Highway Funding: Problems With Highway Trust Fund Information Can Affect
State Highway Funds (Letter Report, 06/29/2000, GAO/RCED/AIMD-00-148).

Pursuant to a congressional request, GAO reviewed the: (1) Department of
the Treasury's process for allocating highway user tax receipts to the
Highway Account of the Highway Trust Fund; (2) Federal Highway
Administration (FHwA) process for estimating motor fuel usage and the
contributions to the account that are attributable to highway users in
each state; and (3) impact of these processes on the amount of highway
program funds distributed to each state.

GAO noted that: (1) given the Transportation Equity Act for the 21st
Century's increased linkage between the receipts in the Highway Account
of the Highway Trust Fund and the level of highway program funds
distributed to the states, the quality of information developed by the
Treasury and FHwA is more critical than ever before in determining
accurate funding amounts to be distributed to the states each year; (2)
the processes used by Treasury and FHwA to estimate overall receipts and
the portion of those receipts attributable to highway users in
individual states are highly complex and susceptible to error, and the
reliability of the estimates has not been demonstrated; (3) as a result,
there is little assurance that the actual amounts distributed to the
states are accurate, although there is no way of knowing the extent of
over- or under-payments, if any, to individual states, given the
information available from the two agencies; (4) although the Treasury
and FHwA are taking actions to review and improve their estimating
processes, these actions are not sufficient to correct all the
weaknesses; (5) for example, although Treasury has a number of ongoing
efforts aimed at improving its revenue allocation and reporting process
for trust funds, it does not have a comprehensive plan, with timeframes,
that addresses all of the policy and process changes needed to fully
implement those efforts; and (6) furthermore, FHwA's efforts to improve
the data and methodology used in its attribution process do not verify
the accuracy of the information resulting from that process.

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

 REPORTNUM:  RCED/AIMD-00-148
     TITLE:  Highway Funding: Problems With Highway Trust Fund
	     Information Can Affect State Highway Funds
      DATE:  06/29/2000
   SUBJECT:  Excise taxes
	     Internal controls
	     Accounting procedures
	     Federal agency accounting systems
	     Funds management
	     Reporting requirements
	     Federal/state relations
	     Federal aid for highways
	     Trust funds
IDENTIFIER:  Highway Trust Fund
	     IRS Excise Files Information Retrieval System

******************************************************************
** 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/RCED/AIMD-00-148

Appendix I: Relationship Between Highway User Tax Receipts
and Federal Highway Program Funds

16

Appendix II: Treasury's Process for Allocating Tax Receipts
to the Highway Account of the Highway Trust
Fund Is Complex and Error Prone

20

Appendix III: Reliability of FHWA's Attribution Process Has
Not Been Demonstrated

31

Appendix IV: Problems With Treasury's and FHWA's Processes
Can Affect Fund Distributions to the States

46

Appendix V: Scope and Methodology

52

Appendix VI: Comments From the Internal Revenue Service

53

Appendix VII: GAO Contacts and Staff Acknowledgments

59

60

Table 1: Highway User Tax Structure 16

Table 2: Gross Highway User Tax Receipts Credited to
the Highway Account of the Highway Trust Fund,
Fiscal Year 1999 17

Table 3: Factors for Attributing Highway User Taxes to Highway
Users in Each State 33

Table 4: Estimated Motor Fuel Usage and Contributions to the
Highway Account of the Highway Trust Fund Attributable
to Highway Users in Each State, Apportionment Cycle
for Fiscal Year 2000 38

Table 5: Calculation of the Revenue Aligned Budget Authority
Adjustment, Fiscal Years 2000 and 2001 48

DOT Department of Transportation

ExFIRS Excise Files Information Retrieval System

FHWA Federal Highway Administration

GAO General Accounting Office

IRS Internal Revenue Service

ISTEA Intermodal Surface Transportation Efficiency Act of 1991

RABA Revenue Aligned Budget Authority

TEA-21 Transportation Equity Act for the 21st Century

Resources, Community, and
Economic Development Division

B-283480

June 29, 2000

The Honorable Bud Shuster
Chairman
The Honorable James L. Oberstar
Ranking Democratic Member
Committee on Transportation and Infrastructure
House of Representatives

The Transportation Equity Act for the 21st Century (TEA-21)1 authorized
$217.9 billion for highway, mass transit, and other surface transportation
programs for fiscal years 1998 through 2003. TEA-21 continued the use of the
Highway Trust Fund--which is divided into a Highway Account and a Mass
Transit Account--as the mechanism to account for federal highway user tax
receipts that fund various surface transportation programs. Highway user
taxes include excise taxes on motor fuels (gasoline, gasohol, diesel, and
special fuels) and truck-related taxes on truck tires, sales of trucks and
trailers, and the use of heavy vehicles.

TEA-21 also established mechanisms for ensuring that the level of federal
highway program funds distributed to the states would be more closely linked
than before to the highway user tax receipts credited to the Highway Account
of the Highway Trust Fund.2 The process for distributing these tax receipts
has two separate components: (1) the Treasury Department determines the
overall amount of receipts in the Highway Account and (2) the Federal
Highway Administration (FHWA) estimates the portion of the overall amount
that is attributable to each state by using state data on motor fuel usage.
Under TEA-21, billions of dollars in highway program funds--about $13
billion in fiscal year 2000 alone--are distributed to the states on the
basis of information developed by the Treasury and Transportation
departments.

Given the increased linkage between Highway Account receipts and the level
of highway program funds distributed to the states--and the magnitude of the
dollars involved--effective processes for collecting and reporting accurate
information on such receipts are critical to the effective implementation of
TEA-21. As agreed with your offices, this report discusses (1) the Treasury
Department's process for allocating highway user tax receipts to the Highway
Account of the Fund, (2) FHWA's process for estimating motor fuel usage and
the contributions to the account that are attributable to highway users in
each state, and (3) the impact of these processes on the amount of highway
program funds distributed to each state. This letter summarizes our
findings. A more detailed analysis of each of the issues we address is
presented in appendixes II through IV.

Under TEA-21, the linkage between highway user tax receipts in the Fund's
Highway Account and federal highway program funding levels was enhanced by
(1) guaranteeing specific annual funding levels for most highway programs
over a 6-year period on the basis of the projected receipts in the Highway
Account, (2) providing that the guaranteed spending level for each fiscal
year would be adjusted upward or downward if the receipt levels in the
Highway Account increased or decreased from those projected in TEA-21, and
(3) changing some of the formulas for distributing highway program funds to
the states to more closely reflect estimated Highway Account contributions
that are attributable to highway users in each state.

Federal highway user taxes include excise taxes on motor fuels (gasoline,
gasohol, diesel, and special fuels) and truck-related taxes on truck tires,
sales of trucks and trailers, and the use of heavy vehicles. The motor fuel
taxes are generally paid by someone other than the consumer. Oil companies
typically pay a per-gallon tax on the motor fuels at the point where their
fuel is loaded into tanker trucks or rail cars at a terminal.3 Tire
manufacturers pay taxes on truck tires, by weight, and retailers pay taxes
on the sales price of trucks and trailers. Owners of heavy highway vehicles
pay taxes on the use of these vehicles, making this the only highway tax
directly paid by the highway user. The highway user pays the other taxes
indirectly, since the costs of these taxes become part of the purchase price
of the products taxed. Thus, the motor fuel taxes are paid by businesses
rather than by consumers at retail fuel pumps.

The Treasury Department uses a revenue allocation and reporting process to
distribute the highway user taxes to the Fund's Highway Account. Twice a
month, business taxpayers make deposits of excise taxes--including highway
user taxes--generally through Treasury's Electronic Federal Tax Payment
System. Excise taxes are deposited into Treasury's General Fund as received.
Treasury uses a complex and lengthy process--involving four organizations
within the Department--for allocating excise tax receipts to the various
trust funds, including the Highway Trust Fund. The Department uses this
process, in part, because it does not obtain data from taxpayers, at the
time they make semimonthly desposits, on the types of excise taxes that
these deposits are intended to cover.

Since highway user taxes are generally paid by businesses, rather than
consumers, most of the federal motor fuel and truck taxes come from only the
handful of states where those businesses have their corporate headquarters
and pay their taxes. As a result, the Treasury Department does not provide
FHWA with state-level data on highway tax receipts, and FHWA must therefore
estimate these data in order to distribute Highway Account funds to the
states under various highway programs. FHWA estimates state-level
contributions through what it refers to as its "attribution process."
Through this process, it determines each state's share of highway motor fuel
usage on the basis of data provided by the states, and it uses that
information to estimate the amount of contributions to the Highway Account
attributable to each state's highway users. FHWA described this process in a
June 1985 Federal Register notice and, according to FHWA officials, obtained
comments on it from the states and from congressional staff. The information
developed by Treasury and FHWA is used to determine the amounts of funds
distributed to each state under several major highway programs.

TEA-21 requires that billions of dollars in highway program funds be
distributed to the states on the basis of Highway Account receipts
information developed by the Treasury and Transportation departments--about
$13 billion in fiscal year 2000 alone. The $13 billion figure includes (1)
$11.5 billion in highway program funds that were distributed to the states
in fiscal year 2000 on the basis of FHWA's analysis of data on each state's
motor fuel usage and (2) a $1.5 billion increase to the fiscal year 2000
overall highway program funding level originally projected in TEA-21 on the
basis of Treasury's reports on Highway Account receipts. (See app. I for
further information on the linkage between receipts in the Highway Account
of the Fund and the level of highway program funds distributed to the
states.)

Given TEA-21's increased linkage between the receipts in the Highway Account
of the Highway Trust Fund and the level of highway program funds distributed
to the states, the quality of information developed by the Treasury
Department and FHWA is more critical than ever before in determining
accurate funding amounts to be distributed to the states each year. The
processes used by Treasury and FHWA to estimate overall receipts and the
portion of those receipts attributable to highway users in individual states
are highly complex and susceptible to error, and the reliability of the
estimates has not been demonstrated. As a result, there is little assurance
that the actual amounts distributed to the states are accurate, although
there is no way of knowing the extent of over- or under-payments, if any, to
individual states, given the information currently available from the two
agencies. Although the Treasury Department and FHWA are taking actions to
review and improve their estimating processes, these actions are not
sufficient to correct all the weaknesses. For example, although Treasury has
a number of ongoing efforts aimed at improving its revenue allocation and
reporting process for trust funds, it does not have a comprehensive plan,
with time frames, that addresses all of the policy and process changes
needed to fully implement those efforts. Furthermore, FHWA's efforts to
improve the data and methodology used in its attribution process do not
verify the accuracy of the information resulting from that process. This
report makes recommendations to the Treasury and Transportation departments
that are designed to reduce the risk of errors and increase the reliability
of the information used to distribute federal highway program funds to the
states. We provided a draft of this report to the Treasury Department, the
Internal Revenue Service (IRS), and the Transportation Department. IRS did
not fully address whether or when the agency will take actions to implement
our recommendations to the Secretary of the Treasury. Transportation and
FHWA officials agreed with our recommendations to the Secretary of
Transportation and indicated that they would address them as part of the
agency's ongoing efforts to improve its attribution process. All agencies,
including Treasury, made technical clarifications, which we incorporated as
appropriate.

Account of the Highway Trust Fund Is Complex and Error Prone

The Treasury Department's process for allocating the appropriate amounts of
highway user tax receipts to the Highway Account of the Fund is complex and
lengthy, in part, because the Department does not obtain data from the
taxpayer on the amounts deposited for highway user taxes at the time those
deposits are made. Treasury uses a very cumbersome process--involving four
organizations within the Department--to estimate highway user tax receipts,
credit the estimated amounts to the Fund, and subsequently certify the
amounts collected and adjust the amounts credited by analyzing payment and
tax return data.4 Furthermore, our prior reviews have repeatedly identified
errors in the information used in Treasury's process.5 In addition, the
process may result in yearly reports on Highway Trust Fund receipts that do
not contain timely information because, for example, the amounts
attributable to a particular quarter have not been included in the
certification for that quarter. Since Treasury's reports on receipts are
used to calculate yearly adjustments to the funding levels authorized by
TEA-21, problems with the accuracy and timeliness of the information in
these reports can affect the accuracy of the amounts distributed to the
states.

Within Treasury, IRS is taking the lead in considering changes aimed at
improving Treasury's trust fund revenue allocation and reporting process.
IRS has undertaken three major efforts in this area. For example, IRS
recently completed a study that included an assessment of the taxpayer
burden, potential benefits, and feasibility of requiring taxpayers to
report, with their semimonthly deposits, a detailed breakdown of the amounts
deposited for each excise tax, including highway user taxes. The study found
that while the benefits to the government would be high, imposing such a
requirement at this time would be potentially burdensome for payers of
excise taxes and the quality of the data would likely be poor, since
companies do not currently have such detailed information at the time they
make their deposits. The study recommended that IRS (1) implement incentives
at this time to encourage taxpayers to start providing IRS with a detailed
tax breakdown with their tax deposits and (2) do another study in several
years to determine whether taxpayers' capability to provide these detailed
data has improved and whether IRS should require taxpayers to provide the
detailed data at that time. IRS is also designing, as part of its
modernization efforts, a new payments information database that would be
capable of recording payment data, by specific tax type, if provided by
taxpayers with their deposits. Furthermore, IRS is planning to design a new
method for allocating tax receipts to the excise-tax-related trust
funds--including the Highway Trust Fund--using the taxpayer-supplied data if
they become available. If taxpayers do not provide the more detailed data,
the new allocation method will allocate their deposits to the trust funds
using estimated tax breakdowns, as derived from their past tax returns.

