United States General Accounting Office GAO Report to the Chairman, Subcommittee on the District of Columbia, Committee on Government Reform and Oversight, House of Representatives August 1997 DISTRICT OF COLUMBIA PUBLIC SCHOOLS Student Enrollment Count Remains Vulnerable to Errors GAO/HEHS-97-161 United States GAO General Accounting Office Washington, D.C. 20548 Health, Education, and Human Services Division B-275273 August 21, 1997 The Honorable Thomas M. Davis Chairman, Subcommittee on the District of Columbia Committee on Government Reform and Oversight House of Representatives Dear Mr. Chairman: The District of Columbia Public Schools (DCPS) is one of the largest public school districts in the country—ranked 32nd in school year 1993-94,1 on the basis of student count, out of nearly 15,000 districts. During school year 1996-97, DCPS operated 158 elementary and secondary schools and reported its official student enrollment to be 78,648. DCPS’ budget was more than $570 million in fiscal year 1997—16 percent, or $90 million, of which came directly from federal government programs providing grants and funds.2 As early as school year 1989-90, questions about the accuracy of DCPS’ enrollment count arose when the previous year’s enrollment count was approximately 6,400 fewer students than had been reported. Even before school year 1989-90, school system figures were at odds with other independent counts. In 1984 and again in 1986, the Department of Education’s Office for Civil Rights released figures differing from DCPS’ official counts by more than 5,000 and 2,000, respectively. More recent critics have found discrepancies when comparing the reported enrollment with census estimates and with the District’s general population decline. In addition, various reviewers and auditors have questioned several aspects of DCPS’ enrollment count process. These criticisms raise concern because an accurate enrollment count is the cornerstone of a school district’s financial needs assessments. Although in the past DCPS did not receive funds on the basis of the number of students enrolled, new budget initiatives will soon directly link DCPS’ funding to school enrollment. Consequently, a valid enrollment count process and an accurate count are critical to DCPS’ district- and school- level planning, staffing, funding, and resource allocation. 1 The latest year for which official rankings are available. 2 The annual appropriation act passed by the Congress provided the rest; the act makes funds (including locally generated revenues and the federal payment) available to the District of Columbia for obligation and expenditure. Page 1 GAO/HEHS-97-161 DCPS Enrollment Count B-275273 In response to these issues, you asked us to examine the enrollment count process that DCPS used in school year 1996-97 to determine whether the process appeared sufficient to produce an accurate count. In addition, you asked us to examine enrollment count processes used by some other urban school systems and to determine the role of the U.S. Department of Education’s Inspector General in preparing DCPS’ official enrollment count for school year 1996-97. To conduct this review, we obtained relevant documents from and we interviewed DCPS administrative staff, city officials, officials in other urban school districts3 and their state departments of education, officials in the Department of Education, and education experts. To gain an understanding of the process’ implementation, we visited 15 DCPS elementary and secondary schools, randomly selected by school level (elementary, junior high or middle school, and senior high) and city quadrant (Northeast, Northwest, Southeast, and Southwest).4 During our school visits, we interviewed principals, school administrative staff, and teachers and reviewed selected documents. Further information on our scope and methodology appears in appendix I. Even though DCPS changed parts of its enrollment count process in school Results in Brief year 1996-97 to address criticisms, the process remains flawed. Some of these changes, such as the use of an enrollment card to verify attendance, increased complexity and work effort but did little to improve the count’s credibility. Errors, including multiple enrollment records for a single student, remained in the Student Information System (SIS), but DCPS had only limited mechanisms for correcting these errors. For example, though Management Information Services (MIS) personnel maintain SIS, they have no authority to correct errors. Furthermore, DCPS’ practice of allowing principals to enroll unlimited out-of-boundary students increases the possibility of multiple enrollment records for one student. Problems also persisted in the critical area of residency verification. In school year 1996-97, schools did not always verify student residency as 3 The other districts were Boston and Chelsea, Massachusetts; Fairfax County and Arlington County, Virginia; and Montgomery County and Prince George’s County, Maryland. We interviewed officials and examined documents they provided, but we did not confirm whether their processes were implemented as they described. 4 The schools visited were Anacostia High School, Bowen Elementary School, Coolidge High School, C.W. Harris Elementary School, Deal Junior High School, Eliot Junior High School, Miner Elementary School, Montgomery Elementary School, Noyes Elementary School, Phelps Career High School, Randle Highlands Elementary School, Roosevelt High School, Terrell Elementary School, Thomson Elementary School, and Tubman Elementary School. Page 2 GAO/HEHS-97-161 DCPS Enrollment Count B-275273 required by DCPS’ own procedures. Proofs of residency, when actually obtained, often fell short of DCPS’ standards. Moreover, Central Office staff did not consistently track failures to verify residency. Finally, school staff and parents rarely suffered sanctions for failure to comply with the residency verification requirements. In addition, the pupil accounting system failed to adequately track students. SIS allowed more than one school to count a single student when the student transferred from one school to another. In addition, schools did not always follow attendance rules, and SIS lacked the capability to track implementation of the rules. Furthermore, some attendance rules, if implemented, could have allowed counting of nonattending students. Other school districts report that they use several approaches to control errors, such as the ones we identified in DCPS, and to improve the accuracy of their enrollment counts. These include using centralized enrollment and pupil accounting centers and a variety of automated student information system edits and procedures designed to prevent or disallow pupil accounting errors before they occur. Finally, the recently enacted District of Columbia School Reform Act of 19955 requires the enrollment count process to produce enrollment numbers for nonresidents and students with special needs. DCPS (acting on behalf of the District of Columbia Board of Education) and the District of Columbia Financial Responsibility and Management Assistance Authority (Authority) are not in compliance with requirements of this new law, including a specified annual report on the enrollment count and its subsequent independent audit. The U.S. Department of Education helped DCPS develop its Request for Proposals (RFP) for the independent audit of the enrollment count for school year 1996-97, but it had no role in preparing DCPS’ official enrollment count for school year 1996-97. The Authority subsequently decided, however, that auditing the count for school year 1996-97 would be counterproductive. Unlike other U.S. school districts, DCPS, due to its location in the nation’s Background capital, has a unique administrative environment. Because Washington, D.C., is not located in a state, DCPS does not benefit from the oversight and assistance often provided by states. Furthermore, recent organizational changes in both the city and its school system have changed administration of the schools. 5 Title II of P.L. 104-134 (1996). Page 3 GAO/HEHS-97-161 DCPS Enrollment Count B-275273 To reform the District’s school system, the Congress recently passed the District of Columbia School Reform Act of 1995, which includes requirements for counting District students. Counting student enrollment, a process involving several interconnected elements, is usually fundamental to assessing funding needs and required of most other U.S. school districts. DCPS’ enrollment count process in school year 1996-97 was centered in the local schools and modified somewhat to address criticisms. DCPS’ Administrative DCPS lacks the state-level oversight that most other school districts in the Environment Is Unique country have. The state’s role in school operations is an important one. States generally provide guidance to their school districts on important issues, including student enrollment counts. The state determines the rules to be used in the enrollment count—who should be counted, by what method, and when. States also distribute state and federal funds to their districts, usually on the basis of enrollment, and they routinely audit various school district operations, including the enrollment count. The governance of DCPS had been performed for many years by an elected Board of Education. In November 1996, however, the specially appointed District of Columbia Financial Responsibility and Management Assistance Authority (Authority) declared a state of emergency in DCPS and transferred DCPS management—until June 30, 2000—to the Authority’s agents, a nine-member, specially appointed Emergency Transitional Education Board of Trustees.6 In so doing, the Authority transferred to the Board of Trustees “. . . all authority, powers, functions, duties, responsibilities . . .” of the former Board of Education7 (with some exceptions not relevant to this report). Meanwhile, the Authority also replaced DCPS’ superintendent with a Chief Executive Officer/ Superintendent. These changes have resulted in a shift of control from elected officials toward those appointed for a specific purpose: to reform the system. Early reform initiatives have included the administrative reorganization of DCPS and the closure of 11 schools. Even before the Authority’s takeover of DCPS, the Congress, relying on its plenary power to legislate for the District of Columbia, acted directly to reform DCPS. In April 1996, the Congress passed the District of Columbia School Reform Act of 1995, calling for the calculation of the number of 6 The president of the Board of Education is one of nine members of the Board of Trustees. 