District of Columbia Public Schools: Student Enrollment Count Remains
Vulnerable to Errors (Letter Report, 08/21/97, GAO/HEHS-97-161).
Pursuant to a congressional request, GAO examined the enrollment count
process that the District of Columbia Public Schools (DCPS) used in
school year 1996-97, focusing on: (1) whether the process appeared
sufficient to produce an accurate count; (2) enrollment count processes
used by some other urban school systems; and (3) the role of the
Department of Education's Inspector General in preparing DCPS' official
enrollment count for school year 1996-97.
GAO noted that: (1) even though DCPS changed parts of its enrollment
count process in school year 1996-97 to address criticisms, the process
remains flawed; (2) some of these changes increased complexity and work
effort but did little to improve the count's credibility; (3) errors
remained in the Student Information System (SIS), but DCPS had only
limited mechanisms for correcting these errors; (4) problems also
persisted in the critical area of residency verification; (5) in school
year 1996-97, schools did not always verify student residency as
required by DCPS' own procedures; (6) proofs of residency, when actually
obtained, often fell short of DCPS' standards; (7) Central Office staff
did not consistently track failures to verify residency; (8) school
staff and parents rarely suffered sanctions for failure to comply with
the residency verification requirements; (9) the pupil accounting system
failed to adequately track students; (10) SIS allowed more than one
school to count a single student when the student transferred from one
school to another; (11) schools did not always follow attendance rules,
and SIS lacked the capability to track implementation of the rules; (12)
some attendance rules, if implemented, could have allowed counting of
nonattending students; (13) other school districts report that they use
several approaches to control errors and to increase the accuracy of
their enrollment counts; (14) 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; (15) the recently enacted
District of Columbia School Reform Act of 1995 requires the enrollment
count process to produce enrollment numbers for nonresidents and
students with special needs; (16) DCPS (acting on behalf of the District
of Columbia Board of Education) and the District of Columbia Financial
Responsibility and Management Assistance Authority are not in compliance
with requirements of this new law; (17) the Department of Education
helped DCPS develop its request for proposals 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; and
(18) the Authority subsequently decided, however, that auditing the
count for school year 1996-97 would be counterproductive.
--------------------------- Indexing Terms -----------------------------
REPORTNUM: HEHS-97-161
TITLE: District of Columbia Public Schools: Student Enrollment
Count Remains Vulnerable to Errors
DATE: 08/21/97
SUBJECT: School management and organization
Public schools
Students
Data integrity
Attendance records
School districts
Data collection
Management information systems
IDENTIFIER: District of Columbia
DC Public Schools Student Information System
******************************************************************
** This file contains an ASCII representation of the text of a **
** GAO report. Delineations within the text indicating chapter **
** titles, headings, and bullets are preserved. Major **
** divisions and subdivisions of the text, such as Chapters, **
** Sections, and Appendixes, are identified by double and **
** single lines. The numbers on the right end of these lines **
** indicate the position of each of the subsections in the **
** document outline. These numbers do NOT correspond with the **
** page numbers of the printed product. **
** **
** No attempt has been made to display graphic images, although **
** figure captions are reproduced. Tables are included, but **
** may not resemble those in the printed version. **
** **
** Please see the PDF (Portable Document Format) file, when **
** available, for a complete electronic file of the printed **
** document's contents. **
** **
** A printed copy of this report may be obtained from the GAO **
** Document Distribution Center. For further details, please **
** send an e-mail message to: **
** **
** **
** **
** with the message 'info' in the body. **
******************************************************************
Cover
================================================================ COVER
Report to the Chairman, 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
DCPS Enrollment Count
(104870)
Abbreviations
=============================================================== ABBREV
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
Letter
=============================================================== LETTER
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.
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 districts\3 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.
--------------------
\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.
\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.
RESULTS IN BRIEF
------------------------------------------------------------ Letter :1
Even though DCPS changed parts of its enrollment count process in
school 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 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 1995\5 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.
--------------------
\5 Title II of P.L. 104-134 (1996).
BACKGROUND
------------------------------------------------------------ Letter :2
Unlike other U.S. school districts, DCPS, due to its location in the
nation's 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.
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
ENVIRONMENT IS UNIQUE
---------------------------------------------------------- Letter :2.1
DCPS lacks the state-level oversight that most other school districts
in the 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 Education\7 (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 students enrolled in DCPS. The law
requires the District of Columbia Board of Education\8 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 needs\9 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.
--------------------
\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.
\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.
MOST SCHOOL DISTRICTS MUST
COUNT STUDENT ENROLLMENT
---------------------------------------------------------- Letter :2.2
The requirement to count students is common to most other U.S.
school districts. Forty-one of the 50 states\10 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 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 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.
--------------------
\10 Seven states use teacher or instructional units rather than
student count, and two states use other methods to distribute funds.
