S. Department of Agriculture/Food and Nutrition Service, 2011b). Certification Certification is the process by which students are approved as being eligible for free or reduced-price meals. There are two types of certifica- tion: direct certification and the solicitation, submission, and review of applications. School districts, often through their state education agency, directly certify "categorically eligible" students based primarily on their participation in SNAP, TANF, or FDPIR. The 2004 Child Nutrition and WIC Reauthorization Act required that all school districts establish a system of direct certification of students from households that receive SNAP benefits by school year (SY) 2008-2009. Some states or districts also make use of TANF or other program data as part of direct certification. For direct certification, states or districts match lists of students (includ- ing names, addresses, and so on) with the administrative data concerning indi viduals participating in SNAP or other assistance programs. Students

OCR for page 16
THE SCHOOL MEALS PROGRAMS 21 matched in this way are "directly certified" as being eligible for free school meals. Parents are notified that their children are eligible and do not need to file an application. Matching for direct certification is done at least once a year, and beginning in 2011-2012 will be done three times a year.5 Some states and districts conduct direct certification more frequently to identify newly eligible students. For example, Washington State conducts direct certification monthly. In 2009-2010, an estimated 72 percent of students from SNAP-participant households nationwide were certified for free school meals through direct certification without applications. 6 In 2010- 2011, this number increased to 78 percent.7 As a result of errors in record matching or participation in a program for which a state does not perform direct certification, however, some categorically eligible students are not directly certified. Families of such students can establish their categorical eligibility by providing a SNAP, TANF, or FDPIR case number on their application for school meals.8 The application process begins just prior to and at the start of a school year (normally mid-July through early September), when school districts send a letter to the parents of their students describing the school meals programs, inviting them to apply, and providing an application form.9 The application requests information about participation in SNAP or other assistance programs, family composition, and family income. School or district officials review the applications and make a determination as to whether the students listed on the application should be approved for free or reduced-price meals. If an application lists a legitimate case number for SNAP or another approved program, the students are certified as being categorically eligible for free meals. While most applications are submitted at the beginning of the school year, applications and eligibility are in effect from the date of approval through the entire school year and up to 30 operating days into the subse- quent school year until a new eligibility determination is made. A family may submit an application at any time during the year, and it may do so later in the year if, for example, its income has fallen or it has started 5A 2011 interim rule issued by the U.S. Department of Agriculture (USDA) requires direct certification using SNAP records at least three times a year, beginning in 2011-2012. 6State-level direct certification rates for SNAP (excluding Alaska, with a direct certifica- tion rate over 100 percent) ranged from 47 percent to 91 percent. A SNAP direct certification rate may be overstated if the state also directly certifies using TANF or FDPIR records (U.S. Department of Agriculture/Food and Nutrition Service, 2010:14). 7In 2010-2011, state-level direct certification rates ranged from 51 percent to 97 percent (excluding Alaska) (U.S. Department of Agriculture/Food and Nutrition Service, 2011a:14). 8If the family provides a valid SNAP, TANF, or FDPIR case number on the application, they do not need to provide further information about family income. 9Some districts are moving to electronic applications.

OCR for page 16
22 USING ACS DATA TO EXPAND ACCESS TO THE SCHOOL MEALS PROGRAMS TABLE 2-1 Percentage of Enrolled Students by Approval Status for School Meals Programs, Fiscal Years (FY) 2005-2010 Approved for Approved for Must Pay Full Price Fiscal Year Free Meals (%) Reduced-Price Meals (%) for Meals (%) 2010 42.5 8.4 49.1 2009 40.1 8.6 51.3 2008 37.9 8.6 53.5 2007 37.1 8.3 54.6 2006 37.8 8.4 53.8 2005 37.1 8.1 54.8 NOTE: Approval status for the school meals programs includes both the National School Lunch Program (NSLP) and the School Breakfast Program (SBP). SOURCE: Tabulation from the Food and Nutrition Service (FNS) National Data Bank pro- vided to the panel, July 5, 2011. participating in SNAP or TANF, qualifying it for greater benefits under the school meals programs. The distribution of approved students by category for the school meals programs in fiscal years (FY) 2005 through 2010 is shown in Table 2-1. This is the CO distribution in Figure 2-1. It should be noted that not all families with students who are eligible for free or reduced-priced meals submit applications. In 1994, the Food and Nutrition Service (FNS) reported that "available data indicate that between 16 percent and 25 percent of potentially eligible families do not apply for school meals benefits" (U.S. Department of Agriculture/Food and Nutrition Service, 1994a:1-5). Although this may no longer be true in light of incentives (such as the allocation of funds in other programs using the school lunch eligibility percentage) and processes (such as direct certification) for certifying as many eligible students as possible for free meals, more recent estimates are not available. An FNS study enabled a comparison of the distributions of eligible and certified students (U.S. Department of Agriculture/Food and Nutri- tion Service, 1999). It used data from the Current Population Survey (CPS) to estimate the percentage of students who were income-eligible for free and reduced-price meals, providing a survey-based estimate for ET (see Figure 2-1), with eligibility based on annual income data. These estimates were compared with the numbers of students approved as eligible for free or reduced-price meals, CO. Table 2-2, taken from that report, indicates that the number of students certified was growing from 1993 through 1998, whereas the number eligible according to annual income was flat or declining. By 1998, the number of students approved for free meals was 127 percent of the number of students who were estimated to be income- eligible for free meals, and the number of students approved for free or

