Money Laundering: FinCEN's Law Enforcement Support Role Is Evolving
(Letter Report, 06/19/98, GAO/GGD-98-117).

Pursuant to a congressional request, GAO reviewed the Financial Crimes
Enforcement Network's (FinCEN) products and services in support of law
enforcement, focusing on: (1) trends in the types and quantities of
products and services provided by FinCEN to the law enforcement
community; (2) the extent to which FinCEN's products and services have
been considered useful by the law enforcement community in identifying,
developing, or prosecuting money laundering and other financial crime
cases; (3) the extent to which FinCEN evaluates the states' compliance
with applicable controls over access to and use of information when
state law enforcement officials directly access FinCEN's resources; and
(4) FinCEN's efforts to provide Internal Revenue Service (IRS) Form 8300
information to the law enforcement community.

GAO noted that: (1) FinCEN has expanded the types of products and
services it provides to the law enforcement community; (2) in recent
years, FinCEN has issued fewer tactical, strategic, and artificial
intelligence products and has encouraged, trained, and increasingly
relied on federal agencies to use the platform concept and state and
local agencies to use Project Gateway to support cases that do not
require FinCEN's expertise; (3) according to FinCEN, one significant
reason for the decrease in the number of products issued was that its
staffing levels have remained fairly constant over the years, while its
overall mission has expanded to include: (a) responsibility for
promulgating Bank Secrecy Act regulations; and (b) a leadership role in
international efforts to combat money laundering; (4) respondents to
surveys indicated that FinCEN's tactical products have been useful and
assisted law enforcement investigations in ways not previously
identified; (5) officials that had requested few or no tactical products
from FinCEN generally did not indicate dissatisfaction with it; (6) some
respondents were not aware of the various products and services offered,
and FinCEN has neither developed nor widely disseminated general
criteria or guidelines on when law enforcement officials should request
its support; (7) according to federal and state officials, the platform
concept and Project Gateway are useful tools for helping agencies combat
money laundering and other financial crimes; (8) FinCEN data show that
an increasing number of federal, state, and local agencies are using
these self-help mechanisms to support their investigations; (9) while
Project Gateway is designed to enhance the capabilities of state and
local law enforcement agencies, this technological advancement increases
the potential risk that sensitive information could be inappropriately
accessed, used, or disclosed; (10) although FinCEN has established
policies and procedures designed to limit access to and use of
information obtained through the Gateway system, it has not evaluated
the states' compliance with these controls; (11) in an effort to enhance
its investigative support, FinCEN is seeking IRS approval to provide IRS
Form 8300 information to law enforcement officials; and (12) however,
several issues must still be resolved before FinCEN can obtain approval
from IRS to disseminate this information.

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

 REPORTNUM:  GGD-98-117
     TITLE:  Money Laundering: FinCEN's Law Enforcement Support Role Is 
             Evolving
      DATE:  06/19/98
   SUBJECT:  Law enforcement agencies
             Intergovernmental relations
             Law enforcement information systems
             Internal controls
             Banking law
             Financial records
             Money laundering
             Law enforcement
             Confidential communication
IDENTIFIER:  Treasury Project Gateway
             ONDCP High Intensity Drug Trafficking Area Program
             
******************************************************************
** This file contains an ASCII representation of the text of a  **
** GAO report.  Delineations within the text indicating chapter **
** titles, headings, and bullets are preserved.  Major          **
** divisions and subdivisions of the text, such as Chapters,    **
** Sections, and Appendixes, are identified by double and       **
** single lines.  The numbers on the right end of these lines   **
** indicate the position of each of the subsections in the      **
** document outline.  These numbers do NOT correspond with the  **
** page numbers of the printed product.                         **
**                                                              **
** No attempt has been made to display graphic images, although **
** figure captions are reproduced.  Tables are included, but    **
** may not resemble those in the printed version.               **
**                                                              **
** Please see the PDF (Portable Document Format) file, when     **
** available, for a complete electronic file of the printed     **
** document's contents.                                         **
**                                                              **
** A printed copy of this report may be obtained from the GAO   **
** Document Distribution Center.  For further details, please   **
** send an e-mail message to:                                   **
**                                                              **
**                                            **
**                                                              **
** with the message 'info' in the body.                         **
******************************************************************


Cover
================================================================ COVER


Report to the Chairman and Ranking Minority Member, Subcommittee on
General Oversight and Investigations, House Committee on Banking and
Financial Services

June 1998

MONEY LAUNDERING - FINCEN'S LAW
ENFORCEMENT SUPPORT ROLE IS
EVOLVING

GAO/GGD-98-117

FinCEN's Law Enforcement Support Role Is Evolving

(182040)


Abbreviations
=============================================================== ABBREV

  BSA - Bank Secrecy Act
  DEA - Drug Enforcement Administration
  FBI - Federal Bureau of Investigation
  FinCEN - Financial Crimes Enforcement Network
  HIDTA - High Intensity Drug Trafficking Area
  IRS - Internal Revenue Service

Letter
=============================================================== LETTER


B-277990

June 19, 1998

The Honorable Spencer T.  Bachus
Chairman
The Honorable Bernard Sanders
Ranking Minority Member
Subcommittee on General Oversight
 and Investigations
Committee on Banking and Financial Services
House of Representatives

The Honorable Maxine Waters
House of Representatives

Although the magnitude of money laundering is unknown, law
enforcement officials have estimated that between $100 billion and
$300 billion in U.S.  currency is laundered each year.\1 Money
laundering provides the fuel for drug dealers, terrorists, arms
dealers, and other criminals to operate and expand their activities,
which can have devastating social and economic consequences.  In your
March 24, 1997, letter, you asked us to study the regulatory,
international, and law enforcement support roles of the Financial
Crimes Enforcement Network (FinCEN), a Treasury Department
organization established in April 1990 to help combat money
laundering and other financial crimes.  In February 1998, we reported
on FinCEN's regulatory role, i.e., the agency's progress in
promulgating Bank Secrecy Act (BSA) regulations.\2

Also, at an April 1998 hearing held by the Subcommittee, we presented
information on FinCEN's various roles, including its efforts to
administer civil penalties for BSA violations.\3

This report focuses on FinCEN's products and services in support of
law enforcement.  Specifically, this report discusses (1) trends in
the types and quantities of products and services provided by FinCEN
to the law enforcement community; (2) the extent to which FinCEN's
products and services have been considered useful by the law
enforcement community in identifying, developing, or prosecuting
money laundering and other financial crime cases; (3) the extent to
which FinCEN evaluates the states' compliance with applicable
controls over access to and use of information when state law
enforcement officials directly access FinCEN's resources; and (4)
FinCEN's efforts to provide Internal Revenue Service (IRS) Form 8300
information (Report of Cash Payments Over $10,000 Received in a Trade
or Business)\4 to the law enforcement community. 

To determine the trends in and usefulness of FinCEN's products and
services, we focused on five principal types of support provided by
FinCEN: 

  -- Tactical support is designed to provide law enforcement
     officials with leads and other information on individuals,
     organizations, or activities currently under investigation. 
     Such support can help law enforcement officials locate
     individuals, determine asset ownership for seizure/forfeiture
     purposes, and establish links among individuals, businesses, and
     assets. 

  -- Strategic support is designed to provide federal and state law
     enforcement officials with intelligence analysis and reports on
     more broadly scoped topics, such as emerging trends, patterns,
     and issues associated with money laundering and other financial
     crimes.  Examples of strategic support include (1) the analysis
     of currency flows to and from federal banks and (2) the
     assessment of the level of threat in a geographic area to
     support state-level anti-money laundering legislative efforts. 

  -- Artificial intelligence support is designed to use a
     computer-based system that groups or links currency transaction
     reports\5 to identify individuals, businesses, and bank accounts
     possibly involved in money laundering and other financial
     crimes. 

  -- On-site tactical self-help (called a "platform concept") is
     designed to provide employees of other federal agencies with
     direct access to FinCEN's resources (e.g., BSA financial
     database and commercial databases) to conduct their own research
     and analysis. 

  -- Off-site tactical self-help (called "Project Gateway") is
     designed to provide designated state and local law enforcement
     officials with direct, on-line access to financial records filed
     under BSA.  These designated officials conduct on-line inquiries
     for other law enforcement agencies in the state. 

In conducting our study, we analyzed data showing the number of
requests for and/or issuances of each of FinCEN's principal types of
products and services.  To assess the usefulness of FinCEN's tactical
support, we surveyed by mail a stratified, statistical sample of 352
federal officials and a simple random sample of 95 state officials
who requested tactical support from April 1996 to March 1997 and to
whom FinCEN had responded before August 1997.  The officials sampled
were representative of the universe of requesters in 31 federal
agencies and 15 states.  To obtain information about the usefulness
of FinCEN's other products and services (i.e., strategic, artificial
intelligence, platform concept, and Project Gateway), we (1)
distributed surveys to or conducted interviews with the main federal
and state agency users of these products and services and (2)
interviewed FinCEN officials.  For these other products and services,
the federal and state agency users we contacted were not
representative samples of the universe of users. 

Also, to determine why some federal agencies' field offices, High
Intensity Drug Trafficking Areas (HIDTA),\6 and some states requested
little or no tactical support from FinCEN, we surveyed (1) selected
Treasury and Justice agency field offices and (2) each HIDTA.  We
also interviewed officials from selected states that requested little
or no support from FinCEN.  As part of these inquiries, we
ascertained what other (non-FinCEN) sources of data and analysis are
used.  However, it was beyond the scope of our study to evaluate or
compare the capabilities or costs related to FinCEN and non-FinCEN
sources. 

Further, to determine what controls are in place when FinCEN provides
other federal and state agencies with direct access to databases
through the platform concept and Project Gateway, we interviewed
FinCEN officials and reviewed program documentation.  However, we did
not audit or test the effectiveness of these controls. 

Finally, to determine FinCEN's efforts to provide IRS Form 8300
information to the law enforcement community, we interviewed
officials from the Treasury Department, FinCEN, and IRS headquarters,
and reviewed correspondence between FinCEN and IRS headquarters. 

We performed our work from August 1997 to March 1998, in accordance
with generally accepted government auditing standards.  Appendix I
provides further details about our objectives, scope, and
methodology. 

We requested comments on a draft of this report from the Department
of the Treasury, FinCEN, and the Department of Justice.  Their
comments are discussed near the end of this letter, and FinCEN's
written comments can be found in appendix VI. 


--------------------
\1 Money laundering, in general, is the disguising or concealing of
illicit income to make it appear legitimate.  U.S.  criminal anti-
money laundering law encompasses the money generated from numerous
different crimes--e.g., drug trafficking, murder for hire,
racketeering, prostitution, and embezzlement. 

\2 Money Laundering:  FinCEN Needs to Better Communicate Regulatory
Priorities and Time Lines (GAO/GGD-98-18, Feb.  6, 1998). 

\3 Money Laundering:  FinCEN's Law Enforcement Support, Regulatory,
and International Roles (GAO/T-GGD-98-83, Apr.  1, 1998). 

\4 Under the Internal Revenue Code, any person who receives more than
$10,000 in cash in one transaction (or two or more related
transactions) in the course of trade or business generally must file
an information return (IRS Form 8300) with IRS specifying the name,
address, and taxpayer identification number of the person from whom
the cash was received and the amount of cash received. 

\5 Financial institutions and certain types of businesses must file a
currency transaction report with IRS for each deposit, withdrawal,
exchange, or other payment or transfer by, through, or to such
financial institutions or businesses that involves more than $10,000
in currency. 

\6 HIDTAs are regions designated by the Office of National Drug
Control Policy as having critical drug trafficking problems that have
a harmful impact on other areas of the United States.  HIDTA task
forces consisting of federal, state, and local law enforcement
agencies are engaged in drug control activities. 


   RESULTS IN BRIEF
------------------------------------------------------------ Letter :1

Since its inception in 1990, FinCEN has expanded the types of
products and services it provides to the law enforcement community. 
In recent years, FinCEN has issued fewer tactical, strategic, and
artificial intelligence products and has encouraged, trained, and
increasingly relied on federal agencies to use the platform concept
and state and local agencies to use Project Gateway to support cases
that do not require FinCEN's expertise.  For example, while the
number of requests for FinCEN's tactical support decreased from 7,817
in calendar year 1995 to 4,878 in 1997, the number of state queries
for BSA financial data via Project Gateway increased from 35,608 in
fiscal year 1995 to 57,663 in fiscal year 1997.\7

According to FinCEN, one significant reason for the decrease in the
number of products issued was that FinCEN's staffing levels have
remained fairly constant over the years, while its overall mission
has expanded to include (1) responsibility for promulgating BSA
regulations and (2) a leadership role in international efforts to
combat money laundering.  Consequently, FinCEN chose to dedicate
fewer staff to generate tactical, strategic, and artificial
intelligence products. 

Respondents to the surveys we sent to officials from 31 federal
agencies and 15 states indicated that FinCEN's tactical products have
been useful.  For example, we estimated that 97 percent of all
federal requesters from the 31 agencies found that, overall, the
case-specific products they received from FinCEN were useful.  The
survey results indicated that FinCEN's tactical products assisted law
enforcement investigations in various ways, such as providing
investigative leads or listing assets not previously identified. 

Further, officials we contacted from selected federal field offices,
states, and HIDTAs that had requested few or no tactical products
from FinCEN generally did not indicate dissatisfaction with FinCEN. 
Rather, they cited other reasons--such as reliance on in-house
capabilities and the availability of intelligence or analytical
support centers other than FinCEN--for their limited use of FinCEN. 
However, we found that some federal respondents were not aware of the
various products and services offered by FinCEN and that FinCEN has
neither developed nor widely disseminated general criteria or
guidelines on when law enforcement officials should request FinCEN's
support. 

In response to feedback from the law enforcement community, FinCEN
began taking steps in 1997 to improve the usefulness of its strategic
and artificial intelligence products.  For example, FinCEN has (1)
created a new strategic office that plans to help agencies improve
their money-laundering detection and prevention programs, while
providing case-specific support to the law enforcement and regulatory
communities and (2) initiated efforts to "partner" with the law
enforcement community to provide strategic and artificial
intelligence products that are more useful to the relevant agencies. 

