This is the accessible text file for GAO report number GAO-04-213R 
entitled 'Applying Agreed-Upon Procedures: Highway Trust Fund Excise 
Taxes' which was released on November 20, 2003.

This text file was formatted by the U.S. General Accounting Office 
(GAO) to be accessible to users with visual impairments, as part of a 
longer term project to improve GAO products' accessibility. Every 
attempt has been made to maintain the structural and data integrity of 
the original printed product. Accessibility features, such as text 
descriptions of tables, consecutively numbered footnotes placed at the 
end of the file, and the text of agency comment letters, are provided 
but may not exactly duplicate the presentation or format of the printed 
version. The portable document format (PDF) file is an exact electronic 
replica of the printed version. We welcome your feedback. Please E-mail 
your comments regarding the contents or accessibility features of this 
document to Webmaster@gao.gov.

This is a work of the U.S. government and is not subject to copyright 
protection in the United States. It may be reproduced and distributed 
in its entirety without further permission from GAO. Because this work 
may contain copyrighted images or other material, permission from the 
copyright holder may be necessary if you wish to reproduce this 
material separately.

November 20, 2003:

The Honorable Kenneth M. Mead:

Inspector General:

Department of Transportation:

Subject: Applying Agreed-Upon Procedures: Highway Trust Fund Excise 
Taxes:

Dear Mr. Mead:

We have performed the procedures contained in the enclosure to this 
report, which we agreed to perform and with which you concurred, solely 
to assist your office in ascertaining whether the net excise tax 
revenue distributed to the Highway Trust Fund (HTF) for the fiscal year 
ended September 30, 2003, is supported by the underlying records. As 
agreed with your office, we evaluated fiscal year 2003 activity 
affecting distributions to the HTF.

In performing the agreed-upon procedures, we conducted our work in 
accordance with U.S. generally accepted government auditing standards, 
which incorporate financial audit and attestation standards established 
by the American Institute of Certified Public Accountants. These 
standards also provide guidance for performing and reporting the 
results of agreed-upon procedures.

The adequacy of the procedures to meet your objectives is your 
responsibility, and we make no representation in that respect. The 
procedures we agreed to perform include (1) detailed tests of 
transactions that represent the underlying basis of amounts distributed 
to the HTF, (2) review of the Internal Revenue Service's (IRS) 
quarterly HTF certifications, (3) review of the Department of the 
Treasury Financial Management Service (FMS) adjustments to the HTF for 
fiscal year 2003, (4) review of IRS's precertification[Footnote 1] of 
receipts for the third quarter of fiscal year 2003, (5) review of 
certain procedures of the Department of the Treasury Office of Tax 
Analysis' (OTA) process for estimating amounts to be distributed to the 
HTF for the fourth quarter of fiscal year 2003, and other procedures 
including (6) comparison of net excise tax distributions to the HTF 
during fiscal year 2003 and amounts reported in the draft financial 
statements prepared by the Bureau of the Public Debt (BPD) for the HTF 
and the HTF's draft financial statements, (7) detailed tests of 
transactions that represent total IRS tax revenue receipts and refunds, 
and (8) review of key reconciliations of IRS records to Treasury 
records. The enclosure contains the agreed-upon procedures and our 
findings from performing each of the procedures.

We were not engaged to perform, and did not perform, an audit, the 
objective of 
which would have been the expression of an opinion on the amount of net 
excise taxes distributed to the HTF. Accordingly, we do not express 
such an opinion. Had we performed additional procedures, other matters 
might have come to our attention that would have been reported to 
you.[Footnote 2] We completed the agreed-upon procedures on November 7, 
2003.

We provided a draft of this report to IRS and OTA officials for review 
and comment. They agreed with the results and findings presented in 
this report. In response to our findings concerning errors in its 
estimates, OTA stated that it has instituted additional internal 
control procedures to promptly detect and correct any future errors.

This report is intended solely for the use of the Office of Inspector 
General of the Department of Transportation and should not be used by 
those who have not agreed to the procedures and have not taken 
responsibility for the sufficiency of the procedures for their purpose. 
However, this report is a matter of public record and its distribution 
is not limited. Copies are available to others upon request. This 
report is also available at no charge on GAO's Internet Web site at 
http://www.gao.gov. If you have any questions, please call me at (202) 
512-3406.