7 The Board of Education retained an advisory role on general educational policy and remains the chartering authority for Public Charter Schools under the School Reform Act. Page 4 GAO/HEHS-97-161 DCPS Enrollment Count B-275273 students enrolled in DCPS. The law requires the District of Columbia Board of Education8 to do the following: • calculate by October 15 of each year the number of students enrolled in the District’s public schools and students whose tuition in other schools is paid by DCPS funds, including students with special needs9 and nonresident students, in the following categories by grade level if applicable: kindergarten through grade 12, preschool and prekindergarten, adult students, and students in nongrade level programs; • calculate the amount of fees and tuition assessed and collected from nonresident students in these categories; • prepare by October 15 and submit to the Authority, the Comptroller General of the United States, appropriate congressional committees, and others an annual report summarizing those counts; and • arrange with the Authority to provide for the conduct of an independent audit of the count. Within 45 days of the Authority’s receipt of the annual report—or as soon thereafter as is practicable—the Authority is to submit the independent audit report to the appropriate congressional committees. Most School Districts Must The requirement to count students is common to most other U.S. school Count Student Enrollment districts. Forty-one of the 50 states10 use some type of direct student count to assess resource needs and to distribute state funds to their school districts. Enrollment counts also usually determine budgets and resource allocations to the individual schools. Three basic methods are used for counting enrollment. One method— called Enrolled Pupils (often called ENR)—counts all enrolled students on a specified day of the year. Definitions of “enrolled students” vary among districts, but they usually include elements of attendance. That is, students must be in attendance at least once during some preceding time period. ENR is used by 12 states and the District of Columbia. Another similar method is called Pupils in Average Daily Membership (often called ADM). This method, used by 22 states, calculates the average of total enrollment 8 The Resolution that established the Board of Trustees specifically transferred all Board of Education responsibilities regarding the District of Columbia School Reform Act to the newly created Board of Trustees. 9 The law defines a “student with special needs” as a student who is a child with a disability under the Individuals With Disabilities Education Act, or a student who is an individual with a disability under the Rehabilitation Act of 1973 as amended. 10 Seven states use teacher or instructional units rather than student count, and two states use other methods to distribute funds. Page 5 GAO/HEHS-97-161 DCPS Enrollment Count B-275273 figures over a specified time period.11 A third method, called Pupils in Average Daily Attendance (often called ADA), calculates the average total daily attendance over a specified time period. Seven states use this method. Enrollment counts may occur several times throughout the school year in response to both state and local information needs and may use different counting methods depending on the purpose of the count. For example, officials in one district reported that they perform a count about 5 days after school opens, using the ENR method. The district uses this count to make final adjustments to school-level resource allocations for the current school year. On September 30, the district conducts the first of three state-required enrollment counts, also using the ENR method. The state uses this count to assess compliance with state quality standards (such as pupil/teacher ratios) and to estimate enrollment before the March 31 count. On March 31, the district conducts the second state-required count, this time using the ADM method. The state uses this count to distribute state funds. Finally, the district conducts the third state-required enrollment count at the end of the school year, also using the ADM method. The state uses this count as a final report on enrollment for the entire school year. In addition to fulfilling reporting requirements, the school district uses the state-required enrollment counts for local planning and monitoring purposes. States vary in their approach to monitoring and auditing their districts’ enrollment counts. Some states do little monitoring or auditing of their districts’ counts, while others stringently monitor and audit. For example, one state simply reviews district enrollment reports for the fall and spring and contacts districts if large discrepancies exist. In contrast, another state not only conducts an electronic audit of its districts’ spring and fall official enrollment counts, but also visits districts and examines a random sample of student records in detail. School district officials in this state reported that the state withdraws from its districts state funds paid for students improperly enrolled or retained on the rolls. Regardless of when the count is performed or by what method, whether audited or not, accuracy is critical. A student count may be inaccurate if it has problems in any of at least three critical areas: enrollment, residency verification, and pupil accounting. Enrollment and residency verification take place when a student enters the school system. They determine a student’s initial eligibility and therefore who may potentially be included 11 Maryland and Massachusetts require ENR; Virginia requires ADM. Page 6 GAO/HEHS-97-161 DCPS Enrollment Count B-275273 in the count. Pupil accounting refers to the tracking of students after initial enrollment. Monitoring student attendance, status, and transfers in and out of school are all part of pupil accounting, which often involves an automated student database. The pupil accounting system provides the basis for determining continued eligibility to be counted—based upon a student’s attendance—and it helps determine which school may count a particular student in its enrollment. Accuracy of DCPS’ Critics have often charged that the District’s reported official enrollment Enrollment Count Has numbers have been overstated. One reviewer asserted, for example, that Been Questioned results of the 1990 U.S. census suggest that the District’s school-age population in 1990 might have been as much as 13,000 less than DCPS’ official enrollment count. Subsequent reviewers, including a certified public accounting firm, the Office of the District of Columbia Auditor, and us, examined the process that DCPS used to count pupils in school years 1994-95 and 1995-96 and found flaws. These flaws included DCPS’ lack of documentation to support enrollment status and lack of sanctions if false enrollment information was provided. These reviewers also reported that DCPS lacked adequate procedures to verify residency and that the student database had errors, including duplicate records, incomplete transfers, and incorrect enrollment status. For a more detailed discussion of audit findings and recommendations, see appendix II. DCPS’ Enrollment Count DCPS’ process for enrolling, verifying residency of, and tracking students Process in School Year remained centered in the local school in school year 1996-97, while central 1996-97 Remained office staff monitored portions of the process. To respond to past criticisms, DCPS instituted some changes for school year 1996-97, including Centered in the Local new forms, residency verification procedures, and additional preparatory Schools counts. The actual official enrollment count was done manually, and school principals were ultimately responsible for ensuring the accuracy of their schools’ counts. Local Schools Conducted DCPS’ local schools conducted all enrollment activities in school year Enrollment Activities 1996-97 for new and returning students, and the schools’ principals made all determinations about enrollment eligibility. Principals were allowed to enroll students who lived outside school boundaries without limitation. Principals could also temporarily enroll students who had not provided evidence of meeting eligibility criteria, including health certificates and proofs of District of Columbia residency. Upon completion of initial paperwork, the schools’ data entry clerks created an electronic record for Page 7 GAO/HEHS-97-161 DCPS Enrollment Count B-275273 each newly enrolled—or temporarily enrolled—student in SIS.12 The system maintained records for returning students from the previous school year, and the records were updated during the summer with promotion information. Similarly, withdrawals were processed during the summer, and these records were removed from the schools’ rolls. Figure 1 shows the enrollment count process for school year 1996-97. 12 SIS has been used for the past 3 years, replacing the former Student Information Membership System (SIMS). It contains biographical, demographic, academic, and status information on each student enrolled in DCPS. Page 8 GAO/HEHS-97-161 DCPS Enrollment Count B-275273 Page 9 GAO/HEHS-97-161 DCPS Enrollment Count B-275273 Figure 1: DCPS Enrollment Count Process for School Year 1996-97 Summer September Responsibility 1996 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 Transfers/New Students Homeroom SIS SIS SIS SIS Returning Rosters Rosters File File File Student Generated Prepared Updated Updated Updated Records for to Show to Convert With Updated as Preliminary SIS Shows New "No-Shows" Addresses of September Count All Students Enrollees to "Inactive" 2 as "Active" Individual School Student Residency Enrollment Cards Forms Completed Completed and Submitted With Proofs Attendance Rosters Reviewed for "No-Shows" School Provides Daily Membership Counts to Central Office Central Office Initial Monitoring Page 10 GAO/HEHS-97-161 DCPS Enrollment Count B-275273 September October December 20 21 22 23 24 25 26 27 28 29 30 1 2 3 4 5 8 9 10 11 12 Transfers/New Students SIS File SIS Reconciled Rosters Official With Generated Count Enrollment for Performed Cards and Official Updated Count Preliminary Sept. 3 - First Day of School Count Performed Sept. 19 - Preliminary Count Using SIS Oct. 3 - First Official Count Rosters Dec. 11 - Official Count Released Preliminary Counts Residency Forms Official Count Forwarded Forwarded Forwarded Central Office Central Reconciled Office and Released Aggregated Count Numbers Interim Enrollment Monitoring Final Enrollment Monitoring Residency Validation Visits New Procedures in School Year 1996-97 The process in school year 1996-97 incorporated the use of a new enrollment card designed to address auditors’ concerns about validating enrollment status. Students were to complete two copies of the enrollment Page 11 GAO/HEHS-97-161 DCPS Enrollment Count B-275273 card on the first day of attendance, and teachers were to sign and certify the cards. A completed card was to serve as proof that a child had appeared the single day required to be considered enrolled. In addition to serving as proof of enrollment status, the card was to be used to update SIS. Process Required Evidence of In addition to the enrollment card, DCPS’ enrollment process for 1996-97 Residency required all students to provide evidence of District of Columbia residency. If the student provided no evidence, DCPS’ rules allowed the student to enroll, but the student was to be assessed tuition. Tuition for a full-time program for school year 1996-97 ranged from $3,349 to $7,558, depending on grade level. Providing evidence of District of Columbia residency was required as part of revised DCPS procedures for school year 1996-97 to answer critics who charged that DCPS’ process for verifying residency was inadequate. In previous years, only students entering DCPS schools for the first time would have been required to submit proof of residency. A new form, the Student Residency and Data Verification Form, which had been piloted at selected schools during the previous school year, was to be completed for all students during school year 1996-97. Students were expected to have their parents or guardians complete the form and return it to the