School Facilities: Construction Expenditures Have Grown Significantly in
Recent Years (Letter Report, 03/03/2000, GAO/HEHS-00-41).

Pursuant to a congressional request, GAO reviewed construction
expenditures for public school facilities, focusing on: (1) the trends
since 1990 in elementary and secondary school construction expenditures
and how construction expenditures were divided between land, buildings,
and equipment; (2) the trends since 1990 in the amount of expenditures
for elementary and secondary school construction by type of school and
type of construction; and (3) what is known about the amounts and mix of
state and local funding for elementary and secondary school
construction.

GAO noted that: (1) annual construction expenditures for elementary and
secondary schools across the nation grew by 39 percent from fiscal year
(FY) 1990 through FY 1997 to about $25 billion after adjusting for
inflation; (2) this trend, according to reports by the Department of
Education and others, reflects a variety of factors, including higher
enrollments, a strong economy, and an increasing need to replace old
buildings; (3) most of the increase in expenditures was for the
construction of buildings, while expenditures for equipment such as
heating and air conditioning systems increased only slightly during the
8-year period, and spending on land and for the purchase of buildings
such as portable classrooms showed a net decline; (4) average annual
construction expenditures per pupil varied widely from state to state,
ranging from $934 in Nevada to $37 in Connecticut; (5) states with the
largest expenditures per pupil, such as Nevada, tended to also have the
highest enrollment growth rates, and those with the lowest expenditures
per pupil had relatively low enrollment growth rates; (6) most of the
growth in construction expenditures during calendar years 1990 through
1998 was for construction at primary schools and high schools, according
to data on construction contracts; (7) also, most of the contract
spending from 1990 through 1998 was for new facilities and additions to
existing facilities, with significantly less being spent on renovations;
(8) the largest source of funding for school construction is generally
local construction bonds; (9) however, no current and complete database
shows the sources of funding for school construction; (10) as a result,
nationwide data on how amounts and portions of funds are divided between
localities and states are not available; (11) studies show a range in
the degree to which the states rely on local and state funding for
school construction projects; (12) while in most states some combination
of local and state funding supports school construction, 15 states
provided little or no funding in 1998-1999; and (13) Hawaii, where the
school district and the state are coterminous, provides all funding.

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

 REPORTNUM:  HEHS-00-41
     TITLE:  School Facilities: Construction Expenditures Have Grown
	     Significantly in Recent Years
      DATE:  03/03/2000
   SUBJECT:  Construction contracts
	     Construction costs
	     Educational facility construction
	     Public schools
	     Intergovernmental fiscal relations
	     State/local relations
IDENTIFIER:  Proposition 1A (CA)

******************************************************************
** This file contains an ASCII representation of the text of a  **
** GAO Testimony.                                               **
**                                                              **
** 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.                                         **
**                                                              **
******************************************************************

GAO/HEHS-00-41

Appendix I: Scope And Methodology

24

Appendix II: Changes in Annual State School Construction Expenditures From
Fiscal Year 1990 to Fiscal Year 1997

26

Appendix III: Changes in Regional School Construction Expenditures for Land,
Buildings, and Equipment, Fiscal Years 1992-97

28

Appendix IV: Average Annual State School Construction Expenditures per
Student, Fiscal Years 1990-97

29

Appendix V: Comments From the Department of Education

31

33

Table 1: Cumulative Percentage Growth Rates for School Construction Contract
Spending, 1990-98 15

Table 2: Bonds and Their Amounts Approved in Local School Construction Bond
Referendums in 19 States, 1998 19

Table 3: South Carolina School Bond Referendum Results, 1994-98 21

Figure 1: Cumulative Percentage Changes in School Construction Expenditures
and Fall Enrollments, Fiscal Years 1990-97 7

Figure 2: Regional School Construction Growth Rates, Fiscal Years 1991-97 8

Figure 3: School Facilities Acquisition and Construction Expenditures,
Fiscal Years 1992-97 10

Figure 4: Average State Construction Expenditures per Student, Fiscal Years
1990-97 12

Figure 5: Construction Contract Spending, 1990-98 14

Figure 6: Contract Expenditures for Additions, Renovations, and New Schools,
1990-98 16

CCD Common Core of Data

NCES National Center for Education Statistics

Health, Education, and
Human Services Division

B-282314

March 3, 2000

The Honorable William F. Goodling
Chairman, Committee on Education and the Workforce
House of Representatives

Dear Mr. Chairman:

More than 45 million elementary and secondary students attend approximately
86,000 public schools in the United States. While enrollments are growing,
the Department of Education has found that the average public school
building in 1998 was 42 years old, and in 1995 we reported that about a
third of the nation's public schools needed extensive repair or replacement
of one or more buildings.1 Funding for school construction has generally
been the responsibility of local school districts and, more recently, the
states, while only limited federal funding has been available for this
purpose. However, a number of congressional proposals would expand the
federal role in financing school construction. To determine how states and
local school districts have been dealing with the issues facing their public
school facilities, you asked us to determine (1) the trends since 1990 in
elementary and secondary school construction expenditures and how
construction expenditures were divided between land, buildings, and
equipment; (2) the trends since 1990 in the amount of expenditures for
elementary and secondary school construction by type of school and type of
construction; and (3) what is known about the amounts and mix of state and
local funding for elementary and secondary school construction.