Sincerely yours,

Steven J. Sebastian:

Director:

Financial Management and Assurance:

Signed by Steven J. Sebastian:

[End of section]

Enclosure:

Highway Trust Fund Excise Tax Procedures and Results:

I. Detailed tests of transactions that represent the underlying basis 
of amounts distributed to the HTF in fiscal year 2003:

A. Nonrepresentative selection of tax returns from the quarter ended 
September 30, 2002[Footnote 3]:

1. For the quarter ending September 30, 2002, select the 30 largest 
excise tax returns containing excise taxes related primarily to the 
HTF and the Airport and Airway Trust Fund (AATF), on the basis of 
total tax liability[Footnote 4] amount, from IRS's master file.
[Footnote 5]

Description of findings and results:

We selected the 30 largest excise tax returns related primarily to the 
HTF and the AATF from the quarter ended September 30, 2002, for 
testing. The selection was based on the total tax liability amount and 
type of taxes owed, for each return, from IRS's master file.

The total tax liability amount related to these 30 returns was 
approximately $8.6 billion, or 66 percent of the total excise tax 
liability amount-$13 billion-for all excise tax types for the quarter 
ended September 30, 2002.

Of these 30 returns, 22 contained primarily HTF-related taxes and 8 
contained primarily AATF taxes.

2. For each of the 22 returns related primarily to the HTF we 
performed the following procedures, which resulted in our testing 
approximately $6.4 billion in prorated collections[Footnote 6] 
affecting fiscal year 2003 distributions to the HTF: 

a. Trace the liability amount for abstracts[Footnote 7] 59, 60, and 62 
from the tax return to IRS's master file.

Description of findings and results:

The liability amount for abstracts 59, 60, and 62 on the tax return 
agreed with IRS's master file for 22 of the 22 returns.

b. Check the mathematical accuracy of the taxpayer's calculations on 
the tax return for the selected abstracts.

Description of findings and results:

The taxpayer's calculations on all 22 returns were mathematically 
correct.

c. Recompute the prorated collection amount for the selected abstracts 
based on information from the master file and compare this amount to 
the amount from the Collection Certification System audit 
file.[Footnote 8]

Description of findings and results:

The recomputed prorated collection amounts for the three selected 
abstracts agreed with amounts in IRS's Collection Certification System 
audit file for all 22 of the returns.

B. Dollar unit sample (DUS) of transactions from the quarters ended 
December 31, 2002, and March 31, 2003:

1. Sampling:

(a) Obtain excise tax assessments and collection data from IRS's master 
file for the first 6 months of fiscal year 2003. Determine if excise 
tax collections per the master file agree with IRS's general ledger. 
Reconcile total excise tax collections from the master file to total 
excise tax collections from the Collection Certification System audit 
files to determine if they materially[Footnote 9] agree.

Description of findings and results:

Excise tax collections for the first 6 months of fiscal year 2003 per 
the master file materially agreed with IRS's general ledger and with 
total excise tax collections from the Collection Certification System.

(b) Select a random attribute sample of 78 excise tax assessments from 
IRS's master file.[Footnote 10] Compare assessment and receipt 
information for each sample item from the master file to the assessment 
and receipt information in the Collection Certification System to 
determine if assessments and receipts from the master file are 
contained in the Collection Certification System.

Description of findings and results:

For each sample item, assessments and receipts from the master file 
were contained in the Collection Certification System.

(c) To determine if the Collection Certification System properly 
summarized the prorated collections, total the prorated collections 
for selected abstracts[Footnote 11] from the audit files and compare 
these amounts to amounts in the Reports of Excise Tax Collection.
[Footnote 12]

Description of findings and results:

The Collection Certification System properly summarized the prorated 
collections for all of the selected abstracts related to the HTF and 
the AATF. Prorated collections for the above-mentioned trust funds from 
the audit files agreed with the corresponding amounts in the Reports of 
Excise Tax Collection.

(d) Separate the total population of prorated collections from the 
audit files into the following distinct populations: (1) HTF, (2) 
AATF, and (3) other excise tax abstracts. Use DUS to select a sample 
of prorated excise tax collections from the HTF population.

Description of findings and results:

Use of DUS with a confidence level of 80 percent, a test materiality of 
$322 million, and an expected aggregate error amount of $96.6 million 
resulted in a sample of 96[Footnote 13] prorated collections for HTF 
for the first 6 months of fiscal year 2003.

(e) Select samples of prorated excise tax collections from the two non-
HTF populations.

Description of findings and results:

Use of DUS with a confidence level of 80 percent, a test materiality of 
$89 million, and an expected aggregate error amount of $26.7 million 
resulted in a sample of 67[Footnote 14] prorated collections for the 
first 6 months of fiscal year 2003 for the AATF.

A random attribute sample of 45 items from the population of prorated 
tax collections, related to all excise taxes other than the HTF and the 
AATF, was selected for testing.[Footnote 15]

2. Detailed tests of transactions:

(a) For each prorated excise tax collection sampled from the HTF 
population:

* Check to see that the assessment amount on the tax return, for the 
sampled abstract, agrees with the amount recorded in IRS's master file.