Also, according to the federal and state officials we interviewed,
the platform concept and Project Gateway are useful tools for helping
agencies combat money laundering and other financial crimes.  FinCEN
data show that an increasing number of federal, state, and local
agencies are using these self-help mechanisms to support their
investigations. 

While Project Gateway is designed to enhance the capabilities of
state and local law enforcement agencies, this technological
advancement increases the potential risk that sensitive information
could be inappropriately accessed, used, or disclosed.  Although
FinCEN has established policies and procedures designed to limit
access to and use of information obtained through the Gateway system,
it has not evaluated the states' compliance with these controls. 
Because of the sensitivity of the data available, it is important
that the system be used only for legitimate purposes. 

In an effort to enhance its investigative support, FinCEN is seeking
IRS approval to provide IRS Form 8300 information to federal, state,
and local law enforcement officials.  From a law enforcement
perspective, these forms can be instrumental in tracing cash payments
by drug traffickers and other criminals for luxury cars, jewelry, and
other expensive items.  In 1997, IRS initiated a process for law
enforcement and regulatory agencies to access Form 8300 information
directly from IRS.  However, several issues must still be resolved
before FinCEN can obtain approval from IRS to disseminate this
information. 


--------------------
\7 FinCEN's records and reports on support provided to the law
enforcement community contain either calendar year or fiscal year
data, depending on the type of product or service. 


   BACKGROUND
------------------------------------------------------------ Letter :2

FinCEN is a relatively small agency with a fiscal year 1997 budget of
about $23 million and an onboard staffing level (as of December 1997)
of 162 employees.  FinCEN does not initiate or carry out any
investigations on its own.  Rather, by serving as a central source
for financial intelligence information and analysis, FinCEN supports
the investigative and prosecutive efforts of numerous law enforcement
authorities.  These include federal agencies, such as the Federal
Bureau of Investigation (FBI), the Drug Enforcement Administration
(DEA), IRS, and the Customs Service; state police departments and
attorney general offices; local police departments and district
attorney offices; and others, including applicable foreign
authorities. 

In supporting these law enforcement entities, FinCEN administers the
largest financial transaction reporting system in the world, which is
based on recordkeeping and reporting requirements mandated or
authorized under the BSA, as amended.  To supplement the BSA-related
financial database, FinCEN procures access to a variety of
commercially maintained databases, such as Autotrack and Lexis/Nexis,
which can be used to locate individuals, determine asset ownership,
and establish links among individuals, businesses, and assets. 
Moreover, through specific written agreements, FinCEN can access the
investigative-case databases of various federal law enforcement
agencies (e.g., DEA and the Customs Service).  Further, FinCEN
maintains an internal database of the cases or requests it has
supported and the related products it has generated.  This internal
database enables FinCEN to help law enforcement agencies coordinate
their efforts.  Appendix II contains additional information about the
databases used by FinCEN. 

Using these financial, commercial, and law enforcement databases,
FinCEN provides five principal types of support.  Historically,
tactical support in response to law enforcement requests for data and
analysis has been a mainstay of FinCEN's various products and
services.  Tactical products can be database extracts on a single
subject or detailed, in-depth analyses of the financial aspects of
major criminal organizations.  According to a FinCEN official,
routine requests for tactical support are handled on a first-come,
first-served basis, with federal, state, and local requests treated
equally.  The official added that FinCEN can respond to tactical
requests requiring immediate attention (e.g., terrorism cases) within
24 hours and sometimes as quickly as 2 to 3 hours.  At times, FinCEN
may also provide off-site support to investigative teams or task
forces working complex cases.  Further, FinCEN provides tactical
support to assist other types of cases, including classified,
terrorism, and grand-jury investigations. 

As the name implies, rather than focusing on individual cases,
FinCEN's strategic support is designed to provide federal and state
law enforcement officials with intelligence analyses and reports on
longer term or more broadly scoped topics.  An example is the
analysis of trends, patterns, and fluctuations in currency flows to
and from Federal Reserve banks to determine any money-laundering
implications.  Also, as part of its strategic efforts, FinCEN has
developed a "case lab" unit to assist law enforcement agencies in
addressing large-scale money-laundering activity, such as that
conducted by drug cartels.  According to FinCEN's Strategic Plan
(1997-2002), the agency's overriding objective over the next few
years will be to construct a viable model for determining the
magnitude of money laundering. 

FinCEN's artificial intelligence support is designed to help identify
subjects (i.e., individuals and businesses) possibly involved in
money laundering and other types of financial crimes.  This support
relies on computerized technology designed to identify suspicious
transactions by reviewing and correlating (i.e., grouping or linking)
currency transaction reports required by the BSA.  FinCEN has used
this technology to self-generate investigative leads as well as to
respond to specific requests.  In the former instances, FinCEN's
policy is to forward its results to the appropriate law enforcement
agency for consideration. 

A fourth type of support is on-site tactical self-help (called a
"platform concept"), whereby FinCEN encourages and trains agents and
analysts from federal law enforcement and regulatory agencies to
access the BSA financial database and commercial databases
(accessible at FinCEN) to conduct their own research and analysis. 
FinCEN provides office space and database access for these personnel,
who use FinCEN resources on a full- or part-time basis and work only
on cases for their respective agencies.  Most platform users do not
have direct access to law enforcement databases or suspicious
activity reports.\8 A FinCEN official explained that (1) interagency
agreements usually prohibit platform users from accessing law
enforcement databases and (2) most federal agencies that use
suspicious activity reports have their own access to the reports. 
When requested, a FinCEN employee will check these databases for
platform users. 

Finally, the fifth type of support is off-site tactical self-help
(called "Project Gateway")--a joint effort between FinCEN and the IRS
Detroit Computing Center, which receives and processes financial
information reported under the BSA (e.g., currency transaction
reports).  Using custom-designed software, designated state and local
law enforcement officials have direct, on-line electronic access to
BSA records, including suspicious activity reports.  These designated
officials also conduct Gateway queries for other law enforcement
agencies in their state and can also request information from the
Autotrack commercial database.  According to a FinCEN official,
Gateway users have direct access to suspicious activity reports
primarily because state and local law enforcement agencies do not
have their own access to the reports.  Further, through Gateway,
FinCEN can "alert" or notify one agency that another agency has or
had an interest in the same investigative subject (individual or
entity).\9 Each state's access to Gateway is free of charge (except
for telephone charges). 


--------------------
\8 Suspicious activity reports, in general, must be filed by banks
and other depository institutions when they know, suspect, or have
reason to suspect that a crime has occurred or that a transaction is
suspicious. 

\9 Gateway electronically captures the information gathered on
incoming inquiries and automatically compares this information with
prior and subsequent queries from Gateway users.  In addition,
Gateway users can ask FinCEN to match new subjects against other law
enforcement databases to identify potential parallel investigations. 


      FINCEN HAS ISSUED FEWER
      REPORTS WHILE AGENCIES HAVE
      INCREASED THEIR DIRECT USE
      OF THE PLATFORM CONCEPT AND
      PROJECT GATEWAY
---------------------------------------------------------- Letter :2.1

While the types of products and services offered by FinCEN have
expanded, the volume of tactical, strategic, and artificial
intelligence reports has decreased over the past several years. 
While the reasons for the decreases varied, one significant factor
cited by FinCEN was that it assigned fewer staff to support these
activities.  Since 1995, law enforcement officials have increased
their use of FinCEN's platform concept and Project Gateway to conduct
their own research and analysis. 


      TACTICAL SUPPORT REQUESTS
      MAY HAVE PEAKED IN 1995
---------------------------------------------------------- Letter :2.2

FinCEN has been providing tactical support to the law enforcement
community since the agency was established in 1990.  As figure 1
shows, the volume of law enforcement requests to FinCEN for tactical
support increased significantly from calendar year 1991--the first
full year of operation--to 1995.  However, from calendar year 1995 to
1997, the volume of requests decreased by approximately 38 percent. 

   Figure 1:  Requests to FinCEN
   for Tactical Support, 1991-1997

   (See figure in printed
   edition.)

Source:  FinCEN. 

According to a FinCEN official, one primary reason for the decrease
in the number of tactical requests was that, in recent years, FinCEN
stopped accepting certain types of requests--such as requests
involving background checks for employment or security clearance
purposes--that did not directly support law enforcement functions. 
For example, FinCEN stopped supporting certain IRS and U.S.  Postal
Inspection Service investigations in 1996.\10 As a result, FinCEN's
1997 workload was reduced by approximately 3,300 cases.  The official
told us that FinCEN's expertise is not needed to conduct these types
of inquiries and that, as a result, FinCEN has made efforts to
encourage agencies to conduct these types of research on their own or
by using FinCEN's platform concept or Project Gateway, as applicable. 

The official told us that FinCEN stopped accepting certain types of
requests to compensate for reductions in staff assigned to provide
tactical support.  As figure 2 shows, from October 1994 to October
1997, the total number of staff (federal and contractor) responding
to requests for tactical investigative support decreased from 77 to
40.  Of this total, the number of federal staff (i.e., FinCEN
employees and detailees from other agencies) decreased from 53 to 19. 

   Figure 2:  Number of Federal
   and Contractor Staff Responding
   to Agency Requests for Tactical
   Support, October 1994 to
   October 1997

   (See figure in printed
   edition.)

Source:  FinCEN. 

According to the FinCEN official, although the agency's overall staff
levels have remained fairly constant over the years, many tactical
staff have been reassigned to support FinCEN's expanded mission,
which includes (1) responsibility for promulgating BSA regulations
and (2) a leadership role in international efforts to combat money
laundering.\11 The official also told us that several tactical staff
were reassigned in March 1997 to support FinCEN's new Office of
Research and Analysis. 

According to FinCEN officials, FinCEN also began to contact
requesters in 1995 to determine specifically what type of data and
analysis was needed.  These contacts included determining which
databases had already been researched (outside of FinCEN), which
databases FinCEN should research, and what type of report should be
provided (e.g., a database extract or an analytical report).  The
officials told us that, before 1995, FinCEN's tactical support staff
would generally research all of the financial, commercial, and law
enforcement databases the agency could access. 

Further, over the past 2 years, an increasing number of tactical
reports consisted of database extracts versus reports that contained
detailed analysis.  According to FinCEN officials, the percentage of
tactical products that contained detailed analysis decreased from
approximately 25 percent in calendar year 1996 to about 10 percent
during calendar year 1997.  According to FinCEN's Strategic Plan
(1997-2002), the agency intends to provide better tactical support to
its customers by increasing the number of FinCEN analysts capable of
performing complex research and analysis. 


--------------------
\10 The IRS casework consisted of background investigations, and the
U.S.  Postal Inspection Service casework involved investigations of
individuals on workman's compensation to determine if they were
earning income from other sources. 

\11 As of December 1997, 36 (22 percent) of FinCEN's 162 onboard
staff were assigned to regulatory or international efforts. 


      NUMBER OF STRATEGIC REPORTS
      DECREASED SIGNIFICANTLY
---------------------------------------------------------- Letter :2.3

As with tactical support, FinCEN has been providing strategic support
to the law enforcement community since 1990.  As shown in figure 3,
the volume of strategic reports issued by FinCEN has decreased from a
high of 23 in 1992 to 1 in 1997.\12

   Figure 3:  Strategic Reports
   Issued by FinCEN, 1990-1997

   (See figure in printed
   edition.)

Source:  GAO analysis of FinCEN data. 

According to a FinCEN official, fewer strategic reports were
generated in recent years because (1) FinCEN assigned fewer staff to
support strategic projects and (2) the products often did not provide
the type of information law enforcement officials needed to do their
work. 

To increase its emphasis on strategic support and to improve the
usefulness of strategic products, FinCEN created a new Office of
Research and Analysis in March 1997.  Although FinCEN initially
planned to staff this office with 23 analysts, only 14 staff were on
board as of November 1997.  Before this reorganization, FinCEN's
Strategic Analysis Division had five to seven analysts performing
strategic analysis.  Additional perspectives about the usefulness of
FinCEN's strategic products and initiatives are discussed later in
this report. 


--------------------
\12 In addition to reports, FinCEN generates other strategic
products, such as responses to inquiries requiring research and
analysis and information developed for educational or training
purposes.  Until 1994, FinCEN also prepared quarterly publications
entitled FinCEN Updates (for law enforcement staff) and FinCEN Trends
(for regulatory and banking institutions).  These publications
contained articles and statistical summaries dealing with money
laundering and currency reporting. 


      FEWER ARTIFICIAL
      INTELLIGENCE REPORTS
      PROVIDED
---------------------------------------------------------- Letter :2.4

Since March 1993, FinCEN has been using its artificial intelligence
targeting system to help the law enforcement community identify
subjects possibly involved in money laundering and other financial
crimes.  As shown in table 1, the volume of products generated by
FinCEN from artificial intelligence analyses has decreased
significantly over the past 2 fiscal years. 



                                Table 1
                
                   Products Generated by FinCEN From
                Artificial Intelligence Analyses, 1993-
                                  1997

                                                    Number of subjects
Fiscal year                     Number of products          identified
------------------------------  ------------------  ------------------
1993\a                                          27                 276
1994                                            75                 403
1995                                           372                 946
1996                                           129               2,092
1997                                            40                 140
----------------------------------------------------------------------
\a March to September 1993. 

Source:  GAO analysis of FinCEN data. 

