The Tax Fraud Hotline Has Not Been an Effective Source for Criminal Tax Investigations
September 2003
Reference Number: 2003-10-210
This report has cleared the Treasury
Inspector General for Tax Administration disclosure review process and
information determined to be restricted from public release has been redacted
from this document.
September
26, 2003
MEMORANDUM FOR
CHIEF, CRIMINAL INVESTIGATION
FROM: Gordon C. Milbourn III /s/ Gordon C.
Milbourn III
Assistant Inspector General for Audit (Small Business and
Corporate Programs)
SUBJECT: Final Audit Report - The Tax Fraud
Hotline Has Not Been an Effective Source for Criminal Tax Investigations
(Audit # 200210042)
This
report presents the results of our review of the Criminal Investigation (CI) function’s
use of information submitted by informants via the Internal Revenue Service’s
(IRS) Tax Fraud Hotline (referred to as the “hotline”). The overall objective of this review was to
determine whether the CI function effectively considers hotline information as
a source for identifying tax investigations.
The
toll-free telephone hotline number is a prominent part of the CI function’s
outreach to the public; press releases and publications often contain a
reference to the telephone number and encourage its use. While only a portion of the information
reported via the hotline warrants consideration by the CI function, we estimate
that each year the call sites forward to the CI function several thousand
hotline allegations that meet the referral criteria.
In
summary, the CI function did not treat the hotline as a viable source for tax
investigations because often the information captured was limited in detail and
the informants were anonymous.
Therefore, the Lead Development Centers (LDC) gave only limited scrutiny
to hotline leads and rarely assigned hotline allegations to CI field agents for
investigation. In our opinion, hotline
leads can reveal allegations of significant tax violations and thus should be
given consideration comparable to other lead sources.
In
addition, the CI function did not provide specific guidance to ensure that
allegations of refund fraud or questionable return preparer activity reported
to the hotline were used by the Fraud Detection Centers (FDC) to prevent the
issuance of potentially fraudulent refunds.
In our opinion, hotline leads can reveal relevant refund fraud
allegations that could contribute to the FDCs’ fraud detection efforts.
We
recommended that the CI function’s Director, Strategy, revise procedures to
specify that LDC managers prioritize research of informant allegations based on
their relative potential, without giving routine preference to field office
requests, and proactively assess the adequacy of the preparation of forms that
refer hotline information to the LDCs.
To improve the investigative potential of hotline allegations, we
recommended guidance for documenting allegations that meet CI referral criteria
be revised to emphasize that IRS personnel should encourage informants to
identify themselves and allow for subsequent contact by CI function personnel
when necessary. As a supplement to the
toll-free hotline, limited use of Internet electronic mail (e-mail) or web page
fill-in forms could be considered to directly collect informant
communications. To evaluate the
benefits of the hotline, the number and results of investigations should be
tracked. Regarding refund fraud
allegations, we recommended that the Director, Refund Crimes, develop
procedures that establish specific criteria for and guide the referral of
appropriate hotline information to the FDCs.
We also recommended that procedures be revised to specify that FDC
managers proactively assess the adequacy of the preparation of forms that refer
hotline information to the FDCs.
Management’s Response: CI function
management agreed with some of our recommendations. In general, CI function management believes that the hotline call
sites provide only a limited number of informant communications that meet
criminal criteria and establish the basis for initiating a criminal
investigation. CI function management
agreed that they have a shared responsibility to provide instruction and
guidance on potential informant communications that meet criminal criteria and
they will support the Wage and Investment Division, which is the business owner
of the hotline call site function, by providing expertise in interview
techniques and training of call site personnel. CI function management also agreed to update the instructions and
guidance for FDC interaction with the hotline call sites and the proper
handling of informant communications that potentially meet criminal criteria
for refund fraud or questionable return preparer activity.
However,
CI function management did not agree that corrective action was required
regarding the LDC procedures for prioritizing research assignments, that the
Internet e-mail or a web site would be useful in collecting informant data, or
that there would be a benefit to tracking the number and results of
investigations initiated from the hotline and other public outreach
methods. In addition, CI function
management did not agree with the calculated potential savings presented in
Appendix IV. Management’s complete
response to the draft report is included as Appendix V.
Office
of Audit Comment: Since we believe
that hotline leads can reveal allegations of significant tax violations, we do
not agree with the CI function management response that corrective actions are
not warranted for some recommendations.