Description of findings and results:

The assessment amounts on the tax returns agreed with the amounts 
recorded in IRS's master file for all of the sampled abstracts.

* Check the mathematical accuracy of the taxpayers' calculations on the 
tax returns for the related abstract.

Description of findings and results:

The taxpayers' calculations on the tax returns for the related 
abstracts were mathematically correct for all of the sampled abstracts.

* Recompute the prorated collection amount based on information from 
the master file and compare this amount to the sample items selected 
from the Collection Certification System audit file.[Footnote 16]

Description of findings and results:

The recomputed prorated collection, based on information from the 
master file, agreed with the amounts for all of the sampled items.

(b) Perform detailed testing on the two samples of prorated collections 
from the non-HTF populations to determine if they contain any HTF 
excise tax collections.

Description of findings and results:

The two samples of prorated collections from the non-HTF populations 
did not contain any HTF excise tax collections.

(c) Evaluate the results of conducting steps (a) and (b).

Description of findings and results:

As noted in the results from steps (a) and (b), we found no errors.

II. Review of IRS's quarterly HTF certifications:

A. Receipt certifications:

Perform the following steps on IRS's HTF receipt certifications for the 
quarters ended September 30, 2002, December 31, 2002, and March 31, 
2003:

1. Inspect the certification letters[Footnote 17] for authorizing 
signatures.

Description of findings and results:

The HTF certification letters for all three quarters had authorizing 
signatures.

2. Determine if evidence exists that the supervisor or another analyst 
checked the certification letters and supporting worksheets.

Description of findings and results:

There was evidence that another analyst and a supervisor checked the 
certification letters and supporting worksheets for all three quarters.

3. Recalculate the totals on the certification letters to determine if 
they are mathematically correct.

Description of findings and results:

The totals on the certification letters for all three quarters were 
mathematically correct.

4. Trace the certified amounts for diesel fuel tax (abstract 60), 
gasoline tax (abstract 62), and tax on 10 percent gasohol (abstract 
59)[Footnote 18] from the certification letters back to the Reports of 
Excise Tax Collection[Footnote 19] and Treasury 90 Report.[Footnote 20]

Description of findings and results:

The certified amounts for diesel fuel tax (abstract 60), gasoline tax 
(abstract 62), and tax on 10 percent gasohol (abstract 59) per the 
certification letters agreed with the related Reports of Excise Tax 
Collection for all three quarters.

However, on IRS's receipt certification for the quarter ended December 
31, 2002, the IRS analyst used excise tax credit information from the 
Treasury 90 Report for the quarter ended September 30, 2002, rather 
than the Treasury 90 Report for the quarter ended December 31, 2002. As 
a result, IRS understated certified receipts to the Highway Account by 
approximately $12.4 million and understated certified receipts to the 
Mass Transit Account by approximately $1.7 million. IRS discovered the 
error after it had sent the certification letter to the FMS, and FMS 
had already recorded the true-up adjustment for using the certification 
letters. IRS corrected this error on the receipt certification for the 
following quarter ended March 31, 2003. Therefore, this error had no 
net effect on fiscal year 2003 distributions to the HTF.

5. Review the distribution rates used by IRS to determine whether the 
distribution rates for diesel fuel tax (abstract 60), gasoline tax 
(abstract 62), and tax on 10 percent gasohol (abstract 59) agree with 
the applicable laws.[Footnote 19]

Description of findings and results:

The distribution rates used by IRS for diesel fuel tax (abstract 60), 
gasoline tax (abstract 62), and tax on 10 percent gasohol (abstract 59) 
agreed with the applicable laws in effect during all three quarters.

6. Review the Reports of Excise Tax Collection used in the 
certification to determine if they contain significant[Footnote 22] 
collections from prior quarters.

Description of findings and results:

The Reports of Excise Tax Collection supporting IRS's certification to 
HTF for the quarter ended September 30, 2002, contained approximately 
$261 million in net HTF excise tax collections related to prior 
quarters. This consisted of approximately $262 million from the quarter 
ended June 30, 2002 and ($1 million)[Footnote 23] from prior quarters. 
IRS attributed the delay in classifying the June quarter amounts to two 
taxpayers who did not file timely and five tax returns with processing 
delays due to taxpayer or IRS errors.

The Reports of Excise Tax Collection supporting IRS's certification to 
HTF for the quarter ended December 31, 2002, contained approximately 
$330 million in net HTF excise tax collections related to prior 
quarters. Of this amount, approximately $253 million was from the 
quarter ended September 30, 2002. IRS attributed this delay to three 
taxpayers who did not file timely and two tax returns with processing 
delays due to taxpayer or IRS errors.