In conducting this study, we obtained and analyzed construction expenditure
and enrollment data from Education's National Center for Education
Statistics (NCES) and construction contract expenditure data collected by a
private firm.2 We also identified and reviewed research on the sources of
financing school construction and conducted a telephone inquiry of each of
the 50 state education agencies to determine the availability of data on
school construction funding. We conducted our work between February and
December 1999 in accordance with generally accepted government auditing
standards. (Appendix I describes our scope and methodology in more detail.)

Annual construction expenditures for elementary and secondary schools across
the nation grew by 39 percent from fiscal year 1990 through fiscal year 1997
to about $25 billion after adjusting for inflation.3 This trend, according
to reports by Education and others, reflects a variety of factors, including
higher enrollments, a strong economy, and an increasing need to replace old
buildings. Most of the increase in expenditures was for the construction of
buildings; expenditures for equipment such as heating and air conditioning
systems increased only slightly during the 8-year period, and spending on
land and for the purchase of buildings such as portable classrooms showed a
net decline. Average annual construction expenditures per pupil varied
widely from state to state, ranging from $934 in Nevada to $37 in
Connecticut. States with the largest expenditures per pupil, such as Nevada,
tended to also have the highest enrollment growth rates, and those with the
lowest expenditures per pupil had relatively low enrollment growth rates.

Most of the growth in construction expenditures during calendar years 1990
through 1998 was for construction at primary schools and high schools,
according to data on construction contracts. Also, most of the contract
spending from 1990 through 1998 was for new facilities and additions to
existing facilities, with significantly less being spent on renovations.

The largest source of funding for school construction is generally local
construction bonds. However, no current and complete database shows the
sources of funding for school construction. As a result, nationwide data on
how amounts and portions of funds are divided between localities and states
are not available. Studies show a range in the degree to which the states
rely on local and state funding for school construction projects. While in
most states some combination of local and state funding supports school
construction, 15 states provided little or no funding in 1998-99. Hawaii,
where the school district and the state are coterminous, provides all
funding.

In 1995-97, we issued seven reports on the facilities in American schools in
which we reported that it would cost about $112 billion to bring them into
good overall condition and that about a third (or 25,000) of all public
schools needed extensive repair or replacement of at least one building.4 In
addition, a recent Education report estimated that at least 2,400 new public
schools will be needed by 2003 to accommodate rising enrollments stemming
from the so-called baby boom echo.5

Traditionally, funding for school construction has been a local school
district responsibility, with some states adding financial support. Indirect
federal support for the construction of schools and other capital
facilities, such as roads, hospitals, and parks, is provided through the tax
system. The interest income individuals and businesses earn on state and
local debt is excluded from their taxable income in computing their federal
income tax due. This exclusion lowers the interest rate on state and local
debt, a reduction in effect paid for by the federal tax revenue not
collected on the excluded interest earnings.6

The federal government has provided money to help offset the effect of
federal activities, such as in making federal Impact Aid payments, improving
accessibility for the disabled, and managing hazardous materials.7 The
Education Infrastructure Act of 1994 was enacted to provide federal
assistance in addressing school infrastructure problems, but the program has
not been funded. The Taxpayer Relief Act of 1997 authorized tax credits for
holders of qualified zone academy bonds whose proceeds can be used for
school infrastructure, such as public school rehabilitation and repair. In
1999, the legislation extended the authority to issue these bonds to 2000
and 2001 and increased the total authorization for them. In its fiscal year
2000 budget, the administration proposed a school construction and
modernization tax incentive, which if enacted would support $22 billion in
interest-subsidized bonds over a 2-year period to help build and repair
schools.

Several proposals introduced in the 106th Congress would, if enacted,
provide new federal financing for school construction. For example, one
proposal would allow the Federal Home Loan Bank to guarantee $500 million
annually in public bonds for school construction. However, by December 1999,
the Congress had passed none of these proposals, and localities and states
continue to provide nearly all school construction financing.

Increased Significantly in the 1990s

Education's data indicate that annual school construction expenditures
(including those for acquired and constructed buildings, land, and
equipment) increased nationally by 39 percent from fiscal year 1990 through
fiscal year 1997. Overall, the largest increase was in the cost of
constructing new buildings and additions to existing ones rather than
purchasing existing buildings, land, and equipment. Education's data also
show that the states varied widely in their construction expenditures per
student during this 8-year period.

1990 Through Fiscal Year 1997

In fiscal years 1990-97, total school construction expenditures increased
from about $17.8 billion to about $24.7 billion. This 39 percent increase
was more substantial than the rise in enrollment, which grew about 12
percent during the same period. As figure 1 illustrates, a large part of the
growth in construction expenditures came in 1996 and 1997. Whereas the
1990-95 expenditure growth was about 12 percent, it was twice that, or 24
percent, in 1995-97. (See appendix I for a discussion of data limitations.)

Figure 1: Cumulative Percentage Changes in School Construction Expenditures
and Fall Enrollments, Fiscal Years 1990-97
Note: The percentages for construction expenditures are based on numbers
that were adjusted to constant 1997 dollars.

Source: U.S. Department of Education.

Regionally, the growth in construction expenditures was highest in the
Midwest, at 63 percent. As figure 2 illustrates, the growth figures in the
Northeast, South, and West were somewhat lower, at 35, 27, and 41 percent,
respectively. (See appendix II for more details.)