Unless procedures are revised to encourage a more balanced consideration
of informant allegations received via the telephone hotline, we believe that
field office informant leads will continue to receive routine preference. In our opinion, the limited use of automated
methods for collecting information would be a forward-looking approach to
improving the effectiveness of public informants as a source for tax
investigations. The absence of specific
management information system codes for the various public informant sources
hinders an objective assessment of the investigative value of hotline
allegations. Where appropriate, we have
included in the report our comments related to management’s response. While we still believe our recommendations
are worthwhile, we do not intend to elevate our disagreement concerning these
matters to the Department of the Treasury for resolution.
While
CI function management agreed with the related recommendation, they did not
agree with the calculated potential savings presented in Appendix IV of our draft
audit report. CI function management
responded that the CI Refund Crimes Section conducted an analysis of the 28
accounts where we identified that the absence of account controls allowed
refunds totaling $93,480 to be issued after the date allegations were reported
to the hotline. The CI function
analysis proved that 2 of the 28 alleged fraudulent refunds totaling $6,030
were false. Since the results of the CI
function’s analysis of all accounts, which was conducted during the draft
report response period, proved that only two refunds were false, we adjusted
the potential savings presented in Appendix IV to reflect the potential benefit
relating to only those two accounts.
Copies
of this report are also being sent to the IRS managers who are affected by the
report recommendations. Please contact
me at (202) 622-6510 if you have questions or Daniel R. Devlin, Assistant
Inspector General for Audit (Headquarters Operations and Exempt Organizations
Programs), at (202) 622-8500.
Appendix I – Detailed Objective, Scope, and Methodology
Appendix II – Major Contributors to This Report
Appendix III – Report Distribution List
Appendix IV – Outcome Measures
Appendix V – Management’s Response to the Draft Report
The Internal Revenue Service’s (IRS) Criminal Investigation (CI) function conducts investigations of potential criminal tax violations within three interdependent categories: legal source tax crimes, illegal source financial crimes, and narcotics-related financial crimes. These investigations all should have a positive effect on tax compliance.
One of the CI function’s strategies since the issuance of the Webster Report in 1999 has been to increase the number of legal source tax cases that it investigates and recommends for prosecution. The report emphasized that the CI function needed to refocus its resources towards potential criminal violations affecting the public’s confidence in the tax system and compliance with the tax laws.
Because noncompliance with tax laws cannot easily be identified solely from tax returns and documents filed with the IRS, the CI function must often obtain information from outside sources to generate leads that may result in the initiation of a tax-related investigation. Accordingly, the CI management information system shows that the CI function often receives information from outside sources that indicates a person is violating the tax laws. While about 39 percent of 3,565 legal source Primary Investigations approved during Fiscal Year (FY) 2002 originated from referrals developed within the IRS, about 30 percent originated from information supplied by general public sources, such as informants.
CI function field employees commonly gather information from
informants during the course of their investigative duties. Typically, if a CI group supervisor elects
to evaluate the information in the field, a Primary Investigation will be
initiated. The IRS also has a specific
toll-free telephone number that provides individuals with a way to report
information regarding alleged tax violations and unreported income. The toll-free tax fraud telephone number
(referred to as the “hotline”) is a prominent part of the CI function’s
outreach to the public; press releases and publications often contain a
reference to the telephone number and encourage its use.
During FY 2002, IRS assistors answered over 42 million calls across all toll-free numbers combined. Approximately 282,000 of these calls were answered by the Atlanta, Georgia; Austin, Texas; Fresno, California; and Philadelphia, Pennsylvania, call sites on the hotline number.
Call site Customer Service Representatives (CSR) record the
information supplied by informants on an Information Report Referral (Form
3949). CSRs should attempt to obtain
all the information to complete the Form 3949, but it is particularly important
to include the alleged tax violator’s name and address, the Social Security
Number if known, the tax years, and the specific nature of the violation. The dollar amounts of unreported income or
erroneous deductions are also important.
If the caller lacks specific information to describe the alleged
violation or cannot identify the violator, a Form 3949 cannot be completed.
Call site personnel screen the Forms 3949 to determine if they should be forwarded to the CI function as a lead for a criminal case or to one of the other IRS functions for Examination, Collection, or other civil consideration. The IRS does not maintain statistics to show how many Forms 3949 were completed, in what proportions they were distributed to the various IRS functions, or how the CI function’s Lead Development Centers (LDC) disposed of the ones they received. However, call site studies conducted in Calendar Year 2000 indicated most were forwarded for Examination, not CI, consideration.