The Reports of Excise Tax Collections supporting IRS's certification to 
the HTF for the quarter ended March 31, 2003, did not contain 
significant prior quarter collections.

7. Heavy vehicle use taxes, which go to the HTF, are reported on Form 
2290 and are not included in the Collection Certification System. 
Trace the heavy vehicle use tax amounts from the Highway Account 
certification letters to the master file.

Description of findings and results:

Heavy vehicle use tax per the Highway Account certification letters 
agreed with the master file for all three quarters.

B. Refund/credit reclassification[Footnote 24]

Perform the following steps on IRS's HTF refund/credit certifications 
for the quarters ended December 31, 2002, March 31, 2003, June 30, 
2003, and September 30, 2003:[Footnote 25]

1. Inspect the certification letters for authorizing signatures.

Description of findings and results:

The certification letters for all four quarters had authorizing 
signatures.

2. Determine if evidence exists that a supervisor or another analyst 
checked the certification letters and accompanying schedules.[Footnote 
26]

Description of findings and results:

There was evidence that another analyst and a supervisor checked the 
certification letters and accompanying schedules for all four quarters.

3. Recalculate the totals on the certification letters and 
accompanying schedules to determine if they are mathematically 
correct.

Description of findings and results:

The totals on the certification letters and accompanying schedules were 
mathematically correct for all four quarters.

4. Trace the refund and credit amounts for gasoline tax and diesel 
tax[Footnote 27] from the schedules accompanying the certification 
letters to other summary refund/credit schedules. These other refund/
credit summary schedules summarize refund and credit data obtained from 
submission processing campuses' records.

Description of findings and results:

The refund and credit amounts for gasoline tax and diesel tax on the 
schedules accompanying the certification letters agreed with the 
amounts on the summary schedules for all four quarters.

III. Review of FMS adjustments:

Perform the following steps on FMS adjustments to the HTF excise tax 
distributions for the quarters ended September 30, 2002, December 31, 
2002, and March 31, 2003.

A. Compare the FMS adjustments made to the HTF for fiscal year 2003 
with original OTA estimates and IRS-certified amounts to see if they 
agree with the supporting schedules.[Footnote 28]

Description of findings and results:

For the FMS adjustments made to the HTF accounts (Highway and Mass 
Transit), the original OTA estimates and IRS-certified amounts agreed 
with the supporting schedules for all three quarters.

B. Recompute the difference between the OTA estimates and final IRS-
certified amounts to see if the amounts agree with the differences 
computed by FMS.

Description of findings and results:

The independently recalculated differences between the OTA estimates 
and the final IRS-certified amounts for the Highway Account agreed with 
the differences computed by FMS for all three quarters.

These amounts were[Footnote 29]

* for the quarter ended September 30, 2002, $636,666,000;

* for the quarter ended December 31, 2002, ($319,825,000); and:

* for the quarter ended March 31, 2003, $205,901,000.

The independently recalculated differences between the OTA estimates 
and the final IRS-certified amounts for the Mass Transit Account agreed 
with the differences computed by FMS for all three quarters.

These amounts were:

* for the quarter ended September 30, 2002, $16,750,000;

* for the quarter ended December 31, 2002, ($17,123,000); and:

* for the quarter ended March 31, 2003, $48,083,000.

IV. Review of IRS precertification for the quarter ended June 30, 
2003[Footnote 30]

A. Determine if evidence exists that the supervisor or another analyst 
checked the results and supporting worksheets.

Description of findings and results:

There was evidence that another analyst and a supervisor checked the 
results and supporting worksheets.

B. Recalculate the totals on the precertification to determine if they 
are mathematically correct.

Description of findings and results:

The totals on the precertification were mathematically correct.

C. Trace the amounts for diesel fuel tax (abstract 60), gasoline tax 
(abstract 62), and tax on 10 percent gasohol (abstract 59)[Footnote 31] 
from the precertification back to the Reports of Excise Tax Collection 
and Treasury 90 Report.

Description of findings and results:

The amounts for diesel fuel tax (abstract 60), gasoline tax (abstract 
62), and tax on 10 percent gasohol (abstract 59) per the 
precertification agreed with the related Reports of Excise Tax 
Collection and Treasury 90 Report.

D. Review the distribution rates used by IRS to determine whether the 
distribution rates for diesel fuel tax (abstract 60), gasoline tax 
(abstract 62), and tax on 10 percent gasohol (abstract 59) agree with 
the applicable laws.

Description of findings and results:

The distribution rates used by IRS for diesel fuel tax (abstract 60), 
gasoline tax (abstract 62), and tax on 10 percent gasohol (abstract 59) 
agreed with the applicable laws in effect during the quarter ended June 
30, 2003.

