This is the accessible text file for GAO report number GAO-08-355 entitled 'Bank Secrecy Act: Increased Use of Exemption Provisions Could Reduce Currency Transaction Reporting While Maintaining Usefulness to Law Enforcement Efforts' which was released on February 21, 2008. This text file was formatted by the U.S. Government Accountability Office (GAO) to be accessible to users with visual impairments, as part of a longer term project to improve GAO products' accessibility. Every attempt has been made to maintain the structural and data integrity of the original printed product. Accessibility features, such as text descriptions of tables, consecutively numbered footnotes placed at the end of the file, and the text of agency comment letters, are provided but may not exactly duplicate the presentation or format of the printed version. The portable document format (PDF) file is an exact electronic replica of the printed version. We welcome your feedback. Please E-mail your comments regarding the contents or accessibility features of this document to Webmaster@gao.gov. This is a work of the U.S. government and is not subject to copyright protection in the United States. It may be reproduced and distributed in its entirety without further permission from GAO. Because this work may contain copyrighted images or other material, permission from the copyright holder may be necessary if you wish to reproduce this material separately. Report to Congressional Committees: February 2008: Bank Secrecy Act: Increased Use of Exemption Provisions Could Reduce Currency Transaction Reporting While Maintaining Usefulness to Law Enforcement Efforts: GAO-08-355: GAO Highlights: Highlights of GAO-08-355, a report to congressional committees. Why GAO Did This Study: To aid law enforcement efforts against financial crimes, under the Bank Secrecy Act (BSA) depository institutions must file the Treasury Department’s Financial Crimes Enforcement Network’s (FinCEN) currency transaction report (CTR) form on their customers’ cash transactions of more than $10,000. While FinCEN’s regulations allow institutions to exempt certain customers, over 15 million CTRs were filed in 2006. Public Law 109-351 directed GAO to report on (1) the usefulness of CTRs to law enforcement; (2) depository institutions’ costs of meeting CTR requirements; and (3) ways to encourage use of exemptions to avoid unnecessary CTRs. Among other things, GAO obtained data from FinCEN on CTRs and exemptions from 2004 to 2006, surveyed 115 state and local law enforcement agencies and 680 depository institutions, held structured interviews with officials of federal agencies and depository institutions, and reviewed relevant laws and regulations. What GAO Found: According to federal, state, and local law enforcement officials, CTRs provide unique and reliable information essential to a variety of efforts, and recent advances in technology have enhanced law enforcement agencies’ ability to use CTR data by integrating it with other information. In addition to supporting specific investigations, CTR requirements aid law enforcement by forcing criminals attempting to avoid reportable transactions to act in ways that increase chances of detection through other methods. Linking law enforcement’s use of CTRs to specific outcomes is difficult, however, because agencies do not track their use of CTRs, which are typically one of many information sources used in investigations. FinCEN does not routinely publish summary information on law enforcement uses of CTR data—as it does for other data required under the BSA—that could help depository institutions understand the value of CTRs. While fewer than 30 of the largest U.S. depository institutions accounted for over half of new CTRs filed during the period GAO examined, all of the nation’s approximately 17,000 institutions incur some costs to meet CTR requirements. Institutions must have processes and staff in place to identify when and if a CTR is required, as well as the ability to aggregate same-day cash transactions by or on behalf of the same person; file CTRs correctly; and, if desired, establish and maintain exemptions for certain customers. Institutions GAO contacted were generally unable to quantify these costs, in large part because they use the same processes and staff for other purposes. While automation has made CTR tasks less difficult, almost all institutions reported that they have not completely automated all steps, such as reviews of CTRs by institution officials. GAO’s work identified a number of factors that deter use of exemptions, as well as opportunities for increasing their use, thereby reducing the number of CTRs that are likely of little or no value to law enforcement efforts. As reasons for not exempting eligible customers, institutions cited uncertainty about the documentation required to demonstrate that some customers are in fact eligible, along with concern that federal banking regulators (who examine institutions for compliance with CTR requirements) would find fault. Institutions also cited as deterrents the need to meet FinCEN’s regulatory requirements to (1) file an exemption form, and annually review the supporting data, particularly for hundreds of customers that are specifically exempted by statute; and (2) biennially renew eligibility for some customers—a process that as a practical matter duplicates the required annual reviews for those customers. Institution officials indicated that additional guidance from FinCEN, as well as Web-based material to help train their staff in making exemption determinations, could increase the use of exemptions. Removing regulatory deterrents and providing additional guidance and Web-based material could help depository institutions avoid filing unnecessary CTRs without harming law enforcement efforts. What GAO Recommends: GAO recommends that the Secretary of the Treasury direct FinCEN to consider routinely publishing summary information on CTR use, revise certain regulations that deter exemptions, and provide additional guidance and Web-based material to help depository institutions interpret exemption requirements. FinCEN concurred with our regulatory and guidance recommendations and stated that it will consider options for providing feedback on CTR use. To view the full product, including the scope and methodology, click on [hyperlink, http://www.GAO-08-355]. To view the results of GAO's surveys, click on GAO-08-385SP. For