We conducted this audit between November 2002 and July 2003 and contacted CI function personnel at the National Headquarters in Washington, D.C., at the 12 nationwide CI function LDCs, and at the Atlanta, Austin, and Fresno CI Fraud Detection Centers (FDC). We contacted personnel at the 3 Wage and Investment (W&I) Division hotline call sites and obtained copies of Forms 3949 prepared during 1 week’s time from each site. We also made an on-site visit to the Austin LDC and call site. The audit was conducted in accordance with Government Auditing Standards. Detailed information on our audit objective, scope, and methodology is presented in Appendix I. Major contributors to the report are listed in Appendix II.
CI function personnel advised us that the hotline process was not effective because often the information captured was limited in detail and the informants were anonymous, which does not allow CI function personnel an opportunity to gather additional details from the informant. CI function personnel also advised us that the call sites sometimes refer information that does not meet CI criteria. Therefore, the LDCs rarely assigned hotline allegations to CI function field agents for investigation. However, our review indicates that hotline leads can reveal significant fraud allegations and thus should be given consideration comparable to leads from other sources.
Hotline leads often received only limited scrutiny even
though significant allegations were made
Internal Revenue Manual (IRM) screening criteria are designed so that only significant allegations, involving greater than $50,000 per year in unreported income, are forwarded to the LDCs for consideration. An LDC manager can select for development and analysis those leads that are consistent with the CI function’s plans and strategies and appear to have some criminal potential. The LDC can subsequently decide that a lead warrants assignment to a field agent for investigation or that it should be forwarded to a civil function for consideration.
In describing the general process conducted when screening
Forms 3949, several LDC personnel outlined a set of characteristics they would
consider when deciding if a lead would be developed, such as the location of
the subject, the number of tax years involved, the amount of tax effect, and
the reasonableness of and the ability to prove or disprove the allegation.
At our request, LDC managers annotated their disposition of 121 Forms 3949 we identified in our 1-week sample that met the CI referral criteria. The LDCs assigned case numbers to 8 allegations selected for research and indicated that another 10 might be researched in the future. The LDCs indicated they forwarded the remaining Forms 3949 to the Compliance functions; 78 without conducting any research and 25 after doing limited research.
We recognize such an assessment draws on the professional experience of CI function personnel and is facilitated if an informant provides detailed and relevant information. However, five LDCs noted that workload priorities affected their ability to select many, if any, Forms 3949 for research or noted that they declined to research the Forms 3949 with simple statements that the allegations were not specific enough or were anonymous.
The following examples, each of which was from an anonymous
informant, illustrate the context of some of the forms that the LDCs forwarded to
the Examination function without conducting research or after conducting
limited research. While the substance
of allegations forwarded to the CI function varied, we believe these are
examples of allegations that merited further research:
· A company president allegedly had underreported income by $400,000 by paying personal expenses from company funds. Comments from the informant noted on the Form 3949 indicated the location of accounting information within the company office. The informant indicated he or she learned of the violation through his or her work. The LDC noted that, although superficially the allegation had merit, it did not contain specifics that would justify conducting research. Our research of IRS account information showed that the corporation’s latest tax return reported over $9 million in gross receipts.
· A real estate broker allegedly had not reported commission income of $500,000 per year since 1997. Comments from the informant noted on the Form 3949 indicated that commissions were split into multiple small payments. The LDC noted that the information was vague. Our research of IRS account information showed that the individual had not filed a tax return in 4 of the past 5 years and had reported about $30,000 in self-employed income in the year for which a return has been filed.
· A personal services business owner allegedly collected $70,000 per month but reported only $20,000 per month, allegedly sending cash for investment overseas. Three business locations were noted on the Form 3949. The LDC noted that it could not handle such information due to other programs. Our research of IRS account information showed that the individual’s latest tax return reported about $122,000 in gross business receipts.
· An owner or partner of 2 manufacturing businesses allegedly had underreported income by $3 million over 7 years by expensing the same equipment on both businesses’ tax returns. Comments from the informant noted on the Form 3949 indicated that the bookkeeper quit after refusing to participate in forging the records and that alleged bankruptcy fraud had also occurred. The LDC noted that the allegation appeared false. Our research of IRS account information showed that each business entity had in excess of $1 million in assets noted on its most recent tax returns.
In our opinion, the CI function may have disregarded allegations which, after additional research, suggested that criminal tax activity may have occurred. The CI function would have to apply additional investigative techniques, starting with detailed LDC research, to fully evaluate the merits of these allegations.