FinCEN officials provided several reasons for the decrease in
artificial intelligence products.  First, they said the artificial
intelligence system was redesigned for fiscal year 1996 to focus on
more complex cases, which they defined by the number of subjects per
case.  FinCEN data indicated there was an average of 2.5 subjects per
case in fiscal year 1995 versus an average of 16.2 subjects per case
in fiscal year 1996.  Second, they said that the majority of the
artificial intelligence products generated in fiscal years 1993 to
1995 were self-initiated, and that FinCEN self-initiated very few
artificial intelligence products in fiscal years 1996 and 1997, in
part because federal law enforcement agencies reportedly did not take
any action on many of the former products.  Third, they said that
FinCEN reduced the number of full-time analysts working on artificial
intelligence cases from three in fiscal year 1995 to the equivalent
of one-half full-time analyst in fiscal year 1997.  In an effort to
increase the number and usefulness of investigative leads and
products generated, FinCEN assigned three full-time staff to support
artificial intelligence cases in early fiscal year 1998. 


      FEDERAL AGENCIES HAVE
      INCREASED THEIR USE OF
      FINCEN'S PLATFORM CONCEPT
---------------------------------------------------------- Letter :2.5

In an effort to expand the availability and use of its tactical
support resources, FinCEN developed the platform concept in 1994,
offering employees of federal law enforcement agencies space at
FinCEN and access to BSA financial data and commercial databases so
they could conduct their own research and analysis.  As table 2
shows, law enforcement agencies have increased their use of
platforms.  As of April 1998, 51 personnel representing 31 agencies
were using platforms on a full- or part-time basis at FinCEN. 



                                Table 2
                
                   Use of FinCEN Platforms, 1995-1997

                                                       Number of cases
Calendar year                   Number of agencies           supported
------------------------------  ------------------  ------------------
1995                                             5                 575
1996                                            15                 779
1997                                            28               1,477
----------------------------------------------------------------------
Source:  GAO analysis of FinCEN data. 

According to data provided by FinCEN, platform users supported 1,477
investigative cases during fiscal year 1997.  These data show that
the most frequent platform users during this period were the Defense
Criminal Investigative Service; the Capitol Police; the Bureau of
Alcohol, Tobacco and Firearms; and the Washington/Baltimore HIDTA. 


      STATES HAVE INCREASED THEIR
      USE OF PROJECT GATEWAY
---------------------------------------------------------- Letter :2.6

Also in 1994, FinCEN initiated Project Gateway to provide state and
local law enforcement agencies with remote access to BSA financial
data.  In 1997, FinCEN expanded Gateway's remote research capability
by providing states with (1) access to suspicious activity reports to
support ongoing investigations; (2) the capability to identify
potential subjects (i.e., new leads) by proactively searching BSA
records, including currency transaction reports and suspicious
activity reports (generated within the users' own state); and (3)
access to information from one commercial database (i.e., Autotrack)
via Gateway's bulletin board system. 

As shown in figure 4, the states' use of Gateway has steadily
increased since 1995--the first full year when all 50 states had
access.  According to FinCEN data, the most frequent Gateway users
during fiscal year 1997 were California, Illinois, Pennsylvania, and
Texas.  Wyoming was the only state that did not use Gateway during
this period. 

   Figure 4:  Use of Project
   Gateway

   (See figure in printed
   edition.)

Source:  GAO analysis of FinCEN data. 

According to FinCEN's records, the 57,663 Gateway queries in fiscal
year 1997 supported 3,327 law enforcement cases.  Approximately 84
percent of the BSA reports reviewed during these queries were
currency transaction reports.  Also, 21 states made a total of 3,440
automated queries of Autotrack during fiscal year 1997. 

Further, the number of times FinCEN used Gateway to alert or notify
one agency that another agency had an interest in the same
investigative subject (individual or entity) increased from 356 in
fiscal year 1996 to 920 in fiscal year 1997. 


   OVERALL, FINCEN'S PRODUCTS AND
   SERVICES WERE CONSIDERED USEFUL
------------------------------------------------------------ Letter :3

Responses to the surveys we sent to officials from 31 federal
agencies and 15 states indicated that FinCEN's tactical products have
been useful in identifying, developing, and prosecuting cases
involving money laundering and other financial crimes.  Also, FinCEN
has taken steps to address concerns with and improve the usefulness
of its strategic and artificial intelligence products, such as
attempting to "partner" with the law enforcement community in
developing these products.  Further, according to the federal and
state officials we interviewed, the platform concept and Project
Gateway are excellent tools for helping agencies combat money
laundering and other financial crimes. 


      OVER 95 PERCENT OF TACTICAL
      SUPPORT USERS SURVEYED FOUND
      THE PRODUCTS USEFUL
---------------------------------------------------------- Letter :3.1

To assess the usefulness of FinCEN's tactical support, we surveyed by
mail a stratified, statistical sample of 352 officials from 31
federal agencies and a simple random sample of 95 officials from 15
states who requested tactical support from April 1996 through March
1997 and to whom FinCEN had responded before August 1997.\13 Based on
the results of these surveys, we estimated that 97 percent of all
requesters from the 31 federal agencies found that, overall, all
types of products they received (i.e., database extracts, analytical
reports, and responses to expedited requests) were useful in
supporting their investigations.  Similarly, we estimated that 98
percent of the requesters from the 15 states found that, overall, the
products they received were useful.\14

The survey results indicated that FinCEN's tactical products assisted
law enforcement investigations in various ways.  As shown in figure
5, we estimated that the majority of the requesters from the 31
federal agencies found that the tactical products (1) verified or
confirmed information already known, (2) saved them time and money,
(3) provided investigative leads that were previously unknown, or (4)
identified assets that were previously unknown.  We also estimated
that the majority of the requesters from the 15 states found that the
tactical products assisted their investigations in various ways (see
app.  IV for additional details on state responses). 

   Figure 5:  Specific Ways in
   Which Federal Users Said That
   FinCEN's Tactical Products
   Assisted Their Investigations

   (See figure in printed
   edition.)

Note:  Respondents were asked to check "Yes," "No," or "Too soon to
know" for each category (e.g., "Saved time and money").  The total
number of estimated responses for each category (i.e., "n") does not
equal 2,379 because some respondents did not provide answers (see
app.  III, question 5). 

Source:  GAO survey of federal law enforcement officials who
requested tactical support from April 1996 through March 1997 and
received a response before August 1997. 

Additionally, on the basis of the survey responses, we estimated that
84 percent of the federal officials received products that responded
to their requests in a complete and thorough manner.  We estimated
that another 14 percent found that some, but not all, of the products
addressed their requests in a complete and thorough manner.\15 Also,
we estimated that 81 percent of the federal officials found the
products to be "very" (29 percent) or "somewhat" (52 percent)
timely,\16 and an estimated 97 percent would "definitely" (78
percent) or "probably" (19 percent) request tactical support from
FinCEN in the future.\17

Although we estimated that 81 percent of the requesters from the 31
federal agencies found FinCEN's tactical products timely, the most
frequent comment or suggestion regarding how FinCEN's tactical
products could be improved concerned the timeliness of responses.  We
estimated that 17 percent of the federal officials who received
tactical products would suggest that FinCEN should respond to
requests in a more timely manner.  Although the number of requests
received by FinCEN decreased during fiscal year 1997 and FinCEN has
taken steps to better focus the scope of each request, the average
turnaround time--i.e., the number of calendar days from when a
request is received at FinCEN to the day the product is mailed to the
requester--for routine requests increased to 2 to 3 months during
most of fiscal year 1997. 

According to a FinCEN official, the primary reason for this lengthy
turnaround time was because many staff responding to tactical
requests were reassigned in March 1997 to support FinCEN's new Office
of Research and Analysis.  Consequently, the number of FinCEN staff
supporting tactical requests decreased from 19 in October 1996 to as
few as 7 during the summer of 1997 before increasing back to 19 in
October 1997. 

The official told us the average turnaround time for routine requests
was reduced to about 3 weeks in September 1997--about 6 months after
the sample period used for our questionnaire--by authorizing overtime
and temporarily having more staff do casework.  He noted that FinCEN
had 22 tactical staff on board as of February 1998, which should help
prevent backlogs and adverse turnaround times in the future.  He
added that FinCEN has also authorized overtime to address any
backlogs that may occur. 


--------------------
\13 Appendix I contains additional information about our sampling
methodology.  The federal and state surveys are presented in
appendixes III and IV, respectively. 

\14 Because the mail survey results for federal officials came from a
sample of 352 officials out of a universe of 2,379 officials in 31
agencies and the survey results for state officials came from a
sample of 95 officials out of a universe of 504 officials in 15
states, all results for both surveys are subject to sampling errors,
along with other potential sources of errors associated with surveys,
such as nonresponses and question misinterpretation.  All percentage
estimates for the survey of federal officials have 95-percent
confidence intervals of no more than plus or minus 10 percentage
points.  All percentage estimates for the survey of state officials
have 95-percent confidence intervals of no more than plus or minus 13
percentage points. 

\15 See appendix III, question 8. 

\16 See appendix III, question 6. 

\17 See appendix III, question 9. 


      FINCEN IS SEEKING TO IMPROVE
      STRATEGIC SUPPORT
---------------------------------------------------------- Letter :3.2

According to a FinCEN official, the feedback FinCEN received from the
law enforcement community indicated that many of the strategic
reports FinCEN issued from 1990 to 1996 were of limited use to law
enforcement agencies, regulators, and banking institutions.  He noted
that many law enforcement officials told FinCEN the reports were
informative but that many of the reports were not effective or useful
in supporting investigative efforts and, some of them did little more
than repackage what law enforcement officials had previously reported
to FinCEN.  The official also told us that, in the past, FinCEN had
not taken an active partnership role with law enforcement agencies,
regulators, and financial institutions to help make the products more
useful to them. 

In response to law enforcement's observations regarding the
usefulness of FinCEN's strategic products, FinCEN reorganized and
reallocated its resources in March 1997 to create a new Office of
Research and Analysis.  This office plans to generate products that
help agencies improve their money-laundering detection and prevention
programs, while providing case-specific support to the law
enforcement and regulatory communities.  For example, according to a
FinCEN official, since mid-1997, FinCEN has analyzed suspicious
activity reports and provided selected federal field offices with
target lists of potential suspects.  He added that this office also
intends to take an active partnership role with law enforcement
agencies, regulators, and financial institutions in developing future
strategic products. 

To assess the usefulness of strategic products, we interviewed
officials who received state-specific money-laundering threat
assessment reports in 1996 or 1997--Georgia (April 1996), Louisiana
(April 1996), and Florida (June 1997).\18 Generally, FinCEN's threat
assessment reports focus on specific industries, businesses,
financial entities, and geographic locations exhibiting unusual
currency movements that may be indicative of money-laundering
activity. 

According to officials from Georgia and Louisiana, the threat
assessment reports FinCEN prepared for their states were informative
but, at the time of our review, had resulted in limited or no
investigative actions.  However, according to a Louisiana state
official, FinCEN's 1996 report contributed to the state enacting
comprehensive money-laundering legislation and expanding its
statutory wiretap authority. 

The June 1997 South Florida Money Laundering Threat Assessment--a
joint effort between FinCEN and the South Florida HIDTA Task
Force--was the first partnership effort and first report issued by
the Office of Research and Analysis.  According to a HIDTA official
in Florida, as of January 1998, this report had not led to any new
investigations.  However, he told us the report provides information
that will help the law enforcement community in the region reallocate
its resources.  He added that the partnership meetings associated
with the report have been useful in bringing together officials from
the law enforcement, regulatory, and banking communities to discuss
money-laundering issues. 


--------------------
\18 These threat assessments were three of the four strategic reports
FinCEN issued from January 1996 through December 1997 (see table 2). 
The fourth report addressed emerging cyberpayment technology and its
potential use in criminal activities. 


      RESPONDENTS FOUND ARTIFICIAL
      INTELLIGENCE PRODUCTS USEFUL
---------------------------------------------------------- Letter :3.3

To assess the usefulness of FinCEN's artificial intelligence
products, we attempted to contact 24 law enforcement officials who,
according to FinCEN's records, received the 85 artificial
intelligence reports FinCEN generated from April 1996 to March 1997. 
The 13 officials we eventually were able to contact received 51 (60
percent) of the 85 artificial intelligence products.\19 The 51
products supported 15 separate cases.\20

According to these officials, the information contained in FinCEN's
artificial intelligence products was useful in various ways.  For
example, in reference to the 15 cases, the products identified
potential subjects that were previously unknown (9 cases), provided
investigative leads that were previously unknown (11 cases),
identified assets that were previously unknown (6 cases), and
verified or confirmed information already known (12 cases). 

According to the 13 officials, the information contained in the
artificial intelligence products led to five preliminary
investigations, and three new cases were opened.  One official told
us the artificial intelligence analyses compiled by FinCEN provided
several investigative leads, which indicated criminal activity
involving people and/or businesses located nationally and
internationally.  Three of the officials told us they did not use the
products at all.  Overall, 12 of the 13 officials indicated they
would definitely or probably request additional products in the
future. 

In an effort to provide artificial intelligence products that are
more useful to the relevant agencies, FinCEN is attempting to
"partner" with the law enforcement community.  For example, in
November 1997, FinCEN analysts began working directly with agents
from U.S.  Customs Service field offices to help them generate,
understand, and use artificial intelligence products.  According to a
FinCEN official, the agency intends to provide continuing support and
updates to these agents.  This initiative has not been in effect a
sufficient period of time to assess its usefulness. 


--------------------
\19 Of the 13 officials, 6 requested the products to both obtain
information and leads on subjects already under investigation and
identify potential subjects, previously unknown, who may have been
involved in money laundering or other financial crimes.  Of the
remaining seven officials, two requested the products to obtain
information and leads on subjects already under investigation, three
others requested them to identify potential subjects, and the final
two did not indicate a reason why. 

\20 Two of the 13 officials received artificial intelligence products
that were used to support 2 separate and unrelated cases. 


      ACCESS TO PLATFORMS WAS
      USEFUL TO FEDERAL
      INVESTIGATIVE AGENCIES
      CONTACTED
---------------------------------------------------------- Letter :3.4

As shown earlier in table 2, both the number of agencies using
platforms and the number of cases supported increased significantly
from 1995 to 1997.  To assess the usefulness of FinCEN's platform
concept, we interviewed officials from three agencies that were among
the top four user agencies in 1997--the Defense Criminal
Investigative Service; the Bureau of Alcohol, Tobacco and Firearms;
and the Washington/Baltimore HIDTA.  According to FinCEN's records,
these agencies accounted for 971 (66 percent) of the 1,477
investigative cases that were supported by platforms in 1997. 