Figure 2: Regional School Construction Growth Rates, Fiscal Years 1991-97
Source: U.S. Department of Education.

Buildings

From fiscal year 1992 to fiscal year 1997, the largest increase in annual
school construction expenditures was for the construction of buildings.8
Expenditures for such construction increased by 81 percent, from about $9.2
billion in fiscal year 1992 to about $16.7 billion in fiscal year 1997.
Expenditures for purchasing acquired buildings, such as portable classrooms,
dropped by 42 percent, from about $5.5 billion in fiscal year 1992 to about
$3.2 billion in fiscal year 1997. While annual expenditures for land
decreased by about 12 percent, expenditures for equipment increased by about
29 percent. (See fig. 3.)

Figure 3: School Facilities Acquisition and Construction Expenditures,
Fiscal Years 1992-97
Note: In billions of constant 1997 dollars.

Source: U.S. Department of Education.

Some regions were exceptions to these national patterns. For example, the
South had an increase in annual expenditures for acquired buildings from
fiscal year 1992 to fiscal year 1997 of about 57 percent, while these
expenditures in other regions declined. (Appendix III provides more
details.)

Average annual school construction expenditures per student from fiscal year
1990 through fiscal year 1997 varied widely. Per-student expenditures in the
state with the highest expenditures were more than 25 times greater than in
the state with the lowest expenditures. For example, Nevada had the highest
average annual expenditure at $934 per student. In contrast, Connecticut,
the state with the lowest average annual expenditure, spent $37 per student,
or about $900 less than Nevada. As a whole, the nation spent an average of
about $473 per student per year over the 8-year period. Figure 4 shows how
all the states compared with the U.S. average. (Appendix IV provides more
detailed state-by-state data.)

Figure 4: Average State Construction Expenditures per Student, Fiscal Years
1990-97
Note: In constant 1997 dollars.

Source: U.S. Department of Education.

In general, the states that had the highest per-student annual expenditures
for school construction also tended to have the highest average enrollment
growth rates, and the states that had the lowest per-student expenditures
had significantly lower enrollment growth rates. For example, the average
enrollment growth rate of the 10 states with the largest per-student
expenditures was about 22 percent from 1990 through 1997. In contrast, the
10 states with the lowest per-student expenditures had an average enrollment
growth rate of about 6 percent.

According to F.W. Dodge's data, contract expenditures for the construction
of primary and high school facilities increased substantially in calendar
years 1997 and 1998. F.W. Dodge's data are somewhat narrower than
Education's in that they do not include expenditures for land, acquired
buildings, and equipment, but they do provide more detail about the types of
buildings and kinds of schools. F.W. Dodge's data indicate that most
contract expenditures went for new buildings and additions at primary and
high schools. The renovation of existing buildings received a smaller
portion, as did all kinds of construction projects at middle schools.9

at an Accelerated Rate in 1997 and 1998

Annual contract expenditures for constructing primary and high schools grew
at an accelerated pace in 1997 and 1998.10 Expenditures for primary schools
had an average annual decline of 1.2 percent from 1990 to 1996 but grew at
an average annual rate of 17.4 percent in 1997 and 1998. Annual spending for
high schools had a more consistent growth rate during this period.
Expenditures for high schools grew at an average annual rate of 6.5 percent
from 1990 to 1996 and then increased at an average annual rate of 10.7
percent in 1997 and 1998. Construction spending for middle schools grew at
an average annual rate of 9.5 percent from 1990 to 1997 but declined
somewhat in 1998. (See fig. 5.)

Figure 5: Construction Contract Spending, 1990-98
Note: In billions of constant 1998 dollars.

Source: F.W. Dodge Inc.

The increases for school construction contract spending from 1990 through
1998 were highest in the South for all three types of schools: 55.1 percent
for primary schools, 94.1 percent for middle schools, and 143.1 percent for
high schools. In contrast, the Northeast had a 3.8 percent decline in
spending for primary schools, and the West had a 7.4 percent decline in
spending for middle schools. (See table 1.) In general, the high increases
in the South were influenced by substantial increases in spending in some
states such as Georgia and Texas.

Table 1: Cumulative Percentage Growth Rates for School Construction Contract
Spending, 1990-98

 Kind of school Northeast  Midwest  South  West Total
 Primary        -3.8%      46.7%    55.1%  7.5% 28.4%
 Middle         60.1       84.7     94.1   -7.4 59.4
 High           59.9       79.1     143.1  29.0 79.3

Source: F.W. Dodge Inc.

and Additions

In calendar years 1990-98, about $58.2 billion was spent for building new
facilities, $45 billion for adding to existing facilities, and $21.3 billion
for renovating existing facilities. The distribution of contract
expenditures changed somewhat from 1990 to 1998. For example, although
renovations were only about 17 percent of total contracted construction
spending from 1990 through 1998, this kind of construction project had the
greatest percentage increase--82 percent. Increases were 50 percent for
additions and 42 percent for new facilities during the same period. As
figure 6 shows, whereas contract spending for additions and new facilities
grew at a fairly gradual rate, virtually all the growth in annual spending
for renovations occurred after 1996. According to some industry experts and
others, the substantial growth in renovating schools since 1996 reflects an
increased awareness among the general public and state and local officials
of the need to repair badly deteriorated school buildings.