CI function field investigations rarely began from hotline
leads, even though other leads investigated were not always persuasive
During FY 2002, the CI function approved and assigned to CI function field agents 690 Primary Investigations that had a General Public source code. Based on a review of a statistical sample of 90 of the 690 leads, we estimate that no more than 47 were based on information received via the hotline. In contrast, we estimate the CI function received several thousand Forms 3949 during the year from hotline calls that met CI referral criteria.
The leads that originate from information received by CI
function field agents are not necessarily any more persuasive than some leads
that originate from hotline calls. We
reviewed the source documents for the sample leads assigned to CI function
field agents where the lead emanated from the general public. We determined that three leads originated
from a hotline call. For the remaining
87 leads that emanated from the field, we determined the following:
· For 31 leads where information about the informant’s identity was shown, 6 listed the informant as being anonymous, even though some CI function managers consider anonymous informants as a barrier to selecting hotline leads for assignment to CI function field agents.
· For 30 leads where the estimated unreported income amount was shown, 4 involved amounts that were less than the $50,000 per year minimum that applies to hotline leads during screening.
· Some allegations were described in non-specific terms, such as unreported income from selling stolen lumber, unreported income from musical performances, or employment tax noncompliance involving 25 employees.
We realize that the lack of an informant identity, the
amount of unreported income, or the description of an allegation on a source
document may not reflect all that a CI function field office knew regarding the
potential for a criminal case. However,
because the leads from other sources may not always be more persuasive than
leads from the hotline, we believe the CI function needs to give comparable
consideration to hotline leads rather than giving preferential treatment to
leads from other sources.
Also, even though the CI function has concerns regarding the effectiveness of the hotline process, it nonetheless widely communicates to the public that the toll-free hotline number is a way to report suspected tax fraud and is thus an important component of tax compliance. Therefore, the CI function needs to enhance the viability of the hotline by taking appropriate actions on hotline calls.
The CI function is not required to, and generally has not, offered electronic mail (e-mail) or Internet-based web pages as alternative methods for the general public to report suspected fraud. The CI function’s Office of Refund Crimes does offer an e-mail address to tax practitioners as an option for reporting refund fraud. Other IRS functions have used e-mail to key in on high-interest topics. For example, the public can submit information relating to potentially improper tax shelter activity to the Office of Tax Shelter Analysis using a specific IRS-provided e-mail address.
In conclusion, allegations reported on the hotline,
which the IRS promotes as the primary way for the public to communicate this
type of information, are often dismissed without any data research that might
tend to support or disprove the allegation.
In our opinion, the CI function has not effectively balanced the
consideration of information provided by informants via the hotline with
similar information received directly by CI function field personnel.
The CI function’s Director, Strategy, should:
1. Revise LDC procedures to specify that LDC managers prioritize research assignments of all informant allegations based on their relative potential, without giving routine preference to research requests solely because CI function field offices make them.
Management’s Response: CI function management responded that no specific corrective actions are required at this time because LDC managers already perform this function as they review informant communications that come from various sources. CI function management also indicated that this recommendation may be affected by the pending redesign for the LDCs.
Office of Audit Comment: We made this recommendation to encourage a more balanced consideration of informant allegations received via the telephone hotline. In our opinion, the CI function’s propensity to select field office informant leads for investigation in essence resulted in the LDCs prioritizing hotline leads only to the extent necessary to supplement field office inventories and usually not in a manner that utilized research to clarify on a broader scope whether hotline allegations might have had significant investigative potential. Unless procedures are revised to encourage similar consideration, we believe that field office research requests will continue to receive routine preference. CI function management did not specify how the pending redesign for the LDCs might affect this recommendation.
2.
Coordinate
as necessary with the applicable operating division functions to revise the IRM
guidance for documenting the receipt of informant allegations to emphasize
that, when the CI referral criteria are met, IRS personnel should be persistent
when encouraging the informant to give his or her name and telephone number, to
allow a subsequent contact in case CI function personnel want to obtain
additional information. The CI function
should revise the guidance to clearly illustrate the
appropriate manner in which CSRs should invite an informant’s continued
participation.
Management’s Response: CI function management agreed that they have a shared
responsibility to provide instruction and guidance on potential informant
communications that meet criminal criteria.
The W&I Division, as the business owner of the hotline call site
function, will assume the lead role in reviewing and revising hotline call site
procedures and IRM guidance. The CI
function will support the W&I Division by providing expertise in interview
techniques and training to call site personnel regarding contacts with
informants.