more information, contact David G. Wood at (202) 512-6878 or woodd@gao.gov. [End of section] Contents: Letter: Results in Brief: Background: CTR Requirements Are Useful to Law Enforcement Efforts in a Variety of Ways, but Measuring Their Impact Is Difficult: Financial Institutions Incur Some Costs to Meet Requirements Regardless of the Number of CTRs They File: Uncertainty about Required Documentation and Some Regulatory Requirements May Unnecessarily Discourage Use of Exemptions: Conclusions: Recommendations for Executive Action: Agency Comments and Our Evaluation: Appendixes: Appendix I: Objectives, Scope, and Methodology: Appendix II: Law Enforcement Agencies' Use of CTR Data: Appendix III: Information on CTRs Filed from 2004 to 2006 : Appendix IV: Additional Information on Depository Institutions' Use of Exemptions: Appendix V: Comments from the Financial Crimes Enforcement Network: Appendix VI: Comments from Federal Banking Regulators: Appendix VII: GAO Contact and Staff Acknowledgments: Related GAO Products: 99: Tables Tables: Table 1: Statutory and Regulatory Provisions That Determine Which Customers May Be Exempted: Table 2: CTR Views by Agencies, Fiscal Years 2004 through 2006: Table 3: Exemption Violations Cited in BSA Examinations by FDIC, Federal Reserve, and OCC, 2005 and 2006: Table 4: Federal Agencies We Interviewed and Number of CTR Views, Fiscal Year 2006: Table 5: Number of CTRs Filed by Asset Size Category of Depository Institutions in 2006: Table 6: Survey Population and Response Rate: Table 7: Number and Percentage of CTRs Filed by Institution Type, 2004- 2006: Table 8: Mean and Median Numbers of CTRs Filed by Size of Institution, 2004-2006: Table 9: Number of CTRs Based on Aggregated Transactions, 2004-2006: Table 10: Number and Percentage Amount of Cash-In Transactions Recorded by CTRs, 2004-2006: Table 11: Number and Percentage of Cash-Out Transactions Recorded by CTRs, 2004-2006: Table 12: Number and Percentage of Phase I Exemptions Filed by Depository Institutions, 2006: Table 13: Number and Percentage of Phase II Exemptions Filed by Depository Institutions, 2006: Table 14: FinCEN's CTR and Exemption-Related Enforcement Actions on Depository Institutions, 1997-2006: Figures: Figure 1: Extent to Which State and Local Agencies Found That CTRs Assisted Them in Verifying Known or Obtaining Previously Unknown Information: Figure 2: Purposes for Which State and Local Agencies Reported Using CTRs and How They Rated CTRs' Usefulness: Figure 3: CTRs Filed in 2006 by Banks and Credit Unions, by Asset Size: Figure 4: General Process for Filing CTRs: Figure 5: Extent to Which Steps in the CTR Process Were Automated at Surveyed Institutions: Figure 6: Median Time, in Minutes, to Accomplish Each CTR Filing Step: Figure 7: Factors That Surveyed Institutions Reported as of Great or Very Great Importance to Their Decision to Exempt Phase I and Phase II Customers: Figure 8: Percentage of Depository Institutions with Customers Eligible for Phase I and II Exemptions and Extent to Which They Filed Exemptions in 2006: Figure 9: An Overview of the Reveal Data Mining System: Figure 10: States That Viewed CTR Data, by Number of CTRs Viewed in the Gateway Program, Fiscal Year 2006: Figure 11: CTRs Filed in 2006, by Institution Asset Size: Figure 12: CTRs Filed in 2006, by County: Figure 13: Factors That Institutions Considered to Be of Very Great or Great Importance When Deciding Not to Exempt a Phase I Eligible Customer: Figure 14: Factors That Institutions Considered to Be of Very Great or Great Importance When Deciding Not to Exempt a Phase II-Eligible Customer: Abbreviations: BSA: Bank Secrecy Act: CBRS: Currency Banking and Retrieval System: CMIR: Report of International Transportation of Currency or: Monetary Instruments: CTR: currency transaction report: DEA: Drug Enforcement Administration: FBI: Federal Bureau of Investigation: FDIC: Federal Deposit Insurance Corporation: FinCEN: Financial Crimes Enforcement Network: HIDTA: High Intensity Drug Trafficking Area: HIFCA: High-Intensity Money Laundering and Related Financial Crimes Area: ICE: Immigration and Customs Enforcement: IRS: Internal Revenue Service: NCUA: National Credit Union Administration: OCC: Office of the Comptroller of the Currency: OCDETF: Organized Crime Drug Enforcement Task Force: OTS: Office of Thrift Supervision: TECS: Treasury Enforcement Communications System: SAR: Suspicious Activity Report: Letter February 21, 2008: The Honorable Christopher J. Dodd: Chairman: The Honorable Richard C. Shelby: Ranking Member: Committee on Banking, Housing, and Urban Affairs: United States Senate: The Honorable Barney Frank: Chairman: The Honorable Spencer Bachus: Ranking Member: Committee on Financial Services: House of Representatives: To assist law enforcement agencies in their efforts to combat money laundering, the financing of terrorist activities, and other crimes, financial institutions are required to provide the federal government with information on customers engaging in certain currency transactions under the Bank Secrecy Act (BSA).[Footnote 1] Among other things, the BSA--administered by the U.S. Department of the Treasury's Financial Crimes Enforcement Network (FinCEN)--and its implementing regulations require financial institutions to file currency transaction reports (CTR) when their customers make large cash transactions, currently defined by regulation as those exceeding $10,000.[Footnote 2] In 2006, the government received about 15 million CTRs, the vast majority of which were filed by depository institutions (banks, thrifts, and credit unions). To reduce the number of CTRs with limited usefulness to law enforcement efforts, in 1994 Congress enacted provisions allowing depository institutions to exempt two broad categories of customers that meet specified criteria.