\11 Maryland and Massachusetts require ENR; Virginia requires ADM.
ACCURACY OF DCPS' ENROLLMENT
COUNT HAS BEEN QUESTIONED
---------------------------------------------------------- Letter :2.3
Critics have often charged that the District's reported official
enrollment numbers have been overstated. One reviewer asserted, for
example, that 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
PROCESS IN SCHOOL YEAR
1996-97 REMAINED CENTERED IN
THE LOCAL SCHOOLS
---------------------------------------------------------- Letter :2.4
DCPS' process for enrolling, verifying residency of, and tracking
students remained centered in the local school in school year
1996-97, while central office staff monitored portions of the
process. To respond to past criticisms, DCPS instituted some changes
for school year 1996-97, including new forms, residency verification
procedures, and additional preparatory 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
ENROLLMENT ACTIVITIES
-------------------------------------------------------- Letter :2.4.1
DCPS' local schools conducted all enrollment activities in school
year 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 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.
Figure 1: DCPS Enrollment
Count Process for School Year
1996-97
(See figure in printed
edition.)
(See figure in printed
edition.)
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 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.
--------------------
\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.
PROCESS REQUIRED EVIDENCE
OF RESIDENCY
-------------------------------------------------------- Letter :2.4.2
In addition to the enrollment card, DCPS' enrollment process for
1996-97 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 THE LOCAL
SCHOOLS
-------------------------------------------------------- Letter :2.4.3
In addition to enrollment and residency verification procedures,
local 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.
MONITORING ACTIVITIES
INCLUDED USE OF
ENROLLMENT CARDS AND
IDENTIFYING NONRESIDENTS
-------------------------------------------------------- Letter :2.4.4
Monitoring activities for school year 1996-97 focused exclusively on
overseeing the schools' implementation of the enrollment card and on
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
STUDENTS
-------------------------------------------------------- Letter :2.4.5
As previously mentioned, for school year 1996-97, DCPS used the ENR
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 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.
--------------------
\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).
SCHOOL YEAR 1996-97 ENROLLMENT
PROCESS CHANGES DID NOT ENSURE
ACCURACY
------------------------------------------------------------ Letter :3
DCPS' new student enrollment card was intended to document that
students had met the 1-day attendance requirement for inclusion in
the official enrollment count. Although the card may have met this
requirement in some respects, it appears to have burdened both school
and DCPS staff and may not offer much advantage over more traditional
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.
ENROLLMENT CARD APPEARS TO
BE BURDENSOME AND
INEFFICIENT
---------------------------------------------------------- Letter :3.1
Teachers and school staff reported that DCPS' new enrollment card was
burdensome and difficult to implement. Each child, on the first day
of attendance, had to complete and sign two separate copies\15 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 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.
--------------------
\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.
LACK OF INTERNAL CONTROLS
ALLOWED MULTIPLE ENROLLMENT
RECORDS
---------------------------------------------------------- Letter :3.2
The procedures that DCPS used for enrolling students in school year
1996-97 allowed multiple records to be entered into SIS for a single
student. 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 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.
--------------------
\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.
ENROLLMENT PROCEDURES ALLOW
COUNT OF STUDENTS NOT
COUNTED IN SOME OTHER
DISTRICTS
---------------------------------------------------------- Letter :3.3
DCPS' official enrollment count of 78,648 included not only regular
elementary and secondary students, but also other categories of
students excluded from enrollment counts in other districts when the
counts are used for funding purposes. For example, DCPS included in
its enrollment 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.
--------------------
\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.
OTHER DISTRICTS REPORT USING
VARIOUS METHODS TO HELP
IMPROVE ENROLLMENT ACCURACY
---------------------------------------------------------- Letter :3.4
In contrast with the DCPS process, students in the Boston and
Chelsea, Massachusetts, school districts enroll at central Parent
Information Centers (PIC), which are separate and independent from
the schools, officials told us. Individual schools in these two
districts cannot enroll 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 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.
--------------------
\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.
NEW RESIDENCY VERIFICATION
PROCEDURES WERE NOT COMPLETELY
IMPLEMENTED
------------------------------------------------------------ Letter :4
During school year 1996-97, District of Columbia schools had features
that attracted nonresidents. Elementary schools in the District had
free all-day prekindergarten and kindergarten, and some elementary
schools had 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 school day's hours to accommodate working
parents--the program began 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 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.
--------------------
\22 Audit reports of the Oct. 1995 count also criticized DCPS for
failure to verify student residency.