E. Review the Reports of Excise Tax Collection used in the 
precertification to determine if they contain significant[Footnote 32] 
collections from prior quarters.

Description of findings and results:

The Reports of Excise Tax Collection supporting IRS's precertification 
for the quarter ended June 30, 2003, did not contain significant prior 
quarter collections.

F. Review the Collection Certification System information to determine 
whether IRS omitted any significant returns[Footnote 33] from the 
precertification. If so, report for the Highway Account and the Mass 
Transit Account: (1) the average amount of HTF-related excise taxes 
from these taxpayers' returns that were included in IRS's certification 
from the four previous quarters and (2) the amount of HTF-related 
excise taxes from these taxpayers' returns that were included in IRS's 
certification for the quarter ended June 30, 2002.

Description of findings and results:

IRS did not omit any returns of historically significant excise tax 
taxpayers from its precertification to the HTF.

G. Heavy vehicle use taxes, which go to the HTF, are reported on Form 
2290 and are not included in the Collection Certification System. Trace 
these amounts from the Highway Account precertification letter to the 
master file.

Description of findings and results:

Heavy vehicle use tax per the Highway Account precertification agreed 
with the master file.

V. Procedures performed on excise tax distributions to the HTF for the 
quarter ended September 30, 2003:

A. Determine if OTA's process for identifying and incorporating into 
its trust fund estimates[Footnote 34] the effect of new legislation on 
excise tax receipts was in place during fiscal year 2003.

Description of findings and results:

OTA's process for identifying and incorporating into its trust fund 
estimates the effect of new legislation on excise tax receipts was in 
place during fiscal year 2003. OTA prepares a tax rate table[Footnote 
35] to capture information relating to legislation that affects tax 
rates, tax basis, accounts, and deposit rules in effect during the tax 
period.

B. Determine if there is evidence of review of the transfer forms and 
supporting schedules.

Description of findings and results:

There was evidence that another OTA economist reviewed the transfer 
forms and supporting schedules for the semimonthly transfers affecting 
distributions to the HTF for the quarter ended September 30, 2003.

C. Recalculate the totals on the transfer forms to determine if they 
are mathematically correct.

Description of findings and results:

The totals on the transfer forms affecting distributions to the HTF for 
the quarter ended September 30, 2003, were mathematically correct.

D. Trace the transfer amounts for diesel fuel tax (abstract 60), 
gasoline tax (abstract 62), tax on 10 percent gasohol (abstract 59), 
and heavy vehicle use tax,[Footnote 36] from the transfer forms, 
through the supporting schedules and back to the related source 
documents.[Footnote 37]

Description of findings and results:

The transfer amounts for diesel fuel tax (abstract 60), gasoline tax 
(abstract 62), tax on 10 percent gasohol (abstract 59), and heavy 
vehicle use tax from the transfer forms did not agree with the 
supporting schedules and source documents on the first two semimonthly 
estimates affecting distributions to the HTF for the quarter ended 
September 30, 2003. An error on one of OTA's electronic spreadsheets 
resulted in the collection percentages for all tax types being 
miscalculated. After we brought this to its attention, OTA corrected 
the spreadsheet error so it did not affect the remaining estimates for 
the quarter. However, since the effect of this error on the total 
quarterly estimate would be only a fraction of a percent, OTA did not 
correct for the effects of this error on the two affected semimonthly 
estimates. Consequently, OTA understated its estimate for these four 
taxes to the HTF by approximately $4.2 million.

In addition, there was another error affecting distributions to the HTF 
that was not directly related to the tax types selected for review. An 
error on OTA's electronic spreadsheets resulted in the omission of the 
estimated amounts for retail tax on trucks (abstract 33) from its 
estimates to the Highway Account for the first four semimonthly periods 
of the quarter. As a result, OTA understated its estimates to the HTF 
by approximately $305 million. OTA corrected for the effects of this 
error through its final semimonthly estimate affecting fiscal year 2003 
distributions to the HTF. Consequently, this error had no net effect on 
fiscal year 2003 distributions to the HTF.

VI. Other procedures:

A. Compare total fiscal year 2003 excise taxes distributed to the HTF 
with drafts of (1) HTF's fiscal year 2003 financial statements and (2) 
BPD's fiscal year 2003 financial statements for the HTF to determine 
if they agree.