OCR for page 16
THE SCHOOL MEALS PROGRAMS 23 TABLE 2-2 Numbers of Students Eligible for the National School Lunch Program (NSLP) from Two Sources: (1) Current Population Survey (CPS) Estimates Based on Annual Income and (2) NSLP Certifications for Free and Reduced-Price Meals (in thousands), 1993-1999 Free Meals Free and Reduced-Price Meals CPS Certified/ CPS Certified/ Income- NSLP- Eligible Income- NSLP- Eligible Year Eligible Certified (%) Eligible Certified (%) 1999 12,464 15,876 127 18,928 19,260 102 1998 13,128 15,965 122 19,190 19,067 99 1997 13,461 15,799 117 19,416 18,762 97 1996 13,382 15,415 115 19,727 18,273 93 1995 13,655 14,920 109 20,030 17,577 88 1994 13,718 14,396 105 19,609 16,952 86 1993 13,924 13,792 99 19,750 16,273 82 SOURCE: U.S. Department of Agriculture/Food and Nutrition Service (1999:3, 5). reduced-price meals was 102 percent of the number of students who were estimated to be income-eligible for free or reduced-price meals, indicating the possibility of over certification in the school meals programs. As noted in a study by the National Research Council of the National Academy of Sciences, "results like this contributed to the Improper Payments Act of 2002, which requires that various federal agencies identify and reduce erroneous payments in their programs" (National Research C ouncil, 2009:14).10 Subsequent research found that at least some of the differ- ence between income eligibility estimated from the CPS and approval status under the school meals programs could be due to how income relative to poverty is measured (annual or monthly) and to changes in monthly income from the time of application to the time of verification (U.S. Department of Agriculture/Economic Research Service, 2006b). In response to the Improper Payments Act, FNS funded the Access, Participation, Eligibility, and Certification (APEC) Study in 2004 to obtain national estimates of the amounts and rates of erroneous payments in the NSLP and SBP (U.S. Department of Agriculture/Food and Nutrition Service, 2007b). Erroneous payments may be due to certification errors attributable to household misreporting or administrative mistakes or to 10As discussed in Chapters 3 and 5 of the 2009 National Research Council report, estimates of eligibility based on annual income are likely to be too low, given that families may have 1 or more months of low income that would qualify them for free or reduced-price meals even when their annual income exceeded the income eligibility limits.