While these efforts may improve the accuracy and timeliness of Treasury's
reporting on the receipts in the Highway Account of the Fund, IRS does not
have a comprehensive plan, including time frames, for addressing all of the
policy and process changes needed to implement its efforts, and full
implementation is still years away. For example, IRS has no formal plan,
including time frames, for carrying out its study's recommendations that the
agency (1) use incentives at this time to encourage taxpayers to provide
detailed data with their deposits and (2) reexamine, in several years,
whether taxpayers have the capability to provide these data and whether IRS
should require them to do so. Furthermore, IRS' plan for developing and
implementing the new database and revenue allocation method does not address
what changes may be needed--as a result of changes being planned in the
revenue allocation method--in Treasury's methods for certifying and
adjusting the amounts allocated to the trust funds on the basis of analyses
of tax return and payment data. IRS officials told us that the agency is
tentatively planning to implement the new database and revenue allocation
method in about 4 years.

FHWA's "attribution" process--which estimates each state's relative highway
motor fuel usage from state tax data and uses that information to estimate
the relative contributions to the Highway Account from each state's highway
users--also has significant weaknesses that raise concerns about its
reliability. For example, the state data on motor fuel usage submitted to
FHWA for use in the attribution process have not been thoroughly and
independently reviewed to verify their accuracy, reliability, and
consistency across all states. Furthermore, FHWA's methodology for analyzing
the state motor fuel data is susceptible to error, since it (1) has never
been fully documented or independently reviewed; (2) is extremely
complicated, involving nearly 200 formulas that are needed to accommodate
all of the differences in states' methods for taxing and reporting on motor
fuels; and (3) has been repeatedly adjusted over several decades in response
to changing state tax laws and federal program requirements. Finally, the
responsibility for collecting, analyzing, and evaluating the source data
submitted by the states and for carrying out the attribution methodology
rests with only two officials in FHWA headquarters, thus increasing the need
for strong oversight.

FHWA has a number of ongoing efforts to review the quality of the state data
and improve its attribution process, but these efforts do not fully address
our concerns. For example, FHWA performs trend analyses of the state data to
ensure consistency in how they are reported from year to year, and it also
conducts periodic reviews of each state to review the overall methodology
used to develop the key motor fuel figures reported to FHWA. While we view
these as positive efforts to improve reporting by the states, these efforts
do not verify the accuracy and reliability of the state data submitted to
FHWA. FHWA is also addressing issues and areas for improving the reporting
system for state motor fuel data with the help of a multiagency committee.
The scope of FHWA's work with this committee, which was initially focused on
a small percentage of the total data used in the attribution process, has
evolved to include broader concerns about the attribution process. For
example, FHWA plans to issue a paper describing its attribution methodology
to better inform state officials about how their state motor fuel data are
used in the attribution process, and the agency has also conducted outreach
efforts to state officials by holding several workshops in cities across the
country. However, FHWA still needs to address additional issues that we have
raised about the reliability of the state motor fuel data and the
attribution methodology.

In addition to FHWA's efforts, IRS is developing a data system--known as the
Excise Files Information Retrieval System (ExFIRS)--that may provide
information that would be useful for FHWA's attribution process.
Specifically, the system may provide data on highway usage of motor fuel by
federal taxpayers in each state. FHWA could use these data as a tool to
validate the state motor fuel data currently used in its attribution
process. Since there are some possible limitations in the reliability of the
ExFIRS information, however, FHWA and IRS would need to evaluate the
information when it becomes available.

Distributions to the States

Problems with the reliability of Treasury's and FHWA's processes for
estimating the overall level of highway user tax receipts, and the portion
of those receipts attributable to the highway users in individual states,
can affect the amounts that are distributed to the states under the
following two methods. First, Transportation uses Treasury's revenue reports
to calculate any changes to the overall amounts of highway program funds
that go to the states each year, as authorized by TEA-21.6 Second, FHWA
distributes funds to the states for several major highway programs using
formulas that rely to a large extent on state-level estimates of motor fuel
usage and contributions to the Highway Account developed from FHWA's
attribution process. Since the amount of money distributed to the states
using Treasury's and FHWA's information is large--about $13 billion in
fiscal year 2000 alone--the potential impact of problems with the accuracy
and timeliness of that information are significant. Given that the
reliability of the estimates developed by Treasury and FHWA has not been
demonstrated, there is little assurance that the actual amounts distributed
to the states are accurate.

Given TEA-21's increased linkage between the receipts in the Highway Account
of the Fund and the level of highway program funds distributed to the
states, the quality of the information developed by the Treasury Department
and FHWA on such receipts is critical to the effective implementation of
TEA-21. Taken together, the processes used by the Treasury Department and
FHWA are highly complex and susceptible to error, and the reliability of the
estimates resulting from those processes has not been demonstrated. As a
result, there is little assurance that the actual amounts distributed to the
states are accurate. The actions being taken by both the Treasury Department
and FHWA to improve their estimating processes are not sufficient to correct
the weaknesses and reduce the risk of errors. Treasury is several years away
from implementing its efforts to improve its revenue allocation and
reporting process, and it has not yet developed a comprehensive plan, with
time frames, for carrying out those efforts. Such a plan is crucial to
ensure that the Department places adequate priority on examining ways to
improve estimates of highway user tax receipts, which are used to make
allocations to the Highway Trust Fund. With regard to FHWA's attribution
process, the state motor fuel data and the extremely detailed and
complicated methodology used by FHWA to analyze those data have not been
thoroughly and independently reviewed to verify their accuracy, reliability,
and consistency across all states. Moreover, the methodology is not
adequately documented and is the responsibility of just two FHWA officials.
Finally, given the complexity of the activities reviewed in this report, the
fact that two major federal departments (including four organizations within
the Treasury Department) share responsibility for addressing the issues we
have identified, and the magnitude of the dollars involved, the cognizant
congressional oversight committees need to be annually informed of
Treasury's and FHWA's actions to improve the reliability of the information
used to distribute federal highway funds to the states.

To reduce the risk of errors and increase the reliability of the information
used to distribute highway program funds to the states, we recommend the
following:

ï¿½ The Secretary of the Treasury should place greater emphasis and priority
on the agency's efforts to improve its process for allocating receipts to
the Highway Trust Fund and other excise-tax-related trust funds by
developing a comprehensive plan that addresses all of the policy and process
changes that will be needed to fully implement these efforts. In particular,
the plan should include steps and time frames for

ï¿½ evaluating and deciding whether to use incentives as a near-term method
for encouraging taxpayers to provide detailed data--at the time of
deposit--on the specific types of excise taxes for which deposits are made;

ï¿½ reexamining taxpayer capabilities to provide these detailed data in
several years and deciding whether to require the data from taxpayers at
that time, considering the potential burden to taxpayers and the potential
benefits to the government; and

ï¿½ determining what changes may be needed--as a result of changes being
planned in the revenue allocation method--in Treasury's methods for
certifying and adjusting the amounts allocated to the trust funds on the
basis of analyses of tax return and payment data.

ï¿½ The Secretary of Transportation should direct the Administrator of FHWA to
improve the reliability of the attribution process by

ï¿½ ensuring that detailed, independent verifications of motor fuel data are
performed for each state, perhaps on a rotational basis over a period of
years;

ï¿½ fully documenting FHWA's current methodology for analyzing the state motor
fuel data used in the attribution process to explain its contents and
rationale and to specify the procedures for carrying it out;

ï¿½ conducting an independent, comprehensive review of this methodology; and

ï¿½ evaluating the potential reliability of IRS' ExFIRS data, once they become
available, for use as a tool to validate the state motor fuel data currently
used in the attribution process.

ï¿½ The secretaries of the Treasury and Transportation should report to the
cognizant highway-related authorizing committees in Congress--the Committee
on Transportation and Infrastructure in the House of Representatives and the
Committee on Environment and Public Works in the Senate--within 12 months,
and annually thereafter, on their progress in implementing our
recommendations and on the extent to which their processes for collecting
and reporting information on Highway Account receipts are producing
accurate, reliable data for use in distributing highway program funds to the
states.

We provided the Treasury Department, IRS, and the Transportation Department
with copies of a draft of this report for their review and comment. Treasury
officials, including the Director of the Office of Tax Analysis, deferred to
IRS for a position on our recommendations to the Secretary of the
Treasury--on the need to develop a comprehensive plan for improving
Treasury's trust fund revenue allocation and reporting process--because they
said that IRS has the lead role in considering changes aimed at improving
that process. Also, Treasury officials said that they are considering our
recommendation that the Secretary of the Treasury annually report to the
cognizant highway-related authorizing committees in the Congress. The
officials also provided minor technical clarifications, which we
incorporated into the report.

The Commissioner of Internal Revenue provided written comments on a draft of
this report, which, along with our response, appear in appendix VI. IRS
suggested some clarifications, which we made, in our discussion of the
agency's processes for collecting excise taxes and certifying the amounts
collected. While these changes improved the accuracy of this discussion,
they did not affect our conclusions and recommendations on Treasury's
process for allocating tax receipts to the Highway Trust Fund. IRS also
stated that initial allocations to the Highway Trust Fund will continue to
be based on estimates, which will require adjustment on the basis of factual
data as such data become available. We agree. However, we believe that
Treasury may have the opportunity to improve the quality of the estimates on
which initial allocations are based if it obtains tax type information from
taxpayers when they make their deposits. While IRS did not take exception to
our recommendations, it did not fully address whether or when the agency
will take actions to implement them.

To obtain Transportation's comments on the draft of this report, we met with
agency officials, including the Director of FHWA's Office of Highway Policy
Information. These officials agreed with our recommendations, indicating
that they would be addressed as part of FHWA's overall efforts to improve
its attribution process. In addition, the officials asserted that FHWA uses
the best available information provided by the states, along with a
"reasonable, logical, and consistent process," to produce what they believe
are reproducible results and the best possible estimates of federal highway
taxes attributable to each state. Although we do not dispute FHWA's intent
to develop a "reasonable, logical, and consistent" attribution process, our
review showed that the reliability of that process has not been demonstrated
and, as a result, there is little assurance that the actual amounts of
highway program funds distributed to the states are accurate. While we
recognize that state motor fuel data may be the best available information
source for estimating the Highway Account receipts attributable to highway
users in each state, we believe that until FHWA (1) performs detailed
verifications of each state's motor fuel data and (2) fully documents and
assesses the reliability of its attribution methodology, the agency cannot
ensure that its attribution process yields reliable information and the
"best possible" estimates for use in distributing highway program funds to
the states. These officials also made clarifying comments, which we
incorporated as appropriate.

The officials also emphasized that FHWA's attribution process was based on
the agency's interpretation of relevant statutes. They noted that the
process was reviewed through a Federal Register notice in 1985, which
allowed for broad public comment. They also said that, when TEA-21 was
enacted, they recognized the need to review the attribution process, address
any concerns from the states, and provide any necessary improvements. The
FHWA officials also believe--on the basis of the results of a series of
outreach sessions they held with the states--that the states have few
specific concerns with the overall accuracy of the attribution process.
Finally, these officials told us that they have identified a number of ways
that the process could be improved and are continuing their efforts with the
states to identify and implement any necessary modifications to the
attribution process. Since our report already recognized that FHWA obtained
comments on its attribution process and is working with the states to
improve the process, we did not revise the report further.

In addition, the officials stated that their confidence in the results of
their attribution analysis is reinforced by its relationship with an
"available, independently derived, alternative method for assessing highway
use." According to FHWA officials, this alternative method involved
comparing state-level data on "vehicle miles traveled" and motor fuel usage.
However, the data on vehicle miles traveled have limitations as a tool for
assessing the reliability of the motor fuel data currently used in the
attribution process, and the comparison does not address our concern that
the reliability of the attribution data and methodology has not been
demonstrated. Therefore, we did not make changes to the report.

Our work was performed at several offices within the Treasury and
Transportation departments. (See app. V for a detailed description of our
scope and methodology.) We conducted our work from August 1999 through June
2000 in accordance with generally accepted government auditing standards.

We are sending copies of this report to the Honorable Rodney E. Slater,
Secretary of Transportation; the Honorable Kenneth R. Wykle, Administrator,
Federal Highway Administration; the Honorable Lawrence H. Summers, Secretary
of the Treasury; the Honorable Charles O. Rossotti, Commissioner of Internal
Revenue; and the Honorable Jacob Lew, Director, Office of Management and
Budget. We will make copies available to others on request. If you have any
questions about this report, please call me at (202) 512-3650 or Gregory D.
Kutz at (202) 512-3406. GAO contacts and staff acknowledgements are listed
in appendix VII.

Phyllis F. Scheinberg
Associate Director,
Transportation Issues

Relationship Between Highway User Tax Receipts and Federal Highway Program
Funds

The Transportation Equity Act for the 21st Century (TEA-21) continued the
use of the Highway Trust Fund as the mechanism for accounting for federal
highway user tax receipts that fund various surface transportation programs.
Established in 1956, the Fund has been divided since 1983 into two accounts:
a Highway Account and a Mass Transit account. Receipts from the Highway
Account are used to fund highway programs, through which billions of dollars
are distributed to the states annually for the construction and repair of
highways and related activities.