Description of findings and results:

The $34.1 billion of fiscal year 2003 excise taxes distributed to the 
HTF agreed with the amount reported on the draft HTF financial 
statements but did not agree with the amount reported on the draft BPD 
fiscal year 2003 financial statements for the HTF. The BPD fiscal year 
2003 financial statements for the HTF reported excise tax distributions 
to the HTF of $33.7 billion. The difference is due to FMS's $388.6 
million downward adjustment for the quarter ended June 30, 2002--a 
$378.9 million adjustment to the Highway Account and a $9.7 million 
adjustment to the Mass Transit Account--which FMS recorded in December 
2002. The trust fund's administrator included this transaction as part 
of the fiscal year 2002 distributions to the HTF because the HTF had 
time to record the adjustment on its fiscal year 2002 financial 
statements, which were issued in January 2003. However, BPD included 
this transaction as part of the fiscal year 2003 distributions because 
it was recorded after the November 1, 2002, issuance date of BPD's 
fiscal year 2002 financial statements for the HTF.

B. Procedures performed as part of the fiscal year 2003 IRS financial 
statement audit:

1. From IRS's master files for the first 8 months of fiscal year 2003, 
use DUS to select statistical samples of (1) total tax revenue 
receipts and (2) refunds. For each sample item, test that the 
collection or refund amount, tax period, and tax class[Footnote 38] 
from source documentation agree with those recorded in IRS's master 
files.

Description of findings and results:

Detailed testing of 169 revenue receipts and 50 refund sample 
transactions showed that the collection or refund amount, tax period, 
and tax class from source documents agreed with those recorded in IRS's 
master files.

2. Review selected IRS submission processing campuses' monthly Treasury 
SF-224 reconciliations to determine if IRS-reported revenue receipts 
were properly classified and reconciled to Treasury FMS records. For 
refunds, review selected IRS submission processing campuses' monthly 
Treasury SF-224 reconciliations to determine if IRS-reported total 
refunds (all tax classes) materially[Footnote 39] reconciled to 
Treasury FMS records.[Footnote 40]

Description of findings and results:

Tax revenue receipts reported by selected IRS submission processing 
campuses through the monthly Treasury SF-224 reconciliation process 
were properly classified and materially agreed with Treasury FMS 
records.

Total refunds reported by the selected IRS submission processing 
campuses through the monthly Treasury SF-224 reconciliation process 
materially agreed with Treasury FMS records.

3. Perform procedures to determine whether tax revenue receipt 
balances by tax class, including excise taxes, per IRS's general 
ledger, materially agree with IRS's master files and Treasury records. 
For refunds, perform a comparison of total refund balances between the 
master file, the general ledger, and Treasury records.

Description of findings and results:

Tax receipt balances for all tax classes, including excise taxes, per 
IRS's general ledger, materially agreed with IRS's master files and 
with Treasury records.

Refund balances per IRS's general ledger materially agreed with the 
master file and with Treasury records.

(196003):

FOOTNOTES

[1] To accommodate the Department of Transportation's accelerated 
reporting date for fiscal year 2003, IRS performed a precertification 
of excise tax collections for the quarter ended June 30, 2003. The data 
are for information purposes only and the precertification does not 
constitute an official certification.



[2] In our report on the results of our audit of IRS's fiscal year 2003 
financial statements, we noted a material weakness in IRS's financial 
reporting process (U.S. General Accounting Office, Financial Audit: 
IRS's Fiscal Years 2003 and 2002 Financial Statements, GAO-04-126, 
November 13, 2003). A component of this process includes IRS's ability 
to allocate excise tax collections to the appropriate trust funds at 
the time deposits are made. This condition affects the adequacy of the 
distributions of federal excise tax revenue to recipient trust funds 
and is a continuation of an issue that we have reported on in prior 
years. 



[3] Since certifications are not completed until 6 months after the end 
of the quarter, the certification and corresponding FMS adjustment for 
the quarter ended September 30, 2002, were completed in March 2003, and 
thus affected fiscal year 2003 distributions to the HTF.



[4] Although the certifications are based on amounts collected, we used 
the tax liability amounts to identify the taxpayers paying the largest 
amounts of excise taxes. Our review shows that these taxpayers 
generally pay their excise taxes in full each quarter.



[5] The master file is a detailed database containing taxpayer 
information.



[6] IRS certifies to trust funds the amount of excise taxes collected. 
Because there are occasions in which taxpayers have not fully paid 
their tax liability, IRS must allocate the amount of payments actually 
received among the different excise taxes reported on the taxpayer's 
return. IRS's Collection Certification System prorates a taxpayer's 
payments proportionately among all taxes reported as owed on the tax 
return. For example, if a corporation reports that it owes $4 million 
for gasoline tax, $2 million for diesel fuel tax, and $1 million for 
gasohol tax on its Form 720 Quarterly Federal Excise Tax Return, but 
has paid IRS only $3.5 million at the time IRS performs its 
certification, the program prorates the $3.5 million in the following 
manner: $2 million to gasoline tax, $1 million to diesel fuel tax, and 
$500,000 to gasohol tax.