OCR for page 16
24 USING ACS DATA TO EXPAND ACCESS TO THE SCHOOL MEALS PROGRAMS noncertification errors in counting and claiming payment for reimburs- able meals. The study used a complex sample design to survey school districts, schools, and students. The APEC study provided baseline estimates of erroneous payments for the 2005-2006 school year. It also provided parameters for estimation models to allow FNS staff to update estimates of erroneous payments. The study found that 77.5 percent of all certified students and denied appli- cants were correctly certified or denied meal benefits, whereas 22.5 per- cent were certified in error or erroneously denied benefits. The study also found that overcertification was more common than undercertifica- tion: the percentage of students certified for a higher level of benefits than that for which they were eligible (the overcertification rate) was 15 percent; the percentage of students either certified for a lower level of benefits than that for which they were eligible or erroneously denied benefits (the undercertification rate) was 7.5 percent. More detailed results from the APEC study are discussed later in this chapter. Verification In addition to special studies, such as the APEC study, the accuracy of the certification process is examined through a requirement for school districts to verify a sample of NSLP applications annually. Typically, a school district is required to conduct an annual verification of 3 percent or 3,000 (whichever is smaller) of the applications approved and on file as of October 1 of the current school year.11 Verification is to be completed by November 15 of the current school year. Samples are to be selected from "error prone" applications, those from families whose reported monthly income is within $100 of a school meals eligibility threshold (130 percent or 185 percent of the applicable poverty guideline). The households that submitted the applications selected for verification are required to submit documentation of income for any point in time between the month prior to application and the time of verification. School districts make at least one follow-up attempt with households that fail to respond. Students in households that fail to provide the required documentation are removed from eligibility. Results of verification studies are reported annually on form FNS-742. Data for each school district are reported through state agencies to FNS regional offices, which upload the data to FNS head quarters, where they are maintained. 11In some states, the state agency conducts the verification.

OCR for page 16
THE SCHOOL MEALS PROGRAMS 25 TABLE 2-3 Official National School Lunch Program (NSLP) Participation (average daily meals, divided by 0.927) and Percentage of Lunches Served by Approval Category, Fiscal Years (FY) 2005-2010 Percentage of Lunches Served NSLP Fiscal Year Participation Free Reduced Price Full Price 2010 31,746,374 55.4 9.5 35.1 2009 31,311,515 52.0 10.1 37.9 2008 31,015,551 49.6 10.1 40.3 2007 30,629,762 48.9 10.0 41.2 2006 30,128,292 49.0 9.8 41.2 2005 29,645,759 49.2 9.7 41.1 SOURCE: Tabulation from the Food and Nutrition Service (FNS) National Data Bank pro- vided to the panel, July 5, 2011. Participation Any student attending a school that participates in the school meals programs may obtain a meal for free or at the reduced price, if so approved, or by paying the full price for the meal. As noted earlier, cashiers assess which meals meet the nutritional requirements of the NSLP and SBP and, for qualifying meals, record each student's approval status (free, reduced price, full price) in a way that does not overtly identify the student's sta- tus. Meal counts by category are aggregated for each month for the school, the school district, and the state. This process provides the meal counts maintained in school records, which are also reported at the state level to FNS via form FNS-10. FNS defines participation to be the 9-month (September-May) average of each month's average daily meals served, divided by an attendance factor of .927 to account for absenteeism. This yields an estimate of the expected number of meals that would be served if students were never absent. Table 2-3 is from a special tabulation from the FNS National Data Bank that was provided to the panel in 2011. It shows participation in the NSLP by year and the percentage of meals served that were free, reduced- price, or full-price. The percentage distribution is MO in Figure 2-1.12 Another way of analyzing participation is to calculate a rate for each meal category (see Table 2-4). Dividing participation (average daily num- ber of meals served divided by .927) in a category by the total number 12The factor .927 is used by FNS to estimate what participation would be if students were never absent. FNS derives participation estimates by applying the assumption that all stu- dents, including those who are eligible for free and reduced-price meals, attend school at the same rate of .927. The panel did not use this factor in any of its analyses.