3.
Revise LDC procedures to specify that LDC managers
co-located with the call sites should be proactive in assessing the adequacy of
the preparation of Forms 3949 by call site personnel and in providing specific
feedback when necessary to improve trends in completing the Forms.
Management’s Response: The W&I Division, as the business owner of the hotline call
site function, will assume the lead role in assuring the adequate preparation
of informant communication forms. The
CI function will support the W&I Division by providing training to call
site personnel on techniques for soliciting high quality and relevant
information from informants. CI
function management also indicated that this recommendation may be affected by
the pending redesign for the LDCs.
4.
Consider
whether the limited use of Internet e-mail or web page fill-in forms would be
an effective supplement to the toll-free hotline in directly collecting
informant communications for specific high-interest topics. For example, an e-mail address could be
promoted as a way to report alleged promoters of the Reparations Credit
scheme. This would also serve to
consolidate such information for the CI function’s
consideration.
Management’s Response: CI function management did not agree that the use of e-mail or an Internet web site would be useful in collecting informant data. CI function management believes that the toll-free hotline call sites are at present the best way to ensure a dialogue between the informant and IRS personnel that provides an opportunity to obtain clarifying information as well as obtain the identity of the informant for follow-up purposes. CI function management believes that concerns that were discussed in a prior call site study regarding the security of taxpayer information in the possible use of an e-mail address for informant communications are still applicable. CI function management also cites the call site study’s conclusion that electronic collection systems would require additional resources to review and analyze the informant communications received.
Office of Audit Comment: This recommendation does not suggest the replacement of the
hotline call site process with an electronic system. We believe that e-mail or Internet access should be considered as
a supplement for topics of high interest where the CI function desires
information from the public. In that
context, we would anticipate that offering an additional method of
communication would only require similar CI function resources as would be necessary
to review such information if it was recorded by the call sites on Forms 3949
and routed to the CI function. Since
the Office of Refund Crimes already offers an e-mail address to tax
practitioners as an option for reporting refund fraud, we believe that security
issues can be adequately addressed. In
our opinion, the limited use of automated methods for collecting information
would be a forward-looking approach to improving the effectiveness of public
informants as a source for tax investigations.
5.
Evaluate
the benefits of the hotline and other public outreach methods by tracking the
number and results of investigations by establishing more definitive source
codes within the CI function’s management information system for the various
General Public sources of information.
Management’s Response: CI function management does not perceive any tangible benefits from this recommendation.
Office of Audit Comment: Having only a single management information system code for the various General Public sources of information does not
allow the CI function to evaluate the comparative volumes or results of the
investigations that originate from allegations reported to the hotline,
gathered by field agents, submitted in written correspondence, or received from
any other particular public informant source.
In our opinion, this hinders an objective assessment of the
investigative value of hotline allegations, or any of the other public sources,
derived from data maintained in the management information system.
The CI function had not provided specific guidance to the call sites regarding the screening of Forms 3949 for refund fraud or questionable return preparer allegations. Our review showed that hotline leads can reveal relevant refund fraud allegations that, absent specific guidance, would likely be forwarded to the Examination function rather than be brought to the immediate attention of a CI FDC.
Hotline screening has not addressed refund fraud issues,
even though that is a prominent CI program
The CI function provided informal guidance on a local basis to the call sites regarding refund fraud issues, but that did not result in those allegations being promptly or regularly forwarded to the FDCs. The FDC Questionable Refund Program (QRP) and Return Preparer Program (RPP) focus on refund returns with a broader goal of identifying more extensive refund fraud schemes.
Refund fraud allegations tended to be overlooked for CI
function purposes in the Form 3949 screening process because the IRM criteria
for forwarding forms to the CI function focus on a high-dollar tax effect
(i.e., more than $50,000 per year in unreported income). That would usually exclude information that
questions individual tax return refunds, which are typically only a few
thousand dollars each. The FDC managers
advised us that some QRP and RPP leads are received from the LDCs, but they are
not normally received from call sites or the Examination function in large
numbers or on a regular basis.
Due to the relevance of this condition during the Tax Year 2002 Filing Season, we notified the CI function’s Director, Refund Crimes, of this observation on February 28, 2003, via e-mail. As a result, some FDC managers discussed with their call site counterparts what actions were needed to guide the referral of refund fraud information to the FDCs. Some FDC managers expressed concerns, however, that a change in procedures could cause an unmanageable volume of Forms 3949 to be forwarded to the FDCs.