[7] The abstract numbers identify the tax type (e.g., gasoline and 
ticket tax) and are used as the basis for determining the distribution 
of the excise taxes to the various trust funds. Abstract numbers are 
preprinted on the Form 720 Quarterly Federal Excise Tax Return and are 
used by the taxpayer to report excise tax assessments. If the return 
was related to the HTF, we selected (1) tax on 10 percent gasohol 
(abstract 59), (2) diesel fuel tax (abstract 60), and (3) gasoline tax 
(abstract 62). If the return was related to the AATF, we selected (1) 
tax on transportation of persons by air (abstract 26), (2) tax on use 
of international air facilities (abstract 27), and (3) tax on 
transportation of property by air (abstract 28). The tax amounts 
related to the selected abstracts for each trust fund are the largest 
tax amounts reported on the taxpayer's excise tax return and make up 
over 86 percent of the total amount certified to the HTF and over 89 
percent of the total amount certified to the AATF each quarter.



[8] The Collection Certification System produces what IRS refers to as 
"audit files." These audit files contain the individual prorated 
collections, by abstract and taxpayer identification number, that make 
up the certified total amounts for each abstract.



[9] For the purpose of this reconciliation, material is defined as 1 
percent of the total Form 720-related excise tax collections, related 
to the quarters ended December 31, 2002, and March 31, 2003. For fiscal 
year 2003, the materiality amount was $203 million for the two quarters 
combined.



[10] For this sample, if one or no errors were found in testing the 78 
items, we would be 90 percent confident that the error rate in the 
population would not exceed 5 percent.



[11] The selected abstracts include the following: (1) tax on 10 
percent gasohol (abstract 59), (2) diesel fuel tax (abstract 60), (3) 
gasoline tax (abstract 62), (4) tax on transportation of persons by air 
(abstract 26), (5) tax on use of international air facilities (abstract 
27), (6) tax on transportation of property by air (abstract 28), and 
(7) tax on aviation fuel for commercial use (abstract 77). The tax 
amounts for the three HTF-related abstracts make up over 86 percent of 
the total amount certified to the HTF and the tax amounts for the four 
AATF-related abstracts make up over 95 percent of the total amounts 
certified to the AATF each quarter.



[12] The Report of Excise Tax Collection contains prorated collections, 
classified by abstracts that serve as the basis for IRS's quarterly 
trust fund certifications. 



[13] The planned sample size using DUS was 143 items. DUS selects 
dollars versus specific transaction items by dividing the population by 
dollar intervals. The dollar interval for the HTF was $118 million. 
Accordingly, any item with a dollar value matching or exceeding the 
sampling interval would be selected, whereas items less than the 
sampling interval might not be selected. For example, an item of $236 
million would cover two dollar-intervals, but represent one sample 
item. Due to large dollar items covering more than one interval, the 96 
unique sampled transactions selected represent 143 dollar-intervals.



[14] The planned sample size using DUS was 132 items. As explained in 
footnote 13, DUS selects dollars versus specific transaction items by 
dividing the population by dollar intervals. The dollar interval for 
the AATF was $33 million. Due to large dollar items covering more than 
one interval, the 67 unique sampled transactions selected represent 132 
dollar-intervals.



[15] For this sample, if no errors are found in testing the 45 items, 
we would be 90 percent confident that the error rate in the population 
would not exceed 5 percent.



[16] The purpose of this test is to determine whether the Collection 
Certification System prorates correctly. This test is not intended to 
determine whether amounts provided to the system are correct.



[17] IRS prepares two certification letters for HTF each quarter: one 
for the Highway Account and the other for the Mass Transit Account.



[18] The certified amounts for diesel fuel tax (abstract 60), gasoline 
tax (abstract 62), and tax on 10 percent gasohol (abstract 59), along 
with the heavy vehicle use tax (traced separately), make up over 89 
percent of the total amount certified to the HTF each quarter.



[19] IRS uses data from two of these reports, covering sequential 
processing intervals, for each quarterly certification. Collections are 
classified by abstract on the report when the related Form 720 tax 
return has been recorded in IRS's master file during the processing 
interval covered by the report. The second of the two reports used may 
contain collections related to previous quarters not classified by 
abstract until the current quarter because the related return was not 
recorded on the master file until the current quarter.



[20] During fiscal year 2002, IRS changed its process of recording and 
summarizing excise tax credits related to taxpayers' Form 720 tax 
returns. As a result, IRS now obtains this excise tax credit 
information from the applicable Treasury 90 Report. The Treasury 90 
Report summarizes excise tax credit information and is produced 
quarterly by IRS submission processing campus systems. IRS has 10 
submission processing campuses that receive and process tax returns and 
payments.



[21] IRS calculates certified collections to the Highway Account and 
the Mass Transit Account using the total prorated collection amount, 
tax rate, and distribution rates applicable to each account.