According to these officials, access to financial and commercial
databases using FinCEN's platforms is an important and useful tool in
helping their respective agencies initiate cases and conduct
investigations.  For example, a Defense Criminal Investigative
Service analyst told us she uses a platform to support every new case
initiated by the agency and, at the request of field offices,
prepares detailed research reports.  The analyst commented that she
would like FinCEN to provide access to additional sources of data,
such as suspicious activity reports and law enforcement databases via
platforms.  Also, an official from the Washington/Baltimore HIDTA
told us that use of platforms is the HIDTA's primary means for
obtaining commercial data.  The official noted that having access to
platforms has enabled analysts to respond to requests for information
in a timely manner. 

According to a FinCEN official, providing agencies with access to
financial and commercial data via platforms should allow FinCEN to
focus its efforts on more complicated cases that require FinCEN's
expertise and analytical support. 


      DIRECT ACCESS TO GATEWAY WAS
      USEFUL TO STATE
      INVESTIGATIVE AGENCIES
      CONTACTED
---------------------------------------------------------- Letter :3.5

As shown earlier in figure 4, 49 states used the Gateway system in
fiscal year 1997 to make a total of 57,663 queries for BSA financial
data.  To assess the usefulness of the Gateway system, we interviewed
the state coordinator\21 for each of the four states that made the
most Gateway queries in fiscal year 1997--California, Illinois,
Pennsylvania, and Texas.  According FinCEN data, these states made
22,876 (or 40 percent) of the total 57,663 Gateway queries during
this period. 

According to these officials, remote access to BSA financial data via
the Gateway system is useful for helping their states combat money
laundering and other financial crimes.  The officials also told us
Gateway's expanded capabilities were useful, including (1) remote
access to commercial data and (2) proactive searches of BSA financial
records, including currency transaction reports and suspicious
activity reports, to generate targets in their state for possible
money laundering and other financial crimes investigations. 

In August 1997, FinCEN began providing additional agencies in some
states with direct access to the Gateway system.  According to
FinCEN's records, 310 state law enforcement officials had Gateway
user accounts as of the end of fiscal year 1997.  Also, selected law
enforcement officials from DEA have been granted access to Gateway. 
According to a FinCEN official, other federal agencies can join
Project Gateway if they agree to (1) identify the case file under
which Gateway queries are run, (2) reasonably share case information
with other agencies, and (3) execute appropriate legal agreements. 
He added that the case information is needed to help FinCEN assist
Gateway users in coordinating their investigative efforts. 

According to a FinCEN official, in 1998, the agency intends to expand
Gateway's capabilities to include additional commercial databases
and, when available, data on money services businesses.\22


--------------------
\21 The state coordinator is an employee of the state agency named in
a letter of agreement (regarding Gateway access) between FinCEN and
the state.  This employee is the state's principal representative for
the Gateway program and is responsible for coordinating all
Gateway-query requests for all using agencies in the state. 

\22 Under FinCEN's proposed definition, money services businesses
would include money transmitters, currency dealers or exchangers,
check cashers, and issuers of traveler's checks and money orders. 
Under a provision of the Money Laundering Suppression Act of 1994,
these entities are to be registered with the Treasury Department. 


   SOME FEDERAL FIELD OFFICES,
   STATES, AND HIDTAS SAID THEY
   USE ALTERNATIVE SOURCES OF
   INFORMATION
------------------------------------------------------------ Letter :4

Our analysis of FinCEN's records revealed that some federal field
offices, states, and HIDTAs requested little or no tactical support
from FinCEN during the period we reviewed, April 1996 to March 1997. 
To determine the reasons why this support was not requested, we
surveyed the U.S.  investigative units of 129 Treasury and Justice
field offices\23 and each of the 22 HIDTAs that had been designated
as of December 1997.  We also interviewed officials from 9 of the 10
states that, according to FinCEN's records, made either one or no
requests for FinCEN's tactical support from April 1996 to March 1997. 
In summary, respondents from many of the federal field offices,
states, and HIDTAs surveyed said they have in-house capabilities or
use sources other than FinCEN for tactical support.  Also,
respondents from some federal field offices said that agents and
investigators are not aware of the products and services offered by
FinCEN. 


--------------------
\23 See appendix V for a copy of the survey, which was sent to all of
the U.S.  investigative units of the IRS (33 offices), the U.S. 
Customs Service (20 offices), the FBI (56 offices), and the DEA (20
offices). 


      MANY AGENCIES CONTACTED SAID
      THEY HAVE IN-HOUSE
      CAPABILITIES OR USE SOURCES
      OTHER THAN FINCEN
---------------------------------------------------------- Letter :4.1

From our surveys and interviews, the most frequently mentioned reason
why many of the federal field offices, states, and HIDTAs requested
few, if any, tactical products from FinCEN was the availability and
use of in-house capabilities and sources other than FinCEN for
financial, commercial, and law enforcement data and analysis. 
Dissatisfaction with FinCEN's tactical products was rarely cited as a
reason for the limited use of FinCEN. 

Of the IRS, Customs, FBI, and DEA field offices that responded to our
survey, 6 percent noted that they did not request FinCEN's tactical
support "over the past year." An additional 42 percent noted that
they requested tactical support 5 times or fewer during this period
(see app.  V, question 1).  Table 3 shows the various reasons why
FinCEN was not the primary source for financial, commercial, or law
enforcement data or analysis for the federal field offices. 



                                Table 3
                
                   Reasons Many of the Federal Field
                  Offices Surveyed Said They Requested
                   Little or No Tactical Support From
                                 FinCEN

                                                Number of
                               Number of         Justice
                            Treasury offices     offices
                            ----------------  --------------
                                                                 Total
                                                              response
Responses                      IRS   Customs     FBI     DEA         s
--------------------------  ------  --------  ------  ------  --------
We have all or most of the      23        16      38      14        91
 capabilities we need in-
 house.
It is easier to access the      17         7      22      11        57
 data or get analytical
 support from other
 sources.
Other sources provide more       2         3       5       4        14
 comprehensive data or
 analysis than FinCEN.
Other sources are more          19         9      28      12        68
 timely than FinCEN.
We are not aware of the          8         5       8       7        28
 products and services
 provided by FinCEN.
======================================================================
Total field offices             31        17      51      17       116
 responding
----------------------------------------------------------------------
Source:  GAO survey of Treasury and Justice field offices (see app. 
V, question 6). 

Also, according to the survey responses, the IRS and Customs
investigative field offices had direct access to BSA financial
databases, while the FBI and DEA offices primarily used grand jury
subpoenas to obtain financial data.  Almost all of the field offices
noted they had access to one or more commercial databases.  Most of
the offices also indicated they had access to a variety of sources
for law enforcement information outside of FinCEN, such as the
National Crime Information Center and the National Law Enforcement
Telecommunications System.  Also, several offices noted their office
and/or agency had staff and other resources to provide analytical
support for money laundering and other financial crime
investigations. 

Several offices indicated that they had access to intelligence or
investigative support centers (other than FinCEN) that provide
financial, commercial, and/or law enforcement data and analysis. 
These sources include, for example, the FBI's Savannah (Georgia) and
Butte (Montana) Information Technology Centers, the South Florida
Investigative Support Center (part of the South Florida HIDTA), the
National Drug Intelligence Center (Johnstown, Pennsylvania), and the
El Paso Intelligence Center (El Paso, Texas). 

In commenting on a draft of this report, Department of Justice
officials noted that while many agencies have in-house capabilities
or use sources other than FinCEN for financial, commercial, and law
enforcement data and analysis, FinCEN still plays an important role
in facilitating the sharing of financial information and is a leader
in the financial database field. 

Officials we interviewed from the nine states that made either one or
no requests for FinCEN's tactical support from April 1996 to March
1997 cited several reasons for not requesting more support.  These
reasons include (1) the lack of authority to conduct investigations
into money laundering; (2) the lack of resources to pursue financial
crimes; and (3) the use of other sources of information, such as
FinCEN's Gateway system and direct access to commercial databases. 

Sixteen of the 22 HIDTAs surveyed responded to our questionnaire. 
Survey results showed that 4 of the 16 respondents were designated as
HIDTAs in 1997 and had not performed the type of money laundering or
financial crime investigations that FinCEN supports.  Of the 12
HIDTAs that were performing such investigations, 8 noted that (1) the
HIDTA had all or most of the capabilities it needed in-house or (2)
it was easier to access data or get analytical support from other
sources.  Of these 10 respondents, 6 noted that other sources were
more timely than FinCEN and 2 (of the 6) indicated that the HIDTA was
not aware of the products and services provided by FinCEN. 


      SOME POTENTIAL USERS SAID
      THEY WERE NOT AWARE OF
      FINCEN'S PRODUCTS AND
      SERVICES
---------------------------------------------------------- Letter :4.2

Our survey responses and interviews indicated another reason for the
limited use of FinCEN's tactical support.  That is, many agents and
investigators located in federal field offices were not aware of the
products and services offered by FinCEN.  For example, as shown in
table 3, 28 (24 percent) of the 116 federal field offices that
responded to our survey noted that their agents and investigators
were not aware of FinCEN's products and services. 

FinCEN officials told us the agency has informed federal and state
officials about its products and services on several occasions.  For
example, in mid-1996, FinCEN initiated a three-phased approach to
inform and obtain feedback from Treasury Department officials.  The
three phases were (1) telephone surveys; (2) on-site visits with U.S. 
Customs Service; Bureau of Alcohol, Tobacco and Firearms; and IRS
field offices in six cities and the U.S.  Customs Service in a
seventh city; and (3) training on how to use FinCEN resources at
events such as Special-Agent-in-Charge meetings, internal training,
and new agent training.  One Treasury Department component, the U.S. 
Secret Service, declined FinCEN's offer to participate in the
outreach program.  FinCEN followed this effort with telephone or fax
surveys to approximately 100 officials from other federal agencies. 

According to FinCEN's records, the agency's tactical staff also
provided information about FinCEN at meetings, training events, or
conferences 51 times in fiscal year 1996 and 48 times in fiscal year
1997.  This included presentations to or meetings with state police
departments, other intelligence centers (e.g., the El Paso
Intelligence Center and the National Drug Intelligence Center), and
federal agencies' asset forfeiture personnel.  Also, according to
FinCEN's records, from February to October 1997, FinCEN officials
visited 24 states to brief state officials about FinCEN's support. 

FinCEN drew several conclusions from its 1996 efforts to inform and
obtain feedback from federal agency officials about its products and
services.  Specifically, FinCEN concluded that it should seriously
consider (1) providing potential requesters with a one-page summary
of FinCEN's capabilities, missions, and programs and (2) advising
potential customers about the capabilities of new or enhanced
commercial and law enforcement databases. 

In March 1998, FinCEN updated its site on the Internet to include
information on the types of support it provides to the law
enforcement community and to offer viewers the opportunity to send
comments or questions via electronic mail.  Also, in his April 1,
1998, statement before the House Banking and Financial Services'
Subcommittee on General Oversight and Investigation, the Treasury
Assistant Secretary (Enforcement) said that FinCEN will work to
further communicate its capabilities to its potential customers. 

While FinCEN has taken several actions to better ensure that
potential customers are aware of the availability of FinCEN's various
products and services, it has not developed written criteria or
guidance specifying the types of cases that FinCEN can best support. 
According to a FinCEN official, tactical support is generally best
suited for (1) cases that involve large criminal organizations, are
of significance to the jurisdiction, or require expert financial
analysis and (2) cases where support is not available from in-house
sources (e.g., requests for BSA data).  The official told us that for
cases or investigations that do not require FinCEN's expertise or
analysis, other sources of data and analysis can be used, including
FinCEN platform concept and Project Gateway.  He also mentioned that
FinCEN staff are available to train agencies to use their own
resources to conduct research and analysis.  The FinCEN official
added that, to his knowledge, FinCEN has never turned down a request
to support an ongoing law enforcement investigation that required
FinCEN's capabilities. 


   KEY CONTROLS OVER GATEWAY NOT
   ROUTINELY EVALUATED
------------------------------------------------------------ Letter :5

FinCEN has established several controls to reduce the risk of
unauthorized access to and use of sensitive information obtained via
Project Gateway.\24 As a primary set of controls, FinCEN has
established policies and procedures that states are to follow in
accessing and using Gateway information.  However, FinCEN has not
evaluated the states' compliance with these controls.  Although
FinCEN has taken some steps to review Gateway queries, this review
does not include a direct assessment of the states' controls over
Gateway information. 


--------------------
\24 It was beyond the scope of our study to evaluate or test these
controls. 


      GATEWAY ACCESS TO SENSITIVE
      INFORMATION MAKES EVALUATING
      CONTROLS IMPORTANT
---------------------------------------------------------- Letter :5.1

FinCEN's Gateway system provides designated state and local agency
officials with direct, on-line access to sensitive BSA information. 
For example, these officials have direct access to currency
transaction reports that reflect transactions with financial
institutions that exceed $10,000.  Gateway users also have direct
access to suspicious activity reports that are filed by banks and
other depository institutions when they know, suspect, or have reason
to suspect that a crime has occurred or that a transaction is
suspicious. 

Security management of the Gateway system is a joint effort between
the IRS Detroit Computing Center and FinCEN.  The Detroit Computing
Center is responsible for ensuring that the host system that contains
the BSA data is secured to the level required by the Treasury
Department's policy for systems that process sensitive data.  The
Center's Project Gateway Users Guide specifies additional access
controls to BSA information, including (1) password requirements and
(2) disconnecting users from the Gateway host computer after 5
minutes of inactivity.  The Center also records Gateway user activity
and provides this information to FinCEN. 