school with proofs of residency attached. Schools were to give students 3 days to complete and submit the form and proofs. Within 10 days, the school was to provide one copy of the form to the Nonresident Tuition Enforcement Branch of the Central Office along with a list of those students for whom residency had not been verified. The Nonresident Tuition Enforcement Branch was responsible for assessing and collecting tuition. Pupil Accounting Also Resided In addition to enrollment and residency verification procedures, local in the Local Schools schools also tracked student attendance, status, and transfers in school year 1996-97. Each of DCPS’ schools had online access to school data, and the schools’ data entry personnel (or enrollment clerks) were responsible for ensuring data were accurate and up to date. The MIS Branch, however, in the Central Office, managed the overall database. Classroom or homeroom teachers took attendance once a day, and data entry staff recorded it in SIS. Transfers were often done electronically, with transfer procedures initiated by the losing school and completed by the gaining school, although a manual back-up transfer process was also available. Page 12 GAO/HEHS-97-161 DCPS Enrollment Count B-275273 Monitoring Activities Included Monitoring activities for school year 1996-97 focused exclusively on Use of Enrollment Cards and overseeing the schools’ implementation of the enrollment card and on Identifying Nonresidents identifying nonresidents. During the early part of the school year, DCPS’ Central Office staff visited each of the schools three times to monitor enrollment cards. Eighteen members of the Central Office staff were temporarily reassigned to monitor the cards. Staff paid the first monitoring visit within the first 2 weeks of school and focused on the extent to which schools were following the process, that is, distributing and completing enrollment cards and filing them in the appropriate locations. Staff paid the interim monitoring visit before the official enrollment count and manually tallied students, comparing the enrollment cards, SIS reports, and the preliminary count documents. Staff paid the final monitoring visit after the October 3 count and were to verify that names on the enrollment cards matched those on SIS homeroom rosters. Nonresident students of the District of Columbia were to be identified through local schools’ monitoring of the completed data verification forms. The Nonresident Tuition Enforcement Branch was to investigate cases the schools identified. In addition, staff from this branch were to visit the schools to survey cars transporting students to and from school, identifying all out-of-state license plates. The monitors were also to review enrollment cards and residency verification forms to determine if the forms indicated residency issues. The branch was to investigate all identified cases and assess tuition for students found not meeting the District’s residency requirements. DCPS Used ENR to Count As previously mentioned, for school year 1996-97, DCPS used the ENR Students method to count its students—counting all enrolled students on a single day—October 3, 1996. Students did not have to attend school on this day to be included in the count because enrollment records were counted— not actual students. DCPS defined an “enrolled student” as any student who had appeared at school at least once—and who had not withdrawn from DCPS—between the beginning of the school year on September 3, 1996, and October 3, 1996, the day of the count. DCPS’October 3, 1996, count was conducted manually by each homeroom teacher using homeroom rosters prepared from SIS. School staff compiled the count, classroom by classroom, and recorded the numbers on the school’s official report. The Central Office received the schools’ reports, and schools’ data were aggregated by the Office of Educational Page 13 GAO/HEHS-97-161 DCPS Enrollment Count B-275273 Accountability (OEA),13 which prepared the official enrollment report.14 Each school’s principal was to ensure not only the accuracy of the school’s manual count, but also the enrollment, residency, and pupil accounting data that supported it. DCPS’ policy for the October 3, 1996, count called for unspecified rewards and sanctions to be applied on the basis of the extent to which staff maintained and reported accurate, up-to-date information. Beyond the official October count, DCPS also performed other counts throughout the year using this same process. These included official counts in December and again in February. The February count aided in computing projections for school year 1997-98. In addition to these counts, DCPS began two new preparatory counts this year. Each school took daily enrollment counts and communicated them by telephone to the Central Office every morning for the first 11 days of the school year. In addition, in September, each school completed a preliminary count using forms established for the official October 3 count. DCPS’ new student enrollment card was intended to document that School Year 1996-97 students had met the 1-day attendance requirement for inclusion in the Enrollment Process official enrollment count. Although the card may have met this Changes Did Not requirement in some respects, it appears to have burdened both school and DCPS staff and may not offer much advantage over more traditional Ensure Accuracy methods of documenting attendance, such as teachers’ attendance tracking. Perhaps even more importantly, the card alone did not ensure that enrollment records were correct before the count. The card did not address a critical problem—one revealed by prior audits—a lack of internal controls of the student database. This problem allowed multiple records to be created for a single student. Furthermore, DCPS continued to include in its enrollment some categories of students often excluded in official enrollment counts used for funding purposes in other states. In contrast to DCPS procedures, officials in other school districts reported using various strategies for ensuring accuracy and minimizing duplicate records. 13 OEA was dissolved in winter 1997 by the DCPS’ new Chief Executive Officer/Superintendent as part of his administrative reorganization. Many of OEA’s functions were transferred to the office of the Chief Academic Officer. 14 District of Columbia Public Schools, FY 1996-97 Official Membership, October 3, 1996, DCPS (Washington, D.C.: Dec. 1996). Page 14 GAO/HEHS-97-161 DCPS Enrollment Count B-275273 Enrollment Card Appears Teachers and school staff reported that DCPS’ new enrollment card was to Be Burdensome and burdensome and difficult to implement. Each child, on the first day of Inefficient attendance, had to complete and sign two separate copies15 of the card. However, many students—primarily the very young, disabled, or non- English speaking—could not complete the card themselves because they could not read or write at all or do so in English. In these cases, teachers had to complete the enrollment cards, although the students were asked to sign the cards when possible. Teachers, particularly in the primary grades, reported that completing the cards was troublesome for them, adding to their paperwork burden. Furthermore, the legitimacy of a child’s signature as a method of validation—particularly when the child cannot read or write—is questionable.16 In addition, the enrollment card did not contain vital enrollment information needed by the schools, such as emergency contact numbers. Consequently, it could not substitute for other enrollment forms that schools had been using. Several of the schools we visited augmented the enrollment card with other forms to obtain needed information. Consequently, the busy school staff had to complete and manage multiple forms to collect and maintain basic enrollment data. Moreover, the procedures that DCPS established for completing the enrollment card were difficult to implement after the first days of the school year. The procedures, which required the teacher to certify the student’s signature, were designed for the initial few days of school when an entire class enrolled together and could complete the form in the teacher’s presence. No provision had been established for students arriving later, who normally enroll at the school office. School staff in the schools we visited reported that they could not sign the card for the teacher, and obtaining the teacher’s signature and certification for these late enrollments was sometimes difficult. As a result, the process sometimes failed when enrollment cards for late enrollees were not completed or signed and certified by teachers. Finally, DCPS officials reported that Central Office monitoring for implementation of the new enrollment card was labor intensive. Enrollment card monitoring efforts did not use statistical sampling. Instead, we were told, monitors visited all the schools on three separate 15 One of the cards was to stay with the teacher, and one was forwarded to the school’s office to be maintained in files by classroom. 16 The six districts we visited have similar attendance requirements but do not require this type of validation, depending instead upon attendance data to prove that the student was present. Page 15 GAO/HEHS-97-161 DCPS Enrollment Count B-275273 occasions, often reviewing 100 percent of the enrollment records. To perform this task, monitoring teams were formed, without regard to their normal responsibilities, from available staff within the former OEA, according to DCPS officials. During our review, we could not confirm the extent of these enrollment card monitoring visits because DCPS could not provide us with any of the monitoring reports prepared on the basis of these visits. Lack of Internal Controls The procedures that DCPS used for enrolling students in school year Allowed Multiple 1996-97 allowed multiple records to be entered into SIS for a single student. Enrollment Records When school staff entered a new record, the SIS processing procedure automatically queried the database for any matching names and dates of birth. If a match occurred—as would be the case if the student had previously enrolled in a DCPS school—SIS informed the person entering the data that a record already existed for an individual with that name and date of birth. SIS, however, also provided the option of overriding the system and creating a new record for the student. DCPS officials reported that some data entry personnel were choosing this override capability and creating the new record. With safeguards overridden and additional records created, two schools could have each had access to a separate record for the same individual, allowing both schools to count the student. DCPS’ mechanisms for resolving this error were limited. Although Central Office MIS personnel maintained SIS, they had no authority to correct the errors once detected. Only the local school had such authority. MIS personnel had limited influence over the schools to ensure that corrections were made quickly and accurately, according to DCPS officials. Furthermore, while duplicate record checks were done, officials told us, the checks were not done on a regular, routine schedule. In addition, individuals who had helped with data quality control in the past as well as those who had monitored attendance were moved in early 1997 to facilities without office telephones or data lines.17 DCPS’ practice of allowing schools to