Figure 6: Contract Expenditures for Additions, Renovations, and New Schools,
1990-98
Note: In billions of constant 1998 dollars.

Source: F.W. Dodge Inc.

Available

There are no complete and current national data on how much funding for
school construction is available annually to each local school district.
While many state education agencies keep records of the number and dollar
amount of local school construction bonds that voters approve or disapprove
each year, most states do not. In addition, there is a wide variance among
states in the degree to which they rely on local compared with state funding
for school construction projects. Even within states, the amount of state or
local funding can vary significantly from year to year. Although a number of
states rely on just one of these funding sources, most states use a
combination of state and local funding.

While we found information on annual school construction expenditures, we
found no national data specifying how these expenditures were financed.
After reviewing the available literature and data sources, we concluded that
the best source of information on how school construction is funded in each
state is a study of public school finance programs in the United States and
Canada that was being readied for publication when we completed our review
in December 1999.11 Preliminary information the editors provided to us shows
that in the 1998-99 school year 32 states provided $10.3 billion in aid for
school facility construction and an additional $228 million for debt
service. Three states (Colorado, Delaware, and Wisconsin) indicated that
state aid is provided for school construction but that this aid is part of
the basic support program and could not be separately accounted for, so no
dollar amounts were given. Fifteen states indicated either no support for
facilities or minimal funding for which no numbers were given. Those states
are Idaho, Louisiana, Michigan, Mississippi, Missouri, Nebraska, Nevada,
North Carolina, North Dakota, Ohio, Oklahoma, Oregon, South Dakota,
Virginia, and Wyoming. The report does not contain similar data on local
funding.

Local general obligation bonds are an important mechanism for financing
school construction. However, we found no comprehensive database of the
number or dollar amount of local school construction bond referendums that
were voted on or passed in 1998. A financial services company collected and
maintained a national database of school bond referendums but it is
incomplete, and we found that it contained only a small portion of the bond
referendums in most of the 19 states from which we obtained bond data.
Therefore, we chose to rely on the results of our survey of all 50 states'
education agencies.

Although most of the 50 states we contacted in our telephone survey did not
maintain comprehensive records of funds that localities provided for school
construction, 19 states provided us data on the number and dollar amount of
local school bond referendums that passed or failed in 1998.12 Table 2
summarizes for each of the 19 states the number and dollar amount of
measures that passed. It also shows that 455 measures were passed, totaling
more than $9 billion and representing 54 percent of both the number and the
dollar amount of the local school construction bond referendums that were
voted on in these states in 1998.

Table 2: Bonds and Their Amounts Approved in Local School Construction Bond
Referendums in 19 States, 1998

                                               Dollar     Dollars approved
                           Number approved     amount      as a percent of
    State       Number     as a percent of   approved     the total dollar
               approved
                          all referendums                   amount of all
                                            (millions)      referendums
 Arizona      12          86%               $253         95%
 Arkansas     36          80                103          81
 California   72          58                3,742        64
 Colorado     19          73                553          87
 Georgiaa     2           100               122          100
 Idaho        4           40                28           44
 Iowa         14          67                89           72
 Michigan     44          41                799          36
 Minnesota    46          74                422          77
 Mississippi  6           55                77           53
 Nebraska     10          45                89           67
 North
 Carolina     3           50                96           53
 Ohio         44          39                532          35
 Oregon       14          30                389          34
 South
 Carolina     4           29                258          26
 Utah         3           100               72           100
 Washington   28          51                773          58
 Wisconsin    93          56                655          48
 Wyoming      1           100               0.5          100
 Total        455         54%               $9,052b      54%

aOn March 18, 1997, all school systems in Georgia became eligible to use
special purpose local option sales tax revenue for school construction
projects if the voters approved a referendum. Only two school systems have
since proposed bond referendums in which the bonds would be repaid in part
or in whole with property tax revenue. Both were voted on and passed in 1998
and are included in this table.

bThe sum of the figures in this column exceeds this total because of
rounding.

The kinds of information on school construction bonds each state maintained
varied. For example, in Oregon, which finances all school construction with
local bond issues, the state department of education maintains data on all
local school construction bond referendums. These data include the name of
the county or school district sponsoring the bond, the date of the vote, the
dollar amount of the measure, and whether it passed or failed.

The Wisconsin office of public instruction maintains local school
construction bond data similar to Oregon's. It also records the purpose of
each bond. For example, its database showed that bonds' proceeds were to be
used for such purposes as a new high school, additions to an elementary
school, a running track, and a pool facility at a middle school. However,
the database does not break out how much of the funds approved in an
individual bond referendum were to be used for each purpose when a bond was
approved for multiple purposes. For example, the $29,695,000 bond referendum
passed for the River Falls School District on December 8, 1998, was for
building a new high school, making additions to an elementary school, and
renovating a middle school and a high school.

Unknown

Data are insufficient to determine on a national basis what portion of
school construction is financed locally and what portion is financed by the
states. The mix of funding between state and local governments varies
considerably from state to state. Even within a given state, the mix can
change dramatically from year to year, especially when localities approve
large bond referendums or when state appropriations or bond referendums are
approved.