FinCEN's Gateway Security Plan specifies additional controls designed
to ensure that BSA information and other sensitive data, including
information and data transmitted to FinCEN and the Detroit Computing
Center from the states, are safeguarded or protected appropriately by
Gateway users.  Among other things, this plan presents policies and
procedures that designated state officials are to follow in (1)
verifying requesters' validity and identity, (2) determining that
requests are made pursuant to bona fide criminal investigations or
other authorized purposes, and (3) maintaining records of Gateway
requests. 

A basic internal control objective for any management information
system is to protect data and programs from unauthorized access,
modification, and disclosure.  Organizations can help protect their
data by establishing controls to prevent unauthorized access and by
monitoring the access activities of individuals to help identify any
significant problems and deter individuals from inappropriate and
unauthorized activities.  According to our recent study of
information security management best practices,\25 evaluating the
effectiveness of controls and preparing summary reports for
management attention are important elements in ensuring that controls
are operating as intended. 


--------------------
\25 Executive Guide:  Information Security Management--Learning From
Leading Organizations (GAO/AIMD-98-21, Nov.  1997 Exposure Draft). 


      STATES' COMPLIANCE WITH
      CONTROLS NOT EVALUATED
---------------------------------------------------------- Letter :5.2

FinCEN has not evaluated the states' compliance with specified
policies and procedures designed to control access to and use of
information obtained through the Gateway system.  The Gateway
Security Plan contains provisions for FinCEN officials to inspect
states' records, including FinCEN staff visits to the states, to
determine if the Gateway system and the BSA data have been misused. 
However, although Project Gateway has been operational since 1994,
FinCEN had not inspected any states' records, scheduled any FinCEN
staff visits to the states, or developed any audit plans for on-site
evaluations of the states' compliance with applicable policies and
procedures, as of May 1998. 

However, FinCEN has taken some steps--based on limited sampling and
self-reporting questionnaires--to determine if selected Gateway cases
were done for official purposes and were requested by authorized
individuals.  That is, in October 1997, FinCEN screened fiscal year
1997 user data recorded by the IRS Detroit Computing Center to
identify cases that appeared to be irregular, such as those requested
by agencies that do not usually conduct criminal investigations.  In
doing so, FinCEN identified 76 cases from 23 states that appeared to
be irregular.  To obtain more information about these 76 cases, and
98 additional cases that were randomly selected from 39 states,
FinCEN distributed questionnaires to applicable state and local
agency officials.  Nine agencies did not respond to the
questionnaires. 

According to FinCEN's audit report dated May 15, 1998, the
self-reporting questionnaires indicated that the Gateway cases
sampled, for which responses were received, were made in connection
with a criminal investigation and were requested by an authorized law
enforcement official.  However, the audit report indicated that 13
agencies did not follow the proper procedures for redissemination of
Gateway information.  According to a FinCEN official, while they
proved useful, the self-reporting questionnaires do not constitute a
direct assessment of state and local compliance with controls over
Gateway information. 

FinCEN and IRS plan to increase security of the Gateway
telecommunications network by January 1999.  Even with these
improvements, control over access to BSA information at state
locations will still largely depend on the states' verification that
individual user requests for information have been properly
authorized.  Therefore, evaluating the states' compliance with
applicable policies and procedures will remain an important element
in ensuring that controls are operating as intended. 


   IRS FORM 8300 INFORMATION NOT
   ACCESSIBLE THROUGH FINCEN
------------------------------------------------------------ Letter :6

In an effort to enhance its investigative support, FinCEN is seeking
approval from IRS to provide IRS Form 8300 information (Report of
Cash Payments Over $10,000 Received in a Trade or Business) to the
law enforcement community.  Several federal and state law enforcement
officials we interviewed also told us FinCEN could enhance its
investigative assistance by providing access to selected information
from these forms.  Although IRS has taken steps to implement 1996
legislation authorizing the Secretary of the Treasury to disclose
Form 8300 information to federal, state, and local agencies,\26
several issues must still be resolved before FinCEN can obtain
approval from IRS to disseminate this information. 

Section 6050I of the Internal Revenue Code, in general, requires any
person engaged in a trade or business (other than financial
institutions required to report under the BSA) who receives more than
$10,000 in cash, in a single transaction or a series of related
transactions, to file a report with the Secretary of the Treasury. 
The Secretary of the Treasury requires this report to be filed on an
IRS Form 8300.  The form provides a paper trail that can help
identify assets acquired with illegal funds, as well as help identify
a lifestyle that is not commensurate with an individual's known
sources of legitimate income. 

The Form 8300 is designed to identify cash transactions in excess of
$10,000 between retail merchants (e.g., automobile dealers, boat
dealers, furriers, etc.) and their customers, much like the BSA
currency transaction report is designed to identify deposits,
withdrawals, exchanges, or other payments in excess of $10,000
between financial institutions and their customers.  The Form 8300
and currency transaction report provide very similar information
(e.g., name, address, Social Security number, and amount of the
transaction). 

The Form 8300 was originally created and is still used to assist IRS
in identifying individuals who might be attempting to evade taxes. 
Given that the requirement for filing the Form 8300 is contained in
the Internal Revenue Code, Form 8300 information is tax return
information and, as such, is confidential and may not be disclosed to
any persons or used in any manner not authorized by the Internal
Revenue Code.  Authorized disclosures of Form 8300 information are
subject to the procedural and recordkeeping requirements of Internal
Revenue Code section 6103.  For example, section 6103(p)(4)(E)
requires recipient agencies to file a report with the Secretary of
the Treasury that describes the procedures established and utilized
by the agency for ensuring the confidentiality of returns or return
information.  IRS requires that agencies requesting return
information, such as Form 8300 information, file a "Safeguard
Procedures Report," and that the report be approved by IRS.\27

In comparison, the currency transaction report and other BSA
information are used to target large currency transactions that may
be suspicious, support investigative cases, assist in tax examination
and collection, and support other law enforcement functions.  Under
FinCEN's authority and oversight, Treasury agencies (i.e., Customs
Service; IRS; Secret Service; and the Bureau of Alcohol, Tobacco and
Firearms) have direct, on-line access to BSA information.  If they
are jointly working a case, law enforcement agencies outside of
Treasury can access this information from one of the Treasury
agencies.  For other law enforcement agencies, FinCEN is the official
contact point within Treasury to request BSA information.  Also,
designated state and local officials have direct, on-line access to
BSA information via Project Gateway.  These officials conduct Gateway
queries for other law enforcement agencies in the state. 

According to FinCEN, agencies may receive BSA information directly
from FinCEN or via Project Gateway only after they have clearly
specified the purpose of their request, and after the identity and
authority of the requester have been confirmed by FinCEN or a
designated state official.  FinCEN regulations provide that BSA
information shall be received in confidence and shall not be
disclosed to any person except for official purposes relating to the
investigation, proceeding, or matter in connection with the
information. 


--------------------
\26 The act, entitled the "Taxpayer Bill of Rights 2," P.L.  104-168,
110 Stat.  1452 (1996), amended section 6103 of the Internal Revenue
Code. 

\27 IRS requires that Safeguard Procedures Reports contain
information on, among other things, the agency official authorized to
request Form 8300 information; the anticipated need for and use of
the information; the location, storage, and disposal of the data;
limiting access to the data; and computer security. 


      LAW ENFORCEMENT USE OF FORM
      8300 INFORMATION
---------------------------------------------------------- Letter :6.1

Although originally intended primarily to assist IRS for tax
administration purposes, Form 8300 information is also used to
support law enforcement investigations.  The Anti-Drug Abuse Act\28

of 1988 provided a special temporary rule permitting IRS to disclose
these information returns to other federal agencies for the purpose
of administering statutes not related to tax administration.  The
special rule, originally set to expire in 1990, was extended 2 years
and then expired in 1992. 

Our earlier work addressed the usefulness of Form 8300 information. 
In 1991, for example, we reported on the usefulness of Form 8300
information in conducting investigations of tax evasion and other
criminal activity, such as money laundering.\29 We reported that law
enforcement officials believed that Form 8300 information could be
instrumental in tracing cash payments by drug traffickers and other
criminals for luxury cars, jewelry, and other expensive items. 

In 1992, we reported that federal law enforcement officials regarded
Form 8300 information as extremely useful and a critical complement
to BSA reports.\30 We noted that, in an attempt to obtain Form 8300
information, several states required that copies of the form also be
filed with the state.  We recommended to Congress that (1) the
temporary and soon-to-expire authority of the Secretary of the
Treasury to disclose Form 8300 information filed under section 6050I
of the Internal Revenue Code be made permanent and (2) the Secretary
of the Treasury be allowed to disclose these returns to state law
enforcement agencies.  The IRS supported this recommendation and
noted that, if the disclosure provisions were amended, IRS would work
closely with the Treasury Department to provide access to the states. 

In 1996, the act entitled the Taxpayer Bill of Rights 2 reinstated,
on a permanent basis, the Secretary of the Treasury's authority to
disclose Form 8300 information to other federal agencies.  The
Secretary was further authorized to disclose Form 8300 information to
state, local, and foreign agencies.  The 1996 Act provided, in
general, that any such disclosure is to be made on the same basis,
and subject to the same conditions, as apply to disclosures of
information on BSA currency transaction reports.  However, the 1996
Act also required, in general, that such disclosures of Form 8300
information be subject to the disclosure and safeguard policies and
guidelines under section 6103(p) of the Internal Revenue Code. 
Consequently, according to an IRS official, Form 8300 information
must be controlled like other IRS tax return information. 

According to an IRS headquarters official, in May 1997, IRS initiated
a process for federal, state, local, and non-U.S.  law enforcement
agencies and regulatory agencies to access Form 8300 information
directly from IRS.  The official told us that, before being permitted
to obtain Form 8300 information, the requesting agency must file and
IRS headquarters must approve a Safeguard Procedures Report that
specifies how the agency plans to comply with applicable safeguard
requirements.  The IRS official added that, after an agency's plan is
approved, it can request Form 8300 information as needed from any IRS
Criminal Investigation Division field office. 

The IRS official explained that Safeguard Procedures Reports are
approved on an agency-by-agency basis.  That is, one approved report
covers all requests from the respective federal agency, such as the
U.S.  Customs Service or the FBI.  A separate report is required for
each state and local agency (e.g., police department and district
attorney's office).  The official noted that, as of February 1998,
five federal and two state law enforcement agencies had been approved
to obtain Form 8300 information from IRS Criminal Investigation
Division field offices. 


--------------------
\28 P.L.  100-690, 102 Stat.  4181 (1988). 

\29 Money Laundering:  The Use of Cash Transaction Reports by Federal
Law Enforcement Agencies (GAO/GGD-91-125, Sept.  25, 1991). 

\30 Money Laundering:  State Efforts to Fight It Are Increasing but
More Federal Help Is Needed (GAO/GGD-93-1, Oct.  15, 1992). 


      SEVERAL ISSUES MUST STILL BE
      RESOLVED BEFORE FINCEN CAN
      OBTAIN APPROVAL TO
      DISSEMINATE FORM 8300
      INFORMATION
---------------------------------------------------------- Letter :6.2

Although IRS has taken measures to provide Form 8300 information to
law enforcement agencies, several issues must still be resolved
before FinCEN can obtain approval from IRS to directly disseminate
this information or provide it to state and local agencies via
Project Gateway. 

Two days after the passage of the 1996 Act that authorized the
Secretary of the Treasury to disclose Form 8300 information to other
agencies, including law enforcement agencies, FinCEN issued a news
release commenting, in part, substantially as follows: 

  -- With the passage of this law, FinCEN anticipates being able to
     make the highly useful Form 8300 information available as part
     of its support to both federal and state law enforcement. 
     Specifically, FinCEN anticipates being granted authority to (1)
     provide Form 8300 information quickly and easily to federal law
     enforcement agencies that do not have direct access to this
     information and (2) provide Form 8300 information directly
     on-line to state law enforcement through its Project Gateway. 

FinCEN and IRS officials began discussing FinCEN's dissemination of
Form 8300 information soon after passage of the 1996 Act.  In June
1997, FinCEN and IRS officials met to discuss the difficulties IRS'
proposed operating rules for disseminating Form 8300 information (as
outlined in the Safeguard Procedures Report) would create for FinCEN
programs.  According to a FinCEN official, this meeting was followed
by a September 11, 1997, memorandum from FinCEN's Director to the
Acting Commissioner of IRS, in which FinCEN noted that the particular
safeguard procedures identified as appropriate by IRS (1) were
virtually impossible for law enforcement agencies to satisfy
efficiently and (2) appear to subject Form 8300 information to the
same safeguards as information generated by individual or corporate
tax returns.  The official noted that the memorandum proposed a
high-level meeting to determine if the issues could be resolved. 

Although FinCEN has not yet been granted approval by IRS to
disseminate Form 8300 information, an IRS administrative action
regarding such dissemination is presently under consideration by
Treasury.  According to an IRS official, in October 1997, IRS
proposed to Treasury that certain IRS implementing regulations be
modified to allow FinCEN to act on behalf of the IRS Commissioner to
disseminate Form 8300 information.\31 The IRS official noted,
however, that such dissemination would still be subject to IRS
disclosure and safeguard policies and guidelines that implement
section 6103 of the Internal Revenue Code.  According to a Treasury
official, Treasury plans to consider the IRS proposal in late 1998. 

Regarding Project Gateway, a FinCEN official called our attention to
a September 22, 1997, memorandum from FinCEN's Office of Legal
Counsel to the IRS Office of Disclosure, in which FinCEN explained
its position as to why Gateway's existing security procedures (for
electronic disclosure of BSA information to state authorities) could
be read to satisfy most of the safeguard procedures IRS was requiring
for disseminating Form 8300 information.  According to the official,
the memorandum noted that IRS procedures appeared to contemplate only
manual dissemination and that Project Gateway was apparently not
considered when the procedures were written by IRS.  The official
added that, in this memorandum, FinCEN requested that IRS accept a
certification by FinCEN with respect to all recipients of
electronically disseminated Form 8300 information rather than apply
IRS procedures to each recipient of such information. 