SCHOOLS FAILED TO VERIFY AND
REPORT RESIDENCY FOR ALL
STUDENTS AS REQUIRED BY DCPS
POLICY
---------------------------------------------------------- Letter :4.1
Most of the schools we visited failed to comply with the new
residency verification process. As discussed previously, all
students' parents or legal guardians had to complete a Student
Residency and Data Verification Form (residency form) and provide at
least two proofs of residency. 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 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
EFFECTIVELY MONITOR OR
ENFORCE RESIDENCY
VERIFICATION PROCEDURES
---------------------------------------------------------- Letter :4.2
Monitors for student residency, in general, did not report the level
of school and student noncompliance that we observed in our review.
For the nine schools for which we could directly assess
compliance,\23 with few exceptions, proofs of residency were missing
for large portions of the 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 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.
--------------------
\23 We could assess compliance only for those schools for which we
could track the proof of residency to the residency form.
OTHER DISTRICTS ALSO VERIFY
RESIDENCY
---------------------------------------------------------- Letter :4.3
Like DCPS, all the other districts reported that all new students
must verify 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.
SYSTEM FOR PUPIL ACCOUNTING DID
NOT ENSURE AN ACCURATE COUNT
------------------------------------------------------------ Letter :5
The foundation of the pupil accounting system--SIS--lacked adequate
safeguards to ensure that students were accurately tracked when they
transferred from one school to another. Furthermore, some schools
did not follow attendance rules, affecting later counts and
projections. These rules, if implemented, may have allowed some
students who no longer attended to be included in the school's count.
SIS PROCESS FOR TRANSFERRING
STUDENTS ALLOWED MORE THAN
ONE SCHOOL TO INCLUDE A
STUDENT IN ITS ACTIVE
ENROLLMENT
---------------------------------------------------------- Letter :5.1
The student transfer process may have allowed a single student to be
enrolled in at least two schools simultaneously.\24 During most of
the 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, however, the losing school was to submit the student's
record to a computer procedure that allowed both the losing and
gaining school to 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 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.
--------------------
\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.
\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.
LIMITATIONS IN CAPABILITY TO
TRACK ATTENDANCE COULD HAVE
AFFECTED ACCURACY OF STUDENT
DATABASE
---------------------------------------------------------- Letter :5.2
During school year 1996-97, two attendance rules directly affected
student status and therefore the number of students eligible to be
counted.
First, schools were to reclassify as inactive, or in this case as a
"no-show," 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 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.
SIS provided 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
CENTRALIZED PROCESSING OR
AUTOMATED DATABASE
PROCEDURES TO CONTROL
POSSIBLE ERRORS IN PUPIL
ACCOUNTING
---------------------------------------------------------- Letter :5.3
Other districts we visited reported using essentially the same
approaches for controlling errors in tracking student transfers as
they use for controlling enrollment and residency verification. For
example, in Boston, all student transfers take place through the
PICs, where a limited number of staff may process the transfers. The
schools lack the authority or ability to transfer students.
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 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.
DCPS, THE BOARD OF TRUSTEES,
AND THE AUTHORITY HAVE NOT
COMPLIED WITH THE DISTRICT OF
COLUMBIA SCHOOL REFORM ACT
------------------------------------------------------------ Letter :6
The District of Columbia School Reform Act of 1995 imposed enrollment
count reporting and audit requirements upon DCPS, the District of
Columbia Board of Education--all of the responsibilities of which
have been delegated to the Board of Trustees--and the Authority. The
Reform Act requires the District's schools to report certain kinds of
information. The schools did not collect all the information
required to be reported, and the official enrollment count that was
released did not comply with the Reform Act's requirements. In
addition, the Reform Act requirements to 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 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.
--------------------
\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.
CONCLUSIONS
------------------------------------------------------------ Letter :7
Although DCPS has tried to respond to criticisms raised by previous
audits, 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.
MATTER FOR CONGRESSIONAL
CONSIDERATION
------------------------------------------------------------ Letter :8
Because the enrollment count will become the basis for funding DCPS,
the Congress may wish to direct DCPS to report separately, in its
annual reporting of the enrollment count, those students
-- 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.
RECOMMENDATIONS
------------------------------------------------------------ Letter :9
We recommend that the DCPS Chief Executive Officer/Superintendent do
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.
-- 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.
AGENCY COMMENTS
----------------------------------------------------------- Letter :10
DCPS' Chief Executive Officer/Superintendent stated that DCPS concurs
with 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 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.
--------------------------------------------------------- Letter :10.1
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
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
SCOPE AND METHODOLOGY
=========================================================== Appendix I
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.
DCPS SCHOOL SAMPLE
--------------------------------------------------------- Appendix I:1
We visited 15 randomly sampled DCPS elementary and secondary schools
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
Total
--------------
Middle/ Senior
Elementary junior high Percen
City quadrant schools high schools schools Number t
-------------------------- ---------- ------------ ---------- ------ ------
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.