School construction in most states is financed with a combination of state
and local funding, although the mix varies considerably from state to state.
A 1998 study by the Education Commission of the States describes school
facilities funding as primarily locally funded in 18 states, shared funding
in 20 states, and primarily state funding in 13 states.13 In Hawaii (which
has one school district that includes the entire state) all school
facilities projects are funded entirely by the state. In our survey of all
50 states, officials in only 9 states informed us that their state did not
appropriate any funds for school construction.

Within states, the amount of state or local bond measures passed or state
appropriations approved to finance school construction can vary
significantly from year to year. For example, as shown in table 3, the bond
referendums local school districts approved in South Carolina rose from $120
million in 1994 to $402 million in 1996 and then declined to $258 million in
1998.

Table 3: South Carolina School Bond Referendum Results, 1994-98

      Bonds approved                     Bonds rejected

 Year Number of        Dollar amount     Number of       Dollar amount
      districts        (millions)        districts       (millions)
 1998 4                $258.2            11              $720.3
 1997 13               320.3             1               108.0
 1996 11               401.7             0               0
 1995 4                176.3             3               178.5
 1994 5                119.7             3               116.5

Several states and localities have passed large school construction bond
measures and several state legislatures have approved appropriations in
recent years that have increased their respective roles in financing school
construction. For example, in November 1998, California voters passed
Proposition 1A, a state school bond initiative that provided $6.7 billion
for the construction and repair of K-12 schools over 4 years.14 Proposition
1A authorized the state to provide funding for 50 percent of the costs
associated with building new schools and 80 percent of the costs associated
with modernizing existing schools. This increased the state's portion of
funding for modernization projects by 30 percentage points from the 50
percent level at which the state allocation board funded previous
modernization programs.

In commenting on a draft of this report, the Department of Education agreed
with our finding that school construction expenditures increased in the
1990s. (The comments are printed in appendix V.) Education noted that our
report did not attempt to determine the effect that increased school
construction expenditures in recent years had on the $112 billion we had
previously estimated was needed to bring schools into good overall
condition.15 Our objective was to report on trends in school construction
funding rather than on how well state and local governments were meeting
their school facilities needs. Additional data beyond those collected for
this report would be needed for such an analysis, including updated data on
school conditions.

Education also suggested that we expand our discussion of the role of
federal tax policy in funding school construction--specifically, the
exemption of interest earned on state and local bonds and qualified zone
academy bonds. We have done so.

Education said that our statement that school construction expenditures rose
39 percent from 1990 to 1997 while school enrollment rose 12 percent during
the period might be misinterpreted. Its concern was that readers might
conclude that a 12 percent increase in school construction expenditures
could accommodate a 12 percent increase in enrollment. Our only point is
that school construction expenditures increased much faster than enrollment
during the period. We do not know whether the increase in construction
expenditures was sufficient to meet construction needs. Making a judgment
about this would have required collecting much more data. Education also
provided us with a number of technical comments that we incorporated as
appropriate.

We are sending copies of this report to the Ranking Minority Member of the
Committee, the Chairman and Ranking Minority Member of the Committee on
Health, Education, Labor, and Pensions of the U.S. Senate, the Secretary of
Education, other appropriate congressional committees and members, and
others who are interested.

If you have any questions, please call me or Joseph J. Eglin, Jr., Assistant
Director, at (202) 512-7215. Major contributors to this report include
Charles M. Novak and Charles H. Shervey.

Sincerely yours,
Barbara D. Bovbjerg
Associate Director, Education, Workforce,
and Income Security Issues

Scope And Methodology

To determine the total expenditures per state for K-12 construction and how
the expenditures were divided between land, buildings, and equipment, we
analyzed annual state-by-state construction expenditure data the Department
of Education's National Center for Education Statistics (NCES) obtained from
state education agencies through its annual Common Core of Data (CCD)
surveys. The U.S. Census Bureau collects these data for NCES. We analyzed
these data to identify national and regional trends in school construction
expenditures from fiscal year 1990 through fiscal year 1997.16 We adjusted
data for fiscal year 1990 through fiscal year 1996 to constant 1997 dollars,
using the gross domestic product price index . For our regional analyses, we
used Census regions. Appendix II shows which states are in which region.

To identify trends in school construction expenditures for land, buildings,
and equipment, we used fiscal year 1992 through fiscal year 1997 because the
data for previous years did not contain the breakouts needed for this kind
of analysis. That is, the data for 1992-97 contained a breakout between
constructed buildings and acquired buildings such as portable classrooms,
but the data for fiscal years 1990 and 1991 did not.

For our analysis of school construction expenditures per student, we used
enrollment data Education also collects through its CCD surveys. To make our
computations, we matched fall enrollment data with construction expenditure
data from the same starting year. For example, we matched fall 1989
enrollment data with fiscal year 1990 construction data because they both
began in 1989. We computed the average annual school construction per capita
expenditure for each state by dividing the average of annual school
construction expenditures for fiscal year 1990 through fiscal year 1997 by
the annual average enrollments over the same time period.

Because some data were missing or incomplete, Education adjusted data for
some states and we did for others to improve comparability among states or
fiscal years. We did not verify the data for accuracy or attempt to obtain
missing data; however, Education reviews and edits each state's data for
completeness and accuracy. If Education finds potential omissions or errors,
it returns the data to the state for correction or verification of its
accuracy. Even though Education took these measures, some data were still
missing for some states.