At the time of our review, IRS had not issued a formal response to
FinCEN's September 22, 1997, memorandum.  However, in March 1998, an
IRS official told us that FinCEN would not be granted approval to use
Project Gateway, as it currently operates, to provide state and local
agencies with Form 8300 information.  The official explained that (1)
the current Gateway system does not meet Treasury Department
requirements for systems that handle sensitive information and (2)
IRS regulations do not allow for the type of redisclosures that would
occur with Gateway.  That is, authorized Gateway users would not be
permitted to act as conduits or middlemen in providing Form 8300
information to other state and local agencies.  The official added
that, even without these barriers, IRS regulations would not allow
FinCEN to certify safeguard procedures on behalf of all recipient
agencies.  Rather, each agency that requested Form 8300 information
via Project Gateway would be required to file a Safeguard Procedures
Report. 

In summary, there are several issues that must be resolved before
FinCEN can disseminate Form 8300 information or provide it to state
and local agencies via Project Gateway.  At this point, it is not
clear if these issues can be resolved administratively among FinCEN,
IRS, and Treasury or if Congress may need to revisit these issues in
the future. 


--------------------
\31 IRS regulations provide, in general, that the Commissioner of
Internal Revenue may authorize another federal agency, to which
returns and return information have been disclosed by the IRS, to
further disclose such returns and return information to certain other
agencies. 


   CONCLUSIONS
------------------------------------------------------------ Letter :7

Over the years, FinCEN has faced a challenge of defining its support
role in providing the law enforcement community with value-added
products and services in a timely manner.  In and of itself, this is
a significant challenge, and given the existence of various
intelligence and investigative support centers and other sources of
information and financial analysis, it is likely to be a continuing
challenge.  Also, compounding this challenge, since its inception in
1990, FinCEN's staffing levels have remained fairly constant, while
its mission has expanded beyond its original law enforcement support
role to include responsibility for promulgating BSA regulations and
assuming a leadership position in international efforts to combat
money laundering. 

Nonetheless, to FinCEN's credit, the agency has expanded its law
enforcement support role from providing tactical and strategic
products to include use of an artificial intelligence system and
implementation of self-help services (i.e., the platform concept and
Project Gateway).  However, due in part to reallocations of staff to
other mission functions, the expansion of FinCEN's line of products
and services has, at times, experienced some growing pains.  These
include, for example, occasional backlogs and lengthy response times
for routine tactical support requests and reductions in the number of
strategic and artificial intelligence products generated. 

Similarly, the expansion or evolution of FinCEN's line of products
and services inherently entails some "role-definition" issues.  To
the extent feasible, for instance, FinCEN considers its expertise
best used to support nonroutine cases or analytically intensive
cases, which may include tactical and strategic as well as artificial
intelligence support.  In recent years, however, while users reported
that all types of FinCEN's tactical support have been useful, an
increasing number of FinCEN's tactical reports consisted of database
extracts versus detailed analysis.  And, the number of analytical
products FinCEN generated from strategic and artificial intelligence
analyses decreased due, in part, to users' concerns about the
usefulness of these products. 

To better focus its efforts and ensure usefulness, particularly
regarding strategic and artificial intelligence products, FinCEN has
begun various "partnering" arrangements with its customers.  Also, to
better ensure that its tactical staff are available to support
traditional law enforcement needs, in 1995, FinCEN discontinued
accepting requests for background and regulatory investigations. 

FinCEN is also taking steps to ensure that its tactical staff are
used to support complex or significant rather than routine cases. 
However, FinCEN has not developed and disseminated general criteria
or guidelines concerning the types of cases that FinCEN can best
support and the most appropriate uses for FinCEN's resources.  We
believe such guidance would help FinCEN better utilize its tactical
resources.  In a similar vein, while FinCEN had taken steps to inform
the law enforcement community about its support, some federal law
enforcement officials said they were not aware of the various
products and services offered by FinCEN.  (FinCEN has since taken
additional steps to inform the law enforcement community about its
products and services.)

FinCEN's efforts to promote self-help services have been well
received by the users of those services we interviewed.  They
commented very favorably about the platform concept and Project
Gateway.  These comments are corroborated by statistics showing that
use of these systems has steadily increased.  In turn, however, the
increased use of these systems underscores the importance of FinCEN's
controls to ensure that only authorized personnel access the systems
and use them for legitimate purposes.  Because FinCEN has not
evaluated states' compliance with Gateway security policies and
procedures, it cannot be sure the controls are working as intended. 
Periodically evaluating the states' compliance with these controls
and preparing summary reports to management would help FinCEN
identify improvements or deterioration in control effectiveness,
reassess the related risks, and take appropriate action. 

FinCEN and IRS have been unable to resolve administrative issues
related to FinCEN's dissemination of IRS Form 8300 information.  Law
enforcement officials believe this information can be useful in
tracing cash payments by drug traffickers and other criminals for
luxury cars, jewelry, and other expensive items.  We recognize the
long-standing concerns regarding the protection of taxpayer
information, although we note that the Form 8300 contains the same
basic information as is contained in the currency transaction report,
which is more readily available to law enforcement.  IRS has
initiated a process for law enforcement agencies to access Form 8300
information directly from IRS.  However, regarding FinCEN's
dissemination of Form 8300 information, additional issues must still
be resolved to address IRS' and Congress' concerns about protection
of taxpayer information.  In addressing these issues, consideration
must also be given to the value of Form 8300 information to law
enforcement efforts in reducing money laundering.  It is not clear if
these issues can be resolved administratively among FinCEN, IRS, and
Treasury or if Congress may need to revisit these issues in the
future. 


   RECOMMENDATIONS TO THE
   SECRETARY OF THE TREASURY
------------------------------------------------------------ Letter :8

We recommend that the Secretary of the Treasury take action to
identify and resolve any administrative issues related to FinCEN's
dissemination of Form 8300 information to the law enforcement
community.  If FinCEN is given approval to disseminate Form 8300
information, we recommend that, prior to such dissemination, the
Secretary ensure that sufficient access and disclosure controls over
FinCEN's programs and systems are in place and operating as intended. 
That is, these controls should be sufficient to protect against
unauthorized access to and disclosure of sensitive information. 


   RECOMMENDATIONS TO THE ACTING
   DIRECTOR, FINCEN
------------------------------------------------------------ Letter :9

We recommend that the Acting Director, FinCEN

  -- incorporate into the agency's communications with potential
     customers general criteria or guidance on the types of cases
     that FinCEN can best support and the most appropriate uses of
     FinCEN's capabilities.  At a minimum, for example, such criteria
     or guidance could be included in information on FinCEN's site on
     the Internet. 

  -- develop and implement a program for on-site evaluations of the
     states' compliance with control policies and procedures related
     to Project Gateway.  Such a program should include periodic
     testing of the controls and summary reports for management
     attention. 


   AGENCY COMMENTS AND OUR
   EVALUATION
----------------------------------------------------------- Letter :10

We requested comments on a draft of this report from the Department
of the Treasury, FinCEN, and the Department of Justice.  On May 29,
1998, the Director of Treasury's Office of Finance and Administration
informed us, via telephone, that the Department declined to provide
comments on the draft report or on the recommendation to the
Secretary of the Treasury related to FinCEN's dissemination of IRS
Form 8300 information. 

In a letter dated May 20, 1998, FinCEN's Acting Director provided us
written comments (see app.  VI).  The Acting Director did not concur
with our draft report's recommendation that FinCEN incorporate into
the agency's communications with potential customers specific
criteria or guidance on the types of cases that FinCEN can best
support and the most appropriate uses of FinCEN's capabilities.  The
Acting Director commented that specific guidelines are impractical
because (1) cases differ widely in circumstance and detail and (2)
guidelines may result in perceived limitations in the inherent
flexibility of FinCEN's products and services.  Also, the Acting
Director noted that a more effective and efficient way to communicate
is through the generalized meetings and other outreach efforts that
FinCEN conducts, although those efforts are already very ambitious
for a small agency like FinCEN. 

The intent of our draft recommendation was not that FinCEN develop or
communicate inflexible or rigid guidelines for its customers. 
Rather, given the availability of various other intelligence and
investigative support centers that are alternate sources of
financial, commercial, and law enforcement data and analysis, we
believe that FinCEN's capabilities would be most fully and
appropriately utilized if the agency widely communicates generally
applicable criteria or guidance on the types of cases that it can
best support.  For example, a FinCEN official told us that tactical
support is generally best suited for (1) cases that involve large
criminal organizations, are of significance to the jurisdiction, or
require expert financial analysis and (2) cases where support is not
available from in-house sources. 

In our opinion, this kind of general guidance could be more widely
disseminated to the law enforcement community, at a minimum, by being
included in information on FinCEN's site on the Internet.  As
previously noted, FinCEN updated its Internet site in March 1998 to
include information on the types of support it provides to the law
enforcement community.  However, the update did not include general
criteria or guidance on the types of cases that FinCEN can best
support or on the most appropriate uses of FinCEN's capabilities. 
Accordingly, with some modifications to the language in this report
for clarification purposes, we are still recommending that FinCEN
better communicate to potential customers general criteria or
guidance on the types of cases that it can best support and the most
appropriate uses of its capabilities. 

On the other hand, the Acting Director concurred with our
recommendation that FinCEN develop and implement a program for
on-site evaluations of the states' compliance with control policies
and procedures related to Project Gateway.  Toward this end, the
Acting Director noted that FinCEN had recently (1) reported on an
early-1998 audit of Gateway use and activity and (2) issued a
comprehensive Gateway Coordinator's Handbook, which presents
up-to-date rules, procedures, formats, dissemination instructions,
and warning notices. 

Also, the Acting Director noted that FinCEN is beginning to develop
guidelines for FinCEN staff to use in reviewing the compliance of
state coordinators with policies and procedures.  In the near future,
according to the Acting Director, FinCEN will (1) use these
guidelines to conduct audits in selected states and (2) share audit
results with the state coordinators to provide a lessons-learned
approach. 

On May 8, 1998, we received written comments from Justice that
indicated that the draft was reviewed by representatives of the
Criminal Division, DEA, FBI, and the Executive Office for U.S. 
Attorneys.  Generally, the Department concurred with the substance of
the draft report.  Justice provided technical comments and also made
suggestions for possible expansion of our discussion related to
various topics in the draft report.  These technical comments and
suggestions have been incorporated in this report where appropriate. 


--------------------------------------------------------- Letter :10.1

We are sending copies of this report to the Chairmen and Ranking
Minority Members of the Senate Banking, Housing, and Urban Affairs
Committee; the Senate Finance Committee; the House Ways and Means
Committee; the Subcommittee on Treasury, Postal Service, General
Government, and Civil Service (Senate Appropriations Committee); and
the Subcommittee on Treasury, Postal Service, and General Government
(House Appropriations Committee); the Secretary of the Treasury; the
Acting Director, FinCEN; the Director, Office of Management and
Budget; the Attorney General; and other interested parties.  Copies
will also be made available to others upon request. 

The major contributors to this report are listed in appendix VII.  If
you or your staffs have any questions about this report, please
contact me on (202) 512-8777. 

Richard M.  Stana
Associate Director
Administration of Justice Issues


OBJECTIVES, SCOPE, AND METHODOLOGY
=========================================================== Appendix I

This report focuses on the Financial Crimes Enforcement Network's
(FinCEN) products and services in support of law enforcement. 
Specifically, this report discusses (1) trends in the types and
quantities of products and services provided by FinCEN to the law
enforcement community; (2) the extent to which FinCEN's products and
services have been considered useful by the law enforcement community
in identifying, developing, or prosecuting money laundering and other
financial crime cases; (3) the extent to which FinCEN evaluates the
states' compliance with applicable controls over access to and use of
information when state law enforcement officials directly access
FinCEN's resources; and (4) FinCEN's efforts to obtain approval from
the Internal Revenue Service (IRS) to provide IRS Form 8300
information (Report of Cash Payments Over $10,000 Received in a Trade
or Business)\1 to the law enforcement community. 

To determine the trends in and usefulness of FinCEN's products and
services, we focused on five principal types of support provided by
FinCEN:  (1) tactical products in support of agencies' ongoing
investigations; (2) strategic products designed to address
longer-term or more broadly scoped topics; (3) artificial
intelligence products designed to provide investigative leads; (4) a
"platform concept," whereby federal agency officials can access
databases and do their own research using FinCEN resources; and (5)
"Project Gateway," which provides state agencies with remote, on-line
access to financial and some commercial data.  The scope of our work
did not include a review of the Suspicious Activity Reporting System,
which is administered by FinCEN. 


--------------------
\1 Under the Internal Revenue Code, any person who receives more than
$10,000 in cash in one transaction (or two or more related
transactions) in the course of trade or business generally must file
an information return (IRS Form 8300) with IRS specifying the name,
address, and taxpayer identification number of the person from whom
the cash was received and the amount of cash received. 


   SCOPE AND METHODOLOGY OF OUR
   WORK REGARDING TRENDS IN THE
   TYPES AND QUANTITIES OF SUPPORT
   PROVIDED BY FINCEN TO THE LAW
   ENFORCEMENT COMMUNITY
--------------------------------------------------------- Appendix I:1

To determine the trends in the types of support provided by FinCEN to
the law enforcement community, we obtained a description of FinCEN's
primary products and services, including the date when each type of
product or service first became available.  Also, to obtain an
understanding of the processes and purposes for each type of product
or service, we interviewed officials of FinCEN's component offices. 
To determine the trends in the quantities of support provided since
1990, including which agencies requested which types of support, we
obtained and analyzed FinCEN workload documents and relevant
information from FinCEN's computerized databases.  Also, to ascertain
reasons for the trends in FinCEN's support, we interviewed FinCEN
officials, Treasury and Justice headquarters' officials, long-term
agency detailees to FinCEN, and state agency officials. 