enroll, without restriction, students who live outside school attendance boundaries increased the possibility of a student’s having multiple enrollment records for school year 1996-97. Students did not have to enroll in the school serving the geographic area where they lived but could enroll in any DCPS school if the principal allowed. For example, a student could have gone first to the school 17 DCPS officials told us that they had no plans to relocate the personnel or install telephone lines before the collocation of DCPS offices in the spring of 1998. Staff members, instead, will continue to be expected to travel to a local school when they need to perform SIS data runs. Page 16 GAO/HEHS-97-161 DCPS Enrollment Count B-275273 serving his or her area, filled out an enrollment card, and been entered into SIS. Subsequently, the student may have gone to another school, filled out another enrollment card, and—if the person entering this record in SIS chose to override the safeguard—been entered into SIS a second time. In addition, some principals reported that schools actively sought to attract out-of-boundary students to increase their enrollment. Enrollment Procedures DCPS’ official enrollment count of 78,648 included not only regular Allow Count of Students elementary and secondary students, but also other categories of students Not Counted in Some excluded from enrollment counts in other districts when the counts are used for funding purposes. For example, DCPS included in its enrollment Other Districts count students identified as tuition-paying nonresidents of the District of Columbia and students above and below the mandatory age for public education in the District, including Head Start participants,18 prekindergarten students (age 4), preschool students (age 0 to 3), and some senior high and special education students aged 20 and older.19 In contrast, the three states we visited reported that they exclude any student who is above or below mandatory school age or who is fully funded from other sources from enrollment counts used for funding purposes. Furthermore, even though the District of Columbia Auditor has suggested that students unable to document their residency be excluded from the official enrollment count, whether they pay tuition or not, DCPS included these students in its enrollment count for school year 1996-97. Other Districts Report In contrast with the DCPS process, students in the Boston and Chelsea, Using Various Methods to Massachusetts, school districts enroll at central Parent Information Help Improve Enrollment Centers (PIC), which are separate and independent from the schools, officials told us. Individual schools in these two districts cannot enroll Accuracy new students, we were told. All enrollment activities, including assignment of all students to schools, take place at PICs. Boston’s PICs were established as a key part of the U.S. District Court’s desegregation plan to alleviate the Court’s concerns about the accuracy of Boston’s reported enrollment numbers and to satisfy the Court’s requirements for credibility and accountability in pupil enrollment, assignment, and accounting. Centralizing student enrollment at PICs has helped reduce errors, according to officials in both districts. For example, staff in Boston have 18 Head Start has its own funding source. 19 The Reform Act requires separate reporting of some of these groups but does not require that they be included in aggregate counts. Page 17 GAO/HEHS-97-161 DCPS Enrollment Count B-275273 specialized in and become knowledgeable about the process. Limiting access to the student database has also helped to reduce errors. For example, in Boston, only six people may enter data into the database. Furthermore, PICs prevent students from being enrolled at two or more schools simultaneously, reducing duplicate counting and preventing schools from inflating their enrollment. In the other four districts we visited, schools—rather than a central site—usually handle student enrollment,20 but they use other safeguards. To enroll, a student goes to the school serving the geographic area in which he or she lives. Out-of-boundary enrollment is not usually allowed.21 In addition, officials in all four of these districts reported having student database safeguards to aid enrollment accuracy. For example, all four districts have procedures and edits in their student databases that automatically block the creation of duplicate enrollment records. If an enrolling student has attended another school in the district, these procedures will not allow a new record to be created once the old record has been located. School staff, officials told us, cannot override this blocking mechanism. In addition, Prince George’s County has a procedure in its student database that automatically checks student addresses with school attendance boundaries as enrollment information is entered. If the address falls outside the enrolling school’s boundaries, the database blocks enrollment. During school year 1996-97, District of Columbia schools had features that New Residency attracted nonresidents. Elementary schools in the District had free all-day Verification prekindergarten and kindergarten, and some elementary schools had Procedures Were Not before- and after-school programs at low cost. One school we visited had before- and after-school care for $25 per week. This program extended the Completely school day’s hours to accommodate working parents—the program began Implemented at 7 a.m. and ended at 6 p.m. In addition, several high schools had highly regarded academic and artistic programs; and some high schools had athletic programs that reportedly attracted scouts from highly rated colleges. Furthermore, students could participate in competitive athletic 20 Officials in the Fairfax County school district said that new students with limited-English proficiency and new students with disabilities enroll and verify residency at central service centers designed to help students with special needs. 21 Arlington County school district officials told us that after enrolling, their students are allowed to seek transfers to out-of-boundary schools for academic reasons. Schools may accept such transfers as long as they do not exceed 4 percent of their total enrollment. Fairfax County and Prince George’s County also allow attendance at out-of-boundary schools under special circumstances, with permission of the central administration. Page 18 GAO/HEHS-97-161 DCPS Enrollment Count B-275273 programs until age 19 in the District, compared with age 18 in some nearby jurisdictions. DCPS established new procedures for school year 1996-97 to detect nonresidents and collect tuition from those who attended DCPS schools, but both school and Central Office staff failed to implement the new procedures completely.22 In addition, DCPS failed to monitor and enforce its new procedures effectively. Schools Failed to Verify Most of the schools we visited failed to comply with the new residency and Report Residency for verification process. As discussed previously, all students’ parents or legal All Students as Required guardians had to complete a Student Residency and Data Verification Form (residency form) and provide at least two proofs of residency. by DCPS Policy Students were told that failure to provide either the completed residency form or proofs would result in an investigation of their residency, and, if appropriate, either tuition payments or exclusion from DCPS. Most of the schools we visited, however, did not obtain completed residency forms for all their students. In fact, only 2 of the 15 schools had—or reported having—residency forms for 100 percent of the student files we reviewed. In addition, schools did not collect all required proofs of residency. Students and their families presented two proofs of residency in only isolated cases, and many students submitted no proofs. In many other cases, the proofs that the schools collected did not meet the standards established by DCPS and printed on the residency form. Although the residency form specified proofs of residency, such as copies of deeds, rental leases, utility bills, or vehicle registrations, as acceptable, schools sometimes accepted proofs such as newspaper or magazine subscriptions, copies of envelopes mailed to the student’s family, stubs from paid utility bills with no name attached, and informal personal notes (rather than leases or rental agreements) from individuals from whom the family reportedly rented housing. We also found some instances in which the names or addresses on the proof did not match those on the form. School staff often complained to us about the difficulty they had trying to get students to return completed residency forms and proofs. Some acknowledged that they placed little emphasis on this effort. Schools we visited also varied in their compliance with the requirements to report residency issues to OEA. Schools were supposed to forward copies of all students’ completed residency forms to OEA. These copies 22 Audit reports of the Oct. 1995 count also criticized DCPS for failure to verify student residency. Page 19 GAO/HEHS-97-161 DCPS Enrollment Count B-275273 were to be attached to a list of students whose residency was considered questionable. Some schools sent copies of their student residency forms along with the list as required. Others sent the proofs with the forms. At least six schools sent no verifications of residency to the Central Office. Some of these implementation issues may have resulted from poorly specified requirements and procedures. For example, though DCPS officials reported to us that the requirements were for at least two proofs of residency, we found no written documentation communicating to the school staff or to the students a requirement for more than one proof. DCPS officials also gave us conflicting information about the number of proofs required. At one meeting, we were told that three proofs were required; at a later meeting, that two to three were required. Similarly, DCPS’ guidance to the schools did not specify how the schools were to maintain their students’ completed residency documentation—or even exactly what documentation was to be maintained. Consequently, schools’ maintenance of residency documents varied considerably. For example, about one-third of the schools we visited maintained the residency forms alphabetically; the remaining schools grouped them by classroom. The schools’ disposition of the proofs of residency varied even more. Eight schools filed proofs of residency with the students’ completed residency forms; one filed the proofs in the students’ permanent (cumulative) record folder; one filed them either with the completed form or in the folder; one placed all proofs in a file drawer without annotating them to permit subsequent identification of the student to whom they belonged; two forwarded all proofs to OEA, along with copies of the completed form; and two schools had no proofs at all for the student records we reviewed. And, because procedures did not provide for the schools to document the proofs on the residency forms, schools not retaining the proofs with the forms could not demonstrate that they had adequately verified residency. Other audits of schools’ compliance with residency verification would face similar obstacles because of the schools’ inability to link student records with proof of residency. Central Office Did Not Monitors for student