Some FDC managers also suggested that the Form 3949 could be revised to distinguish between general criminal activity and refund fraud activity for referral purposes, and that refund fraud referrals should be forwarded directly to the FDCs without passing through an LDC. We believe these initial, varied responses indicate that the CI function needs a more cohesive approach to address this issue. This approach, through national guidance, would make more effective use of potential refund fraud information received by the hotline.
Hotline leads can reveal potential refund fraud that has not
yet been identified through other efforts
The CI function’s QRP and RPP are designed to detect
potentially fraudulent refund returns and prevent the issuance of those
refunds; therefore, it is imperative that relevant hotline allegations are
promptly routed to the CI function.
Of the 1,312 Forms 3949 that the call sites prepared during our 1-week sample, 94 contained information that alleged some type of refund or return preparer fraud:
Ø Preparer scheme – 26.
Ø False tax withholding forms – 25.
Ø Identity theft – 21.
Ø Misuses of Social Security Number – 21.
Ø Reparations Credit scheme claim – 1.
The call sites designated only 18 of these 94 Forms 3949 for
forwarding to the LDCs and marked the other 76 for forwarding to the
Examination or other functions.
We determined that the questionable refund situation had already been identified for some of the alleged violators, but that other individuals had not been identified:
· The CI function had already established account controls on 10 of the taxpayers prior to receipt of the hotline call, which prevented $32,071 from being potentially refunded in error. This suggests that these hotline calls had some merit, since the CI function had already identified a questionable refund situation on these taxpayers.
· In contrast, the absence of account controls on 28 of the taxpayers allowed potentially fraudulent refunds totaling $93,480 to be issued after the date that the allegation was reported to the hotline.
· During the 2003 processing year, 12 tax return preparers that were named in hotline allegations prepared a total of 8,770 tax returns filed as of May 19, 2003. Each preparer had at least an 89 percent refund return rate, and 9 preparers exceeded a 97 percent refund rate.
The QRP and RPP detect large numbers of questionable refunds and schemes. However, in our opinion, the CI function could have neither proved nor disproved allegations of refund fraud on some of the Forms 3949 in our sample, without applying investigative techniques consistent with steps taken on other QRP or RPP situations. The following examples, which were forwarded to the Examination rather than CI function, illustrate the context of some of the allegations reported by either an informant or a tax preparer:
· A taxpayer allegedly attempted to file a tax return with a suspicious Wage and Tax Statement (Form W-2) and dependents with nonidentical names, which were specifically identified in the allegation. Our research of IRS account information showed that a tax return, which claimed 1 of the 2 dependents noted on the Form 3949, resulted in a refund of an Earned Income Tax Credit (EITC), in excess of $3,500, issued 21 days after the hotline call was received.
· A tax preparer allegedly instructed a taxpayer to obtain a child’s Social Security Number to increase his or her refund amount. Our research of IRS account information showed that a tax return claiming a dependent resulted in a refund of an EITC, in excess of $1,900, issued 35 days after the hotline call was received. During the 2003 processing year, the return preparer involved had prepared more than 1,300 tax returns filed as of May 19, 2003.
· A disabled taxpayer suspected he or she was a victim of identity theft regarding the filing of tax returns under his or her Social Security Number. Our research of IRS account information showed that a tax return was filed using the taxpayer’s Social Security Number, reporting business income and claiming 2 dependents, and resulted in a refund of an EITC in excess of $2,500, issued 19 days after the hotline call was received. In the prior year, a nearly identical amount of business income was reported under this same Social Security Number, but two different dependents were claimed.
In conclusion, because the CI function did not provide specific guidance, the call sites forwarded most Forms 3949 of this type for Examination consideration, which would not contribute to the FDCs’ efforts to promptly identify questionable returns and prevent the issuance of potentially fraudulent refunds. In our opinion, it is imperative that criteria and guidance be established on a national level to ensure that the referral of Forms 3949 from the hotline call sites to the FDCs becomes an established routine designed to capture all Forms relevant for the FDCs.
The CI function’s Director, Refund Crimes, should:
6.
Develop national procedures that establish specific
criteria for and guide the referral of appropriate QRP and RPP allegations from
the hotline directly to the FDCs.
Management’s Response: CI function management agreed to update the instructions and guidance for the proper handling of informant communications that potentially meet criminal criteria for QRP and RPP allegations.