OCR for page 16
26 USING ACS DATA TO EXPAND ACCESS TO THE SCHOOL MEALS PROGRAMS TABLE 2-4 National School Lunch Program (NSLP) Participation Rates by Approval Category, Fiscal Years (FY) 2005-2010 Participation Rate (Percentage) Reduced Price Fiscal Year Free Approved Approved Full Price 2010 81.9 73.1 43.7 2009 80.5 72.8 45.9 2008 80.7 72.2 46.4 2007 80.8 73.5 46.2 2006 78.5 70.8 46.3 2005 79.1 71.5 44.9 NOTE: The participation rate is computed as average daily meals served in category divided by the product of .927 and the number of students certified in that category. The factor .927 is intended to account for the fact that not all enrolled students are at school every day. SOURCE: Tabulation from the Food and Nutrition Service (FNS) National Data Bank pro- vided to the panel, July 5, 2011. of enrolled students approved in that category shows consistently higher participation by students approved for free meals (81.9 percent in 2009- 2010), followed by students approved for reduced-price meals (73.1 per- cent in 2009-2010). Students having to pay full price participate at lower rates (43.7 percent in 2009-2010). Additional information on participation is available from the School Nutrition Dietary Assessment Study-III (SNDA-III) (U.S. Department of Agriculture/Food and Nutrition Service, 2007a). The main focus of the study was on assessing the nutritional content of school meals and identi- fying students' and parents' reasons for participation or nonparticipation. The study used the following two definitions of participation: (1) the percentage of enrolled students who took a meal that qualified under the school meals programs on a target day and (2) the percentage who "usu- ally" took such a meal, with "usually" being defined as 3 or more days per week. On a typical day in the 2004-2005 school year, about 62 percent of all students participated in the NSLP and about 18 percent in the SBP according to SNDA-III. Nearly three-quarters of students reported partici- pating in the NSLP on 3 or more days per week, and one-quarter reported participating in the SBP on 3 or more days per week. Parents of students who did not participate in the NSLP reported some of the same reasons for this decision as those given by students--for example, that their child did not like the cafeteria food (68 percent) or preferred to bring a lunch from home (65 percent). Table 2-5, based on SNDA-III, shows the percentage of enrolled stu- dents who participated in the NSLP on a target day in 2004-2005 sepa-

OCR for page 16
38 USING ACS DATA TO EXPAND ACCESS TO THE SCHOOL MEALS PROGRAMS reimbursement amount. Therefore, the reimbursement formula for Pro- vision 2 is Mf Mr Mp Gt2 = R f 0 ,t + R r 0 ,t + R p 0 ,t Mt M0 ,t M0 ,t M0 ,t = 2 BRRt Mt where G2t is the federal government's outlay for reimbursable meals served in month t in Provision 2 schools, in dollars; Rf, Rr, and Rp are reimbursement rates as defined above; Mt is the total number of reimbursable meals served during month t; Mf0,t is the total number of free meals served in month t of the base year; Mr0,t is the total number of reduced-price meals served in month t of the base year; Mp0,t is the total number of full-price meals served in month t of the base year; M0,t = Mf0,t + Mr0,t + Mp0,t is the total number of reimbursable meals served during month t of the base year; and BRRt2 is the blended reimbursement rate for Provision 2 in month t. The ratios in the first version of the equation above are the Provision 2 claiming rates, based on the percentage of meals served in each category in the base year. Like the BRR for the traditional approach, the BRR for Provision 2 varies from month to month. Under Provision 3, meals served by category are estimated by using meals served in the same month of the base year multiplied by a factor reflecting the change in enrollment and the number of operating days relative to the base year. Therefore, the reimbursement formula for Provi- sion 3 is Gt3 = R f M0f,t + R r M0 r , t + R M0 , t p p Mf Mr Mp = R f 0 ,t + R r 0 ,t + R p 0 ,t M0 ,t M0 ,t M0 ,t M0 ,t = BRRt M0 ,t 3 where

OCR for page 16
THE SCHOOL MEALS PROGRAMS 39 G3t is the federal government's outlay in month t for Provision 3 schools, in dollars; Rf, Rr, and Rp are reimbursement rates as defined above; Mf0,t is the total number of free meals served in month t of the base year; Mr0,t is the total number of reduced-price meals served in month t of the base year; Mp0,t is the total number of full-price meals served in month t of the base year; M0,t = Mf0,t + Mr0,t + Mp0,t is the total number of meals served dur- ing month t of the base year; d is a ratio adjustment (ratio of current-year to base-year value) reflecting changes in enrollment and the number of operating days (e.g., if enrollment increased by 5 percent since the base year and the number of operating days were unchanged, the factor would be 1.05); and BRRt3 is the blended reimbursement rate for Provision 3 in month t. The claiming percentages and BRR under Provision 3 are identical to the claiming percentages and BRR under Provision 2. Total reimburse- ments are different, however, because under Provision 2, schools count the number of meals served in each month (Mt), while under Provision 3, schools use the number of meals served in that month of the base year, adjusted only for changes in enrollment and operating days (using d), as an estimate of the meals served in the current month. Under the Community Eligibility Option, reimbursement is based on the total number of meals served, the ratio of the number of identified students26 to the number of enrolled students in the base year (or a year since the base year),27 and a factor specified in the Healthy, Hunger-Free Kids Act of 2010. The factor was set by the act at 1.6, and can be updated by the Secretary of Agriculture beginning with the 2014-2015 school year. 26Identified students are certified as eligible for free meals based on documentation of receipt of benefits or categorical eligibility as described in section 245.6a(c)(2) of Title 7, Code of Federal Regulations. They include students who are directly certified, on the home- less liaison list, income-eligible for Head Start or pre-K Even Start, in residential child care institutions, migrants, runaways, foster children certified through means other than an application, and other nonapplicants approved by local officials. 27The base year immediately precedes a district's implementation of the CEO. Under the CEO, districts may conduct direct certification on a yearly basis. If the most current data show an increase in the percentage of enrolled students who are identified, the district may use that percentage for determining the USDA reimbursement. If the data show a decrease, the district may continue to use the original percentage. (From a memorandum issued by the U.S. Department of Agriculture, Food and Nutrition Service, director of Child Nutrition Division, Cynthia Long, dated May 20, 2011.)