residency, in general, did not report the level of Effectively Monitor or school and student noncompliance that we observed in our review. For Enforce Residency the nine schools for which we could directly assess compliance,23 with few exceptions, proofs of residency were missing for large portions of the Verification Procedures 23 We could assess compliance only for those schools for which we could track the proof of residency to the residency form. Page 20 GAO/HEHS-97-161 DCPS Enrollment Count B-275273 student population. But, most DCPS Daily Activity Reports (monitoring reports) failed to cite the missing proofs, focusing instead on students who lived with someone other than a parent or whose forms indicated a nonresident address or phone number. For example, in one school we visited, we determined that about one-fourth of the students (or 108) did not return a proof to the school. The DCPS monitoring report, however, identified only one student living with a grandmother and two students with nonresident addresses. In another school, we found no proofs, and staff reported that they could not get students to provide proofs. But the monitoring report showed that only two students had nonresident addresses or phone numbers. Moreover, DCPS officials did not provide monitoring reports for 3 of the 15 schools we visited, telling us that it only prepared monitoring reports for schools where issues of nonresidency had been identified on enrollment cards or residency verification forms. At one of the three schools without monitoring reports, we found no proofs of residency on file for any student. Some of the monitors’ failure to detect and report residency problems may have resulted from poorly specified guidance. Instructions to monitors were not specific enough to guide implementation, for example, asking monitors to identify students for whom parents had not “sufficiently documented” residence. Monitoring instructions did not specify what to examine to determine whether residency was documented or what documentation was considered sufficient. Furthermore, despite recommendations of previous audits, monitors had no instructions to review the files to determine whether students had submitted a residency form. Consequently, when monitors failed to compare names on the student roster with those on completed residency forms, DCPS missed a key element in determining school and student compliance. We found forms missing for at least some of the students at 13 of the 15 schools we visited. At one school, the staff estimated that about 25 to 30 percent of the students did not return the residency forms, and, at another school, the staff could not find about one-third of the forms. Despite monitoring efforts and threats of sanctions, DCPS administration did not ensure that the schools completed the residency verification procedures. DCPS conducted no follow-up of schools failing to submit the office copy of the residency form. In addition, on the basis of the reports from the schools we visited, it conducted only minimal follow-up of schools failing to collect adequate proofs. Furthermore, as noted earlier, DCPS conducted no follow-up of those schools failing to collect residency Page 21 GAO/HEHS-97-161 DCPS Enrollment Count B-275273 forms for all students because no one in the Central Office checked to see if all forms had been received. In addition, the Central Office did not consistently apply the established sanctions to the students or their families for failing to submit forms or proofs. As noted earlier, parents and guardians were told that failure to provide proof could result in an investigation, a tuition bill, or exclusion from DCPS. On the basis of our visits to 15 schools, we assessed the degree of student noncompliance as very high. In one school alone, staff estimated that about 80 percent of the students—or about 700 students— did not comply. Yet, for all 158 schools, the Nonresident Tuition Enforcement Branch reported that, as of May 1, 1997, it issued only 469 letters to students requesting them to submit proofs of residency, collected tuition from only 35, and excluded only 156 students from DCPS schools. Action was pending for another 136. DCPS officials in the Nonresident Tuition Enforcement Branch told us that, at the request of one of the assistant superintendents, they were focusing their enforcement action mainly on high school athletes largely because the athletic program may have been attracting nonresidents. Other Districts Also Verify Like DCPS, all the other districts reported that all new students must verify Residency residency upon enrolling. Residency verification occurs either at the individual schools or at central service centers. Officials in Boston and Chelsea reported that the PICs verify residency. Officials in the other four districts told us that all or most new students enroll and verify residency at the school they will attend. School staff verify residency and check to see that the student’s address falls within the attendance boundary of the school. If the parent fails to provide satisfactory proof of residency, the child is not allowed to enroll. Other districts reported relying upon the schools to verify residency for continuing students. For example, officials in Arlington, Fairfax, and Prince George’s counties told us that teachers and principals are expected to monitor continually for students’ possible relocation, and students must provide information on address changes. Schools also often make use of returned mail as a reliable data source for address changes. None of the other districts we visited requires annual residency verification for all students as DCPS does. Page 22 GAO/HEHS-97-161 DCPS Enrollment Count B-275273 The foundation of the pupil accounting system—SIS—lacked adequate System for Pupil safeguards to ensure that students were accurately tracked when they Accounting Did Not transferred from one school to another. Furthermore, some schools did Ensure an Accurate not follow attendance rules, affecting later counts and projections. These rules, if implemented, may have allowed some students who no longer Count attended to be included in the school’s count. SIS Process for The student transfer process may have allowed a single student to be Transferring Students enrolled in at least two schools simultaneously.24 During most of the Allowed More Than One school year, a student’s record could be accessed and modified only by the school in which the student was enrolled. When a student transferred, School to Include a however, the losing school was to submit the student’s record to a Student in Its Active computer procedure that allowed both the losing and gaining school to Enrollment have identical copies of the student’s record. During this process, both schools could enter the student’s status as “active” or “inactive.” The computer procedure provided no safeguards to ensure that the student was only active at one school at a time. Until the losing school completed the computer procedure with a withdrawal code, both schools could have claimed the student as active or enrolled. The possible impact of this vulnerability upon the count may have been sizeable. DCPS officials reported that the number of transfers between schools in the District during school year 1996-97 was well in excess of 20,000. DCPS officials in the MIS Branch, concerned with this problem, performed periodic data runs to detect cases in which students were shown as enrolled in two schools. Resolving these issues and completing the transfers, however, sometimes involved a lengthy delay. We found cases that took as long as 1 to 2 months to resolve. Local schools made all changes—the MIS Branch did not have authority to change the data—and some school staff did not use the electronic transfer procedures. Furthermore, DCPS did not specify a time limit for completing the transfer. In addition, students could also be counted at more than one school when the massive transfers took place at year end during “roll-over”—when students transferred as a group to either middle or high school. During school year 1996-97, well over 6,800 roll-overs took place, and the process was multistaged and generally occurred when students were still enrolled in the elementary or middle schools, officials said. SIS has a programming 24 This problem is separate from the multiple record issue discussed earlier, in which two separate SIS records, with different student identification numbers, were created for a single individual. In the transfer process, only one record exists (in two copies)—and one identification number per student, and the two copies of the record can be accessed by both schools. Page 23 GAO/HEHS-97-161 DCPS Enrollment Count B-275273 anomaly allowing students to have active status in both schools’ databases, according to DCPS officials. Sometimes students were legitimately enrolled in two schools simultaneously, for example, when attending a regular high school program in addition to one of the School-to-Aid-Youth (STAY)25 programs. In these cases, the database of the school with the secondary program— STAY—should have shown the student with the special status of “enrolled” and the student’s regular school should have shown his or her status as “active.” The student should have only been counted at the school where active. School clerks did not use the “enrolled” code properly, however, and, because the status code had no safeguards, the student could be counted at both schools, according to DCPS officials. Limitations in Capability to During school year 1996-97, two attendance rules directly affected student Track Attendance Could status and therefore the number of students eligible to be counted. Have Affected Accuracy of First, schools were to reclassify as inactive, or in this case as a “no-show,” Student Database any student expected to enroll but not actually attending school at least once during the first 10 days of school. Students classified as inactive would not be included in the official enrollment count. No-shows, however, were sometimes not reclassified as inactive as required by the attendance rules. While most schools we visited appeared to be following this rule, at least one school we visited apparently had difficulty changing these students’ status to inactive. At this school, the data entry staff reported that they were having trouble maintaining student status as “inactive” for the no-shows. Some of these students were appearing on their active rolls as late as February, possibly affecting DCPS’ official count. Second, schools were required to change to inactive status those students who showed up for at least 1 day but subsequently accumulated 45 consecutive days of absences. For students who had 45 days of absences, schools reported that they only rarely changed their status to inactive. School officials often told us that they did not change a student’s status unless they could obtain accurate information about the student’s whereabouts, confirming that the student should be dropped from the rolls. School administrators stated reluctance to “give up on a student,” and they viewed changing the student’s status to inactive as such. Unlike the no-show rule, failing to implement the 45-day rule would not have 25 High school students are sometimes sent to a STAY program when they need to make up academic deficiencies. In these cases, the STAY program focuses on the deficiencies, while the student