Financing for the Fund is derived from a variety of federal highway user
taxes including excise taxes on motor fuels (gasoline, gasohol, diesel, and
special fuels) and truck-related taxes on truck tires, sales of trucks and
trailers, and the use of heavy vehicles. The motor fuel taxes are generally
paid by someone other than the consumer. Oil companies pay a per-gallon tax
on motor fuels at the point where it is loaded into tanker trucks or rail
cars at a terminal.7 Tire manufacturers pay taxes on truck tires, by weight,
and retailers pay taxes on the sales price of trucks and trailers. Owners of
heavy highway vehicles pay a tax annually on the use of these vehicles,
making this the only highway tax directly paid by the highway user. The
highway user pays the other taxes indirectly, since the costs of these taxes
become part of the purchase price of the products taxed. Table 1 provides
further details on the highway user tax structure and the rates currently in
effect, as described by the Federal Highway Administration (FHWA).

         Tax type                     Tax rate (as of Aug. 1999)
 Gasoline                  18.4 cents per gallon
 Diesel                    24.4 cents per gallon
 Gasohol (10 percent
 ethanol)a                 13 cents per gallon
 Special fuels
 General rate              18.4 cents per gallon
 Liquefied petroleum gas   13.6 cents per gallon
 Liquefied natural gas     11.9 cents per gallon
 M85 (from natural gas)    9.25 cents per gallon
 Compressed natural gas    48.54 cents per thousand cubic feet
 Tires:
 0 − 40 pounds       No tax
 Over 40 pounds to 70      15 cents per pound in excess of 40 pounds
 pounds                    $4.50 plus 30 cents per pound in excess of 70
 Over 70 pounds to 90      pounds
 pounds                    $10.50 plus 50 cents per pound in excess of 90
 Over 90 pounds            pounds
                           12 percent of retailer's sales price for
 Truck and trailer sales   tractors and trucks over 33,000 pounds gross
                           vehicle weight (GVW) and trailers over 26,000
                           pounds GVW.
                           Annual tax applied to trucks 55,000 pounds and
 Heavy vehicle use         over GVW. Tax equals $100 plus $22 for each
                           1,000 pounds (or fraction thereof) in excess of
                           55,000 pounds. Maximum tax of $550.

aOther rates apply to gasohol blends containing less than 10 percent ethanol
or blends made with methanol.

Source: FHWA.

The Treasury Department credits most motor fuel tax receipts and all
truck-related tax receipts to the Highway Account of the Highway Trust
Fund.8 The Department of Transportation relies on Treasury to allocate
highway user tax receipts to the Fund and report on the amounts credited.
Transportation uses Treasury's reports on the Fund's overall Highway Account
receipts in determining the levels of funding to be distributed to the
states through several major highway programs. Table 2 shows receipts of
these taxes credited by Treasury to the Highway Account of the Fund in
fiscal year 1999.

          Dollars in millions

              Type of tax                Amount of Percentage of
                                          taxes     total taxes
 Gasoline                               $21,373.1  61
 Gasohol                                1,310.7    4
 Diesel                                 8,388.3    24
 Subtotal−all motor fuels         $31,072.1  89
 Truck tires                            416.0      1
 Truck and trailer sales                2,809.9    8
 Heavy Vehicle Use                      813.7      2
 Subtotal−all truck-related taxes $4,039.6   11
 Total−gross highway user taxesa  $35,111.7  100

a An additional $6.7 million in fines and penalties for motor carrier safety
violations was credited to the Fund in 1999.

Source: GAO's presentation of data from Treasury's financial statements for
the Highway Trust Fund for fiscal year 1999.

As shown in table 2, in fiscal year 1999, these taxes generated about $35.1
billion in gross receipts for the Fund. About 89 percent of the total
receipts were derived from the excise taxes on motor fuels, and the
remaining receipts came from the truck-related taxes. The total tax
collections ($35.1 billion), however, were subject to certain refunds,
credits, and transfers9 totaling about $1.3 billion; thus, net taxes totaled
about $33.8 billion in fiscal year 1999.

TEA-21 established guaranteed spending levels for certain highway and
transit programs. Prior to TEA-21, these programs competed for budgetary
resources through the annual appropriations process with most other domestic
programs. In a major change to federal budget rules, TEA-21 guaranteed a
minimum level of spending for these programs. New budget categories were
established for highway and transit spending, effectively establishing a
budgetary "firewall" between those programs and other domestic discretionary
spending programs. Of the $217.9 billion authorized for surface
transportation programs over the 6-year life of TEA-21, about $198 billion
is protected by the budgetary firewall--about $162 billion for highway
programs and $36 billion for transit programs.

Under TEA-21, the amount of highway program funds distributed to the states
is tied to the amount of tax receipts credited to the Highway Account of the
Highway Trust Fund. As part of the budgetary firewall, TEA-21 guaranteed
specific levels of funding for highway programs from fiscal year 1999
through fiscal 2003, on the basis of projected receipts of the Highway
Account. Specifically, TEA-21 set a minimum amount to be available for
obligation each year for the federal-aid highway program as well as several
highway safety programs. The federal-aid highway program--administered by
FHWA--is the umbrella term used to describe various separate highway
programs or activities, such as Interstate Maintenance, the National Highway
System, Surface Transportation, and others.

TEA-21 also provided that beginning in fiscal year 2000, this guaranteed
funding level for each fiscal year would be adjusted upward or downward if
the levels of Highway Account receipts increased or decreased from those
projected in TEA-21. The Treasury Department provides information on Highway
Account receipt levels used in calculating the amount of this adjustment,
called Revenue Aligned Budget Authority (RABA). Any higher funding levels
resulting from the RABA calculation are then distributed to the states for
certain highway programs.10 Thus, the RABA adjustment ensures that highway
program funding levels will change as Highway Account receipt levels change.

Highway account receipts are also used as a factor in distributing, or
"apportioning,"11 funds to each state under several major programs included
in the federal-aid highway program. TEA-21 significantly changed some of the
formulas used to apportion funds to the states. The formulas were changed to
more closely reflect factors related to the use of motor fuel and Highway
Account contributions that are attributed to each state's highway users. For
example, estimated Highway Account contributions are significant factors in
the apportionment formulas for three of the major highway programs--the
Interstate Maintenance, Minimum Guarantee, and Surface Transportation
programs. In addition, data on the highway use of diesel fuel--one of the
factors used to estimate contributions to the Highway Account that are
attributed to highway users in each state--is used in the formula for
apportioning funds for the National Highway System program.

Treasury's Process for Allocating Tax Receipts to the Highway Account of the
Highway Trust Fund Is Complex and Error Prone

The Treasury Department uses a very complex and lengthy process for
allocating highway user excise taxes to the Highway Account of the Highway
Trust Fund, in part, because the Department does not obtain data from
taxpayers, as they make semimonthly deposits, on the specific types of
excise taxes for which deposits are made. Taxpayers make semimonthly
deposits of estimated excise taxes but do not provide a breakdown of the
amounts owed, by tax type, until they file their quarterly tax returns.
Treasury uses a process--involving four separate bureaus and offices--for
making initial estimates of highway user tax receipts, allocating these
amounts to the Fund, and subsequently adjusting the amounts allocated after
analyzing payment and tax return data. Our previous reviews of this process
have found that it is cumbersome and prone to error. In addition, the
process may result in yearly reports on Highway Trust Fund receipts that do
not contain timely information, for example, because the amounts
attributable to a particular quarter have not been included in the
certification for that quarter. The Internal Revenue Service (IRS) is
considering changes aimed at improving its revenue allocation and reporting
process for trust funds. Specifically, IRS has studied the taxpayer burden,
potential benefits, and feasibility of obtaining more detailed excise tax
payment data from taxpayers as deposits are made. IRS is also planning to
develop a new database that would allow the agency to record and process
more detailed tax payment data as well as a new revenue allocation method to
utilize the detailed tax payment data if they become available. However,
IRS' plans for the development of the new database and revenue allocation
method are not comprehensive in that they do not address policy changes
needed to fully implement them, such as the actions that IRS would need to
take to obtain the detailed payment data from taxpayers, if it decides to do
so.

to the Fund

Four different bureaus and offices within Treasury--IRS, the Office of Tax
Analysis, the Financial Management Service, and the Bureau of Public
Debt--share responsibility for allocating excise tax receipts12 to the
Highway Account of the Highway Trust Fund. This process is highly complex
and lengthy, in part, because IRS does not obtain data from taxpayers, as
they make semimonthly deposits, on the types of excise taxes for which the
deposits are made.

Payers of highway user excise taxes on motor fuels, truck tires, and trucks
and trailers make semimonthly deposits to cover their estimated excise tax
liabilities, generally through Treasury's Electronic Federal Tax Payment
System.13 Businesses that make these deposits do not specify which types of
excise taxes they are paying with each semimonthly deposit. They are
required to report the amounts owed for each specific excise tax on a
quarterly tax return due 1 to 2 months after the end of each quarter. When
filing the return, the taxpayer is required to make a final payment to make
up the difference between the total of semimonthly deposits and the reported
total amount owed for the quarter, if the latter amount is greater. Payers
of the heavy vehicle use tax generally file returns annually and make
payments directly to IRS. These payments may be made with the annual returns
or in up to four installments.

Excise taxes received are deposited into Treasury's General Fund. Because
data are not available to determine the amounts of these receipts that
represent highway user taxes, Treasury uses estimates of highway user tax
receipts prepared by its Office of Tax Analysis to make initial allocations
from the Treasury General Fund to the Highway Trust Fund each month.14 After
this initial allocation, IRS certifies quarterly the amounts collected for
highway user taxes that should have been allocated to the Fund on the basis
of tax returns and payment data. However, IRS does not certify collections
for each quarter until about 6 months after the quarter ends. IRS needs this
amount of time to allow for the submission and processing of returns as well
as recording, reviewing, and analyzing payment and liability data. Following
certification, Treasury adjusts the amount initially allocated to the
Highway Trust Fund for that quarter. In fiscal years 1998 and 1999, these
quarterly adjustments, which can be downward or upward, ranged from a
decrease of $711 million (which changed the initial allocation by 9 percent)
to an increase of $917 million (which changed the initial allocation by 11
percent).

Treasury's Financial Management Service and Bureau of Public Debt share
responsibility for making the initial allocations to the Highway Trust Fund,
on the basis of the Office of Tax Analysis' estimates, and subsequent
adjustments to these amounts, on the basis of IRS' certifications. The
Financial Management Service prepares vouchers for these allocations and
adjustments. The Bureau of Public Debt, which maintains accounting records
for the Fund, uses these vouchers to record and process the allocations and
adjustments.

Following the close of each fiscal year, the Bureau of Public Debt prepares
a report on the amount of tax receipts that were allocated to the Fund
during that fiscal year. The Bureau issued final versions of its fiscal year
1998 and 1999 reports in January 1999 and January 2000, respectively. The
Department of Transportation and the Office of Management and Budget use the
Highway Account receipts figures in these reports to determine the amounts
of highway program funds to be distributed to the states.

Error Prone

A number of our reports have cited significant weaknesses in Treasury's
process for allocating and reporting on excise tax receipts to various trust
funds, including the Highway Trust Fund. Specifically, under the current
process, IRS does not certify quarterly receipts that should be credited to
the Highway Trust Fund until about 6 months after the end of the quarter,
and this process is cumbersome and prone to error. Since Treasury's reports
on Highway Trust Fund receipts are based on IRS' certified collections,
problems with the timeliness and accuracy of IRS' certification process may
result in misstatements of receipts in the Fund for a given fiscal year.

We have reported instances in which the amounts attributable to a quarter
have not been included in the certification for that quarter.15 For example,
the amount that IRS certified to the Highway Trust Fund for the third
quarter of fiscal year 1998 included approximately $590 million that was
related to the previous quarter. We continued to find such instances in
fiscal year 1999. For example, the amount that IRS certified to the Highway
Trust Fund for the second quarter of fiscal year 1999 included nearly $700
million from previous quarters. There are a number of reasons why the
amounts attributable to a given quarter may not be included in the
certification for that quarter, but rather in a subsequent quarter's
certifications. These reasons include (1) delays in IRS' processing of tax
returns; (2) the late filing of returns by taxpayers; and (3) subsequent
activities, such as amended returns submitted by taxpayers or IRS
examinations resulting in adjustment or reclassification. IRS has recently
implemented new procedures to address the problems of processing delays and
late-filed returns. These procedures include new processes for expediting
the processing and recording of high-dollar tax returns and for identifying
and contacting taxpayers who have not filed by the due date in an attempt to
obtain returns in time to include associated collections in the
certification for the quarter.

Instances in which the amounts from a given quarter are certified as
collections in a subsequent quarter sometimes result in large adjustments to
initial allocations to the Highway Trust Fund. Such instances, particularly
those that occur in the latter part of the fiscal year, could affect the
amounts distributed to the states for that fiscal year because the
certification would not be reflected in Treasury's report on the receipts in
the Fund for that fiscal year. In such cases, these amounts would be
included in Treasury's report on receipts in the Fund for the next fiscal
year.

We have also reported that ineffective controls over the certification
process--such as a lack of written procedures and ineffective supervisory
reviews--have resulted in misstatements of the amounts certified. Although
IRS implemented improvements in its certification process in response to our
previous reports, we found, in reviewing certifications for fiscal year 1999
and the underlying data, that similar errors have continued. These included
(1) taxpayers' errors on excise tax returns that IRS did not identify when
processing the returns, (2) errors made by IRS when processing excise tax
information, and (3) errors made in IRS' preparation of excise tax
certifications.