[Footnote 3] For these exempted customers, the institutions do not have to file CTRs, because the customers' cash transactions would likely be of little or no value to law enforcement efforts. First, the law required FinCEN to provide appropriate exemptions for customers that are another depository institution; governmental entities, including state and local governments; certain other entities exercising U.S., state, and local governmental authority; and "any business or category of business the reports on which have little or no value for law enforcement purposes," which FinCEN has defined through regulations to generally include companies that are listed on any of three stock exchanges (listed companies) and subsidiaries that are 51 percent or more owned by a listed company. Second, the law authorized FinCEN to establish, through regulation, exemptions for "qualified business customers" that maintain an account at the depository institution, frequently engage in large cash transactions, and meet other criteria specified by regulation. FinCEN's regulations provide that certain qualified business customers may not derive more than 50 percent of their gross revenue from activities or lines of business specifically deemed ineligible, such as the purchase or sale of automobiles or gaming of most kinds. Because FinCEN promulgated the regulations for these two categories in separate rule- making phases, the exemptions are commonly referred to as "Phase I" and "Phase II" exemptions, respectively.[Footnote 4] It is up to the depository institutions as to whether they actually exempt each of their customers who are eligible for exemption; if they do, the institutions must file an exemption form documenting the customer's eligibility and must review and verify eligibility at least once each year. Depository institutions have expressed concerns about the cost and effort required to meet CTR filing requirements--including the steps needed to establish and maintain exemptions for their customers--as well as doubts about the usefulness of CTRs to law enforcement agencies. They note that they are also required, under the BSA, to file with FinCEN Suspicious Activity Reports (SAR) in cases of certain transactions that may involve violations of law or regulation, including money laundering. However, law enforcement officials have maintained that the CTR requirements help deter money laundering and that CTRs provide information that is highly useful to their investigations. Data from CTRs are aggregated and stored electronically in a large database accessible to law enforcement agencies and maintained for FinCEN by the Internal Revenue Service (IRS). The database also includes information about customers for which institutions have filed exemption forms. The Financial Services Regulatory Relief Act of 2006 required that we examine several aspects of CTRs, including their usefulness to law enforcement and the burden on depository institutions filing them, and to determine whether CTR filing rules could be modified without harming law enforcement operations.[Footnote 5] This report discusses (1) the usefulness of CTRs to federal, state, and local law enforcement agencies; (2) the costs to depository institutions of meeting CTR requirements; and (3) factors that affect depository institutions' decisions to exempt or not exempt eligible customers, including opportunities for encouraging use of exemptions while maintaining the usefulness of CTR data to law enforcement agencies. To examine the usefulness of CTRs to federal, state, and local law enforcement agencies, we first obtained data from both FinCEN and IRS indicating the frequency of access to the CTR database by specific agencies. We conducted structured interviews with officials of 12 federal agencies and organizations--including those that most frequently accessed CTR data in 2006, such as FinCEN, IRS, the Federal Bureau of Investigation (FBI), the Drug Enforcement Administration (DEA), and the U.S. Department of Homeland Security's Immigration and Customs Enforcement (ICE). In addition, we used a Web-based instrument to survey all 115 state and local law enforcement agencies that had access to CTR data as of May 2007; our overall response rate was 77 percent. We supplemented the survey by interviewing officials of 12 state and 5 local law enforcement agencies, selected to achieve a mix of agencies that had accessed CTR data frequently and agencies that had not. We asked officials at the law enforcement agencies to identify how information provided by CTRs is useful to their efforts and how technological changes have affected the use of CTR data. To understanding filing trends and obtain information on depository institutions' costs to meet CTR requirements, we first analyzed FinCEN's CTR and exemption data covering 3 calendar years--2004, 2005, and 2006--to identify the numbers of CTRs and exemptions filed by depository institutions of different sizes. For this purpose, we established four size categories (based on the dollar value of institutions'assets) for banks and thrifts and three categories for credit unions.[Footnote 6] To obtain specific information on the costs of meeting CTR requirements, we conducted 35 structured interviews with officials of depository institutions of different sizes. We asked the officials whether they use manual or automated processes and what costs they incur to meet CTR requirements, including the costs of filing individual CTRs and exemption forms. Finally, to identify the factors affecting depository institutions' exemption decisions, as well as opportunities for potentially increasing the use of exemptions, we used a Web-based instrument to survey 680 of the 3,880 depository institutions that filed at least 120 CTRs in 2006, stratified by asset size category. Our overall response rate was 68 percent. When presenting the survey results, all percentage estimates in this report have 95 percent intervals of within plus or minus 8 percentage points of the estimate, unless otherwise noted. This report does not contain all of the results of our surveys of law enforcement agencies and depository institutions, but the surveys and a more complete tabulation of the results can be viewed at [hyperlink, http://www.gao.gov/cgi- bin/getrpt?GAO-08-385SP]. We also analyzed statutory and regulatory filing requirements and interviewed officials and examiners from the five federal banking regulators to obtain their viewpoints on the difficulties, if any, institutions might confront in meeting the CTR and exemption filing requirements.