maintains enrollment for the remaining courses at his or her regular high school. Page 24 GAO/HEHS-97-161 DCPS Enrollment Count B-275273 directly affected the October count. It would have affected, however, subsequent counts and the accuracy of projections from them. The 45-day attendance rule, if implemented, may have allowed some nonattending students to be considered active and enrolled. The rule enabled any student who reported 1 day to be considered enrolled until evidence was obtained that he or she had transferred elsewhere or until 45 days had elapsed. If a student went to another school district without notifying the school, the school would not have known to drop the student from its rolls. Consequently, even if the student appeared only on the first day of school, the 45-day time period would not have expired before the official enrollment count, allowing a student to be counted who no longer attended a DCPS school. This 45-day time period might be considered lengthy by some other nearby districts. Other school districts we visited reported that they have shorter time periods. For example, Virginia law requires that students with 15 or more consecutive days of absence be withdrawn from school, district officials told us. Therefore, neither Arlington County nor Fairfax County counts any student with 15 or more days of consecutive absence. Neither does Boston count any student in this category. SISprovided no safeguards to ensure that the schools followed either the no-show rule or the 45-day rule. It had no feature that would allow students’ status to be automatically changed to inactive on the basis of absences. Nor could SIS identify students with 45 consecutive days of absence—it does not readily permit calculating consecutive days of absence for students throughout the school year. Consequently, quality control or management assistance from the MIS Branch on this issue was not possible. Other Districts Rely on Other districts we visited reported using essentially the same approaches Centralized Processing or for controlling errors in tracking student transfers as they use for Automated Database controlling enrollment and residency verification. For example, in Boston, all student transfers take place through the PICs, where a limited number Procedures to Control of staff may process the transfers. The schools lack the authority or ability Possible Errors in Pupil to transfer students. Accounting In most of the other districts, officials reported that the individual schools handle student transfers. These districts rely on a variety of automatic edits and procedures in their student database systems to prevent such Page 25 GAO/HEHS-97-161 DCPS Enrollment Count B-275273 errors and serve as ongoing checks and balances on the schools. For example, in Arlington, Fairfax, Prince George’s, and Montgomery counties, the student database systems either do not allow a transfer to proceed unless the losing school removes the student from its rolls or automatically removes the student from the losing school as part of the transfer process. The school cannot override these safeguards. In addition, Arlington, Fairfax, and Prince George’s counties reported using two centralized oversight mechanisms for further enhancing accuracy in accounting for student transfers. First, they regularly and frequently check their student databases for duplicate student entries using students’ names and dates of birth as well as identification numbers. These checks also help to safeguard against multiple student entries arising from other sources such as enrollments. Arlington County performs this check every 15 days; Fairfax County, every 2 weeks; and Prince George’s County, daily concerning transfers. Second, if these districts identify duplicates, they notify the school immediately and work with the school to resolve the situation, officials reported. For example, Prince George’s County reports duplicates from transfers to the schools every day; when school staff log onto the computer system in the morning, the first thing that appears is an error screen showing duplicates from transfers as well as any other errors. Prince George’s County officials also review these schools’ error screens and follow up daily. If schools do not respond, according to these officials, database management staff can readily access senior district officials to quickly resolve such problems. In addition, in Arlington, Fairfax, and Prince George’s counties, Boston, and Chelsea, the database staff may make changes to the student database. As in DCPS, all six of the districts we visited reported to us that teachers are responsible for tracking daily attendance and schools for recording attendance data in the student database. Most of the other districts reported that they also use their central student databases to track all student absences as a check on the schools’ tracking. In addition, several districts withdraw students from school after substantially fewer days of consecutive absences than DCPS. For example, in Boston and Arlington and Fairfax counties, students absent 15 days in a row are withdrawn from school. They are therefore not included in school or district enrollment counts. These students must re-enroll if they return. Page 26 GAO/HEHS-97-161 DCPS Enrollment Count B-275273 The District of Columbia School Reform Act of 1995 imposed enrollment DCPS, the Board of count reporting and audit requirements upon DCPS, the District of Trustees, and the Columbia Board of Education—all of the responsibilities of which have Authority Have Not been delegated to the Board of Trustees—and the Authority. The Reform Act requires the District’s schools to report certain kinds of information. Complied With the The schools did not collect all the information required to be reported, and District of Columbia the official enrollment count that was released did not comply with the Reform Act’s requirements. In addition, the Reform Act requirements to School Reform Act independently audit the count have not been met. The Reform Act requires an enrollment count that includes—in addition to data historically reported by DCPS—a report of special needs and nonresident students by grade level and tuition assessed and collected. The official enrollment count report released for school year 1996-97—the first year of the new reporting requirements—failed to provide information on special needs and nonresident students as well as on tuition assessed and collected. DCPS has not provided any evidence that additional documentation was released that would include the required information. Despite October 1996 correspondence from the U.S. Department of Education referring them to the law, DCPS officials repeatedly expressed to us unfamiliarity with the law or the type of information it requires. The Reform Act also stipulates that the Authority, after receiving the annual report, is to provide for the conduct of an independent audit. The Authority, however, had delegated this function to DCPS earlier this year, according to DCPS procurement officials. With that understanding, DCPS’ Procurement Office, with technical assistance provided by the U.S. Department of Education Inspector General’s Office,26 issued a Request for Proposals (RFP). DCPS received proposals in response, and, in early June 1997, the Procurement Office was preparing to make an award. When we queried Authority officials at that time about their role in this effort, however, they reported that they did not know of any DCPS efforts to procure the audit and were preparing to advertise an RFP for the audit. Subsequent correspondence from the Authority indicated that the inadequacies that led to the restructuring of the public school system would make auditing the count counterproductive. In addition, the Authority’s comments in response to our draft report reiterated its notion 26 In response to DCPS’ request for assistance, the U.S. Department of Education Inspector General’s Office reviewed DCPS’ draft RFP for an independent audit of the count. The Inspector General’s Office made recommendations for improvements to the RFP, and DCPS incorporated all of them in the RFP that it issued. In addition, the Department has offered to review audit plans once a contract award has been made and make suggestions for revision if necessary. The U.S. Department of Education played no other role in DCPS’ enrollment count process in school year 1996-97. Page 27 GAO/HEHS-97-161 DCPS Enrollment Count B-275273 that auditing the flawed count would be counterproductive. In short, the Reform Act’s requirements to count and report student enrollment and audit that enrollment count have not been met. Although DCPS has tried to respond to criticisms raised by previous audits, Conclusions its efforts have overlooked larger systemic issues. Consequently, fundamental weaknesses remain in the enrollment count process that make it vulnerable to inaccuracy and weaken its credibility. For example, the lack of internal controls allows multiple records and other errors that raise questions about the accuracy of the database used as a key part of the count. Furthermore, unidentified nonresident students may be included in the count when they avoid detection because DCPS’ sanctions are not enforced. An accurate and credible enrollment count demands a process with stringent accountability and strong internal controls. Moreover, the need to correct DCPS’ problems is more critical now than ever before. Current reform initiatives have heightened public awareness of the issues and increased scrutiny of the process. Meanwhile, new budget initiatives for per pupil accounting will increase this level of scrutiny. Even without the new initiatives, an accurate enrollment count is essential if DCPS is to spend its educational dollars wisely. Because the enrollment count will become the basis for funding DCPS, the Matter for Congress may wish to direct DCPS to report separately, in its annual Congressional reporting of the enrollment count, those students Consideration • fully funded from other sources, such as Head Start participants or tuition-paying nonresidents; • above and below the mandatory age for compulsory public education, such as prekindergarten or those aged 20 and above; and • for whom District residency cannot be confirmed. We recommend that the DCPS Chief Executive Officer/Superintendent do Recommendations the following: • Clarify, document, and enforce the responsibilities and sanctions for employees in all three areas of the enrollment count process—enrollment, residency verification, and pupil accounting. Page 28 GAO/HEHS-97-161 DCPS Enrollment Count B-275273 • Clarify, document, and enforce the residency verification requirements for students and their parents. • Institute internal controls in the student information database, including database management practices and automatic procedures and edits to control database errors. • Comply with the reporting requirements of the District of Columbia School Reform Act of 1995. We also recommend that the District of Columbia Financial Responsibility and Management Assistance Authority comply with the auditing requirements of the District of Columbia School Reform Act of 1995. DCPS’ Chief Executive Officer/Superintendent stated that DCPS concurs with Agency Comments the major findings and recommendations of the audit and will correct the identified weaknesses. He also acknowledged that the enrollment numbers