To determine how school construction expenditures were broken down by grade
or school level (primary school, middle school, and high school) and by kind
of construction (new schools, additions, and renovations), we obtained data
from F. W. Dodge Inc., a division of McGraw-Hill Companies, because
Education data did not contain these breakouts. F. W. Dodge collects data
nationwide for all kinds of private and public construction projects,
including K-12 school facilities. The firm's data measure the value of
contracts public schools or other public entities awarded to private firms
for the construction or renovation of public K-12 school facilities. Unlike
the Education data we used, F. W. Dodge's data do not include expenditures
for land, acquired buildings, or architect and engineering design
activities. Additionally, school expenditures for the use of in-house staff
or other resources of the public school entities for the purposes of
constructing school facilities are included in Education's data but not in
F. W. Dodge's data.

We used F. W. Dodge's data to compute national and regional trends for each
kind of school and each kind of construction project for calendar years
1990-98. We adjusted the data for calendar years 1990-97 to constant 1998
dollars, using the gross domestic product price index. We did not verify the
accuracy of the data we obtained from F. W. Dodge.

To determine how school construction projects are funded, we identified and
reviewed research on school construction financing sources and conducted a
telephone inquiry of each of the 50 state education agencies. Through the
telephone survey, we identified states that collect information from the
local education agencies on voter-approved school construction bonds that
pass and fail and on their 1998 dollar amounts.

We conducted our fieldwork from February through December 1999 in accordance
with generally accepted government auditing standards.

Changes in Annual State School Construction Expenditures From Fiscal Year
1990 to Fiscal Year 1997

                         Annual expenditures                Change
 Region and state       1990            1997           Dollars     Percent
 Northeast
 Connecticut      $17,482,435      $33,257,748      $15,775,313    90.2%
 Maine            115,688,863      44,225,118       -71,463,745    -61.8
 Massachusetts    60,590,794       31,926,031       -28,664,763    -47.3
 New Hampshire    117,724,812      110,912,249      -6,812,563     -5.8
 New Jersey       137,969,243      735,794,720      597,825,477    433.3
 New York         1,425,066,856    1,787,446,267    362,379,411    25.4
 Pennsylvania     1,086,908,888    1,200,977,174    114,068,286    10.5
 Rhode Island     2,939,711        8,016,709        5,076,998      172.7
 Vermont          31,352,439       83,681,866       52,329,427     166.9
 Total            $2,995,724,042   $4,036,237,882   $1,040,513,840 34.7%
 Midwest
 Illinois         $808,822,550     $1,210,214,932   $401,392,382   49.6%
 Indiana          399,266,035      554,053,090      154,787,055    38.8
 Iowa             120,186,443      182,617,511      62,431,068     51.9
 Kansas           169,822,322      81,461,772       -88,360,550    -52.0
 Michigan         518,034,186      1,120,381,516    602,347,330    116.3
 Minnesota        493,522,410      854,746,110      361,223,700    73.2
 Missouri         311,326,902      362,801,611      51,474,709     16.5
 Nebraska         102,635,307      136,915,222      34,279,915     33.4
 North Dakota     24,112,630       26,489,593       2,376,963      9.9
 Ohio             372,635,053      679,989,223      307,354,170    82.5
 South Dakota     31,656,864       36,181,823       4,524,959      14.3
 Wisconsin        240,930,021      605,690,863      364,760,842    151.4
 Total            $3,592,950,722   $5,851,543,266   $2,258,592,544 62.9%
 South
 Alabama          $228,644,407     $287,828,745     $59,184,338    25.9%
 Arkansas         92,358,852       125,903,107      33,544,255     36.3
 Delaware         25,593,095       60,370,550       34,777,455     135.9
 District of
 Columbia         39,446,634       43,940,239       4,493,605      11.4
 Florida          1,716,075,675    1,910,535,292    194,459,617    11.3
 Georgia          610,050,658      861,931,564      251,880,906    41.3
 Kentucky         62,535,737       70,044,741       7,509,004      12.0
 Louisiana        177,560,031      152,737,466      -24,822,565    -14.0
 Maryland         395,093,018      518,801,968      123,708,950    31.3
 Mississippi      81,078,879       156,124,935      75,046,056     92.6
 North Carolina   518,892,880      714,961,739      196,068,859    37.8
 Oklahoma         241,696,773      135,963,740      -105,733,033   -43.7
 South Carolina   277,427,184      390,898,295      113,471,111    40.9
 Tennessee        185,879,267      380,039,655      194,160,388    104.5
 Texas            1,870,901,824    2,446,640,238    575,738,414    30.8
 Virginia         512,689,453      623,842,806      111,153,353    21.7
 West Virginia    36,442,828       96,761,018       60,318,190     165.5
 Total            $7,072,367,197   $8,977,326,098   $1,904,958,901 26.9%
 West
 Alaska           $41,211,270      $126,972,418     $85,761,148    208.1%
 Arizona          455,654,395      467,662,439      12,008,044     2.6
 California       2,000,367,162    2,302,099,441    301,732,279    15.1
 Colorado         258,175,973      620,617,810      362,441,837    140.4
 Hawaii           61,855,731       123,703,629      61,847,898     100.0
 Idaho            42,664,461       145,668,470      103,004,009    241.4
 Montana          56,114,601       33,266,541       -22,848,060    -40.7
 Nevada           205,398,799      311,600,198      106,201,399    51.7
 New Mexico       145,330,637      253,074,817      107,744,180    74.1
 Oregon           118,365,788      347,301,016      228,935,228    193.4
 Utah             103,394,643      280,989,713      177,595,070    171.8
 Washington       634,352,484      761,098,864      126,746,380    20.0
 Wyoming          23,689,363       57,150,184       33,460,821     141.2
 Total            $4,146,575,307   $5,831,205,540   $1,684,630,233 40.6%
 U.S. total       $17,807,617,268  $24,696,312,786  $6,888,695,518 38.7%

Note: In 1997 dollars.