[Footnote 7] We also obtained data on BSA examinations conducted by each of the regulators for 2005 and 2006, particularly data on their citations of depository institutions for noncompliance with CTR requirements. Additional details on our methods are presented in appendix I. We conducted this performance audit from November 2007 through February 2008 in accordance with generally accepted government auditing standards. Those standards require that we plan and perform the audit to obtain sufficient, appropriate evidence to provide a reasonable basis for our findings and conclusions based on our audit objectives. We believe that the evidence obtained provides a reasonable basis for our findings and conclusions based on our audit objectives. Results in Brief: Federal, state, and local law enforcement officials we interviewed and surveyed described a variety of ways in which CTR requirements are useful to their efforts; however, measuring their impact is difficult. Recent advances in technology, along with FinCEN's distribution of BSA data in bulk to certain federal agencies, have enhanced the ability to access and integrate CTR data with information from other sources. According to law enforcement officials, CTRs provide unique and reliable information that is essential to supporting investigations and detecting criminal activities, in part because CTRs provide information that is often unavailable elsewhere or is more objective or up-to-date than that obtained from other sources. In addition to supporting individual cases, law enforcement agencies use aggregated CTR data to help detect patterns or trends; for example, FinCEN analysts routinely analyze CTR data in conjunction with other information to develop "big picture" views of suspicious financial activity. Law enforcement officials noted that CTR requirements also aid their efforts by making it more difficult for criminals to get their illicit proceeds into the financial system and forcing them to act in ways that increase chances of detection--such as smuggling cash or "structuring" their cash transactions to avoid CTRs, which often prompts depository institutions to file a Suspicious Activity Report. Linking law enforcement's use of CTRs to specific impacts is difficult, however, because agencies do not track their use of CTRs, which are typically one of many sources of information used to support investigations. FinCEN does not routinely publish any summary information on law enforcement's use of CTR data-- such as identified trends and case examples--as it does for Suspicious Activity Reports. Although concerns about revealing investigatory sources and methods limit dissemination of detailed information on how law enforcement agencies use CTR data, our interviews with financial institution officials suggest that they would better understand the value of meeting their requirements if the institutions were provided with similar summary information on CTR use. Our analysis of CTR data for 2004 to 2006 shows that a large proportion was filed by a small number of the largest depository institutions--for example, fewer than 30 very large banks accounted for 55 percent of new CTRs during the period--and while these institutions likely incur the greatest expenditure of time and resources to meet CTR requirements, all depository institutions incur some costs regardless of the number of CTRs they file. This is because institutions must have processes and trained staff in place to identify when and if a CTR is required, including the ability to aggregate same-day cash transactions made by or on behalf of the same person, and to file CTRs correctly. While automation has made these tasks less difficult, most institutions reported that their processes still include "manual" steps; for example, most institutions reported that their CTRs are reviewed by branch managers or compliance officers before being sent electronically to FinCEN or by mail to IRS. Institutions we contacted were generally unable to quantify their costs for meeting CTR requirements, in large part because they use the same personnel and processes for meeting other BSA requirements or for other purposes and do not separately account for CTR-related costs. However, they noted that personnel costs include the cost of training staff on meeting CTR requirements, as well as the cost of labor involved in filing CTRs. Reflecting the range of numbers of staff that may be involved, officials provided a wide variance of estimated personnel costs. For example, while one very large bank that filed almost 1 million CTRs in 2006 estimated personnel costs, including tellers and compliance officers, of about $5.4 million, a large bank that filed just under 5,000 CTRs in 2006 estimated personnel costs at $76,000. Officials at institutions with automated processes also cited technology as a significant cost. Our survey of depository institutions and interviews with officials identified a variety of factors that deter the use of exemptions, as well as opportunities for increasing their use without diminishing the usefulness of CTR requirements to law enforcement. Our survey results showed that many financial institutions with customers considered eligible for exemptions do not actually exempt them, but instead continue to file CTRs on the customers' transactions--despite the institutions' recognition that making use of the exemption provisions would enable them to file fewer CTRs. The reasons they cited included uncertainty about the documentation needed to demonstrate that certain customers are in fact eligible for exemptions, accompanied by concern that the federal banking regulators would deem the documentation insufficient and cite them for noncompliance with BSA requirements. For example, depository institutions that chose to use the Phase II exemption relied on various methods--sometimes to a considerable extent--to determine and document the portion of the customer's revenues derived from ineligible activities, including asking the customer for financial statements, tax records, or other documentation such as a letter certifying its revenue sources. While our review of data from the banking regulators showed relatively few violations compared with the number of BSA examinations conducted, we found variations in the types of documentation the regulators find acceptable. For example, officials from two of the banking regulators said that a letter from a customer self-certifying its revenue sources could be acceptable to document eligibility for a Phase II exemption, while officials from two other regulators indicated that such a letter alone would be inadequate documentation. Both the difficulties cited by the institutions and the variation among examiners indicate that further CTR guidance from FinCEN could be helpful in this regard. Other factors discouraging the use of exemptions were the cost and effort involved in meeting FinCEN's