Table I.2
Distribution of Schools in the Sample
Selected for Site Visits
Total
--------------
Middle/ Senior
Elementary junior high Percen
City quadrant schools high schools schools Number t
-------------------------- ---------- ------------ ---------- ------ ------
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.
OTHER URBAN SCHOOL DISTRICTS
--------------------------------------------------------- Appendix I:2
We also interviewed officials in other selected urban school
districts to gather general information about their enrollment count
processes. Table 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
district
Enrollment schools
count for during
school school
year 1996- year 1996- Method used to count
School district 97 97 students
--------------------------------- ---------- ---------- ---------------------
Montgomery County Public Schools, 122,505 181 Enrolled Pupils
Rockville, Maryland
Prince George's County Public 122,831 177 Enrolled Pupils
Schools, Upper Marlboro,
Maryland
Boston School District, Boston, 63,738 125 Enrolled Pupils
Massachusetts
Chelsea Public Schools, Chelsea, 5,302 4 Enrolled Pupils
Massachusetts
Arlington County Public Schools, 17,895 30 Average Daily
Arlington, Virginia Membership
Fairfax County Public Schools, 144,599 232 Average Daily
Fairfax, Virginia Membership
--------------------------------------------------------------------------------
PRIOR CRITICISMS AND AUDITS
========================================================== Appendix II
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
of count, school count, school
Issue year 1994-95 year 1995-96
-------------------------------------------- ------------------ --------------
Enrollment issues
--------------------------------------------------------------------------------
No documentation to support enrollment X X
status
No penalty for providing false enrollment X
information
Lack of oversight or controls to ensure X
information in count was accurate
Nonresidency issues
--------------------------------------------------------------------------------
Poor procedures to validate residency X
Procedures to validate residency not always X
followed
Weak controls to detect and follow-up on X X
nonresidency
Monitoring efforts failed to reconcile Not applicable X
residency forms with rosters
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 X X
current enrollment status
No guidance for withdrawing students and X
excluding them from the count
Other Issue
--------------------------------------------------------------------------------
Too much time elapsed between count and X X
audit
--------------------------------------------------------------------------------
GRIER PARTNERSHIP CRITICISM
OF DCPS' 1990 ENROLLMENT
COUNT
------------------------------------------------------ Appendix II:0.1
In 1995, the Grier Partnership, as part of a study commissioned by
DCPS, asserted that results of the 1990 U.S. census suggested that
the District's 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), 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'
OCTOBER 1994 COUNT
------------------------------------------------------ Appendix II:0.2
The first of several independent audits took place following the
September 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 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.
--------------------
\27 See District of Columbia: Weaknesses in Personnel Records and
Public Schools' Management Information and Controls
(GAO/T-AIMD-95-170, June 14, 1995).
INDEPENDENT AUDITOR'S
FINDINGS REGARDING OCTOBER
1995 COUNT
------------------------------------------------------ Appendix II:0.3
The DCPS Superintendent, after the October 1995 enrollment count,
contracted for an independent audit and validation of the count.\28
In addition to a 100-percent validation of the count, DCPS expected
that the independent auditor would assess the accuracy of DCPS'
Student Information System\29 (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 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.
--------------------
\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.
OFFICE OF THE DISTRICT OF
COLUMBIA AUDITOR FINDINGS
REGARDING THE OCTOBER 1995
COUNT
------------------------------------------------------ Appendix II:0.4
The District of Columbia Auditor, in its audit of the October 5,
1995, enrollment count,\32
found that DCPS needed significantly improved procedures for student
enrollment counts to ensure more reliable and valid counts. The
Auditor's office expressed concerns about the security 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 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.
(See figure in printed edition.)Appendix III
--------------------
\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).
COMMENTS FROM DISTRICT OF COLUMBIA
PUBLIC SCHOOLS
========================================================== Appendix II
(See figure in printed edition.)Appendix IV
COMMENTS FROM THE DISTRICT OF
COLUMBIA FINANCIAL RESPONSIBILITY
AND MANAGEMENT ASSISTANCE
AUTHORITY
========================================================== Appendix II
(See figure in printed edition.)
(See figure in printed edition.)Appendix V
COMMENTS FROM THE U.S. DEPARTMENT
OF EDUCATION
========================================================== Appendix II
(See figure in printed edition.)
GAO CONTACTS AND STAFF
ACKNOWLEDGMENTS
========================================================== Appendix VI
GAO CONTACTS
-------------------------------------------------------- Appendix VI:1
D. Catherine Baltzell, Assistant Director, (202) 512-8001
Dianne Murphy Blank, Evaluator-in-Charge, (202) 512-5654
STAFF ACKNOWLEDGMENTS
-------------------------------------------------------- Appendix VI:2
In addition to those named above, the following individuals made
important contributions to this report: Christine McGagh led
numerous 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.
*** End of document. ***