OCR for page 16
40 USING ACS DATA TO EXPAND ACCESS TO THE SCHOOL MEALS PROGRAMS The factor is intended to estimate the additional number of eligible stu- dents who would have been certified through the traditional application process. The CEO reimbursement formula uses just two rates--free and full price--and is I I GCEOt = R f + R p 1 - Mt E E = BRR CEO Mt where GCEOt is the federal government's outlay in month t for CEO schools, in dollars; Rf is the reimbursement rate for free meals as defined above; Rp is the reimbursement rate for full-price meals as defined above; I/E is the ratio of the total number of identified students (I) to total enrollment (E) as of April 1 of the base year or a subsequent year; j is a factor specified by the Secretary of Agriculture (currently I 1.6), and the product is restricted to being no greater than 100 percent; and E BRRCEO is the blended reimbursement rate for the CEO. Under the CEO, the claiming rate for free meals is the ratio of identi- fied students to enrolled students times the factor. The claiming rate for reduced-price meals is zero. The claiming rate for full-price meals is 1 minus the claiming rate for free meals. Using a socioeconomic survey (as in Philadelphia), a district com- bines the data from the survey with the number of directly certified students to estimate the percentage of enrolled students eligible for free, reduced-price, and full-price meals. The reimbursement formula is Ef rE r pE p Gt survey = R f + R + R Mt E E E = BRR survey Mt where Gtsurvey is the federal government's outlay established for the schools providing free meals to all students; Mt is the total number of reimbursable meals served in month t; Rf, Rr, and Rp are reimbursement rates as defined above;

OCR for page 16
THE SCHOOL MEALS PROGRAMS 41 Ef is the number of enrolled students who have been directly certified or estimated as eligible for free meals based on a survey of students' families; Er is the number of enrolled students who have been estimated as eligible for reduced-price meals based on a survey of students' families; E is the total student enrollment; Ep = E Ef Er is the number of enrolled students who are eligible for full-price meals; and BRRsurvey is the blended reimbursement rate for the schools that provide free meals to all students and use data from the survey to determine reimbursements (roughly two-thirds of the schools in Philadelphia). In this equation, the claiming percentages are the eligibility ratios Ef/E, Er/E, and Ep/E. Under the AEO the panel considered two potential reimbursement equations. The first is modeled after the option that relies on a socio economic survey and uses claiming rates based on eligibility: Ef Er Ep Gt ACS(1) = R f + R r + R p Mt E E E = BRR AEO (1) Mt where GtACS(1) is the federal government's outlay for reimbursable meals served in month t by AEO schools in dollars (the (1) denotes that this is the first version of the AEO proposed for using ACS data, and it uses eligibility estimates alone to define the claiming percentages); Rf, Rr, and Rp are reimbursement rates as defined above; Mt is the total number of reimbursable meals served in month t; Ef/E is the estimated fraction of enrolled students who are eligible for free meals based on the ACS and other sources; Er/E is the estimated fraction of enrolled students who are eligi- ble for reduced-price meals based on the ACS and other sources; Ep/E = 1 Ef/E Er/E is the estimated fraction of enrolled stu- dents who are eligible for full-price meals based on the ACS and other sources; and BRRAEO(1) is the BRR under the assumption that claiming percent- ages are based on student eligibility fractions estimated using the ACS and other information.