regulatory requirements to (1) file an exemption form, and annually review and update the information, particularly for certain customers that are required to be exempted by statute as appropriate; and (2) biennially file a form to document the continued eligibility of customers that have been exempted under the Phase II regulation, which many institutions viewed as redundant in light of the required annual review process. Eliminating these requirements could encourage institutions to make greater use of exemptions. Other opportunities to encourage the use of exemptions include (1) shortening the waiting period--currently a full year under FinCEN's regulations--before exempting certain customers with frequent cash transactions that exceed the $10,000 threshold, and (2) making available from FinCEN Web-based material to help train and guide depository institutions' staff in making exemption determinations. While FinCEN currently provides such material--such as answers to frequently asked questions, rulings, and guidance--for other BSA requirements, the information on CTR exemption requirements is very limited; and about 50 percent of respondents to our survey indicated that the availability of such Web-based material from FinCEN would increase their use of exemptions. Because the transactions of exempt customers are likely to be of little or no value to law enforcement, these actions could avoid the burden of filing some CTRs without harming law enforcement efforts. We are recommending that the Secretary of the Treasury direct FinCEN to consider routinely publishing summary information on law enforcement uses of CTRs, provide additional guidance on the documentation needed to demonstrate eligibility for some customers, revise certain regulations that deter exemptions, and provide Web-based material to help depository institutions interpret exemption requirements. In written comments on a draft of this report, the Director of FinCEN concurred with our recommendations seeking regulatory amendments and those related to guidance and materials to aid industry in making eligibility determinations for CTR exemptions, and said that FinCEN will consider options to provide industry with additional feedback on the use of CTRs by law enforcement. We also received written comments from the Board of Governors of the Federal Reserve System (Federal Reserve), Federal Deposit Insurance Corporation (FDIC), National Credit Union Administration (NCUA), and Office of Thrift Supervision (OTS) that, in a joint letter, reaffirmed their support for effective administration of the BSA and said they believe that streamlining and clarifying the exemption regulations, as we recommend, would be a positive step. Background: According to BSA's objectives, CTRs are to have a "high degree of usefulness" and their uses include criminal, tax, or regulatory investigations or proceedings. In 2001, the USA PATRIOT Act added a fourth purpose: the conduct of intelligence or counterintelligence activities, including analysis, to protect against international terrorism.[Footnote 8] CTRs are intended to provide a paper trail for federal, state, and local law enforcement agencies in their investigations and, thus, potentially hinder using financial institutions as intermediaries for the transfer or deposit of money derived from criminal activity.[Footnote 9] A CTR records account cash withdrawals and deposits, as well as currency exchanges, and wire transfers purchased with cash, when the amount of the transaction is more than $10,000. In addition to the dollar amount of the cash transaction, a CTR records information about the account owner, including the owner's occupation, and the identity of the person actually conducting the transaction (the conductor), if not the account holder. A depository institution must file a CTR for transactions that collectively exceed $10,000 during the course of a day if the institution has knowledge that they are for or on behalf of the same person. The Money Laundering Suppression Act of 1994 provided basic criteria for establishing and maintaining exemptions and authorized Treasury to establish further requirements.[Footnote 10] FinCEN has done so through both regulations and "interpretative letters" that supplement the regulations to provide further guidance. Table 1 summarizes the requirements as outlined in the statute, implementing regulations, and interpretive letters. Table 1: Statutory and Regulatory Provisions That Determine Which Customers May Be Exempted: Statutory provision: Customers eligible for exemptions under Phase I: Another depository institution; Regulatory provision: Customers eligible for exemptions under Phase I: A bank, to the extent of such bank's domestic operations. Statutory provision: Customers eligible for exemptions under Phase I: A department or agency of the United States, any state, or any political subdivision of any state; Regulatory provision: Customers eligible for exemptions under Phase I: The same as statutory provision. Statutory provision: Customers eligible for exemptions under Phase I: Certain other entities exercising governmental authority on behalf of the United States, any state or political subdivision of any state; Regulatory provision: Customers eligible for exemptions under Phase I: The same as statutory provision. Statutory provision: Customers eligible for exemptions under Phase I: Any business or category of business the reports on which have little or no value for law enforcement purposes; Regulatory provision: Customers eligible for exemptions under Phase I: Any entity, other than a bank, whose common stock is listed on the New York, American, or NASDAQ Stock Exchange, with some exceptions (a "listed entity"); and any subsidiary, other than a bank, of any "listed entity" that is organized under U.S. law and at least 51 percent of whose common stock is owned by the listed entity. A nonbank financial institution meeting these criteria may be extended only to the extent of its domestic operations. Statutory provision: Eligibility criteria for business customers under Phase II: Maintains a transaction account at the depository institution, and; Regulatory provision: Eligibility criteria for business customers under Phase II: A commercial enterprise that has maintained a transaction account at the bank for at least 12 months. Statutory provision: Eligibility criteria for business customers under Phase II: Frequently engages