Improving Its Revenue Allocation and Reporting Process

IRS has a number of efforts aimed at improving Treasury's trust fund revenue
allocation and reporting process, but it has not yet developed a
comprehensive plan with milestones for carrying out all of those efforts.
For example, IRS recently completed a study that included an assessment of
the taxpayer burden, potential benefits, and feasibility of requiring
taxpayers to report a detailed breakdown of all excise taxes--including
highway user taxes--when they make deposits of these taxes. IRS is also
designing a new payment information database that would be able to record
payment data, by specific tax type, if provided by taxpayers with their
deposits. Furthermore, IRS is planning to design a new method for making
initial allocations of tax receipts to the excise-tax-related trust
funds--including the Highway Trust Fund--using the taxpayer-supplied data if
they become available. If taxpayers do not provide the more detailed data,
the new method will allocate their deposits to the trust funds using
estimated tax breakdowns derived from their past returns. While these
changes, if implemented, may improve the accuracy and timeliness of
Treasury's reporting on Highway Trust Fund receipts, IRS has not addressed
all of the policy and process changes needed to fully implement its efforts,
and full implementation is still years away. The effectiveness of these
changes largely depends on IRS' obtaining data from taxpayers, at the time
of deposit, on the associated excise taxes. However, IRS has not yet
determined whether or how to seek this information from taxpayers. IRS is
preparing a plan for developing and implementing the new database and
revenue allocation method. This plan includes milestones for completing the
various steps involved and focuses on changes to be made in IRS' information
systems but does not address all of the policy and process changes needed to
fully implement the new database and revenue allocation method.

or How to Obtain Detailed Excise Tax Data From Taxpayers at Time of Deposit

We have raised concerns about IRS' inability to identify revenue received
for each excise-tax-related trust fund, including the Highway Trust Fund, at
the time taxpayers make deposits of these taxes.16 Obtaining this
information could improve the quality of estimates used to make initial
allocations to the trust funds and could reduce the likelihood of
significant adjustments. In response to our concerns, IRS completed a study
in January 1999 to determine the taxpayer burden, potential benefits, and
feasibility of requiring taxpayers to provide detailed tax breakdowns at the
time of deposit.17 The study found that while the perceived benefits to the
government would be high for excise taxes18, imposing such a requirement at
this time would be potentially burdensome for payers of these taxes,
especially companies with large operations and that the quality of these
data would likely be poor. In particular, large companies said they do not
currently have this information available at the time of deposit and they
would have to collect it from multiple plants in various states to comply
with a new requirement. These companies generally make semimonthly deposits
using estimates of the amounts owed, as of the previous quarter, and report
actual amounts owed for each tax type on their quarterly tax returns.
Although the companies interviewed for the IRS study reported that it would
be difficult to comply with a new reporting requirement at this time, they
indicated that they are improving their information systems and may be able
to provide a detailed tax breakdown at the time of deposit in several years.

The American Petroleum Institute officials we interviewed also expressed
concerns about imposing a new requirement on oil companies to provide the
detailed excise tax data at this time. They explained that oil companies
generally use an estimate--one-sixth of their actual total excise tax
liability in the prior quarter19--to determine the amount of their
semimonthly excise tax deposits. They use this method, which allows them to
fulfill IRS' deposit requirements, because they do not have detailed tax
breakdown information on a semimonthly basis.

Given the concerns about imposing a new reporting requirement at this time,
the IRS study recommended that the agency implement incentives, such as a
small rebate, at this time to encourage taxpayers to start providing a
detailed tax breakdown with their deposits by using Treasury's Electronic
Federal Tax Payment System. This system, which most excise taxpayers use to
pay their taxes, allows the taxpayer to provide a detailed tax breakdown
with their deposits. The study noted that the use of this system has
increased rapidly in the past several years and that an increasing number of
companies could develop the capability to provide this detailed information
through the system if IRS provides some encouragement, such as financial
incentives, for them to do so. In order for IRS to develop such incentives,
the study noted that the agency should undertake a study, as soon as
possible, of state incentive programs for business taxpayers that are
already in place, such as those in Pennsylvania and Virginia.

The study also recommended that IRS do another study in 2002 or 2003 to
determine whether taxpayers' capability to provide these detailed data has
improved and whether IRS should require taxpayers to provide this
information with their deposits. IRS officials told us that the agency
intends to reevaluate taxpayers' ability to provide these detailed tax
breakdown data with their deposits, as recommended, but that the agency
would prefer to use incentives, rather than a requirement, if it decides to
obtain this information. They explained that business and taxpayer
organizations are sensitive to IRS' mandating additional information from
taxpayers and that the agency's goal would be to get most taxpayers to
provide the needed information through an incentive system. They noted that
if IRS could get large oil companies to provide information through an
incentive system, it could get detailed information for a high proportion of
Highway Trust Fund tax receipts. However, these officials stated that it is
not possible to determine, at this point, whether or when incentives could
be implemented because IRS must first determine what incentives could be
offered. 20

IRS officials acknowledged that the agency currently has no formal plans,
including time frames, for (1) carrying out the study's recommendation that
IRS examine, as soon as possible, existing state tax incentive programs to
identify possible incentives that IRS could offer to taxpayers or (2)
determining whether to implement incentives to obtain these detailed data
from taxpayers. They also acknowledged that IRS has no formal plans,
including time frames, for reexamining in several years taxpayers'
capabilities for providing these detailed data and determining whether IRS
should require the data from taxpayers at that time, considering the
potential burden on taxpayers and the potential benefits to the government.

Funds Do Not Address All Changes Needed for Full Implementation

IRS is pursuing two related initiatives--including the development of a
payments information database and a revised revenue allocation method--for
improving its process for allocating excise tax receipts to the Highway
Trust Fund as well as other trust funds. Both the payment information
database and the new revenue allocation method are part of an IRS
modernization project, called the Custodial Accounting Project, which is
aimed at developing a new integrated IRS financial information system.21

The first initiative--which IRS started planning in October 1998 as part of
a broad effort to modernize its information systems--involves establishing a
database that can record and process the detailed tax breakdown data
discussed above, should taxpayers provide that information with their
deposits. According to IRS officials, the database will be capable of
recording such detailed tax data and will serve as a repository for all IRS
payment data.22 Although the Electronic Federal Tax Payment System allows
taxpayers to provide detailed tax breakdowns with their deposits, IRS'
information systems are not currently capable of accepting these data.

The other initiative--which IRS started planning in response to a
recommendation in its January 1999 study--involves revising the method used
to make initial allocations of tax receipts to the trust funds to
incorporate the more detailed tax payment information in the database or in
taxpayers' historical records. This new method would make initial
allocations automatically, shortly after the time of deposit, on the basis
of a detailed tax breakdown provided by the taxpayer, if that information is
available. If that information is not available, this new method would
automatically allocate deposits to the trust funds using estimated tax
breakdowns derived from each taxpayer's past tax returns. IRS is developing
this new revenue allocation method in coordination with the Office of Tax
Analysis, the Financial Management Service, and the Bureau of Public Debt.

IRS officials explained that the goal of this new revenue allocation method
is to improve the accuracy and timeliness of allocations to the trust funds.
For example, Treasury officials expect that the new initial allocation
method will be more precise than the current method because it will be based
on taxpayers' actual deposit amounts23 rather than overall estimates of tax
receipts, by tax type, prepared by Treasury's Office of Tax Analysis.
Furthermore, IRS officials explained that--although IRS would still need to
certify the amounts due to the trust funds and make adjustments to initial
allocations that are based on tax returns--the certification and adjustment
methods used could be improved as a result of the development of the new
revenue allocation method and the integrated financial information system.
They explained that the current certification and adjustment methods, which
are performed quarterly through a combination of manual procedures and
computer programs, could be replaced by an automated method that occurs more
frequently. However, it is not yet clear exactly how these methods would
change, and IRS officials noted that any changes would need to be determined
jointly by responsible Treasury organizations. Furthermore, the changes
being considered do not address all of the control weaknesses identified in
our prior reports, such as taxpayers' errors on excise tax returns that IRS
does not identify when processing the returns and errors made by IRS when
processing excise tax return information.

IRS' plan for the development and implementation of the new database and
revenue allocation method is still in the early stages and does not address
all of the factors that will need to be considered to implement these two
initiatives. IRS officials told us that, on the basis of tentative
milestones in this plan, they expect to start implementing the new database
and allocation method in 2004. Although the plan contains specific tasks and
associated time frames, IRS officials acknowledged that the plan focuses on
the technical changes needed in IRS' information systems and does not
address all of the policy and process changes needed to fully implement the
new database and revenue allocation method. In particular, the plan does not
include time frames for examining, in the near-term, possible incentives for
encouraging taxpayers to provide these data and determining whether to
implement incentives. Also, the plan does not include time frames for
reexamining, in several years, taxpayers' capabilities for providing these
detailed data, and determining whether IRS should require the data from
taxpayers at that time. Finally, the plan does not address the need for, or
include time frames for, determining what changes may be needed in
Treasury's methods for certifying the amounts due to the trust funds and
adjusting initial allocations on the basis of these certifications.

Reliability of FHWA's Attribution Process Has Not Been Demonstrated

While Treasury reports on the overall amounts of highway user taxes credited
to the Highway Account of the Fund, FHWA must estimate the portion of the
overall amounts that are attributable to each state--information that is
ultimately used in distributing funds to the states through the
apportionment process used in several major highway programs. FHWA estimates
the amounts attributable to each state through its "attribution" process.
The underlying premise of this process is that federal highway user tax
payments to the Highway Account are received from highway users in each
state in proportion to the highway use of motor fuel in that state relative
to total highway use of motor fuel in all states. Under this process, FHWA
goes through a two-fold process to (1) collect and analyze state motor fuel
information to develop a uniform database across all states and determine
the portion of motor fuel that was used on highways in each state and (2)
use the resulting analyses to estimate the amount of Highway Account
contributions that are attributable to highway users in each state. The
steps related to the state motor fuel analysis used in the attribution
process are complicated, utilizing nearly 200 formulas for analyzing the
data. In designing these formulas, FHWA's intent was to analyze the state
motor fuel data in a manner that ensures equitable treatment of all states.
The large number of formulas reflect the wide variation in the ways that
states (1) tax motor fuels and (2) define, capture, and report certain data
items to FHWA.

Given the complexity of the state motor fuel analysis used in the
attribution process and the fact that it has been developed over decades
without any thorough, systematic, or independent review of its accuracy,
reliability, or consistency, we have concerns about the overall reliability
of the process. Furthermore, FHWA has identified some specific weaknesses in
the motor fuel data and attribution methodology, which it is addressing
through its participation in a multiagency committee aimed at improving the
motor fuel reporting process. In addition, IRS is developing a system that
will provide an alternate source of information on motor fuel that may be
useful as a tool in validating the state motor fuel data currently used in
FHWA's attribution process.

Highway User Tax Data

FHWA must estimate the contributions to the Highway Account that are
attributable to each state's highway users because state-level data on
federal highway tax receipts are not available. Most of the federal highway
user taxes are not collected directly from the consumer. For example, all of
the federal taxes on motor fuels are imposed when the fuel is first removed
from bulk storage. Thus, the taxes are paid by businesses--typically the
major oil companies--rather than by consumers at retail fuel pumps.24 As a
result, most of the federal motor fuel taxes come from the handful of states
where the businesses are headquartered. Similarly, the tax on truck tires is
collected from the manufacturer and, as a result, the truck tire tax
receipts are concentrated in Ohio, the home of the U.S. tire industry.

FHWA estimates state-level contributions to the Fund through its
"attribution" process, which is basically a process of determining each
state's share of highway motor fuel use from state tax data and using that
information to estimate the amount of contributions to the Highway Account
attributable to each state's highway users. The relative percentage of
contributions attributed to each state is then used to calculate a portion
of the amount of federal-aid highway dollars distributed to each state.

FHWA and its predecessor agencies have been collecting state data on motor
fuels since 1919 to determine the highway use of motor fuel and to identify
how states tax motor fuel. FHWA publishes statistics on motor fuel usage and
other highway topics in an annual report.25 FHWA first began using that
information to estimate Fund payments during the 1970s as part of an
analysis of issues related to so-called "donor/donee" states26. Since fiscal
year 1984, FHWA has used state-reported data on motor fuel, supplemented
with information from other sources, to attribute federal tax payments to
the Fund to highway users in each state for the purpose of apportioning
highway program funds. FHWA also uses statistical methods to estimate
missing or unreported state data used in the state motor fuel data analysis
portion of the attribution process. FHWA informed the states of its proposed
attribution methodology in 1983 and solicited their comments on it. FHWA
finalized the methodology and published a description of it in a June 1985
Federal Register notice. According to FHWA officials, the methodology has
generally been supported by the states since that time.

In the attribution process, FHWA calculates ratios of the highway use of
motor fuel in each state to the highway use of motor fuel in all states. As
shown in table 3, FHWA prepares a separate attribution analysis for each
type of motor fuel tax (gasoline, gasohol, and diesel and special fuels),
and it attributes the truck-related taxes (tires, truck and trailer sales,
and heavy vehicle use) to the states using data on the highway use of diesel
and special fuels.

             Type of tax                        Attribution factor
                                       Ratio of highway use of gasoline in
 Gasoline                              each state to highway use of
                                       gasoline in all states.
                                       Ratio of highway use of gasohol in
 Gasohol                               each state to highway use of gasohol
                                       in all states.
                                       Ratio of highway use of diesel and
 Diesel and special fuels              special fuels in each state to
                                       highway use of diesel and special
                                       fuels in all states.