OCR for page 16
42 USING ACS DATA TO EXPAND ACCESS TO THE SCHOOL MEALS PROGRAMS The computations for BRRsurvey and BRRAEO(1) are the same, and these BRRs do not vary from month to month. However, these BRRs are based on different data. A local socioeconomic survey is used to estimate the eligibility-based claiming percentages in the former, and the ACS is used to estimate the eligibility-based claiming percentages in the latter. In light of the differences between the distributions of students by eligibility category in Table 2-1 and the distributions of meals served by eligibility category in Table 2-3, a concern with the above "enrollment- based" reimbursement equation--that is, an equation based on the dis- tribution of enrolled students--is that it might be unfair to districts. Spe- cifically, as illustrated in an example presented by FNS at the panel's first meeting,28 districts might receive smaller reimbursements than they would with a "participation-based" equation--that is, an equation based on the distribution of meals served. Therefore, the panel focused on a more general expression for the AEO reimbursement formula:29 Gt ACS(2) = R C + R C + R C f f r r p p Mt = BRR AEO (2) Mt where GtACS(2) is the federal government's outlay for reimbursable meals served in month t in AEO schools in dollars, with the (2) indicat- ing that this is the second version of the AEO considered by the panel, and it is based on estimated claiming percentages that account for both eligibility and participation; Rf, Rr, and Rp are reimbursement rates as defined above; Mt is the total number of reimbursable meals served in month t; Cf is the claiming rate for free meals, an estimate of the fraction of reimbursable meals served to students eligible for free meals; 28FNS gave a hypothetical example of a school with 70 percent of students eligible for free meals, 10 percent eligible for reduced-price meals, and 20 percent eligible for full-price meals. In this hypothetical school, however, 77.7 percent of meals were served to students eligible for free meals, 10 percent of meals to students eligible for reduced-price meals, and 12.3 percent of meals to students eligible for full-price meals. In this example, the average reimbursement per meal based on the eligibility distribution is $2.17, while the average reimbursement per meal based on the participation (meals served) distribution is $2.36. (In this situation, the school was eligible for the $.02 per meal increment, and the reimbursement rates for free, reduced-price, and full-price meals were $2.70, $2.30, and $.25, respectively.) 29A special case of this formula uses the enrollment percentages from the previous formula to estimate the claiming percentages.

OCR for page 16
THE SCHOOL MEALS PROGRAMS 43 Cr is the claiming rate for reduced-price meals, an estimate of the fraction of reimbursable meals served to students eligible for reduced-price meals; Cp = 1 Cf Cr is the claiming rate for full-price meals; and BRRAEO(2) is the BRR when claiming rates are based on both eligi- bility and participation. As indicated by the formula, the BRR does not vary from month to month. The claiming rate for a category is the estimated fraction of reimburs- able meals that are served to students who are eligible for that category, although meals would be provided free to all students. The three claiming percentages are the MU distribution in Figure 2-1, and, as noted earlier, one objective of the panel was to determine whether there is a reliable and operationally feasible method for estimating this distribution. 30 ERRORS IN METHODS FOR DETERMINING REIMBURSEMENTS Both the traditional method and the special provisions and options have limitations that result in errors in determining reimbursements. The limitations associated with the traditional method are described in the following section. The limitations associated with the special provisions are described in the final section. Traditional Method Currently, the majority of school districts use what we call the "tradi tional" method of operating the school meals programs. As described earlier, at the beginning of the school year, the district initiates a process in which parents are asked to apply for free or reduced-price meals by supplying their income and the number of household members or the information required to establish categorical eligibility (e.g., a SNAP case number).31 In this process, parents of students who are not directly certified need to apply in order for their children to receive the benefits of free or reduced-price meals. If a family that is eligible for these benefits does not apply and is not identified by direct certification, the students have been denied access to free or reduced-price meals to which they are entitled.32 30As shown in Chapter 5, meals served claiming percentages can be expressed in terms of the product of eligibility percentages and participation rates. 31An application does not need to be submitted if a student has been directly certified for free meals. 32This is not counted as a certification error in official statistics, however.