in transactions with the depository institution that are subject to the CTR reporting requirements, and; Regulatory provision: Eligibility criteria for business customers under Phase II: Frequently (defined in an Interpretive Letter as at least 8 times within a 12-month period, excepting certain "seasonal" customers) engages in cash transactions in excess of $10,000. Statutory provision: Eligibility criteria for business customers under Phase II: Meets other criteria which the Secretary determines are sufficient to ensure the purposes of the BSA are carried out; Regulatory provision: Eligibility criteria for business customers under Phase II: Is incorporated or organized under U.S. law, or state law, or is registered as and eligible to do business in the United States or a state, to the extent of its domestic operations, and to the extent that no more than 50 percent of its gross revenues come from activities specified as non-eligible business activities. Statutory provision: Customers not eligible for exemption: FinCEN must establish guidelines, which may include a description of the type of business for which no exemption will be granted; Regulatory provision: Customers eligible for exemptions under Phase I: Businesses for which no exemption as a nonlisted business will be granted are those engaged primarily in; * serving as financial institutions or agents thereof; * purchase or sale of motor vehicles, vessels, aircraft, farm equipment or mobile homes; * practice of law, accountancy, or medicine; * auctioning of goods; * chartering or operation of ships, buses, or aircraft; * gaming of any kind (other than pari-mutuel betting at race tracks); * investment advisory services or investment banking services; * real estate brokerage; * pawn brokerage; * title insurance and real estate closing; * trade union activities; and; * any other activities that may be specified by FinCEN. Source: GAO analysis of 31 U.S.C. § 5313(d) and (e), and 31 C.F.R. § 103.22(d). Note: Phase II exemptions also include a second category referred to as "payroll businesses," which are defined in 31 U.S.C. § 103.22(d)(2)(vii). [End of table] Legislative proposals would alter the basis for establishing exemptions, as well as raise the reporting threshold amount above $10,000. For example, in January 2007, the U.S. House of Representatives passed the Seasoned Customer CTR Exemption Act of 2007 (H.R. 323), which would require the Secretary of the Treasury to (1) prescribe regulations for exempting "qualified customers," including criteria for suspending, rejecting, or revoking exemptions; and (2) periodically review the threshold amount and adjust it for inflation as appropriate. H.R. 1447, the CTR Modernization Act, introduced in March 2007, would raise the threshold amount for insured depository institutions to $30,000. FinCEN's role is to oversee administration of the BSA throughout the federal government. Pursuant to this role, FinCEN, among other things, develops policy and provides guidance to other agencies and analyzes BSA data for trends and patterns. FinCEN relies on the regulators of depository institutions--the Federal Reserve Board (Federal Reserve), Federal Deposit Insurance Corporation (FDIC), National Credit Union Administration (NCUA), Office of the Comptroller of the Currency (OCC), and Office of Thrift Supervision (OTS)--to ensure that depository institutions comply with BSA reporting requirements. In addition to CTRs, depository institutions are required by BSA and its implementing regulations to make available information on their customers' transactions in certain circumstances: * Depository institutions are required to file Suspicious Activity Reports (SAR) with FinCEN if a transaction involves or aggregates at least $5,000 in funds or other assets, and the institution knows, suspects, or has reason to suspect that the transaction is designed to evade any requirements of the BSA.[Footnote 11] * Under Section 314(a) of the USA PATRIOT Act, federal law enforcement agencies, through FinCEN, can reach out to financial institutions to locate accounts and transactions of persons suspected of engaging in terrorism or money laundering.[Footnote 12] FinCEN is responsible for providing these agencies with assistance in educating institutions on their BSA responsibilities. To focus and direct their efforts in supporting the effectiveness of BSA compliance, FinCEN's strategic plan for fiscal years 2006-2008 outlines several goals. For example, to assist law enforcement, the plan calls for FinCEN to reduce the number of CTRs filed on legitimate financial transactions that are of little or no value to law enforcement; and, to assist financial institutions, the plan calls for FinCEN to revise its data collection forms, regulations, and practices to ensure that FinCEN collects the information necessary to meet its mission while minimizing reporting burdens on the financial industry. In addition, FinCEN indicated in its plan that it would consider providing guidance on BSA requirements through written and Web-based materials and by means of a call center to respond to specific questions. FinCEN, together with the IRS, is responsible for managing and storing the BSA data that financial institutions report. Financial institutions that submit CTRs in paper form mail them directly to IRS's Enterprise Computing Center in Detroit. Institutions that submit data electronically transmit them directly to FinCEN, which in turn transmits them to the center. The center collects and stores all BSA data in its Currency Banking and Retrieval System (CBRS).