 Truck-related taxes (truck tires,     Ratio of highway use of diesel and
 truck and trailer sales, heavy        special fuels in each state to
 vehicle use)                          highway use of diesel and special
                                       fuels in all states.

Source: FHWA.

The factors shown in table 3 are used by FHWA to attribute Highway Account
receipts to the states, and that information is incorporated, in turn, in
the apportionment formulas used for distributing certain federal-aid highway
program funds back to the states.

Data to Estimate State-Level Motor Fuel Usage and Contributions to the
Highway Account

In its attribution process, FHWA first performs complicated and lengthy
steps to collect information from the states on the gallons of motor fuels
consumed and to analyze that information in a consistent format for the
purpose of determining each state's share of the highway use of motor fuel.
FHWA's intent is to analyze the state motor fuel data in a manner that
ensures equitable treatment of all states. Once this is done, FHWA uses the
results of the analysis to estimate the relative contributions to the Fund
from each state's highway users. The following is a summary of the key steps
involved in the attribution process and their associated time frames for the
process culminating in the apportionment of funds in fiscal year 2000.

The attribution process begins with the collecting and reporting of
information on the use of motor fuels by each state. To do this, each state
submits a monthly report on the volume of motor fuels used within that state
(FHWA Form 551M, entitled "Monthly Motor Fuel Consumption"). The report is
designed to capture and report many details about the motor fuel, including
the (1) type of fuel used (e.g., gasoline, gasohol, and diesel and other
special fuels27), (2) tax status of the fuel (e.g., amounts that are fully
taxed at the state's prevailing rate, taxed at some other rate, or subject
to exemptions or refunds), (3) reason for tax exemption (e.g., loss or
shrinkage of fuel, federal government use), and (4) reason for tax refund
(e.g., specific classes of nonhighway use of fuel, such as agriculture,
aviation, industrial and commercial, construction, and marine).

States have 60 days from the end of the month to submit the Form 551M for
that month to FHWA. While each state's transportation department generally
takes the lead in providing FHWA with the motor fuel data, the source data
come from the state's revenue department or agency. FHWA provides the states
with detailed instructions for reporting the motor fuel data in chapter 2 of
its publication, A Guide to Reporting Highway Statistics.

Analyzes the Data to Identify Highway Usage of Motor Fuel and to Reconcile
Inconsistent Data

After the states have submitted all of the monthly reports for a given
calendar year, FHWA prepares a summary version of Form 551M for each state;
the summary version contains data for the preceding 12-month calendar
period. FHWA then performs a quality check that involves (1) estimating
state tax receipts by using the summary data on motor fuel gallons and
information about each state's motor fuel tax rates and (2) comparing the
estimated receipts with actual state tax receipts reported to FHWA on
another form.28

FHWA must analyze the monthly motor fuel data submitted by the states to
determine the portion of the motor fuel that was used on the highways and to
develop a uniform database across all states. State data on motor fuel usage
are derived primarily from information from their tax agency, and there is
significant variation among states in the ways that various motor fuels are
taxed and the ways that certain data items are defined, captured, and
reported to FHWA. Generally, state motor fuel taxes are levied on road
users, and exemptions to the tax or refunds of the taxes paid are given for
nonhighway use of the motor fuel. However, not all states offer the same
exemptions or refunds for all classes of nonhighway use, and not all refunds
are actually claimed. Thus, the net volume of motor fuels taxed in a state
is not always the same as the volume consumed on the highways.

Typically, state revenue departments have data on motor fuels that are
exempted, refunded, or taxed at multiple rates. From this information, FHWA
attempts to identify the use of the fuel and thus its place in the
attribution process. To do this, FHWA maintains records on about 90
subcategories in which these data are reported by the states and uses this
information to estimate highway versus nonhighway use of motor fuels and
reconcile any data that were inconsistently reported by the states. FHWA's
methodology for analyzing the motor fuel data submitted by the states has
evolved over several decades and is very complex--involving nearly 200
formulas that analyze the subcategories of motor fuel data.

In many cases, sufficient data do not exist at the state level to determine
the highway versus nonhighway use of the various motor fuels and, in other
cases, FHWA considers some of the motor fuel data submitted by the states to
be inconsistently developed. As a result, FHWA estimates some components of
motor fuel usage--including the government use of gasoline, nonhighway use
of gasoline, and highway use of gasohol--using statistical models and
federal data sources. When FHWA's estimates and the state-submitted data are
both available, FHWA makes a judgment about the relative quality of the data
and selects one or the other or some combination of the two data sets. The
end result of FHWA's analyses is a summary of each state's relative share
(and actual gallons) of the highway use of motor fuel, by fuel type.

FHWA completes its detailed analyses and compiles the first draft of the
summary about 8 months after the end of the calendar year--about 90 days
after all of the states have submitted their motor fuel usage and receipts
data. Thus, FHWA completed its initial analysis of the data on motor fuel
usage for calendar year 1997 during the summer of 1998. According to FHWA
officials, the results of this analysis are then reviewed and modified, in
consultation with the states, where necessary, over the following year. For
example, FHWA performs trend analyses of the data for each state to identify
any significant or unusual changes from prior years' data and, if necessary,
visits states to clarify any discrepancies and adjust the data results. FHWA
finalizes its analysis of each state's relative highway use of motor fuels
about 1 year later. Thus, the agency's analysis of motor fuel data for 1997
was finalized in August 1999. Given the time required for obtaining and
analyzing the state data, the data on motor fuel usage for 1997 were the
most recent data available for use in the apportionment for fiscal year
2000, effective on October 1, 1999.

Contributions to the Highway Account of the Highway Trust Fund That Are
Attributable to Each State

To estimate state-level contributions to the Fund's Highway Account, FHWA
computes each state's relative share of fuel gallons used on the highways
and multiplies this computation by the amount of highway user tax receipts
allocated to the account (as determined by Treasury's revenue reports). To
do this, FHWA first calculates three main data elements on motor fuel
gallons used on highways for each state: (1) total motor fuel gallons, (2)
total diesel and special fuels gallons, and (3) total gasoline and gasohol
gallons. FHWA then computes each state's relative share of highway gallons
by calculating the ratio of the gallons used in the state to the total
gallons used in all states.

In general, the ratio for a given fuel type is used to attribute tax
receipts for that particular fuel type (e.g., the ratio for gasoline and
gasohol use is used to attribute gasoline and gasohol tax payments). The one
exception pertains to the truck-related taxes (truck and trailer sales,
truck tires, and heavy vehicle use), which are attributed to states on the
basis of highway use of diesel and special fuels--a factor that FHWA
considers to be the best available proxy measure for truck use in each
state.

For each of the fuel types, the resulting percent share of gallons used in
each state is then multiplied by the amount of highway user tax receipts
allocated to the Fund's Highway Account, as reported by the Treasury
Department. FHWA generally receives reports on Highway Account receipts from
Treasury about 9 to 10 months prior to the effective date of the
apportionment of funds to the states. For the fiscal year 2000 apportionment
cycle, for example, FHWA used a report on Highway Account receipts for
fiscal year 1998 that it received from the Treasury Department in January
1999. The end result of this analysis is a table showing estimated
contributions to the Highway Account of the Fund attributed to each state
for each of the fuel types.

FHWA finalizes its estimate of state-level data on contributions to the
Highway Account of the Fund about a month before the effective date of the
apportionment. For example, it prepared the estimates for the apportionment
for fiscal year 2000--which became effective on October 1, 1999--in August
1999, using the state data on motor fuel usage for calendar year 1997 and
the Treasury report on Highway Account receipts for fiscal year 1998.

In total, FHWA's overall attribution process takes about 19 months from the
time that states submit their final reports on motor fuel usage for a given
calendar year to the time that FHWA apportions funds to the states on the
basis of its state-level analysis of estimated contributions to the Fund.
The lengthy time period associated with the process is largely due to the
amount of time (about 17 months29) that FHWA, in consultation with the
states, devotes to analyzing the state motor fuel data. For the fiscal year
2000 apportionment cycle, for example, FHWA received the state motor fuel
data in March 1998, obtained Treasury's revenue reports in January 1999, and
apportioned the highway program funds to the states in October 1999. Table 4
contains the results of FHWA's attribution process for the fiscal year 2000
apportionment cycle.

                                            Dollars in thousands

                1997 highway motor fuel          1998 Highway Account receipts attributable to states
                 usage (percent share)

                                 Diesel                                                            Percentage
                                  and                                                               of total
     State                       other                         Diesel and                            Highway
                                special                          other    Truck-related              Account
               Gasoline Gasohol  fuels     Gasoline   Gasohol   special      taxesa    Total taxes  receipts
                                                                 fuels

 Alabama       2.11     0.27    2.28      $372,290    $2,508   $143,192   $75,154      $593,144    2.10
 Alaska        0.16     0.60    0.17      28,938      5,602    10,497     5,509        50,546      0.18
 Arizona       1.62     2.65    2.04      285,473     24,885   128,512    67,450       506,320     1.80
 Arkansas      1.28     -       1.84      225,559     -        116,034    60,901       402,494     1.43
 California    11.09    16.37   7.92      1,958,176   153,742  498,637    261,711      2,872,266   10.19
 Colorado      0.97     5.93    1.22      170,854     55,700   76,695     40,254       343,503     1.22
 Connecticut   1.21     0.23    0.84      213,058     2,194    53,144     27,893       296,289     1.05
 Delaware      0.33     -       0.22      58,294      -        13,786     7,235        79,315      0.28
 District of
 Columbia      0.16     -       0.07      27,548      -        4,704      2,470        34,725      0.12
 Florida       6.22     0.09    3.92      1,098,028   866      246,515    129,385      1,474,794   5.23
 Georgia       3.94     -       4.11      695,181     -        258,727    135,793      1,089,701   3.87
 Hawaii        0.36     -       0.10      63,116      -        6,249      3,280        72,645      0.26
 Idaho         0.55     -       0.76      96,686      -        47,940     25,161       169,787     0.60
 Illinois      2.49     12.25   3.73      439,445     115,024  234,721    123,193      912,383     3.24
 Indiana       2.08     4.08    3.34      367,562     38,297   210,102    110,272      726,233     2.58
 Iowa          0.75     3.79    1.59      132,997     35,556   99,835     52,398       320,786     1.14
 Kansas        1.14     0.18    1.31      200,680     1,714    82,673     43,391       328,458     1.17
 Kentucky      1.85     0.43    2.29      327,269     4,018    144,259    75,714       551,260     1.96
 Louisiana     1.78     0.05    1.82      314,806     474      114,741    60,223       490,244     1.74
 Maine         0.62     -       0.48      109,152     -        30,224     15,864       155,240     0.55
 Maryland      2.05     0.3     1.44      362,064     2,784    90,718     47,613       503,179     1.78
 Massachu-setts2.38     -       1.22      419,402     -        76,558     40,181       536,141     1.90
 Michigan      4.07     1.76    2.82      718,977     16,492   177,277    93,044       1,005,790   3.57
 Minnesota     0.33     13.61   1.73      58,241      127,853  109,178    57,303       352,575     1.25
 Mississippi   1.36     -       1.5       240,142     -        94,342     49,515       383,999     1.36
 Missouri      2.63     0.45    3.04      463,773     4,209    191,323    100,416      759,721     2.69
 Montana       0.42     -       0.61      74,864      -        38,135     20,015       133,014     0.47
 Nebraska      0.55     1.28    1.10      96,911      12,048   69,081     36,258       214,298     0.76
 Nevada        0.60     1.47    0.77      105,465     13,842   48,479     25,444       193,230     0.69
 New Hampshire 0.57     -       0.40      100,253     -        25,259     13,258       138,770     0.49
 New Jersey    3.31     1.30    2.35      584,691     12,190   148,121    77,742       822,744     2.92
 New Mexico    0.64     1.31    1.18      112,655     12,270   74,421     39,059       238,405     0.85
 New York      4.83     1.91    3.13      853,127     17,947   197,153    103,476      1,171,703   4.16
 North Carolina3.19     2.28    2.93      562,521     21,369   184,523    96,848       865,261     3.07
 North Dakota  0.27     0.32    0.48      47,991      2,996    30,098     15,797       96,882      0.34
 Ohio          2.98     9.87    4.71      526,217     92,663   296,654    155,699      1,071,233   3.80
 Oklahoma      1.62     -       1.94      286,445     -        122,233    64,154       472,832     1.68
 Oregon        1.29     -       1.38      228,199     -        86,956     45,639       360,794     1.28
 Pennsyl-vania 3.85     4.57    4.28      679,942     42,897   269,324    141,355      1,133,518   4.02
 Rhode Island  0.36     -       0.15      63,298      -        9,470      4,971        77,739      0.28
 South Carolina1.92     -       1.8       338,738     -        113,324    59,478       511,540     1.81
 South Dakota  0.21     1.07    0.49      37,235      10,065   30,738     16,132       94,170      0.33
 Tennessee     2.55     0.02    2.68      450,456     176      168,842    88,617       708,091     2.51
 Texas         8.71     2.79    8.04      1,537,578   26,229   505,829    265,486      2,335,122   8.28
 Utah          0.85     -       1.02      150,225     -        64,025     33,604       247,854     0.88
 Vermont       0.29     -       0.27      52,057      -        16,896     8,868        77,821      0.28
 Virginia      2.85     2.30    2.88      502,635     21,566   181,540    95,282       801,023     2.84
 Washington    2.08     2.34    1.62      367,961     21,968   101,858    53,460       545,247     1.93
 West Virginia 0.77     0.01    0.87      135,323     115      54,933     28,832       219,203     0.78
 Wisconsin     1.50     4.14    2.19      265,232     38,904   137,821    72,335       514,292     1.82
 Wyoming       0.28     0.01    0.90      49,162      82       56,452     29,629       135,325     0.48
 Total         100.00   100.00  100.00    $17,656,892 $939,245 $6,292,751 $3,302,761   $28,191,649 100.00

aThe truck-related taxes are attributed to each state's highway users on the
basis of the highway use of diesel and other special fuels in each state.