   SCOPE AND METHODOLOGY OF OUR
   WORK REGARDING THE EXTENT TO
   WHICH FINCEN'S PRODUCTS AND
   SERVICES HAVE BEEN USEFUL TO
   THE LAW ENFORCEMENT COMMUNITY
--------------------------------------------------------- Appendix I:2

FinCEN generates a variety of products and services to support many
different agencies or organizations.  Therefore, to the extent
practical, we used mail surveys and in-person and telephone
interviews to collect information on the usefulness of the full range
of FinCEN's support.  This included support of money laundering and
other financial crimes investigations as well as support of
regulatory or background checks for employment, security clearance,
and business or professional licensing purposes.  We surveyed or
interviewed officials from

  -- the principal federal law enforcement agencies within the
     Treasury Department (i.e., IRS; the Bureau of Alcohol, Tobacco
     and Firearms; the Customs Service; and the Secret Service);

  -- the principal federal law enforcement agencies within the
     Justice Department (i.e., the Federal Bureau of Investigation
     (FBI), the Marshals Service, the Drug Enforcement Administration
     (DEA), U.S.  Attorney Offices, and the Immigration and
     Naturalization Service);

  -- other federal agencies that were high-volume requesters of
     FinCEN's products and services, such as the Postal Inspection
     Service, the Department of Agriculture, the Air Force Office of
     Special Investigations, and the Naval Criminal Investigative
     Service (see table I.1 at the end of this app.  for a complete
     list of the high-volume requesters we surveyed); and

  -- state requesters of FinCEN support. 

During our inquiries, we obtained case-specific information (if
available and appropriate) as well as opinions, views, and anecdotal
information from managers and agents of the applicable agencies and
organizations.  We focused on relatively recent support--that is,
support generated since January 1996.  We obtained opinions and views
on support provided before January 1996 but could not provide a
statistically valid or meaningful comparison of the usefulness of
FinCEN today versus FinCEN in the past. 

The following sections provide additional details regarding our scope
and methodology for determining the extent to which each of FinCEN's
five principal types of products and services have been useful to the
law enforcement community in identifying, developing, investigating,
or prosecuting cases,


      TACTICAL SUPPORT
------------------------------------------------------- Appendix I:2.1

We focused on requests for tactical products that FinCEN received or
logged in during the 12-month period from April 1996 through March
1997 and responded to before August 1997.  To obtain general or
broad-based indications of the usefulness of FinCEN's full range or
types of tactical support, we surveyed by mail a stratified,
statistical sample of 352 federal agency officials who requested
various tactical work products (i.e., database extracts, analytical
reports, and expedited reports).\2 The 352 federal officials were
selected from 2,379 officials in 31 agencies that, according to
FinCEN's records, made 4,409 (95 percent) of the 4,626 total federal
requests for FinCEN's tactical support during the 12-month period
(see table I.1 at the end of this app.).  We did not survey officials
from the 34 agencies that made the remaining 217 (5 percent) federal
tactical requests because of time constraints and the logistics
involved in coordinating surveys with these agencies. 

We received usable responses from 251 (71 percent) of the 352 federal
officials.  The responses were weighted to represent the 2,379
federal officials who made the 4,409 requests for tactical support
during the survey period.  Unless otherwise noted, the estimates of
percentages from this sample in the report have 95-percent confidence
intervals of no more than plus or minus 10 percentage points. 

Also, we surveyed by mail a simple random sample of 95 state agency
officials who requested FinCEN's tactical support during the 12-month
period.\3 The 95 state officials were from 15 states that made 18 or
more requests in the period.  According to FinCEN's records, these 15
states made 813 (75 percent) of the 1,077 total requests made through
all state coordinators for FinCEN tactical support during the
12-month period (see table I.2 at the end of this app.).  State
coordinators were able to identify 504 requesters of FinCEN's
tactical support in these 15 states.  We randomly selected 95 users
for our survey.  After making at least 2 attempts to contact
nonrespondents by telephone, we received usable responses from 54 (57
percent) of the 95 state officials.  The percentage estimates from
this sample in the report have 95-percent confidence intervals of no
more than plus or minus 13 percentage points and are projectable to
all of the requesters in the 15 states. 

Further, to obtain additional perspectives on the usefulness of
FinCEN's tactical support, we conducted in-person or telephone
interviews with (1) officials from the FBI and the Bureau of Alcohol,
Tobacco and Firearms who were on detail to FinCEN at the time of our
review and (2) senior FinCEN officials responsible for overseeing
tactical support. 

In addition to these surveys and interviews, we determined why some
federal and state agencies and offices submitted few if any requests
for FinCEN's tactical support.  Based on our review of FinCEN's
records, we determined that many federal field offices, states, and
High Intensity Drug Trafficking Areas (HIDTA) requested little or no
tactical support from FinCEN during the period we reviewed, April
1996 through March 1997.  To determine the reasons why: 

  -- We surveyed by mail all U.S.  field offices of the IRS (33
     offices), the U.S.  Customs Service (20 offices), the FBI (56
     offices), and the DEA (20 offices) (see app.  V for a copy of
     the questionnaire).  Of the total 129 field offices, we received
     responses from 116 (90 percent). 

  -- Also, we interviewed officials from 9 of the 10 states that,
     according to FinCEN's records, made either one or no requests
     for FinCEN's tactical support from April 1, 1996, through March
     31, 1997.\4

  -- Further, we surveyed all 22 HIDTAs that had been designated by
     the Office of National Drug Control Policy at the time of our
     review to determine the extent to which FinCEN products are used
     to support their money laundering and other financial crimes
     investigations.  We received responses from 16 HIDTAs.  We did
     not project these responses to the 22 HIDTAs. 

In our mail surveys and interviews, we ascertained what other
(non-FinCEN) sources of financial, commercial, and/or law enforcement
information and analyses are used by the federal field offices, state
agencies, and HIDTAs. 

Further, to determine how FinCEN informed the law enforcement
community about the availability of FinCEN's products and services,
we (1) reviewed the results of FinCEN's May 1996 outreach efforts,
which were designed to obtain feedback from federal law enforcement
agencies; (2) reviewed FinCEN's records of participation at meetings,
training events, and conferences at the federal and state levels; and
(3) interviewed FinCEN officials. 


--------------------
\2 See appendix III for a copy of the questionnaire. 

\3 See appendix IV for a copy of the questionnaire. 

\4 The tenth state--South Dakota--did not respond to our telephone
inquiries. 


      STRATEGIC SUPPORT
------------------------------------------------------- Appendix I:2.2

To determine the usefulness of FinCEN's strategic support, we
conducted telephone interviews with law enforcement officials who
received or used three of the four strategic products FinCEN prepared
from January 1996 through December 1997 (i.e., the Louisiana,
Georgia, and South Florida money-laundering threat assessments).\5
Also, we interviewed senior FinCEN officials about changes in
strategic support, including how the newly formed Office of Research
and Analysis was functioning. 


--------------------
\5 The fourth report addressed emerging cyberpayment technology and
its potential use in criminal activities. 


      ARTIFICIAL INTELLIGENCE
      SUPPORT
------------------------------------------------------- Appendix I:2.3

To determine the usefulness of FinCEN's artificial intelligence
products, we attempted to survey by mail all federal and state law
enforcement officials identified in FinCEN's database as having
received artificial intelligence products from April 1996 through
March 1997.  First, we determined that FinCEN had provided 85
artificial intelligence products during this period.  Next, we
attempted to locate and survey the end user for each product.  We
identified and mailed a survey questionnaire to 13 officials who were
end users for 51, or 60 percent, of the 85 products.\6 If more than
one product was associated with the same investigative effort, we
requested that only one questionnaire be completed for the associated
products.  We received responses from all 13 officials.\7

Further, we reviewed FinCEN data to determine the extent to which
FinCEN self-initiated artificial intelligence products (versus
products requested by law enforcement agencies).  We also interviewed
FinCEN officials to ascertain reasons for variations in the number of
self-initiated products over the years, including the extent to which
the self-initiated products were used by the law enforcement
community. 


--------------------
\6 According to FinCEN's records, 12 individuals requested the
remaining 34 artificial intelligence products.  We contacted by
telephone 8 of the 12 requesters who told us they could not readily
identify the person(s) who eventually used the products.  We made at
least two attempts to contact the other four individuals by
telephone. 

\7 Two of the 13 officials received artificial intelligence products
that were used to support 2 separate and unrelated cases.  Therefore,
we received a total of 15 questionnaires. 


      PLATFORM CONCEPT
------------------------------------------------------- Appendix I:2.4

We conducted in-person or telephone interviews with officials from
three agencies that, according to FinCEN officials, were frequent
platform users--the Washington/Baltimore HIDTA; the Bureau of
Alcohol, Tobacco and Firearms; and the Defense Criminal Investigative
Service.  According to FinCEN's year-end records, the three agencies
we contacted were among the four agencies that most frequently used
platforms in 1997.\8 The three agencies accounted for 971 (66
percent) of the 1,477 investigative cases that were supported by
platforms in 1997.  We asked these officials about their use of
platforms, including the number and type of cases they support, the
databases they access, and their views about the usefulness of the
platform concept. 


--------------------
\8 The Capitol Police was the fourth agency. 


      PROJECT GATEWAY
------------------------------------------------------- Appendix I:2.5

We conducted telephone interviews with officials\9 from the four
states that made the most Gateway queries in fiscal year
1997--California, Illinois, Pennsylvania, and Texas.  These 4 states
made 22,876 (40 percent) of the total 57,663 Gateway queries in
fiscal year 1997.  We also interviewed officials from two
states--Montana and Wyoming--that used Gateway fewer than 50 times in
fiscal year 1997 to determine reasons for the low use.  We
interviewed FinCEN officials to determine future Gateway plans and to
identify efforts to provide Gateway access for non-Treasury federal
agencies. 


--------------------
\9 The state officials we interviewed were responsible for
coordinating the respective state's requests for access to Bank
Secrecy Act data via Gateway. 


   SCOPE AND METHODOLOGY OF OUR
   WORK REGARDING THE EXTENT TO
   WHICH FINCEN EVALUATES THE
   STATES' COMPLIANCE WITH
   APPLICABLE CONTROLS OVER
   PROJECT GATEWAY
--------------------------------------------------------- Appendix I:3

To determine how FinCEN controls access to and use of information
obtained via Project Gateway, we (1) interviewed FinCEN officials and
Gateway users and (2) reviewed FinCEN's Gateway Security Plan (July
1996) and the IRS Detroit Computing Center's Project Gateway Users
Guide (May 1997).  Also, we interviewed FinCEN officials to determine
the extent to which FinCEN had evaluated the states' compliance with
applicable controls over access to and use of information obtained
via Project Gateway.  It was beyond the scope of our study to audit
or test the controls or safeguards FinCEN maintains over Project
Gateway. 


   SCOPE AND METHODOLOGY OF OUR
   WORK REGARDING FINCEN EFFORTS
   TO PROVIDE IRS FORM 8300
   INFORMATION TO THE LAW
   ENFORCEMENT COMMUNITY
--------------------------------------------------------- Appendix I:4

To determine the potential usefulness of IRS Form 8300 information
for law enforcement purposes, we reviewed previous GAO reports and
interviewed federal and state law enforcement officials.  To
determine what disclosure and safeguard requirements apply to the
dissemination of Form 8300 information and Bank Secrecy Act data, we
(1) reviewed applicable provisions in the 1996 act that authorized
the use of Form 8300 information for law enforcement purposes, (2)
interviewed IRS officials and reviewed applicable sections of the
Internal Revenue Code, and (3) interviewed FinCEN officials and
reviewed applicable documentation. 

Also, to determine how IRS currently makes Form 8300 information
available to federal, state, local, and non-U.S.  law enforcement
agencies, we (1) interviewed IRS officials and (2) reviewed safeguard
and reporting requirements contained in IRS' guidance for completing
a Safeguard Procedures Report.  To determine the status of FinCEN's
efforts to obtain approval from IRS to provide IRS Form 8300 to the
law enforcement community, we interviewed FinCEN, IRS, and Treasury
Department officials. 



                               Table I.1
                
                 Number of Federal Agency Requests for
                  FinCEN Tactical Support, April 1996-
                               March 1997

                                     Number of    Category  Cumulative
Agency                                requests     percent     percent
----------------------------------  ----------  ----------  ----------
Justice agencies we surveyed
----------------------------------------------------------------------
Federal Bureau of Investigation            662
U.S. Marshals Service                      323
Drug Enforcement Administration            202
U.S. Attorneys                             128
Immigration and Naturalization              56
 Service
Department of Justice                       26
======================================================================
Total for Justice agencies we            1,397       30.2%       30.2%
 surveyed

Treasury agencies we surveye
----------------------------------------------------------------------
Internal Revenue Service                   413
Bureau of Alcohol, Tobacco and             341
 Firearms
U.S. Customs Service                       216
U.S. Secret Service                         34
Department of the Treasury                  29
======================================================================
Total for Treasury agencies we           1,033        22.3        52.5
 surveyed

Other federal agencies we su
----------------------------------------------------------------------
Central Intelligence Agency                508
Naval Criminal Investigative               226
 Service
U.S. Postal Inspection Service             175
Department of Agriculture                  168
Air Force Office of Special                117
 Investigations
Department of Labor                         98
Defense Criminal Investigative              92
 Service
Small Business Administration               77
Federal Deposit Insurance                   74
 Corporation
Department of Housing and Urban             60
 Development
Army Criminal Investigation                 58
 Command
Office of the Secretary of the Air          53
 Force
Department of Health and Human              47
 Services
Federal Trade Commission                    41
Securities and Exchange Commission          41
U.S. Probation and Parole Office            33
Department of Education                     31
Railroad Retirement Board                   30
Department of Commerce                      27
Social Security Administration              23
======================================================================
Total for other federal agencies         1,979        42.8      95.3\a
 we surveyed

Federal agencies we did not
----------------------------------------------------------------------
Department of Energy                        19
Department of State                         19
Environmental Protection Agency             17
U.S. Army (other)\b                         15
Department of the Interior                  13
Veterans Administration                     11
Food and Drug Administration                11
Office of the Comptroller of the            10
 Currency
U.S. Information Agency                      9
Government Printing Office                   9
Fish and Wildlife Commission                 8
Department of Transportation                 8
National Aeronautics and Space               8
 Administration
Nuclear Regulatory Commission                7
Office of Foreign Asset Control              7
U.S. Coast Guard                             6
Various other agencies\c                    40
======================================================================
Total for federal agencies we did          217         4.7
 not survey
======================================================================
Grand total                              4,626                  100.0%
----------------------------------------------------------------------
\a For our survey questionnaire to federal agency officials (see app. 
III), we selected a sample from the 31 agencies listed above; these
agencies made 95 percent of the total federal requests for tactical
support during the survey period (April 1996-March 1997). 