[22] For this test, "significant" is defined as $170 million. This 
represents approximately 2 percent of the quarterly total certified to 
the HTF.



[23] Negative amounts occur due to taxpayer amendments or IRS 
corrections.



[24] IRS performs a quarterly reclassification of excise tax refunds 
and credits originally entered into its master file as a personal or 
corporate refund/credit. IRS refers to these reclassifications as 
"refund/credit certifications." These amounts do not represent the 
total excise tax refund/credit activity to the trust funds. Other 
routine excise tax refunds and credits (e.g., overpayments), which are 
claimed on taxpayers' Form 720 excise tax returns, are included in 
IRS's excise tax receipt certification to trust funds.



[25] In order to meet certain reporting deadlines, IRS-certified 
refunds and credits for the fourth quarter of fiscal year 2003 as of 
September 5, 2003.



[26] IRS attaches a separate schedule to the HTF refund/credit 
certification letter that includes the detailed excise tax amounts that 
support the total amount shown on the letter. IRS compiles the amounts 
on these schedules from submission processing campus systems and its 
Interim Revenue Accounting Control System. IRS has 10 submission 
processing campuses that process tax returns and taxpayer receipts. 



[27] The certified refund/credit amounts for gasoline and diesel make 
up at least 91 percent of the total certified refund/credit amount for 
the HTF.



[28] An FMS accountant compiles this schedule, called "Subsidiary 
Quarterly Account of Estimates and Actual Related Taxes Appropriated to 
the Highway Account." The schedule computes the difference between IRS 
certified amounts and the OTA estimate for excise taxes, individually 
and in total, that relate to the Highway Account. A similar schedule is 
prepared for the Mass Transit Account. The schedules, along with OTA 
transfer forms and IRS certifications, support the FMS adjustment.



[29] A positive amount indicates that the FMS adjustment increased 
excise taxes distributed to the trust fund. A negative amount, shown in 
parentheses, indicates that the FMS adjustment decreased excise taxes 
distributed to the trust fund. Since the adjustment amount is the 
difference between OTA's estimate and IRS's certified amount, it may be 
significantly affected by IRS's ability to certify receipts in the 
appropriate quarter.



[30] To accommodate the Department of Transportation's accelerated 
reporting date for fiscal year 2003, IRS performed a precertification 
of excise tax collections for the quarter ended June 30, 2003. The data 
are for information purposes only and the precertification does not 
constitute an official certification.



[31] The amounts for diesel fuel tax (abstract 60), gasoline tax 
(abstract 62), and tax on 10 percent gasohol (abstract 59), along with 
the heavy vehicle use tax (traced separately), make up over 89 percent 
of the precertified total to the HTF.



[32] For this test, "significant" is defined as $170 million, which 
represents approximately 2 percent of the precertified total to the 
HTF.



[33] For this test, "significant returns" are defined as those from 
taxpayers with a total quarterly excise tax liability equal to or 
greater than $10 million during each of the prior four quarters. Tax 
returns related specifically to HTF from taxpayers with liabilities 
equal to or greater than $10 million have, in the aggregate, 
historically accounted for over 91 percent of distributions certified 
to HTF.



[34] OTA makes semimonthly estimates of excise tax collections for 
transfer to trust funds. There are five semimonthly estimates for the 
quarter ended September 30, 2003, which affect fiscal year 2003 
distributions to the HTF.



[35] OTA communicates this information to interested parties at 
Treasury, the Federal Highway Administration, the Federal Transit 
Administration, and the Department of Transportation. IRS uses the tax 
and distribution rates from this table in its subsequent certification 
of collections to trust funds.



[36] The transfer amounts for diesel fuel tax (abstract 60), gasoline 
tax (abstract 62), tax on 10 percent gasohol (abstract 59), and heavy 
vehicle use tax make up over 92 percent of the total amount transferred 
quarterly to the HTF.



[37] The source documents include the IRS report of excise taxes used 
to derive the percentages applied to reported receipts, the Daily 
Treasury Statement, the Monthly Treasury Statement, and the excise tax 
rate tables.



[38] IRS assigns a tax class number to specific types of taxes. Excise 
taxes are tax class 4.

[39] For the purpose of this procedure and procedure VI.B.3, we define 
material as $20 billion. This represents approximately 1 percent of the 
total tax revenue receipts collected by IRS in fiscal year 2003.

[40] IRS maintains records of refund balances by tax class in its 
master file and reports this information monthly to Treasury on the 
SF-224. Treasury provides IRS with a Statement of Differences 
(TFS-6652), which reports differences between total refunds reported 
by IRS on the SF-224 and the total refunds per Treasury records.