Source: GAO's presentation of data from FHWA.

Demonstrated

On the basis of our review and recent work by FHWA, we have a number of
concerns regarding both the motor fuel data submitted by the states and
FHWA's methodology for analyzing those data for use in its attribution
process. Some of these concerns involve the overall reliability of the state
data and FHWA's methodology, and FHWA's efforts to improve its attribution
process only partly address these concerns. Other concerns relate to
specific weaknesses in certain aspects of the state data and FHWA's
methodology, and FHWA has formed a committee that is helping the agency take
actions to address these concerns.

of the Attribution Data and Methodology

The state motor fuel data submitted to FHWA have not been thoroughly and
independently reviewed to test and verify their accuracy, reliability, and
consistency across all states, which raises questions about the overall
reliability of the state data and FHWA's attribution methodology. Although
the original source data on motor fuel tax receipts prepared by the state
revenue agencies may be audited by the states, according to FHWA officials,
FHWA does not trace the individual data elements submitted on Form 551M to
their supporting documentation to verify the underlying support. Since the
source data provided by the state revenue agencies must be reformatted and
analyzed by officials in the state transportation agencies and by FHWA,
there are significant opportunities for errors to occur. Given the wide
variation in state tax methods and data, FHWA would probably have to review
and verify each state's summary tax records. While the potential magnitude
of any errors that would be identified as a result of such detailed reviews
is unknown, conducting such reviews would provide an indication of the scope
of any inaccuracies or assurance that the motor fuel data submitted by the
states are sound. At the same time, we recognize that since the costs
involved with conducting such reviews may be high, FHWA would have to
consider various options for doing them. One option might be for FHWA to
conduct the reviews on a rotational basis over a period of years.

Another concern about the reliability of FHWA's attribution process pertains
to the complexity of FHWA's methodology, the way it has been developed, and
the fact that it has never been fully documented or independently reviewed.
Because of the wide variation in the ways that states tax motor fuels--and
the ways that certain data items are defined, captured, and reported to
FHWA--FHWA's methodology for analyzing the motor fuel data submitted by the
states is extremely complicated. The methodology involves a spreadsheet
program that has nearly 200 separate formulas needed to analyze the state
data in a manner that is consistent across all states. Furthermore, the
methodology has been repeatedly adjusted over several decades in response to
changing state tax laws and federal program requirements. Moreover, the
methodology has never been fully documented to explain the content and
rationale for the formulas it contains and to specify the procedures for
carrying it out. The methodology now contains some factors that are not used
because FHWA officials consider them unreliable. Finally, the attribution
methodology has never been subject to a comprehensive, independent review to
validate its reliability and provide assurance that the information
resulting from the methodology is accurate.

In addition, FHWA considers some of the motor fuel data submitted by the
states to be inconsistent with that provided by other states--such as data
on the gallons consumed by various industries for nonhighway purposes--so it
estimates those gallons using other federal data sources. FHWA officials
told us that they use their judgment in deciding which figure to use in the
attribution process--sometimes they choose to use the state-submitted data,
sometimes they choose to use FHWA's own estimates, and sometimes they use a
figure that combines both data sources. According to FHWA officials, once
FHWA decides how it will handle a particular state's data, it makes note of
that decision and follows it consistently from year to year until it makes a
judgment that further changes are warranted.

Finally, all of the responsibility for collecting, analyzing, and evaluating
the motor fuel data--which are submitted by 50 states and the District of
Columbia--rests with only two officials in FHWA headquarters. Each state
submits hundreds of data elements for review by these two officials, thus
increasing the risk that errors could occur during the attribution process.
Moreover, since the attribution methodology has never been clearly
documented, FHWA management does not have an adequate basis for providing
oversight of the attribution process or delegating the responsibility for
this process to other individuals in the event that the two officials depart
or are otherwise no longer available to handle this responsibility.

FHWA's efforts to validate the state motor fuel data and its own methodology
only partially address questions about the overall reliability of the
agency's attribution process. First, FHWA performs trend analyses to compare
each state's data submission with comparable data submitted in earlier years
and identify any major discrepancies. It then works with the state to
determine the cause of the discrepancy and make any needed adjustments. This
is a helpful quality control technique to ensure consistency from year to
year, but it does not verify the accuracy of the underlying data submitted
by the states.

Second, FHWA has a policy to conduct motor fuel reviews in each state every
3 years. These reviews are designed to identify and review the overall
methodology used by the states to develop the key motor fuel figures
reported to FHWA. Officials in FHWA's division offices in the states conduct
these reviews, using guidance from FHWA headquarters. FHWA renewed its
efforts to conduct these reviews in June 1997, and as a result, about 83
percent of the states have undergone a motor fuel review during the past 3
years. We view this as a positive effort to improve state reporting, but it
does not verify the accuracy and ensure the reliability of the state data
submitted to FHWA. To accomplish a detailed verification, FHWA would need to
trace the figures on the Form 551M submitted by the states to the supporting
documentation prepared by the state transportation and revenue agencies, and
the agency would have to verify the accuracy of the underlying support.

Address Specific Weaknesses in the Attribution Data and Methodology

FHWA has identified a number of weaknesses in the state motor fuel analysis
portion of its attribution process, which the agency is discussing with the
Motor Fuel Reporting Information Committee. This committee--which includes
representatives from FHWA, state revenue and transportation agencies, the
American Association of State Highway and Transportation Officials, and the
Federation of Tax Administrators--was established by FHWA in September 1998
to address issues and areas for improving the reporting system for state
motor fuel data and FHWA's attribution methodology.

FHWA has identified and is addressing a number of specific weaknesses in the
data and methodology used in the attribution process, including problems
with the reporting on government entities' use of fuel, fuel loss
allowances, and gasohol usage. FHWA officials told us that the impact of
these problems is relatively small--representing only 2 to 3 percent of the
total state motor fuel data. Nonetheless, FHWA believes the problems are
significant enough to require discussion with the committee and action by
the agency. Although the scope of FHWA's work with this committee was
initially focused on these limited areas, it has evolved to include broader
concerns about the attribution process. For example, FHWA also plans to
issue a paper describing its attribution methodology to better inform state
officials about how the motor fuel data are used to estimate contributions
to the Highway Account of the Fund that are attributable to each state.
Furthermore, FHWA has conducted significant outreach efforts to state
officials through this committee by holding workshops in Washington, D.C.;
Atlanta, Georgia; Philadelphia, Pennsylvania; and Denver, Colorado to
discuss issues related to state motor fuel reporting and other aspects of
the attribution process.

Three of the weaknesses--involving problems with FHWA's treatment of motor
fuel usage by government versus private and commercial vehicles--are
relatively significant. The first weakness is that the state data on diesel
fuel usage may not be accurate because states have a problem identifying and
quantifying the amount of diesel fuel that is used by private and commercial
entities versus government entities. FHWA guidance directs states to submit
data on only the private and commercial use of diesel fuel because many
states are unable to provide information on government use of such fuel.
However, many states cannot split out the government uses from the private
and commercial uses, so they incorrectly include government use in their
data submission for diesel fuel. This, in turn, affects the accuracy of
FHWA's attribution of contributions to the Highway Account of the Fund
associated with the use of diesel fuel--a factor used to apportion funds for
the Interstate Maintenance and National Highway System programs.

The second weakness involves an inconsistency in the way FHWA attributes
government entities' use of gasoline versus diesel fuel to the states for
the purpose of estimating contributions to the Highway Account of the Fund.
FHWA's attribution methodology includes government use of gasoline fuel but
excludes government use of diesel fuel in attributing motor fuel usage to
each state. FHWA decided to exclude government use of diesel from its motor
fuel data collection because many states were unable to provide this
information. FHWA officials stated that this inconsistency in FHWA's
attribution methodology has caused considerable confusion in state
reporting; for example, many states include the wrong numbers on the Form
551M or provide unnecessary and confusing information to FHWA.

The third weakness is that FHWA's use of diesel fuel as a factor in
apportioning funds for one of the major highway programs--the National
Highway System program--is not consistent with TEA-21's requirements. TEA-21
identifies total highway use of diesel fuel as a factor for apportioning the
National Highway System funds. In practice, however, FHWA excludes
government use of diesel fuel and calculates the apportionment factor using
only private and commercial vehicle use of diesel fuel. Current FHWA
reporting procedures instruct states to exclude government use of diesel
fuel--that is, use by federal, state, and local government entities--in
their reports to FHWA. Again, FHWA decided to exclude government use of
diesel fuel from its motor fuel data collection because many states were
unable to provide this information. FHWA is exploring alternative solutions
to this problem, which include requiring states to develop the information
or developing estimates at FHWA by using modeling techniques.

FHWA is exploring ways of addressing all of the weaknesses it has
identified, and it will develop an action plan--in consultation with the
committee--with a schedule for implementing the corrective actions. FHWA
plans to publish the proposed implementation schedule in a Federal Register
notice, solicit comments, and finalize the schedule in December 2000. During
2001 and beyond, FHWA plans to revise its guidance to the states on
compiling motor fuel data, to conduct training sessions on motor fuel data
requirements for officials in state transportation and revenue agencies and
FHWA division offices, and to take other actions that are included in its
plan.

Attribution Process

IRS--in partnership with the states, industry representatives, and FHWA--is
developing a fuel-tracking system that will monitor the movement of motor
fuel through the distribution chain to determine if the proper federal tax
is being paid. This system--known as the Excise Files Information Retrieval
System (ExFIRS)--will include data on the gallons of motor fuel, by type,
that are sold to retailers and wholesalers in each state. The ExFIRS data
project is being jointly administered by IRS and the Department of
Transportation through a Memorandum of Understanding.

According to IRS officials responsible for ExFIRS, the data will provide
details that could be used to analyze highway versus nonhighway use of motor
fuels in each state. Specifically, ExFIRS will provide data on the final
destination state for motor fuel when it is removed from bulk storage tanks
at the terminal.30 Thus, the ExFIRS data could be used as a tool in
validating the state motor fuel data currently used in FHWA's attribution
process.

One potential limitation with using the ExFIRS data in the attribution
process--that would need to be evaluated by FHWA and IRS--is that the data
on the destination state come from the terminal owner, as reported by the
truckers or rail carriers, and are not independently verified by IRS. This
raises the possibility that, in some cases, the truckers or rail carriers
may divert the gasoline to a state other than the "destination state" that
they reported to the terminal operator in an attempt to evade the higher
motor fuel taxes imposed by the true destination state where the fuel was
actually sold. However, such diversions would also affect the state-reported
data on motor fuels currently used by FHWA in its attribution process.
Moreover, IRS is developing a compliance technique that may reduce the
likelihood that this type of evasion will occur. This technique, known as
"fuel fingerprinting", is designed to identify the terminal source for fuel
that is sampled from a retail or wholesale establishment. According to the
IRS official responsible for ExFIRS, this technique, when fully developed,
will probably be useful for determining whether the fuel sampled came from
an in-state terminal source or was diverted across state lines.

IRS is developing the requirements and procedures for ExFIRS, and, according
to the officials responsible for the system, the agency can begin
implementing it by October 2000. Once the system is implemented, taxpayers
will start reporting data to IRS. According to those officials, if the
agency meets all of its current milestones for implementing ExFIRS, IRS will
be able to report on a full year's worth of motor fuel data (i.e., fiscal
year 2001) by November 2001.

Problems With Treasury's and FHWA's Processes Can Affect Fund Distributions
to the States

Problems with the reliability of the processes used by the Treasury
Department and FHWA for estimating highway user tax information can affect
(1) the calculation of any changes to the overall amounts of highway program
funds that go to the states each year, as authorized by TEA-21, and (2) the
amounts that are apportioned to individual states under several major
programs. FHWA uses Treasury's revenue reports and the results of its
attribution process to distribute highway program funds to the states in two
ways. First, FHWA distributes any additional funds (or fund reductions)
called Revenue Aligned Budget Authority (RABA). The amounts of these
additions to or reductions in funds are determined by analyzing changes in
receipts in the Fund on the basis of information provided by the Treasury
Department. Second, FHWA apportions funds to the states for several major
highway programs by using formulas that rely to a large extent on the
state-level estimates of motor fuel usage and contributions to the Highway
Account of the Fund developed during the attribution process. The amount
distributed to the states on the basis of information developed by the
Treasury Department and FHWA is large--totaling about $13 billion in fiscal
year 2000 alone--so the potential impact of problems with the accuracy and
timeliness of that information are significant. Although we have concerns
about the reliability of the estimates developed by Treasury and FHWA, there
is no way of knowing the extent of over- or under-payments, if any, to
individual states, given the information currently available from the two
agencies.

Calculation

The yearly adjustment to the highway program's funding levels authorized in
TEA-21, called RABA, is based on available Treasury Department information
on Highway Account receipts. Problems with the accuracy and timeliness of
the information in Treasury's reports on Highway Account receipts can thus
affect the amounts distributed to the states under the RABA provision. As
discussed in appendix II, Treasury's reporting on these receipts has
contained errors and is not timely. In response to concerns raised by the
Transportation Department, Treasury revised its fiscal year 1998 and 1999
revenue reports to incorporate the most recent information available on
Highway Account receipts. Nonetheless, this action does not fully address
our concerns about the accuracy and timeliness of Treasury's reporting on
those receipts, which could potentially affect the RABA calculations.