[Footnote 13] For fiscal year 2007, the IRS estimated the total cost of processing CTRs to be about $7 million, including about $3.5 million to convert CTRs submitted on paper to an electronic format. IRS examiners and investigators access BSA data directly through IRS's Intranet, while FinCEN has a direct connection to the Enterprise Computing Center. Staff at other law enforcement agencies can access BSA data via the Internet, and certain federal agencies also periodically receive bulk data downloads of BSA data for use at their agencies, as described later in this report. CTR Requirements Are Useful to Law Enforcement Efforts in a Variety of Ways, but Measuring Their Impact Is Difficult: Federal, state, and local law enforcement officials we interviewed and surveyed said that information in CTRs provided unique and reliable information essential to a variety of efforts and that recent advances in technology, along with FinCEN's distribution of BSA data in bulk, have enhanced their ability to use and analyze CTR information. Law enforcement officials stated that, in addition to supporting specific investigations, CTR requirements aid their efforts by making it more difficult for criminals to get their illicit proceeds into the financial system and forcing them to act in ways that increase chances of detection. Linking law enforcement's use of CTRs to specific outcomes is difficult, however, because agencies do not track their use of CTRs, which are typically one of many sources of information used to support investigations. FinCEN does not routinely publish any summary information on law enforcement's use of CTR data as it does for Suspicious Activity Reports, such as identified trends and case examples. Although concerns about revealing investigatory sources and methods limit dissemination of detailed information on how law enforcement agencies use CTR data, our interviews with depository institution officials suggest that they would better understand the value of meeting their requirements if the institutions were provided with similar summary information on CTR use. CTRs Provide Unique Information for Investigating Cases and Detecting Criminal Activities: In part due to advances in technology that have enhanced access to, and analysis of, CTR data, law enforcement officials use CTR data to help investigate a variety of crimes, including tax evasion, customs violations, and drug trafficking. They use CTR data both "reactively"- -that is, to support existing investigations of one or more suspects-- and "proactively"--to analyze patterns or trends that can serve as the basis for initiating new investigations. Technological Advances Have Increased Access and Analytic Capability: In 1993, we reported that CTRs were not used to their full extent by law enforcement agencies because the large volume of reports made meaningful analysis difficult and access to the data, particularly at the state level, was limited and cumbersome.[Footnote 14] However, access to BSA data at the federal, state, and local levels has improved and technological advances have made meaningful analysis of large BSA data sets possible. Consistent with its strategic goal of facilitating information sharing through electronic means, FinCEN has increased access to CTR (and other BSA) data in two ways. First, FinCEN began providing selected federal agencies with access to "bulk" CTR data--essentially all of the data resulting from CTRs. In 2004, FinCEN first provided the FBI with bulk transfer of data, and during 2005 and 2006 FinCEN agreed to provide two federal agencies--the Secret Service and ICE--and a multiagency program established by the Department of Justice (the Organized Crime Drug Enforcement Task Force, or OCDETF Fusion Center) with access to a bulk data set.[Footnote 15] Receiving these data in bulk, rather than accessing the database remotely and querying it for specific records, allows agencies to conduct more sophisticated analyses by combining the BSA data with other data sets, as can be seen in the following examples. * The FBI has combined bulk BSA data into its Investigative Data Warehouse, a collection of more than 50 multisource data sets that includes counter terrorism data. According to the FBI, access to BSA bulk data has significantly increased its usage of CTR data (the bureau reported data that indicated approximately 194,000 CTR views from 2004 through 2006 of the downloaded data). According to FBI officials, about 40 percent of all FBI terrorism subjects appeared on CTRs that were filed between January 1, 2000, and June 30, 2006; further, CTR data were the most viewed data in the warehouse. * OCDETF's Fusion Center integrated bulk BSA data with drug, financial, and gang-related investigative data provided by several other federal, state, and local law enforcement agencies. According to OCDETF officials, as of June 2007, CTR data had appeared in 61 percent of the Fusion Center's analytical products. * ICE has combined BSA data with import and export data for selected countries to help identify and detect discrepancies or anomalies in international commerce that might indicate trade-based money laundering. Second, FinCEN improved Internet access to CTR data. FinCEN provides and grants access using its "Gateway" program, through which law enforcement staff may access the database using a system known as WebCBRS.[Footnote 16] With WebCBRS, users can download large volumes of CTR data--up to 20,000 CTRs on a single query--and export it to a spreadsheet application, such as Excel. This allows users to more readily conduct proactive analyses, such as identifying transaction trends by categories. Most of the law enforcement officials with whom we spoke, as well as officials of state and local agencies we surveyed, confirmed that WebCBRS is more user friendly than its predecessor and has greatly improved their ability to search for and analyze CTR data.[Footnote 17] (More detailed information about the technological advances enabling greater use of CTR data, along with examples of use in specific investigations, is presented in appendix II. Our survey of law enforcement agencies and a more complete tabulation of the results can be viewed at [hyperlink, http://www.gao.gov/cgi-bin.getrpt?GAO-08- 385SP.]) Perhaps reflecting improvements in the ability to access and analyze CTR data, the number of agencies using CTR data has increased, and officials at some agencies noted that they have incorporated a search of CTR data as a routine part of their investigations. For example, from 2004 through 2006, the number of agencies that viewed CTR data through the Gateway program increased from 109 to 136 and, as of October 2007 requests from an additional 110 agencies for access to CTR data were pending FinCEN's review. Agencies Value CTRs as a Source of Unique, Reliable, and Timely Data: Officials from law enforcement agencies we interviewed emphasized that CTRs are important because they provide information that is often unavailable from other sources, or is more objective or up-to-date than that obtained from other sources. They cited ways in which CTRs provide more comprehensive or timely information about a suspect's banking transactions than they can obtain using other provisions of law. More specifically, law enforcement