Office of Audit Comment: While CI function management agreed with the recommendation relating to this finding, they did not agree with the calculated potential savings presented in Appendix IV. CI function management responded that the CI Refund Crimes Section conducted an analysis of the 28 accounts where we identified that the absence of account controls allowed refunds totaling $93,480 to be issued after the date allegations were reported to the hotline. The CI function analysis proved that 2 of the 28 alleged fraudulent refunds totaling $6,030 were false. Since the results of the CI function’s analysis of all accounts proved that only two refunds were false, we adjusted the Outcome Measure presented in Appendix IV to reflect the potential benefit relating to only those two accounts.
7.
Revise the FDC IRM guidance to specify that FDC managers
co-located with the call sites should be proactive in assessing the adequacy of
the preparation and referral of Forms 3949 by call site personnel for QRP and
RPP purposes and in providing feedback when necessary to improve trends in
Forms.
Management’s Response: CI function management agreed to revise the instructions and guidance relevant to the FDC interaction with the hotline call sites. The FDC will conduct regular reviews and report findings as needed to hotline call site management.
Appendix I
Detailed Objective, Scope,
and Methodology
The overall objective of this review was to determine whether the Criminal Investigation (CI) function effectively considers information submitted by informants via the Internal Revenue Service’s (IRS) Tax Fraud Hotline (referred to as the “hotline”) as a source for identifying tax investigations. We conducted the following steps to accomplish this objective:
I.
Determined whether the CI
function provided sufficient guidance to the hotline call sites to aid in their
screening and distribution of information among the IRS Compliance functions
and ensured that items with criminal potential were forwarded to the CI
function for consideration.
A.
Contacted IRS personnel
regarding the processing of information received via the toll-free hotline at
the CI National Headquarters offices of the Director, Strategy, and the
Director, Refund Crimes, in Washington, D.C.; the Wage and Investment Division
call sites at the Atlanta, Austin, and Fresno locations; the 12 nationwide CI
Lead Development Centers (LDC); and the CI Fraud Detection Centers (FDC) at the
Atlanta, Austin, and Fresno locations.
We made an on-site visit to the Austin LDC and call site.
B. Reviewed copies of the population of 1,312 Information
Report Referrals (Form 3949) prepared by the 3 call sites during 1 week (Austin
and Fresno for the week of January 13-17, 2003, and Atlanta for the week of
February 3-7, 2003) to determine volumes, distribution patterns, and whether
the allegations met the CI function’s referral criteria of at least $50,000 per
year in unreported income. In cases
where a dollar amount was not estimated, we determined whether the allegations
appeared to have criminal potential. A
total of 225 of the 1,312 Forms 3949 were marked for forwarding to the CI
function; we believe 140 of the 225 clearly met the CI referral criteria. We judgmentally selected Forms 3949 prepared
during 1 week’s time at each call site because the population for Forms 3949
prepared during any certain period could not be determined or retrieved. The IRS does not record or track individual
Forms 3949 or maintain overall statistics regarding the preparation of Forms
3949 by the hotline call sites. A
sample period greater than 1 week was not feasible because of the relatively
large volume of Forms 3949 prepared.
Sampling for an equal duration from each hotline call site was made to
allow equal representation from each location.
II.
Determined whether the CI
function applied reasonable and appropriate lead development or evaluative
techniques to information referrals received from the hotline call sites.
A.
Obtained a description from
10 of the 12 LDCs of the methods they use in screening Forms 3949 received from
the call sites.
B.
Distributed to the 12 LDCs
(based on the geographical location of the alleged violator) copies of the 129
Forms 3949 meeting CI referral criteria that we obtained in step I.B. above
which were marked for forwarding to the CI function by the Austin and Fresno
call sites. CI management
provided us with their disposition decision on 121 of these 129 Forms. We did not distribute 11 Forms from 1 of the
call sites because they arrived in our office subsequent to the distribution of
the 129 Forms to the LDCs. We also did
not receive a response from one of the LDCs for eight Forms originally distributed.
C.
Distributed to the 10 FDCs
(based on the geographical location of the alleged violator) copies of 90 of
the 94 Forms 3949 obtained in step I.B. above that contained refund fraud type
allegations and researched IRS tax account information, where appropriate, to
identify Tax Year 2002 refunds issued, refunds prevented, or tax returns
prepared for the parties reported on the Forms 3949.
III.
Determined whether the CI
function’s management information system was effective in providing data that
allowed management to evaluate the program results attributable to hotline
information referrals.
A.