The RABA adjustment is based on a comparison between currently available
information on Highway Account receipts and projections of those receipts
contained in TEA-21. TEA-21 used projections of Highway Account receipts to
develop guaranteed highway funding levels for fiscal years 1999 through
2003. Beginning in fiscal year 2000, these guaranteed levels were to be
adjusted upward or downward each year on the basis of actual Highway Account
receipts and new projections of these receipts.31

The RABA adjustment to the funding levels authorized in TEA-21 is based on
actual receipts from 2 years prior to the fiscal year, as reported by
Treasury, plus revised Treasury receipt projections for the fiscal year in
question. For example, for fiscal year 2000, TEA-21 requires that this
adjustment be calculated by comparing (1) actual Highway Account receipts
for fiscal year 1998 with the TEA-21 projection of these receipts and (2)
revised projections of Highway Account receipts for fiscal year 2000 with
the TEA-21 projection of these receipts. The sum of these differences
becomes the RABA funding level. Table 5 shows how this adjustment was
calculated for fiscal years 2000 and 2001. As shown, for fiscal year 2000,
the RABA adjustment, included in the President's budget, added $1.5 billion
to the $28.1 billion estimate of Highway Account receipts in TEA-21. For
fiscal year 2001, the adjustment added $3.1 billion to the $28.5 billion
estimate of Highway Account receipts. FHWA apportions any additional RABA
funds to the states on October 15 of each fiscal year--about 2 weeks after
apportioning the amount of highway program funds for the fiscal year that
was authorized in TEA-21. This RABA adjustment ensures, for the first time,
that highway program funding levels will change as Highway Account receipt
levels change.

                            Dollars in millions

                    Comparison of
                   actual Highway
                                      Comparison of revised
 Fiscal year that Account receipts    projection of Highway
  funds are made    from 2 years      Account receipts for  Revenue Aligned
    available           prior         the fiscal year with      Budget
                    to the fiscal                              Authority
                  year with TEA-21    TEA-21 projection of    adjustment
                     projection          these receipts
                  of these receipts
                  FY 1998 actual      Revised projection of
                  Highway             FY 2000
                  Account receipts    Highway Account
                  $23,135             receipts $28,551

 FY 2000          TEA-21 projection   TEA-21 projection of
                  of FY1998           FY 2000
                  Highway Account     Highway Account       $971 + $485 =
                  receipts 22,164     receipts 28,066       $1,456

                  Difference $971     Difference $485
                  FY 1999 actual      Revised projection of
                  Highway             FY 2001
                  Account receipts    Highway Account
                  $33,815             receipts $30,368

 FY 2001          TEA-21 projection   TEA-21 projection of
                  of FY1999           FY 2001
                  Highway Account     Highway Account       $1,196 + $1,862
                  receipts 32,619     receipts 28,506       = $3,058

                  Difference $1,196   Difference $1,862

Legend

FY = fiscal year

Note: Actual Highway Account receipts are net receipts (excluding fines and
penalties) after the deduction of transfers and refunds.

Source: Department of Transportation.

To include the amount of the RABA adjustment for the next fiscal year in the
President's budget for that fiscal year, the Office of Management and Budget
and the Office of the Secretary in the Department of Transportation rely on
information on Highway Account receipts supplied by the Treasury Department.
The Bureau of Public Debt provides them with the prior fiscal year's Highway
Account receipts in a report on Highway Trust Fund receipts for the fiscal
year, and the Office of Tax Analysis provides them with a projection of
Highway Account receipts for the next fiscal year.

Problems with the accuracy and timeliness of information in Treasury's
reports on Highway Account receipts can affect the amounts distributed to
the states, since these reports are used in calculating the RABA adjustment.
In developing the RABA calculations for fiscal years 2000 and 2001, the
Department of Transportation raised concerns with the Treasury Department
regarding its reporting on Highway Account receipts. In its yearly reports
on Highway Trust Fund receipts, the Bureau of Public Debt has generally used
estimates of these Highway Account receipts, rather than certified figures,
for the third and fourth quarters of the fiscal year because of the 6-month
lag in IRS' certification of quarterly collections. The Bureau has generally
finalized its reports on Highway Trust Fund receipts for each fiscal year by
December--about the time that IRS issues its third-quarter certification.
Transportation requested that the Bureau incorporate the third-quarter
certification in its fiscal year 1998 and 1999 reports, so that the RABA
calculation could be based on the most recently available information. In
the case of the report for fiscal year 1998, Transportation was particularly
concerned because the third-quarter certified collection amount for that
fiscal year was unusually large. As noted in appendix II, this certified
amount included approximately $590 million that was related to the previous
quarter.

In response to the Transportation Department's concerns and considering
applicable accounting standards, the Bureau of Public Debt revised its
reporting on the Highway Account receipts for fiscal years 1998 and 1999 to
reflect the third-quarter certifications. While this action helped ensure
that the most recent available information was used in the calculation of
the RABA adjustments for fiscal years 2000 and 2001, it did not address the
underlying issues discussed earlier in this report regarding the accuracy
and timeliness of Treasury's reporting on Highway Trust Fund receipts. For
example, in the future, if significant amounts attributable to the third or
fourth quarter of a given fiscal year are certified as collections in
subsequent quarters, the RABA calculation for that fiscal year would be
affected because those amounts would not be reflected in Treasury's report
on Highway Trust Fund receipts for that fiscal year. The amounts would be
reflected in the report for the following fiscal year. Therefore, although
we do not have information on the extent of over- or under-payments to the
states, if any, it is evident that problems with the accuracy and timeliness
of information in Treasury's revenue reports could affect the accuracy of
the amounts distributed for a particular fiscal year.

Can Affect the Amounts of Highway Program Funds Apportioned to Individual
States

Given the concerns about the reliability of the data and methodology used in
FHWA's attribution process, the reliability of the estimates resulting from
that process has not been demonstrated, and there is little assurance that
the actual amounts distributed to individual states are accurate. FHWA uses
the information developed in its attribution process to apportion large
amounts to states--about $11.5 billion in fiscal year 2000 alone. Although
we have concerns about the accuracy of the relative amounts apportioned to
individual states, there is no way of knowing the extent of over- or
under-payments, if any, to individual states because FHWA has not
systematically verified the accuracy of the state motor fuel data or its own
methodology used in the attribution process.

Under TEA-21, the state motor fuel data used in FHWA's attribution process
has considerably greater importance in apportioning highway program funds
than under the predecessor legislation known as ISTEA.32 For example, under
TEA-21's formulas for the three programs and the "Minimum Guarantee"
provision discussed below, state motor fuel data directly affected the
apportionment of about $11.5 billion in federal highway program funds in
fiscal year 2000.33 In contrast, under ISTEA, an average of about $1.2
billion of highway program funds was apportioned each year on the basis of
state motor fuel data during the fiscal year 1992 through fiscal 1997
period, according to FHWA officials.

On October 1 of each fiscal year, FHWA's Office of Budget and Finance
apportions funds for several major highway programs to each state on the
basis of various formulas. These formulas, established by TEA-21, involve an
array of data and factors combined into a series of mathematical
calculations that are applied to the highway programs' funding levels
authorized in TEA-21.34 While the apportionment factors influence the
relative funding that is distributed to each state, those factors do not
influence each year's overall funding levels for the highway programs which
are estimated in TEA-21 and then annually adjusted through the RABA
calculation on the basis of Highway Account receipts information.

The attribution data on each state's estimated motor fuel usage and
contributions to the Highway Account of the Highway Trust Fund are used for
apportionment purposes in three major highway programs--the Interstate
Maintenance, Surface Transportation, and National Highway System programs.
In addition, the data on attributed contributions are used to implement the
"Minimum Guarantee" provision of TEA-21. The attribution data for each state
are applied to the apportionment formulas in the following ways:

ï¿½ Commercial vehicle contributions are used as a factor in distributing 33.3
percent of Interstate Maintenance program funds to the states. This factor
is defined as the sum of contributions from the motor fuel taxes on diesel
and other special fuels and the truck-related taxes on tires, truck and
trailer sales, and heavy vehicle use.

ï¿½ Highway Account contributions are used as a factor in distributing 35
percent of the Surface Transportation Program's funds. This factor is
defined as the sum of contributions from all motor fuel and truck-related
taxes.

ï¿½ Total diesel fuel used on highways is a factor in distributing 30 percent
of the National Highway System program's funds. FHWA excludes the public use
of diesel fuel and calculates this apportionment factor using only private
and commercial vehicle use of diesel fuel.

ï¿½ Highway Account contributions are used as a factor in calculating the
"Minimum Guarantee" provision in TEA-21, which states that each state is
guaranteed that its share of apportionments will be at least 90.5 percent of
its percent share of contributions to the Highway Account.

FHWA uses the results of its motor fuel attribution analysis to calculate
each state's relative share of motor fuel consumed on highways. It then
multiplies the resulting percent share of gallons by the amount of highway
tax receipts allocated to the Fund's Highway Account to estimate each
state's attributed contributions to the Fund.

Scope and Methodology

Our approach in reviewing Treasury's and FHWA's processes, and the impact of
those processes on the level of highway program funds distributed to the
states, was to (1) document the major steps in these processes and the times
when these steps occur; (2) identify and evaluate any issues regarding the
accuracy, timeliness, or reliability of the data and the methodologies used
in Treasury's and FHWA's processes; and (3) describe and assess any efforts
by those agencies to improve their data or processes. To do this, we
reviewed pertinent agency documentation and interviewed knowledgeable agency
officials at several offices within the Treasury and Transportation
departments. Within Treasury, our work was performed at IRS, the Office of
Tax Analysis, the Bureau of Public Debt, and the Financial Management
Service. Within Transportation, our work was performed at several offices in
FHWA and at the Office of the Secretary.

With regard to our work on Treasury's revenue allocation process, we
incorporated the results of our prior work addressing Treasury's processes
for allocating tax receipts to federal excise-tax-related trust funds,
including the Highway Trust Fund. With regard to our work on FHWA's
attribution process, we reviewed the ongoing efforts of the Motor Fuel
Reporting Information Committee--an organization of officials in FHWA and
state transportation and revenue agencies that is identifying issues and
improvements needed in the motor-fuel reporting system. Additionally, we
visited FHWA field offices and state revenue and transportation agencies in
two states--Florida and North Carolina--to develop an understanding of
states' role in providing data used in estimating states' contributions to
the Fund. We chose these two states because they are members of the Motor
Fuel Reporting Information Committee. Also, an FHWA official responsible for
the attribution process told us that officials in these two states were very
knowledgeable about the motor-fuel reporting process and thus their insights
would be useful to our work. We conducted our work from August 1999 through
June 2000 in accordance with generally accepted government auditing
standards.

Comments From the Internal Revenue Service

The following are GAO's comments on the Internal Revenue Service's letter
dated June 16, 2000.

1. We agree that "payment" and "deposit" are separate terms since
semimonthly deposits are based on estimates of tax liability, while final
payments are due with the quarterly return on the return due date. We have
revised the report language as appropriate.

2. We agree that initial allocations to the trust funds are and will
continue to be estimates. As stated in our report, once the new method for
making initial allocations to the trust funds is in place, IRS will still
need to certify the amounts due to the trust funds and make adjustments to
initial allocations on the basis of tax returns. We also made revisions to
the report to clarify that we are not implying that IRS certify collections
of highway user tax receipts on the basis of deposits. While these changes
improved our description of the allocation process, we believe that
opportunities exist for improving the quality of the estimates on which
initial allocations to the Highway Trust Fund and other excise-tax-related
trust funds are based. Our report points out that obtaining more detailed
data from taxpayers at the time they make tax deposits could improve these
estimates.

3. We agree that there are a number of reasons why the amounts attributable
to a given quarter are certified as collections in a subsequent quarter. To
clarify this point, we have revised our description of such instances and
our explanation of the reasons they occur. As stated in our report, such
instances can affect Treasury's yearly reports on Fund receipts and
consequently the amounts distributed to the states, particularly those that
occur in the latter part of the fiscal year.

4. IRS did not fully address our recommendation that Treasury develop a
comprehensive plan for implementing its efforts to improve the process for
allocating receipts to the excise-tax-related trust funds. While IRS
indicated that it will develop comprehensive plans for the Custodial
Accounting Project--which includes these efforts--it did not provide
information on when these plans will be developed.

5. IRS did not fully address our recommendation that this comprehensive plan
include milestones for evaluating, as a near-term measure, the use of
incentives for encouraging taxpayers to start providing a detailed excise
tax breakdown with their deposits. We believe that waiting until mid-2003 to
study incentives will not allow enough time for the agency to implement an
incentives program, if it decides to do so, prior to the planned deployment
of its new payment information database which, according to IRS, will have
the capability to record this detailed data. Therefore, we did not change
the report's recommendation.

6. IRS did not fully address our recommendation that the comprehensive plan
include milestones for reexamining taxpayer capabilities to provide detailed
data in several years and deciding whether to require the data from
taxpayers. While IRS indicated that it intends to perform these tasks in
four years, we believe that the agency needs to perform these tasks before
the deployment of its new payment information database and revenue
allocation method, planned for 2004, in order to ensure that IRS will be
ready to fully implement these efforts. Furthermore, these tasks, with
appropriate milestones, should be included in IRS' comprehensive plan for
its database and revenue allocation efforts. Therefore, we did not change
the report's recommendation.