Source: U.S. Department of Education.

Changes in Regional School Construction Expenditures for Land, Buildings,
and Equipment, Fiscal Years 1992-97

            Northeast      Midwest         South          West           Total
 Land
 1992       $830,851,194   $293,686,114    $1,097,299,352 $463,061,676   $2,684,898,336
 1997       684,752,698    592,172,395     435,413,393    656,996,570    2,369,335,056
 $ change   -$146,098,496  $298,486,281    -$661,885,959  $193,934,894   -$315,563,280
 % change   -17.6%         101.6%          -60.3%         41.9%          -11.8%
 Acquired
 buildings
 1992       $377,305,846   $1,510,445,138  $1,596,674,555 $2,066,150,709 $5,550,576,248
 1997       61,312,426     173,529,065     2,499,989,895  476,101,115    3,210,932,501
 $ change   -$315,993,420  -$1,336,916,073 $903,315,340   -$1,590,049,594-$2,339,643,747
 % change   -83.7%         -88.5%          56.6%          -77.0%         -42.2%
 Constructed
 buildings
 1992       $1,578,305,207 $1,702,595,003  $3,910,307,078 $2,048,296,154 $9,239,503,442
 1997       3,105,293,185  4,184,360,008   5,268,577,360  4,164,409,859  16,722,640,412
 $ change   $1,526,987,978 $2,481,765,005  $1,358,270,282 $2,116,113,705 $7,483,136,970
 % change   96.7%          145.8%          34.7%          103.3%         81.0%
 Equipment
 1992       $456,169,378   $422,720,811    $607,243,905   $374,106,895   $1,860,240,990
 1997       184,879,573    901,481,798     773,345,450    533,697,996    2,393,404,817
 $ change   -$271,289,805  $478,760,987    $166,101,545   $159,591,101   $533,163,827
 % change   -59.5%         113.3%          27.4%          42.7%          28.7%

Note: In constant 1997 dollars.

Source: U.S. Department of Education.

Average Annual State School Construction Expenditures per Student, Fiscal
Years 1990-97

                                              Average       Average
         State
                              Total           per year     per student
                                                            per year
 Alabama                $1,551,579,473    $193,947,434     $265
 Alaska                 739,850,611       92,481,326       759
 Arizona                4,305,612,605     538,201,576      773
 Arkansas               916,681,883       114,585,235      258
 California             17,513,021,641    2,189,127,705    417
 Colorado               3,294,148,874     411,768,609      667
 Connecticut            144,622,543       18,077,818       37
 Delaware               333,504,567       41,688,071       399
 District of Columbia   226,294,798       28,286,850       352
 Florida                14,143,070,060    1,767,883,757    877
 Georgia                5,137,812,623     642,226,578      523
 Hawaii                 703,550,019       87,943,752       491
 Idaho                  762,889,360       95,361,170       410
 Illinois               5,378,199,606     672,274,951      357
 Indiana                3,785,099,551     473,137,444      490
 Iowa                   1,218,014,087     152,251,761      308
 Kansas                 734,255,420       91,781,927       203
 Kentucky               802,972,138       100,371,517      155
 Louisiana              1,142,491,267     142,811,408      180
 Maine                  566,549,318       70,818,665       330
 Maryland               2,989,958,884     373,744,861      491
 Massachusetts          342,573,635       42,821,704       49
 Michigan               5,798,446,152     724,805,769      450
 Minnesota              5,263,465,426     657,933,178      825
 Mississippi            881,727,371       110,215,921      218
 Missouri               2,883,718,330     360,464,791      420
 Montana                351,480,150       43,935,019       275
 Nebraska               1,115,640,217     139,455,027      493
 Nevada                 1,734,638,927     216,829,866      934
 New Hampshire          585,950,609       73,243,826       399
 New Jersey             3,460,640,456     432,580,057      379
 New Mexico             1,458,048,463     182,256,058      575
 New York               13,297,370,342    1,662,171,293    614
 North Carolina         4,447,661,879     555,957,735      491
 North Dakota           212,047,429       26,505,929       223
 Ohio                   3,955,509,329     494,438,666      274
 Oklahoma               1,475,607,506     184,450,938      308
 Oregon                 1,534,340,325     191,792,541      378
 Pennsylvania           8,624,372,007     1,078,046,501    623
 Rhode Island           47,892,249        5,986,531        41
 South Carolina         2,112,733,556     264,091,695      414
 South Dakota           352,063,648       44,007,956       321
 Tennessee              1,967,850,022     245,981,253      286
 Texas                  18,039,309,919    2,254,913,740    631
 Utah                   1,677,754,867     209,719,358      452
 Vermont                361,417,654       45,177,207       449
 Virginia               4,104,146,867     513,018,358      494
 Washington             6,147,436,619     768,429,577      854
 West Virginia          764,308,148       95,538,519       303
 Wisconsin              3,315,586,301     414,448,288      496
 Wyoming                315,188,234       39,398,529       395
 U.S. total and average $163,019,105,967  $20,377,388,246  $473

Note: In constant 1997 dollars.