for school year 1996-97 are subject to question for the reasons we cited— especially because the enrollment count credibility hinges almost entirely on the written verification provided by local administrators. No substantial checks and balances, no aggressive central monitoring, and few routine reports were in place. In addition, virtually no administrative sanctions were applied, indicating that the submitted reports were hardly reviewed. DCPS’ comments appear in appendix III. The Authority shared DCPS’ view that many findings and recommendations in this report will help to correct what it characterized as a flawed student enrollment process. Its comments did, however, express concerns about certain aspects of our report. More specifically, the Authority was concerned that our review did not discuss the effects of the Authority’s overhaul of DCPS in November 1996. It also commented that our report did not note that the flawed student count was one of the issues prompting the Authority to change the governance structure and management of DCPS as noted in its report, Children in Crisis: A Failure of the D.C. Public Schools. Although we did not review the Authority’s overhaul of DCPS or the events and concerns leading to that overhaul, we have revised the report to clarify the Authority’s transfer of powers and responsibilities from the District of Columbia Board of Education to the Emergency Board of Trustees. The Authority was also concerned about the clarity of our discussion of the District of Columbia School Reform Act, suggesting that we enhance this discussion to include the portion of the Reform Act that addresses the Page 29 GAO/HEHS-97-161 DCPS Enrollment Count B-275273 funding of the audit. We have clarified in the report that the relevant responsibilities of the Board of Education—including that of funding the audit—were transferred to the Emergency Board of Trustees. Finally, the Authority questioned statements made in our report about its role in preparing an RFP for an audit. Specifically, it disputes our statement that the Authority was “. . . unaware of any of DCPS’ efforts to produce the audit and were preparing to advertise an RFP for the audit.” In disputing our statement, the Authority asserts that this is a misrepresentation of a conversation between a new employee of the Authority who would have known nothing about the Authority’s contracting process and our staff. We disagree that this misrepresents our conversations with Authority staff. In preparing to meet with the Authority the first time, we spoke with a more senior, long-time member of the Authority’s staff about the audit issues who referred us to the new staff member as the expert on District education issues. When we met with the new staff member, she stated that she had reviewed the act and had spoken with other staff who were preparing to develop an RFP. Furthermore, after meeting with this new staff member, we met a second time with other Authority staff present. At both meetings, Authority staff expressed unfamiliarity with DCPS’ efforts to produce an audit. The Authority’s comments appear in appendix IV. The U.S. Department of Education, in commenting on our draft report, noted that its Office of Inspector General had no role in preparing DCPS’ enrollment count for school year 1996-97 but provided some clarifications about correspondence between it and DCPS regarding an audit of the count. We have revised the report where appropriate. Education’s comments appear in appendix V. We are sending copies of this report to the U.S. Department of Education; the Office of the Chief Executive Officer/Superintendent, District of Columbia Public Schools; the District of Columbia Financial Responsibility and Management Assistance Authority; appropriate Page 30 GAO/HEHS-97-161 DCPS Enrollment Count B-275273 congressional committees; and other interested parties. Please call Carlotta Joyner, Director, Education and Employment Issues, at (202) 512-7014 if you or your staff have any questions about this report. Major contributors to this report are listed in appendix VI. Sincerely yours, Richard L. Hembra Assistant Comptroller General Page 31 GAO/HEHS-97-161 DCPS Enrollment Count Contents Letter 1 Appendix I 34 DCPS School Sample 34 Scope and Other Urban School Districts 35 Methodology Appendix II 36 Prior Criticisms and Audits Appendix III 41 Comments From District of Columbia Public Schools Appendix IV 42 Comments From the District of Columbia Financial Responsibility and Management Assistance Authority Appendix V 44 Comments From the U.S. Department of Education Page 32 GAO/HEHS-97-161 DCPS Enrollment Count Contents Appendix VI 46 GAO Contacts 46 GAO Contacts and Staff Acknowledgments 46 Staff Acknowledgments Tables Table I.1: Distribution of Schools in the Population 34 Table I.2: Distribution of Schools in the Sample Selected for Site 35 Visits Table I.3: Other School Districts Visited 35 Table II.1: Summary of Audits and Observations Conducted of 36 School Year 1994-95 and 1995-96 Official Enrollment Counts Figure Figure 1: DCPS Enrollment Count Process for School Year 10 1996-97 Abbreviations ADM Average Daily Membership DCPS District of Columbia Public Schools ENR Enrolled Pupils MIS Management Information Services OEA Office of Educational Accountability PIC Parent Information Center RFP request for proposal SIMS Student Information Membership System SIS Student Information System STAY School-to-Aid-Youth Page 33 GAO/HEHS-97-161 DCPS Enrollment Count Appendix I Scope and Methodology We designed our study to gather information about DCPS’ enrollment count process for school year 1996-97 and the process used by other selected urban school districts. To do so, we visited DCPS administrative offices, interviewed administration officials, and reviewed documents. We also visited randomly selected DCPS schools unannounced, interviewing school faculty and staff and reviewing student records. In addition, we interviewed officials in other urban school districts, officials in the U.S. Department of Education and the District of Columbia, and other experts in the field. We did our work between October 1996 and June 1997 in accordance with generally accepted government auditing standards. We visited 15 randomly sampled DCPS elementary and secondary schools DCPS School Sample to review documents and interview faculty and staff about DCPS’ enrollment count process. We selected these schools from a list of 158 elementary and secondary schools provided to us by school district officials. We focused our review on regular elementary and secondary schools and excluded the two School-to-Aid-Youth (STAY) programs, two educational centers, and one elementary art center. Therefore, our final population included 153 schools. Fifteen schools were randomly selected by city quadrant (Northeast, Northwest, Southeast, and Southwest) and by level of school (elementary, middle/junior high, and senior high). Table I.1 shows the population distribution, and table I.2 shows the sample distribution for schools visited. Table I.1: Distribution of Schools in the Population Elementary Middle/junior Senior high Total City quadrant schools high schools schools Number Percent Northeast 33 8 6 47 31 Northwest 42 9 10 61 40 Southeast 32 6 2 40 26 Southwest 4 1 0 5 3 Total Number 111 24 18 153 100 Percent 73 16 12 100 Note: Percentages may not add to 100 percent due to rounding. Page 34 GAO/HEHS-97-161 DCPS Enrollment Count Appendix I Scope and Methodology Table I.2: Distribution of Schools in the Sample Selected for Site Visits Elementary Middle/junior Senior high Total City quadrant schools high schools schools Number Percent Northeast 2 1 1 4 27 Northwest 3 1 2 6 40 Southeast 3 0 1 4 27 Southwest 1 0 0 1 7 Total Number 9 2 4 15 100 Percent 60 13 27 100 Note: Percentages may not add to 100 percent due to rounding. We also interviewed officials in other selected urban school districts to Other Urban School gather general information about their enrollment count processes. Table Districts I.3 shows the districts we visited with their enrollment count, counting method, and number of schools for school year 1996-97. We did not visit schools or interview school faculty or staff in these other districts. Table I.3: Other School Districts Visited Number of Enrollment district count for schools during school year school year Method used to count School district 1996-97 1996-97 students Montgomery County Public Schools, Rockville, Maryland 122,505 181 Enrolled Pupils Prince George’s County Public Schools, Upper Marlboro, 122,831 177 Enrolled Pupils Maryland Boston School District, Boston, Massachusetts 63,738 125 Enrolled Pupils Chelsea Public Schools, Chelsea, Massachusetts 5,302 4 Enrolled Pupils Arlington County Public Schools, Arlington, Virginia 17,895 30 Average Daily Membership Fairfax County Public Schools, Fairfax, Virginia 144,599 232 Average Daily Membership Page 35 GAO/HEHS-97-161 DCPS Enrollment Count Appendix II Prior Criticisms and Audits Critics have charged that DCPS’ reported enrollment numbers are overstated. Questions raised about the credibility of DCPS’ enrollment count have led to a series of reviews and audits. This appendix discusses in detail these efforts, which varied in scope and involved the efforts of several organizations. Table II.1 summarizes these efforts. Table II.1: Summary of Audits and Observations Conducted of School Year 1994-95 and 1995-96 Official Enrollment Counts Audit/observation Audits of count, of count, school school Issue year 1994-95 year 1995-96 Enrollment issues No documentation to support enrollment status X X No penalty for providing false enrollment information X Lack of oversight or controls to ensure information in count was accurate X Nonresidency issues Poor procedures to validate residency X Procedures to validate residency not always followed X Weak controls to detect and follow-up on nonresidency X X Monitoring efforts failed to reconcile residency forms with rosters Not applicable X Pupil accounting issues Enrollment status incorrect X Transfers not completed X Duplicate records X X Questions of system security and reliability X X Database not routinely updated to reflect current enrollment status X X No guidance for withdrawing students and excluding them from the count X Other Issue Too much time elapsed between count and audit X X Grier Partnership Criticism In 1995, the Grier Partnership, as part of a study commissioned by DCPS, of DCPS’ 1990 Enrollment asserted that results of the 1990 U.S. census suggested that the District’s Count total school-age population in 1990 might have been as much as 13,000 less than DCPS reported in its official enrollment count. Grier also expressed concern about the apparent relative stability of DCPS’ official enrollment count in the face of the District’s declining resident population. Limitations to the methodology the Grier Partnership used, however, may have caused the apparent differences to be overstated. For example, Grier did not include some subgroups—preschool (Head Start), Page 36 GAO/HEHS-97-161 DCPS Enrollment Count Appendix II Prior Criticisms and Audits prekindergarten, and kindergarten students—that DCPS routinely includes in its official count. Even if these groups had been included in the estimates, using census data to estimate public school enrollment can be problematic. For example, the Census Bureau reports that estimates generated from its official files undercount some groups. From