7. IRS did not fully address our recommendation that the comprehensive plan
include a milestone for determining what changes may be needed--as a result
of changes being planned in the revenue allocation method--in Treasury's
methods for certifying and adjusting the amounts allocated to the trust
funds on the basis of analyses of tax return and payment data. While IRS'
letter states that the Secretary of the Treasury is considering this issue,
we continue to emphasize that this milestone should be included in IRS'
comprehensive plan for its database and revenue allocation efforts.
Therefore, we did not change the report's recommendation.

GAO Contacts and Staff Acknowledgments

Phyllis F. Scheinberg (202) 512-3650

Gregory D. Kutz (202) 512-3406

In addition to the individuals named above, Paul Caban, Rita A. Grieco, Judy
L. Guilliams-Tapia, Ted Hu, Anne T. McLeod, Sara Ann Moessbauer, Steven J.
Sebastian, and Ronald E. Stouffer made key contributions to this report.

Related GAO Products

Highway Trust Fund: Financial Condition as of September 30, 1997
(GAO/RCED-98-171R, Apr. 24, 1998).

Highway Trust Fund: Possible Impact if It Had Financed All Highway
Expenditures (GAO/RCED-98-78R, Feb. 6, 1998).

Federal Highway Programs: Status of Federal Highway Programs in the Absence
of Reauthorization (GAO/T-RCED-98-38, Nov. 4, 1997).

Highway Funding: The Federal Highway Administration's Funding Apportionment
Model (GAO/RCED-97-159, June 5, 1997).

Highway Trust Fund: Financial Status and Outlook (GAO/T-RCED-96-169, May 16,
1996).

Highway Funding: Alternatives for Distributing Federal Funds (GAO/RCED-96-6,
Nov. 28, 1995).

Tax Systems Modernization: Results of Review of IRS' March 7, 2000,
Expenditure Plan (GAO/AIMD-00-175, May 24, 2000).

IRS Modernization: Long-term Effort Under Way, But Significant Challenges
Remain (GAO/T-GGD/AIMD-00-154, May 3, 2000).

IRS Modernization: Business Practice, Performance Management, and
Information Technology Challenges (GAO/T-GGD/AIMD-00-144, Apr. 10, 2000).

Internal Revenue Service: Results of Fiscal Year 1999 Financial Statement
Audit (GAO/T-AIMD-00-104, Feb. 29, 2000).

Financial Audit: IRS' Fiscal Year 1999 Financial Statements (GAO/AIMD-00-76,
Feb. 29, 2000).

Applying Agreed-Upon Procedures: Highway Trust Fund Excise Taxes
(GAO/AIMD-00-82R, Feb. 25, 2000).

Internal Revenue Service: Custodial Financial Management Weaknesses
(GAO/AIMD-99-193, Aug. 4, 1999).

Tax Systems Modernization: Results of Review of IRS' Initial Expenditure
Plan (GAO/AIMD/GGD-99-206, June 15, 1999).

Financial Audit: IRS' Fiscal Year 1998 Financial Statements (GAO/AIMD-99-75,
Mar. 1, 1999).

Agreed-Upon Procedures: Highway Trust Fund Excise Taxes (GAO/AIMD-99-71R,
Feb. 25, 1999).

Excise Taxes: Internal Control Weaknesses Affect Accuracy of Distributions
to the Trust Funds (GAO/AIMD-99-17, Nov. 9, 1998).

Internal Revenue Service: Immediate and Long-Term Actions Needed to Improve
Financial Management (GAO/AIMD-99-16, Oct. 30, 1998).

Internal Revenue Service: Remaining Challenges to Achieve Lasting Financial
Management Improvements (GAO/T-AIMD/GGD-98-138, Apr. 15, 1998).

Financial Audit: Examination of IRS' Fiscal Year 1997 Custodial Financial
Statements (GAO/AIMD-98-77, Feb. 26, 1998).

Agreed-Upon Procedures: Excise Taxes (GAO/AIMD-98-78R, Feb. 26, 1998).

Tax Systems Modernization: Blueprint Is a Good Start But Not Yet
Sufficiently Complete to Build or Acquire Systems (GAO/AIMD/GGD-98-54, Feb.
24, 1998).

Corrections to the Federal Highway Trust Fund (B-275490, Dec. 5, 1996).

Tax Systems Modernization: Actions Underway But IRS Has Not Yet Corrected
Management and Technical Weaknesses (GAO/AIMD-96-106, June 7, 1996).

Tax Administration: Diesel Fuel Excise Tax Change (GAO/GGD-96-53, Jan. 16,
1996).

Tax Systems Modernization: Management and Technical Weaknesses Must Be
Corrected If Modernization Is To Succeed (GAO/AIMD-95-156, July 26, 1995).

Financial Management: Important IRS Revenue Information Is Unavailable or
Unreliable (GAO/AIMD-94-22, Dec. 21, 1993).

(348183)

Table 1: Highway User Tax Structure 16

Table 2: Gross Highway User Tax Receipts Credited to
the Highway Account of the Highway Trust Fund,
Fiscal Year 1999 17

Table 3: Factors for Attributing Highway User Taxes to Highway
Users in Each State 33

Table 4: Estimated Motor Fuel Usage and Contributions to the
Highway Account of the Highway Trust Fund Attributable
to Highway Users in Each State, Apportionment Cycle
for Fiscal Year 2000 38

Table 5: Calculation of the Revenue Aligned Budget Authority
Adjustment, Fiscal Years 2000 and 2001 48
  

1. P.L. 105-178 (June 9, 1998).

2. The Treasury Department credits most of the highway motor fuel tax
receipts and all of the truck-related tax receipts to the Highway Trust
Fund's Highway Account. The remaining motor fuel tax amounts are credited to
the Highway Trust Fund's Mass Transit Account and to three other funds--the
Leaking Underground Storage Tank Trust Fund, the Aquatic Resources Trust
Fund, and the General Fund.

3. A terminal is a fuel storage and distribution facility that receives fuel
supplies by pipeline or vessel and from which fuel may be disbursed into a
truck, rail car, or other means of transfer.

4. In this report, we use the term "allocation" to refer to the processes
for transferring amounts from the General Fund to the Highway Trust Fund, on
the basis of initial estimates of highway user tax receipts, and for
subsequently adjusting the amounts transferred, on the basis of certified
collections.

5. See, for example, Financial Audit: IRS' Fiscal Year 1999 Financial
Statements (GAO/AIMD-00-76 , Feb. 29, 2000), Applying Agreed Upon
Procedures: Highway Trust Fund Excise Taxes (GAO/AIMD-00-82R , Feb. 25,
2000), Internal Revenue Service: Custodial Financial Management Weaknesses
(GAO/AIMD-99-193 , Aug. 4, 1999), Financial Audit: IRS' Fiscal Year 1998
Financial Statements (GAO/AIMD-99-75 , Mar. 1, 1999), Agreed-Upon
Procedures: Highway Trust Fund Excise Taxes (GAO/AIMD-99-71R , Feb. 25,
1999), and Excise Taxes: Internal Control Weaknesses Affect Accuracy of
Distributions to the Trust Funds (GAO/AIMD-99-17 , Nov. 9, 1998).

6. Transportation's Office of the Secretary works with the Office of
Management and Budget to develop this calculation. FHWA executes any
distribution or reduction in funding resulting from this calculation.

7. A terminal is a fuel storage and distribution facility that receives fuel
supplies by pipeline or vessel and from which fuel may be disbursed into a
truck, rail car, or other means of transfer.

8. Some of the motor fuel tax amounts are credited to the Mass Transit
Account of the Highway Trust Fund, the Leaking Underground Storage Tank
Trust Fund, the Aquatic Resources Trust Fund, and the General Fund.

9. Refunds and credits reflect amounts returned to taxpayers because the
motor fuel was ultimately used for a tax-exempt purpose. Transfers reflect
amounts transferred from the Highway Trust Fund to other trust funds as
required by law.

10. If this RABA adjustment lowers the guaranteed funding level for a given
fiscal year, TEA-21 requires that the Department of Transportation reduce
proportionately the amount of funding authorized for the next fiscal year.

11. Apportioning is the distributing of funds using a formula provided in
law.

12. For the purpose of this discussion, the term receipts refers to
semimonthly deposits by taxpayers for estimated tax liabilities as well as
any additional amounts subsequently received for each quarter.

13. The Electronic Federal Tax Payment System allows taxpayers to make tax
deposits electronically. All business taxpayers that have an annual federal
tax liability exceeding $50,000 are required to use this system for making
tax deposits.

14. In preparing these estimates of monthly tax receipts, the Office of Tax
Analysis uses estimated fiscal year tax receipts published in the
President's budget each February and in a subsequent review and update of
the budget each July. The Office develops overall fiscal year estimates
using economic models and then develops monthly estimates by using
additional models to spread the fiscal year receipt estimates over a
12-month period.

15. IRS' certified collection figures represent those receipts that IRS
classifies as highway user tax receipts for each quarter on the basis of
processed tax returns, rather than actual receipts of these taxes in that
quarter.

16. See Financial Audit: IRS' Fiscal Year 1999 Financial Statements
(GAO/AIMD-00-76 , Feb. 29, 2000), Internal Revenue Service: Custodial
Financial Management Weaknesses (GAO/AIMD-99-193, Aug. 4, 1999), and
Financial Audit: IRS' Fiscal Year 1998 Financial Statements (GAO/AIMD-99-75
, Mar. 1, 1999).

17. Tax Type/Fee Code Study−Final Report (prepared for the IRS by
Integration Support Contract, Fairfax, Va., Jan. 1999).

18. For example, the study reported that Federal Highway Administration
officials said that the detailed data would likely increase the
predictability of the excise tax revenue stream and reduce the number of
adjustments made to the data.

19. During any given quarterly period, taxpayers make a total of 6
semimonthly excise tax payments.

20. We have not considered whether IRS would be authorized to provide
taxpayers with financial incentives to report detailed data by tax type at
the time of deposit or whether IRS would require additional statutory
authority to implement such a program.

21. Since 1995, the Congress and GAO have raised concerns about IRS'
capability to effectively build modernized systems. While IRS has taken
actions aimed at addressing these concerns, we have recently reported that
IRS continues to lack this capability and that substantial challenges remain
to establishing it. See, for example, Tax Systems Modernization: Results of
Review of IRS' March 7, 2000, Expenditure Plan (GAO/AIMD-00-175 , May 24,
2000), IRS Modernization: Long-term Effort Under Way, but Significant
Challenges Remain (GAO/T-GGD/AIMD-00-154 , May 3, 2000), IRS Modernization:
Business Practice, Performance Management, and Information Technology
Challenges (GAO/T-GGD/AIMD-00-144, Apr. 10, 2000), Tax Systems
Modernization: Results of Review of IRS' Initial Expenditure Plan
(GAO/AIMD/GGD-99-206 , June 15, 1999), Tax Systems Modernization: Blueprint
Is a Good Start But Not Yet Sufficiently Complete to Build or Acquire
Systems (GAO/AIMD/GGD-98-54 , Feb. 24, 1998), Tax Systems Modernization:
Actions Underway, but IRS Has Not Yet Corrected Management and Technical
Weaknesses (GAO/AIMD-96-106 , June 7, 1996), and Tax Systems Modernization:
Management and Technical Weaknesses Must Be Corrected if Modernization Is to
Succeed (GAO/AIMD-95-156 , July 26, 1995).

22. These payments include taxes, penalties, fees, and other remittances.

23. These amounts will be broken down by tax type on the basis of
information provided by the taxpayer (if available) or in IRS' historical
taxpayer account records.

24. These taxes become part of the price of the product and are ultimately
paid by the highway user.

25. For example, see Highway Statistics `98, U.S. Department of
Transportation, Federal Highway Administration (1999).

26. In some states, known as "donor" states, highway users pay more in
highway user taxes than those states receive in federal highway funds.
Conversely, in states known as "donee" states, highway users pay less in
highway user taxes than those states receive in highway funds.

27. Special fuels include liquefied petroleum gases, compressed natural gas,
and other fuels.

28. FHWA Form 556 entitled "State Motor Fuel Tax Receipts And Initial
Distribution By Collection Agencies."

29. FHWA obtained and analyzed the 1997 state motor fuel data from March
1998 through August 1999--a period of about 17 months.

30. The federal motor fuel taxes are imposed at the terminal rack, the
facility where fuel from bulk storage tanks is loaded into tanker trucks or
rail cars for delivery to retail stations or to bulk users. Bulk users
include the power companies, hospitals, and farmers who use the fuel for
nonhighway purposes, and wholesalers who place the fuel into intermediate
storage, according to an IRS official.

31. If this RABA adjustment lowers the guaranteed funding level for a given
fiscal year, TEA-21 requires that the Department of Transportation reduce
the amount of funding authorized on October 1 of the next fiscal year.

32. Intermodal Surface Transportation Efficiency Act of 1991 (P.L. 102-240).

33. The $11.5 billion included about $1.3 billion of the Interstate
Maintenance funds, $2 billion of the Surface Transportation Program funds,
$1.4 billion of the National Highway System funds, and $6.8 billion in
"Minimum Guarantee" funds.

34. In addition to motor fuel usage and estimated state-level contributions
to the Highway Account, TEA-21's apportionment formulas include such factors
as vehicle miles traveled, population, and highway lane miles.
*** End of document. ***