Source: U.S. Department of Education.

Comments From the Department of Education

Related GAO Products

School Facilities: Reported Condition and Costs to Repair Schools Funded by
Bureau of Indian Affairs (GAO/HEHS-98-47, Dec. 31, 1997 ).

School Facilities: America's Schools Report Differing Conditions
(GAO/HEHS-96-103, June 14, 1996 ).

School Facilities: Profiles of School Condition by State (GAO/HEHS-96-148,
June 24, 1996 ).

School Facilities: Accessibility for the Disabled Still an Issue
(GAO/HEHS-96-73, Dec. 29, 1996 ).

School Facilities: States' Financial and Technical Support Varies
(GAO/HEHS-96-27, Nov. 28, 1995 ).

School Facilities: America's Schools Not Designed or Equipped for 21st
Century (GAO/HEHS-95-95, Apr. 4, 1995 ).

School Facilities: Condition of America's Schools (GAO/HEHS-95-61, Feb. 1,
1995 ).

(104965)

Figure 1: Cumulative Percentage Changes in School Construction Expenditures
and Fall Enrollments, Fiscal Years 1990-97 7

Figure 2: Regional School Construction Growth Rates, Fiscal Years 1991-97 8

Figure 3: School Facilities Acquisition and Construction Expenditures,
Fiscal Years 1992-97 10

Figure 4: Average State Construction Expenditures per Student, Fiscal Years
1990-97 12

Figure 5: Construction Contract Spending, 1990-98 14

Figure 6: Contract Expenditures for Additions, Renovations, and New Schools,
1990-98 16

Table 1: Cumulative Percentage Growth Rates for School Construction Contract
Spending, 1990-98 15

Table 2: Bonds and Their Amounts Approved in Local School Construction Bond
Referendums in 19 States, 1998 19

Table 3: South Carolina School Bond Referendum Results, 1994-98 21
  

1. School Facilities: Condition of America's Schools (GAO/HEHS-95-61 , Feb.
1, 1995).

2. We used data from F.W. Dodge Inc., a division of McGraw-Hill Companies,
for our analyses of expenditures by kind of school and by kind of
construction project because Education's data do not contain such breakouts.
Appendix I explains the two databases we used for our analyses.

3. We converted all Education data for the fiscal years before 1997 to
constant fiscal year 1997 dollars and all F.W. Dodge data for the calendar
years before calendar year 1998 to constant 1998 calendar year dollars.

4. A list of our related products appears at the end of the report.

5. "Baby boom echo" is the term commonly used to describe the children of
the baby boomers, persons born from 1946 to 1964.

6. Dennis Zimmerman, "Tax-exempt Bond Proposals to Increase Public
Elementary and Secondary School Facilities," statement submitted for the
hearing record, Senate Finance Committee, Congressional Research Service
(Washington, D.C.: Mar. 3, 1999), p. 1.

7. Impact Aid provides funds to compensate local school districts affected
by federal activities for revenue lost because of the nontaxable status of
federal property within their jurisdictions and the cost of educating
children who live on or whose parents work on federal property or whose
parents are in the military. Some of these funds may be used, for example,
for the construction of urgently needed minimum school facilities in
districts whose numbers of students increase substantially as a result of
federal activities.

8. Fiscal year 1992 is the earliest year for which the available data
contain the expenditure components used for this analysis.

9. For the purpose of this report, we used the term "middle schools" to also
include junior high schools.

10. We converted F.W. Dodge's data for the years before 1998 to constant
1998 dollars.

11. Catherine C. Sielke and others, eds., "Public School Finance Programs of
the United States and Canada 1998-99," draft report.

12. Three of these states did not keep data on a portion of bonds. Minnesota
did not maintain data on bonds of less than $400,000, and Michigan did not
have data on bonds that were not submitted to the state for approval before
they were voted on. Arkansas did not have data on bonds that did not pass
and were not submitted to the state for approval before they were voted on.
In addition, Iowa was able to provide bond data for only the first 6 months
of 1998.

13. Education Commission of the States, "Making Better Decisions About
Funding School Facilities" (April 1998). In the study, the states total 51
because South Carolina's school facilities are shown as being both primarily
locally funded and using shared funding. Much of the data used in this
analysis was from an earlier study entitled "Public School Finance Programs
of the United States and Canada, 1993-94."

14. Joel Cohen, "School Facility Financing: A History of the Role of the
State Allocation Board and Options for the Distribution of Proposition 1A
Fund," California Research Bureau (Feb. 1999), pp. 2-3 and 31.

15. GAO/HEHS-95-61 , Feb. 1, 1995.

16. For purposes of these data, Education uses a fiscal year that begins on
July 1 and ends on the following June 30, and it uses the year of the ending
date to designate the fiscal year. For example, fiscal year 1990 begins on
July 1, 1989, and ends on June 30, 1990.
*** End of document. ***