\b As listed earlier in this table, the only Army component we
surveyed was the Army Criminal Investigation Command. 

\c This group of requesters includes a total of 18 agencies, each of
which made 5 or fewer requests for FinCEN tactical support during the
survey period.  For example, the Tennessee Valley Authority made five
requests, and the Federal Communications Commission made one request. 

Source:  Developed by GAO based on FinCEN data. 



                               Table I.2
                
                  Number of State Requests for FinCEN
                Tactical Support, April 1996-March 1997

Agency                          Number of requests             Percent
------------------------------  ------------------  ------------------
States we surveyed\a
----------------------------------------------------------------------
Pennsylvania                                   124
California                                     115
New York                                        93
Ohio                                            86
Texas                                           59
Indiana                                         59
Arizona                                         46
Illinois                                        44
Michigan                                        36
Connecticut                                     32
North Carolina                                  32
New Mexico                                      25
Virginia                                        25
Colorado                                        19
Wisconsin                                       18
======================================================================
Total for states we surveyed                   813               75.5%
Other states we did not                        264                24.5
 survey\b
======================================================================
Grand total                                  1,077               100.0
----------------------------------------------------------------------
\a We surveyed 15 states, each of which made 18 or more requests for
FinCEN tactical support during the period April 1996 to March 1997. 

\b This group consists of 35 states (that made 264 total requests)
and the District of Columbia (11 requests).  Of the 35 states, 5
states made 1 request each, and 6 states, plus Puerto Rico, made no
requests. 

Source:  Developed by GAO based on FinCEN data. 


DATABASES AND OTHER INFORMATION
RESOURCES USED BY FINCEN
========================================================== Appendix II

FinCEN's information resources fall into four broad categories: 
financial, commercial, and law enforcement databases, and its own
internal databases. 


   FINANCIAL DATABASE
-------------------------------------------------------- Appendix II:1

The financial database consists of reports that are required to be
filed under the Bank Secrecy Act (BSA).  These reports can be
retrieved from either the U.S.  Customs Service's computer center in
Newington, Virginia, or the Internal Revenue Service's (IRS) Detroit
Computing Center.  Both computer center systems are accessible
on-line at FinCEN through either the Treasury Enforcement
Communications System or the Currency and Banking Retrieval System. 
The financial database is available to bona fide entities\1 for
criminal, civil, regulatory, and foreign investigations.\2 Table II.1
describes the various BSA reports used by FinCEN. 



                               Table II.1
                
                       BSA Reports Used by FinCEN

BSA report                          Description
----------------------------------  ----------------------------------
Currency Transaction Report (IRS    Filed by bank and nonbank
Form 4789)                          financial institutions reflecting
                                    cash transactions of more than
                                    $10,000. The database contains
                                    reports since 1983.

Currency Transaction Report by      Filed by casinos reflecting cash
Casinos (IRS Form 8362) and the     transactions of more than $10,000.
Nevada Currency Transaction Report  The database contains reports
by Casinos                          since 1985.

Report of International             Filed by persons carrying currency
Transportation of Currency or       or certain monetary instruments in
Monetary Instruments (U.S. Customs  aggregate amounts of more than
Service Form 4790)                  $10,000 into or out of the United
                                    States. The database contains
                                    reports since 1989.

Report of Foreign Bank and          Filed annually by U.S. persons
Financial Accounts (Treasury        with interest in, or signature
Department Form 90-22.1)            authority over, bank securities or
                                    other financial accounts in a
                                    foreign country, which exceed
                                    $10,000 in total value at any time
                                    during a calendar year. The
                                    database contains reports since
                                    1985.

Suspicious Activity Report          Filed by financial institutions
(Treasury Department Form 90-       which determine that some activity
22.47)                              is suspicious.\a The database
                                    contains reports since April 1996
                                    and some incomplete history files
                                    dating back to 1990.
----------------------------------------------------------------------
\a FinCEN regulations provide, for example, that a suspicious
activity report shall be filed for a transaction that has no business
or apparent lawful purpose or that is not the sort in which the
particular customer would normally be expected to engage, and the
bank knows of no reasonable explanation for the transaction after
examining the available facts, including the background and possible
purpose of the transaction. 

Source:  FinCEN. 

There are some restrictions on the release of Currency Transaction
Reports, Currency Transaction Reports by Casinos, Reports of
International Transportation of Currency or Monetary Instruments, or
Reports of Foreign Bank and Financial Accounts.  For example, FinCEN
regulations provide that such information made available to other
departments or agencies of the United States, any state or local
government, or any foreign government shall be received in confidence
and shall not be disclosed to any person except for official purposes
relating to the investigation, proceeding, or matter in connection
with the information.  In addition, written approval must be obtained
from the FinCEN Director prior to releasing BSA data to foreign
governments.  Suspicious Activity Reports are only researched and
disseminated for designated user agencies\3 and certain state agency
officials. 


--------------------
\1 Bona fide entities consist of federal, state, and local law
enforcement agencies; regulatory agencies; and foreign governments. 

\2 Foreign requests require written approval of the FinCEN Director. 

\3 At the time of our review, designated user agencies consisted of
the Federal Deposit Insurance Corporation, Federal Reserve,
Comptroller of the Currency, National Credit Union Association,
Office of Thrift Supervision, Federal Bureau of Investigation,
IRS-Criminal Investigation Division, U.S.  Secret Service, U.S. 
Customs Service-Office of Investigations, and Executive Office of the
U.S.  Attorneys or U.S.  Attorneys' Offices. 


   COMMERCIAL DATABASES
-------------------------------------------------------- Appendix II:2

FinCEN procures access to a variety of commercial databases that are
used to locate individuals, determine asset ownership, and establish
links between individuals, businesses, and assets.  These databases
encompass a wide variety of demographic and census information,
including corporate ownership, nationwide telephone directories, real
estate purchases, and courthouse records.  There are no restrictions
on the release of commercial information to bona fide entities for
criminal, civil, regulatory, and foreign investigations.  Table II.2
describes the commercial databases FinCEN had access to as of
February 1998.\4



                               Table II.2
                
                  Commercial Databases Used by FinCEN

Database            Description
------------------  --------------------------------------------------
Autotrack           Provides on-line access to public records from
Plus                selected states, with broad coverage of Florida
                    and Texas (e.g., driver's license number, motor
                    vehicle registration, corporation filings, and
                    real estate ownership) and less broad coverage of
                    public records from New York, New Hampshire,
                    Oregon, and Washington. Also available is a
                    nationwide "National Dossier," which compiles
                    current and former addresses, telephone numbers,
                    and a listing of neighbors.

CBI-Equifax         One of the nation's large credit bureau services.
                    FinCEN, by law, obtains only the "header"
                    information from an individual's credit report,
                    which comprises current and former addresses,
                    possibly a spouse's name, and the month and year
                    of birth. FinCEN does not obtain any information
                    from the credit report relative to the credit
                    history or employment history of an individual.

CDB-InfoTek         Provides access to on-line public records that can
                    assist in locating people and businesses,
                    identifying and verifying assets, exposing and
                    controlling fraud, and uncovering and verifying
                    background information.

Dallas Computer     Provides on-line access to names, dates of birth,
Services            addresses, phone numbers, and driver's license
National Inquiry    information.
System

Dialog              A gateway to over 450 databases that provide
                    access to millions of documents in the area of
                    business news and industry analysis from full-
                    text trade journals, newspapers, and news wires.
                    Domestic and international company directories are
                    also available, many of which include corporate
                    financial statistics.

Dun and Bradstreet  Provides access to business information from
                    millions of companies worldwide. For companies in
                    the United States, a complete business information
                    report is available, which includes a corporate
                    history, financial data, banking relationships,
                    public record filings, and biographical
                    information on key officers. For international
                    companies, directory information includes the
                    names of up to 10 executives. All companies in the
                    database are identified by a unique number, which
                    allows a complete "corporate family tree" to be
                    produced.

Information         One of the major providers of on-line public
America             records, providing access to the corporation and
                    limited partnership filings from most states; an
                    asset locator that searches real property records
                    (including tax assessor records and deed
                    transfers) from over 30 states as well as
                    nationwide Federal Aviation Administration
                    aircraft and Coast Guard watercraft files;
                    nationwide business bankruptcy filings; a
                    nationwide people finder service; and an extensive
                    file of liens and judgments from every state.

Lexis-Nexis         One of the major providers of on-line public
                    records with extensive files of real estate
                    ownership and deed transfers, state corporation
                    and limited partnership filings, personal and
                    business bankruptcies, nationwide liens and
                    judgments, a nationwide person and business
                    locator file, and civil and criminal indices. The
                    service also provides full-text access to more
                    than 2,300 information sources from U.S. and
                    overseas newspapers, magazines, journals,
                    newsletters, wire services, and broadcast
                    transcripts. A comprehensive company library
                    provides numerous files of U.S. company reports,
                    international company reports, bankruptcy filings,
                    business biographical reports, and Securities and
                    Exchange Commission filings.

National            Provides biographical information on stock broker/
Association of      dealer businesses and individuals.
Securities Dealers

PhoneDisc           A set of five CD-ROM discs providing a nationwide
                    telephone directory on both individuals and
                    businesses. This database contains only listed
                    telephone numbers.
----------------------------------------------------------------------
Source:  FinCEN. 


--------------------
\4 According to a FinCEN official, the commercial databases FinCEN
procures access to may vary each year. 


   LAW ENFORCEMENT DATABASES
-------------------------------------------------------- Appendix II:3

Through written agreements outlining the details of database access,
dissemination authority, etc., FinCEN has access to some
investigative databases maintained by other law enforcement agencies. 
FinCEN provides information from these law enforcement databases to
bona fide entities for domestic criminal investigations.  Several
queries of Treasury Department's system (see table II.3) are also
available for domestic regulatory and foreign investigations.  Table
II.3 describes the law enforcement databases FinCEN had access to as
of November 1997. 



                               Table II.3
                
                   Law Enforcement Databases Used by
                         FinCEN, November 1997

Agency              Description\a
------------------  --------------------------------------------------
Department of the   The Treasury Enforcement Communications System is
Treasury            administered by the U.S. Customs Service on behalf
                    of a consortium of over 25 federal agencies, for
                    the purpose of providing a broad scope of
                    information for law enforcement purposes. In
                    addition to providing information on U.S. Treasury
                    law enforcement investigations, it provides access
                    to the National Crime Information Center, the
                    National Law Enforcement Telecommunications
                    Network, the Federal Aviation Administration's
                    Private Aircraft records, and U.S. Custom
                    Service's Automated Commercial Shippers record
                    system of import/export declarations and related
                    international shipping documents.

Drug Enforcement    The Narcotics and Dangerous Drugs Information
Administration      System database contains information on Drug
                    Enforcement Administration cases.

U.S. Postal         The Inspection Service Database Information System
Inspection Service  is a database of all current and closed Postal
                    Inspection Service criminal cases.
----------------------------------------------------------------------
\a According to a FinCEN official, FinCEN does not have access to all
of the investigative case information in the law enforcement
databases, such as details on active cases. 

Source:  GAO analysis of FinCEN data. 


      FINCEN'S INTERNAL DATABASES
------------------------------------------------------ Appendix II:3.1

FinCEN uses its internal databases to index and track inquiries made
on individuals and businesses.  This includes the results of the
intelligence reports prepared by FinCEN as well as information
captured through FinCEN's platform and Gateway programs.  These
databases enable FinCEN to "alert" or notify one agency that another
agency has or had an interest in the same investigative subject,
which can help the agencies coordinate their efforts. 




(See figure in printed edition.)Appendix III
GAO QUESTIONNAIRE:  SURVEY OF
FEDERAL AGENCY OFFICIALS REGARDING
SUPPORT PROVIDED BY FINCEN
========================================================== Appendix II



(See figure in printed edition.)



(See figure in printed edition.)



(See figure in printed edition.)




(See figure in printed edition.)Appendix IV
SURVEY OF STATE AND LOCAL LAW
ENFORCEMENT OFFICIALS REGARDING
SUPPORT PROVIDED BY FINCEN
========================================================== Appendix II



(See figure in printed edition.)



(See figure in printed edition.)



(See figure in printed edition.)




(See figure in printed edition.)Appendix V
USE OF THE FINANCIAL CRIMES
ENFORCEMENT NETWORK:  SURVEY OF
JUSTICE AND TREASURY FIELD OFFICES
========================================================== Appendix II



(See figure in printed edition.)



(See figure in printed edition.)



(See figure in printed edition.)



(See figure in printed edition.)




(See figure in printed edition.)Appendix VI
COMMENTS FROM FINCEN
========================================================== Appendix II



(See figure in printed edition.)



(See figure in printed edition.)


MAJOR CONTRIBUTORS TO THIS REPORT
========================================================= Appendix VII

GENERAL GOVERNMENT DIVISION,
WASHINGTON, D.C. 

Danny R.  Burton, Assistant Director
David P.  Alexander, Senior Social Science Analyst
James M.  Fields, Senior Social Science Analyst
Stuart M.  Kaufman, Senior Social Science Analyst
Barry L.  Reed, Senior Social Science Analyst

OFFICE OF THE GENERAL COUNSEL,
WASHINGTON, D.C. 

Geoffrey R.  Hamilton, Senior Attorney

DALLAS FIELD OFFICE

R.  Eric Erdman, Evaluator-in-Charge
Michael H.  Harmond, Evaluator

*** End of document. ***