the 1990 census, the largest group undercounted was “renters.” Census estimates of pre-primary students enrolled in school are also understated because parents reporting the number of students enrolled in “regular school” often fail to include their pre-primary children. Finally, declines in residency do not necessarily mean declines in school enrollment. Census currently projects a loss of 31,000 in the District’s population over the next 5 years, while projecting an increase in the number of school-aged children. GAO Criticisms of DCPS’ The first of several independent audits took place following the September October 1994 Count 29, 1994, enrollment count. At that time, DCPS organized an internal audit and validation of the count. DCPS randomly selected a sample of students and focused on validating these students’ actual attendance in schools before the enrollment count. We were asked to observe DCPS’ internal audit effort.27 We questioned the reliability of the student database, finding that the database used to enroll and track students—the Student Information Membership System (SIMS)—included students who had not enrolled before the official enrollment count. We also found that transfer students were never removed from SIMS when they transferred. In addition, SIMS had other errors, was not regularly updated, and had at least 340 duplicate student records. We also criticized DCPS’ inability to identify nonresident students and the absence of procedures to validate residency. DCPS estimated that at that time approximately 2 percent of its students were probably undetected nonresidents. DCPS also estimated that this equaled more than $6 million in lost tuition revenues. We consequently recommended that DCPS periodically check SIMS for duplicates and errors, particularly before the official enrollment count, and update it regularly to reflect the changes in the enrollment status of DCPS students. We also recommended that DCPS develop systematic 27 See District of Columbia: Weaknesses in Personnel Records and Public Schools’ Management Information and Controls (GAO/T-AIMD-95-170, June 14, 1995). Page 37 GAO/HEHS-97-161 DCPS Enrollment Count Appendix II Prior Criticisms and Audits procedures at the school level to verify student residency and that schools refer names of nonresident students to DCPS administration for enforcement and collection of nonresident tuition. Independent Auditor’s The DCPS Superintendent, after the October 1995 enrollment count, Findings Regarding contracted for an independent audit and validation of the count.28 In October 1995 Count addition to a 100-percent validation of the count, DCPS expected that the independent auditor would assess the accuracy of DCPS’ Student Information System29 (SIS) and determine if school and headquarters staff had followed DCPS’ policies and procedures. The independent auditor chosen by DCPS conducted a full validation of the enrollment count and examined SIS for duplicates and errors. The auditor failed, however, to determine if DCPS school and headquarters staff consistently implemented the policies and procedures developed by the DCPS administration.30 The independent auditor found several weaknesses in the October 1995 count, including problems with the way the enrollment count was taken and documented by DCPS staff; lack of residency documentation and validation; the questionable accuracy of SIS; and the lack of guidance for withdrawing students and excluding them from the schools’ rolls. For example, a new form, the Student Residency and Data Verification Form, used to document residency, was piloted in some schools during school year 1995-96.31 The auditor found that these forms were sent home to parents but were not always returned to the schools, and the forms were not reconciled to student enrollment reports to determine the number of missing forms. The auditor also found 550 sets of students with the same name and date of birth, that is, duplicate entries in SIS. In addition, the auditor criticized the time lapse—about 4 months—from the October 5, 1995, enrollment count to the audit. This meant that the auditor could not validate the enrollment of some students—students who were no longer in school at the time of the audit and for whom the school 28 Audit of the Official Membership of the District of Columbia Public Schools as of October 5, 1995, F.S. Taylor & Associates, P.C. (Washington, D.C.: Apr. 26, 1996). 29 SIS replaced SIMS as DCPS’ automated student information database. 30 The Office of the District of Columbia Auditor, in its own review and report of the October 1995 enrollment count, criticized the independent auditor’s failure to evaluate the compliance of DCPS’ regulations and guidelines because it was an agreed-upon deliverable in the independent auditor’s contract with DCPS. 31 The Student Residency and Data Verification Form piloted during the 1995-96 school year is the same form used in school year 1996-97 to verify residency for all students. Page 38 GAO/HEHS-97-161 DCPS Enrollment Count Appendix II Prior Criticisms and Audits could provide no documentation demonstrating attendance before the count. To remedy the problem with duplicate database entries, the auditor recommended that DCPS periodically search the database for duplicates and errors before the enrollment count. Because of differences found in SIS and the manually prepared enrollment count report, the auditor also recommended that these two data sources be reconciled periodically to help update SIS. Regarding timing of the audit, the auditor recommended that the audit of the official enrollment count take place closer to the date of the count. And, to facilitate future audits, the auditor suggested that documentation exist to support a student’s attendance in school before the enrollment count. The independent auditor also suggested that after an enrollment count is taken, the staff responsible for monitoring attendance problems have the opportunity to review the enrollment count so they can remove from the count those students who have not attended at least 1 day of school or who have withdrawn from DCPS. Office of the District of The District of Columbia Auditor, in its audit of the October 5, 1995, Columbia Auditor Findings enrollment count,32 found that DCPS needed significantly improved Regarding the procedures for student enrollment counts to ensure more reliable and valid counts. The Auditor’s office expressed concerns about the security October 1995 Count and reliability of SIS, the absence of any penalty for providing false enrollment information, and the lack of oversight or controls to ensure the accuracy of the information reported on the enrollment count. In addition, the Auditor found that SIS was not updated regularly to reflect changes in the enrollment status of students, particularly before the official enrollment count. The Auditor also discussed the weak controls in place to detect nonresidency and the weak procedures to collect nonresident tuition. The Auditor found that DCPS did not maintain records on the number of Student Residence and Data Verification Forms completed and returned by students’ parents, and it did not test the information on these forms or the documents provided to support the forms. As a result, the Auditor reported that according to the DCPS Nonresident Tuition Enforcement 32 Review of the District of Columbia Public Schools’ Official Membership Count Procedures, Office of the District of Columbia Auditor (Washington, D.C.: July 19, 1996). Page 39 GAO/HEHS-97-161 DCPS Enrollment Count Appendix II Prior Criticisms and Audits Branch estimates, about 4,000 to 6,000 DCPS students were nonresidents but did not pay nonresident tuition. Consequently, the Auditor recommended that each local school periodically reconcile SIS-generated reports with the attendance records it maintains. This would allow for adjustments to SIS to include those students who have physically presented themselves in class and removing those who have not presented themselves, withdrawn, or transferred. In addition, the Auditor suggested that unless students could document their residency, including proof of residency, they should be excluded from the official enrollment count. Furthermore, the Auditor suggested that those nonresidents who pay tuition be excluded from the enrollment count. Page 40 GAO/HEHS-97-161 DCPS Enrollment Count Appendix III Comments From District of Columbia Public Schools Page 41 GAO/HEHS-97-161 DCPS Enrollment Count Appendix IV Comments From the District of Columbia Financial Responsibility and Management Assistance Authority Page 42 GAO/HEHS-97-161 DCPS Enrollment Count Appendix IV Comments From the District of Columbia Financial Responsibility and Management Assistance Authority Page 43 GAO/HEHS-97-161 DCPS Enrollment Count Appendix V Comments From the U.S. Department of Education Page 44 GAO/HEHS-97-161 DCPS Enrollment Count Appendix V Comments From the U.S. Department of Education Page 45 GAO/HEHS-97-161 DCPS Enrollment Count Appendix VI GAO Contacts and Staff Acknowledgments D. Catherine Baltzell, Assistant Director, (202) 512-8001 GAO Contacts Dianne Murphy Blank, Evaluator-in-Charge, (202) 512-5654 In addition to those named above, the following individuals made Staff important contributions to this report: Christine McGagh led numerous Acknowledgments site visits, reviewed DCPS’ enrollment count process, and cowrote portions of this report; James W. Hansbury, Jr., performed numerous site visits, reviewed prior audit reports, and summarized those audits. Wayne Dow, Edward Tuchman, and Deborah Edwards assisted with the visits to the schools; Sylvia Shanks and Robert Crystal provided legal assistance; and Liz Williams and Ann McDermott assisted with report preparation. (104870) Page 46 GAO/HEHS-97-161 DCPS Enrollment Count Ordering Information The first copy of each GAO report and testimony is free. Additional copies are $2 each. Orders should be sent to the following address, accompanied by a check or money order made out to the Superintendent of Documents, when necessary. VISA and MasterCard credit cards are accepted, also. Orders for 100 or more copies to be mailed to a single address are discounted 25 percent. Orders by mail: U.S. General Accounting Office P.O. Box 6015 Gaithersburg, MD 20884-6015 or visit: Room 1100 700 4th St. NW (corner of 4th and G Sts. NW) U.S. General Accounting Office Washington, DC Orders may also be placed by calling (202) 512-6000 or by using fax number (301) 258-4066, or TDD (301) 413-0006. Each day, GAO issues a list of newly available reports and testimony. To receive facsimile copies of the daily list or any list from the past 30 days, please call (202) 512-6000 using a touchtone phone. A recorded menu will provide information on how to obtain these lists. For information on how to access GAO reports on the INTERNET, send an e-mail message with "info" in the body to: firstname.lastname@example.org or visit GAO’s World Wide Web Home Page at: http://www.gao.gov PRINTED ON RECYCLED PAPER United States Bulk Rate General Accounting Office Postage & Fees Paid Washington, D.C. 20548-0001 GAO Permit No. G100 Official Business Penalty for Private Use $300 Address Correction Requested
District of Columbia Public Schools: Student Enrollment Count Remains Vulnerable to Errors
Published by the Government Accountability Office on 1997-08-21.
Below is a raw (and likely hideous) rendition of the original report. (PDF)