OCR for page 16
44 USING ACS DATA TO EXPAND ACCESS TO THE SCHOOL MEALS PROGRAMS Even if parents submit an application form for their children, they must complete it correctly. To do so, they must have an accurate under- standing of the program definitions of income and membership in the household. When parents are asked to report the number of household members, for example, they need to know that the count does not include foster children living in their household33 but does include relatives such as aunts or grandparents who are part of a student's economic unit. The parents need to know which forms of income should and should not be included and the correct dollar amounts for included forms. The applica- tion process further requires that parents apply these concepts accurately to their individual family situation. Once an application has been submitted, school or district officials must review it and determine whether the students in the family are eligible for free or reduced-price meals (or must pay full price). Even if the application is completely accurate, errors can be made at this stage in the certification process. Although the required annual verification of a sample of applications may reduce errors in the completion and review of applications, substantial certification errors still remain, as discussed below. Once a student has been approved for free or reduced-price meals or the application for such benefits is denied and the student must pay full price for a meal, the meal counting and claiming process begins. A school must retain daily records of the number of meals served for each eligibil- ity status by linking a reimbursable meal served to a student and then linking that student to his or her certified eligibility status. The school's daily records are compiled and submitted to the school district, and the school district submits them to the state. The state completes form FNS-10, providing the information that FNS uses to determine reimbursements. At each stage of this process, errors may occur. The APEC study (U.S. Department of Agriculture/Food and Nutri- tion Service, 2007b), discussed earlier, found that the certification process is especially prone to error, with approximately 9 percent of total reim- bursements for both the NSLP and SBP considered erroneous because of certification errors. The study reported on two sources of certification error: (1) household reporting errors and (2) administrative errors made by districts in processing applications. It established that 23.2 percent of all certified students and denied applicants had household reporting errors on their forms, while 8.3 percent were subject to administrative 33This statement was true when the panel began its work, but the policy has changed. According to U.S. Department of Agriculture/Food and Nutrition Service (2011b), foster children are now to be counted as part of the household.

OCR for page 16
THE SCHOOL MEALS PROGRAMS 45 error.34 (The two sources of error could occur on the same application and could have been offsetting.) Household reporting error led to over- certification for 13.5 percent of applications and undercertification for 9.7 percent of applications, while administrative error led to overcertifica- tion for 6.2 percent of applications and undercertification for 2.1 percent of applications.35 The most common type of household reporting error was misreporting of total income; this error affected 20 percent of certified students and denied applicants. Eight percent of certified students and denied applicants had errors in the number of household members listed on the form. The most common administrative error was certification of a student as eligible for free or reduced-price meals when the application was incomplete and should have been denied. According to the APEC study (U.S. Department of Agriculture/ Food and Nutrition Service, 2007b:53, vol. 1), roughly 14 percent of those approved as eligible for free meals should have been approved for a status with fewer benefits (8 percent for reduced-price and 6 percent for full-price meals). At the other end of the distribution, 36 percent of students whose applications were denied, and thus were required to pay full price, should have been approved for free or reduced-price meals (19 and 17 percent, respectively). Given the limited income range over which a student qualifies for reduced-price meals, approvals for that cat- egory are the most error prone. Roughly one-third of students approved for reduced-price meals should have been approved for free meals, and 25 percent should have had their applications denied. To quantify the potential effect of certification errors on the distri- bution of students by eligibility status when the traditional method is used, the APEC study compared the distribution of students based on the categories for which they had been approved with the distribution based on their true eligibility status, using the sample of students who had undergone the certification process and either had been certified for free or reduced-price meals or had their applications denied.36 The distri- bution based on approval status was 78 percent free, 17 percent reduced price, and 5 percent full price (U.S. Department of Agriculture/Food and Nutrition Service, 2007b:51), while the distribution based on true eligibil- ity status was 74 percent free, 14 percent reduced price, and 12 percent full price (U.S. Department of Agriculture/Food and Nutrition Service, 34Denied applicants--that is, applicants who are not approved for free or reduced-price meals--can still purchase meals at full price. 35Overcertification occurs when a student is certified for more benefits than those to which she or he is entitled. For example, a student approved for free meals is overcertified if she or he should have been approved for reduced-price or full-price meals. 36Because estimates were not obtained for students who did not apply, these distributions do not pertain to all enrolled students.