Interviewed applicable CI
function personnel and analyzed the CI function’s management information system
to determine if data are maintained on investigations initiated from the
hotline that would allow for the tracking of results.
B. Reviewed source documents for a statistically valid sample
of 110 General Public source code Primary Investigations approved during Fiscal
Year 2002 that were assigned to CI function field agents for investigation, to
estimate the number that were initiated based on an allegation received via the
hotline. The sample size was based on a
population of 690 case records meeting the above criteria extracted from the CI
function’s management information system, a 95 percent confidence level, an
estimated occurrence rate of 3 percent, a precision level of +/-3 percent, and
an additional 5 percent oversampling for contingency purposes. The attribute of whether the case was
initiated based on a hotline lead was determinable for 90 of the 110 sample
cases; 3 cases were initiated from a hotline lead. Using the 95 percent confidence level, the sample occurrence rate
of 3.33 percent (3/90) calculates to a range estimate of between 3 and 47
hotline cases in the population.
Appendix II
Major Contributors to This Report
Daniel R. Devlin, Assistant Inspector General
for Audit (Headquarters Operations and Exempt Organizations Programs)
John R. Wright, Director
Ronald
F. Koperniak, Audit Manager
Diana M. Tengesdal, Audit Manager
Timothy A. Chriest, Senior Auditor
Jeffrey K. Jones, Senior
Auditor
Joseph P. Snyder, Senior Auditor
Appendix III
Commissioner C
Deputy Commissioner for Services and Enforcement SE
Commissioner, Wage
and Investment Division SE:W
Director, Customer
Account Services, Wage and Investment Division
SE:W:CAS
Director, Refund Crimes, Criminal Investigation SE:CI:RC
Director, Strategy, Criminal Investigation SE:CI:S
Chief Counsel CC
National
Taxpayer Advocate TA
Director, Office of Legislative Affairs CL:LA
Director, Office of
Program Evaluation and Risk Analysis
RAS:O
Office of Management Controls OS:CFO:AR:M
Audit Liaisons:
GAO/TIGTA Liaison, Planning and Strategy, Criminal Investigation SE:CI:S:PS
GAO/TIGTA Liaison, Wage and Investment Division SE:W:S:PA
Appendix IV
This appendix presents detailed information on the
measurable impact that our recommended corrective actions will have on tax
administration. These benefits will be
incorporated into our Semiannual Report to the Congress.
Type and Value of Outcome Measure:
· Cost Savings (Funds Put to Better Use) – Potential; $2,678 (see page 11).
· Revenue Protection – Potential; $3,352 (see page 11).
Methodology Used to Measure the Reported Benefits:
We reviewed a judgmental sample of 1,312 Information Report Referrals (Form 3949) prepared by the Customer Account Services call sites in Austin and Fresno for the week of January 13-17, 2003, and in Atlanta for the week of February 3-7, 2003. We identified 94 Forms 3949 alleging 26 preparer schemes, 25 false tax withholding forms, 21 identity thefts, 21 misuses of Social Security Numbers, and 1 Reparations Credit scheme claim. These Forms 3949 were not forwarded directly to the Criminal Investigation (CI) function’s Fraud Detection Centers to prevent the issuance of questionable refunds until the appropriateness of the claim for refund could be validated. Our research of tax account information for individuals identified on these 94 Forms 3949 indicated that the absence of account controls on 28 taxpayer accounts allowed refunds totaling $93,480 to be issued after the date the allegations were reported to the hotline. In response to this finding, the CI Refund Crimes Section conducted an analysis of the 28 accounts and proved that 2 of the 28 alleged fraudulent refunds totaling $6,030 were false. Since the results of the CI function’s analysis of all accounts, which was conducted during the draft report response period, proved that only two refunds were false, we adjusted the reported value to reflect the potential benefit relating to only those two accounts. The Cost Savings portion of the Outcome Measure ($2,678) recognizes that the prevention of erroneous payments of the Earned Income Tax Credit and the Child Tax Credit represent not the protection of tax revenues but rather the protection of budgeted expenditures from appropriated funds. The Revenue Protection portion of the Outcome Measure ($3,352) relates to the prevention of erroneous refunds or efforts to defraud the tax system not attributed to those 2 refundable credits. The outcome measure is limited to the actual number of transactions identified during the testing.
Appendix V
Management’s Response to
the Draft Report
The response
was removed due to its size. To see the
response, please go to the Adobe PDF version of the report on the TIGTA Public
Web Page.