OCR for page 16
46 USING ACS DATA TO EXPAND ACCESS TO THE SCHOOL MEALS PROGRAMS 2007b:53). It was also estimated (U.S. Department of Agriculture/Food and Nutrition Service, 2007b:97) that the gross reimbursement error result- ing from certification errors in the NSLP was 9.4 percent of total reim- bursements (sum of absolute values of overpayments and underpayments divided by total cash and commodity reimbursement). Underpayment due to undercertification offset some of the overpayment due to overcertifica- tion, resulting in a net overpayment of 4.8 percent of total cash and com- modity reimbursements. The APEC study also evaluated noncertification errors, classified as cashier or aggregation37 errors. The study found that the process by which cashiers assess and record whether a meal is reimbursable is a substantial source of erroneous payments, particularly in the SBP, even though most schools had fairly low levels of cashier error. The high aggregate level of cashier error arose from a few large schools having very high levels of this type of noncertification error. However, it was conjectured that automated point-of-sale technology in place in most schools would minimize this type of error. Provisions and Options With the traditional method, the accuracy of reimbursements depends on four factors: 1. the correct certification of students as eligible for free or reduced- price meals (certification error); 2. the correct determination that a meal qualifies for reimbursement (cashier error); 3. the correct classification of each student taking a meal by approval category (free, reduced price, or full price) (cashier error); and 4. the summation of counts of meals served over cashiers and days, transmission of the school's meal counts by category to the school district, the state, and the federal government for reimbursement (aggregation error). For Provisions 2 and 3 and the AEO, these same factors would con- tribute to errors in the base year. The APEC study found that overcertifi- cation rates are higher and erroneous payments due to certification error 37Aggregation error is the sum of three potential errors: (1) the school does not sum meal counts correctly, (2) the school does not report to the district correctly, and (3) the district does not report to the state correctly. The error rate for the first type of error was very small, while the error rates for the second and third types were about 2 percent and 1.5 percent of NSLP reimbursements, respectively. The last two error types typically resulted in an overpayment to the school.

OCR for page 16
THE SCHOOL MEALS PROGRAMS 47 are more common in Provision 2 and 3 schools (in their base years) than in schools using the traditional method. Erroneous payments in the NSLP were approximately 1.75 percent larger for Provision 2 and 3 schools. With these provisions, any overstatement (or understatement) of claim- ing percentages for the base year will persist through subsequent years of their use until a new base year is established. The APEC study did not differentiate between schools in their first base year and subsequent base years, because the sample of Provision 2 or 3 schools was too small. In subsequent base years, there is likely to be more error because after 4 years of not taking applications, parents and school district staff have become less familiar with the application and verification procedures and less skilled in carrying them out. In light of the ongoing provision of free meals, some parents may not understand why applications need to be submitted, and may not submit applications at all or take the time to com- plete them accurately. School food service directors participating in the workshop hosted by the panel expressed concern about such problems arising when a new base year is established. Because the AEO has only one base year at the beginning of the process, the challenges associated with subsequent base years will not obtain. After the first base year, the reimbursements under Provisions 2 and 3 and the AEO include any base-year errors. Under Provisions 2 and 3, a new base year may be established, possibly resulting in increased errors for reasons discussed above. In the years between base years under Provi- sion 2 and after the base year under the AEO, any additional errors due to factors 1 and 3 are eliminated, leaving possible errors in determining that a meal qualifies for reimbursement (factor 2) and in compiling and trans- mitting the information (factor 4). In the years between base years under Provision 3, any additional errors due to any of the factors are eliminated, although there may be aggregation error in reporting to the district and state. In addition, Provisions 2 and 3 are subject to errors due to the fact that claiming percentages or meal counts are fixed and will not reflect changes in eligibility or participation rates, a source of error that does not arise under the traditional method. Because the recommended procedures for implementing the AEO include a base year only at the beginning of the process, the difficulties associated with redoing a base year after sev- eral years of free feeding with no applications will not be encountered. The AEO relies on the ACS for an annual update and thus accounts for changes in a district's socioeconomic conditions, albeit with a lag. The panel is not aware of an analysis of the accuracy of the CEO and its impact on reimbursements. This special provision is new, having been implemented in three states during school year 2011-2012. Potential errors under the CEO include direct certification errors and errors associated with using the factor of 1.6 to account for eligible students who are not

OCR for page 16
48 USING ACS DATA TO EXPAND ACCESS TO THE SCHOOL MEALS PROGRAMS identified through direct certification (or lists used for identification). As with other provisions and options, errors under the CEO also include errors in counting total meals and compiling and transmitting data. While the CEO may not keep up to date with changes in the economic status of the community, the school district has the option of conducting direct certification in any year and using the new results if they would lead to an increase in reimbursement.