officials frequently identified the name of the currency transaction's conductor--the person who actually carries out a cash transaction at a financial institution, but who is not the holder of the affected accounts--as useful information that is unique to CTRs. For example, an FBI official noted a case in which analysis of information obtained by searching the CTR database conductor field provided the agency with the investigative lead needed to track the banking activities of persons who, according to the FBI official, were involved in a cocaine distribution ring. The conductor information was useful because the main person under investigation in the ring had associates open bank accounts in their own names at different banks and then made large currency transactions into these accounts, resulting in CTRs that recorded the main person under investigation as the conductor. Further, FinCEN and other law enforcement officials explained that because multiple individuals may use the same account to conduct transactions, CTRs often could be used to identify unknown persons associated with suspects, thereby expanding the scope of investigations. For example, during a 4-year FBI investigation, analysis of CTRs showed where suspects were banking as they opened and closed accounts, and on which day of the week suspects typically made their deposits, allowing the FBI to better plan its surveillance. Law enforcement officials also noted that CTRs provide a unique source of information on the occupations of account holders that often proves useful. For example, a DEA official reported that analyzing CTRs by the information in the occupation field has allowed him to identify whether medical companies or doctors--lines of businesses that typically would not be dealing in high volumes of cash--were diverting controlled substances for illegal use. Finally, officials from several federal law enforcement agencies, including the Bureau of Alcohol, Tobacco, Firearms, and Explosives, IRS-Criminal Investigation, ICE, and DEA, commented that because depository institutions are required to file CTRs soon after a reportable transaction occurs, the CTR database provides up-to-date information on large cash transactions.[Footnote 18] Many federal, state, and local officials we interviewed commented that CTRs provided them with ready access to information that they could not otherwise obtain in a timely manner. Officials contrasted these useful aspects of information from CTRs with information they may be able to obtain on suspects' banking activities under other provisions of the BSA or other laws: * Suspicious Activity Reports (SAR) also provide useful information; however, depository institutions have some discretion in determining whether a transaction or customer is "suspicious," and therefore the institutions determine whether to file a SAR and, if so, what information to include. Thus, a SAR might not capture the same level of information about specific transactions that a CTR routinely would provide. Further, criminals may use several different banks to conduct their transactions, and a SAR would reflect the suspicious activity only within the bank filing the SAR. * The Section 314(a) process, under which federal law enforcement agencies may reach out to financial institutions to locate accounts and transactions of persons of interest, is reserved for significant money- laundering or terrorist-financing investigations, and agencies may make such requests only upon approval by FinCEN, which limits the number of subjects on the list. Further, according to FinCEN, the request provides lead information only--law enforcement agencies must meet the legal standards that apply to the investigative tool that it chooses to use to obtain documents, such as a subpoena. In addition, officials at FinCEN and the IRS noted that the 314(a) process provides law enforcement access only to transactions conducted within the last 6 months, or accounts held within the last 12 months, while the CTR (and other BSA) data provide access to data going back 10 years. * Obtaining bank records through subpoenas could take months or be difficult. Further, in order to subpoena a specific institution, officials would need to know that a suspect banked at that institution. A majority (55 percent) of the state and local law enforcement officials we surveyed noted that it would be "somewhat more" or "much more" difficult to obtain information from bank records in this fashion than from using CTRs. * Other methods of obtaining information about a suspect's bank accounts--including "mail covers" (copies, obtained from the U.S. Postal Service, of the fronts of envelopes delivered to a suspect), subpoenas for credit reports, and surveillance--are time consuming and less likely to provide needed information about a bank account, according to law enforcement officials. ICE and state officials from New York and Texas noted that their agencies could wait days to obtain mail covers, with no guarantee of receiving one bearing an address of the suspect's financial institution. New York state law enforcement officials said that the next best alternative would be to subpoena a suspect's credit report, a process that could take 30 days. While the credit report may provide useful leads--for example, a suspect's mortgage application--that the agency might then subpoena, the time required would further lengthen the investigation. Agencies Reported Using CTR Data to Support a Wide Variety of Investigative Cases: Most officials of law enforcement agencies we contacted indicated that they most often use CTRs reactively, and many routinely review CTRs at the beginning of each investigation.[Footnote 19] For example, tax investigators in IRS routinely query BSA data to identify CTRs with information that suggests situations such as a business paying employees in cash (and thus not withholding taxes). However, law enforcement agencies typically did not use CTRs in isolation to develop a case; rather, they used CTR data to identify leads for further investigation, in part by comparing CTR information with information from other sources. As explained by law enforcement officials, the information in CTRs is useful in corroborating information contained in other BSA reports. For example, agencies may compare information from the CTR database to that obtained from Reports of International Transportation of Currency or Monetary Instruments (CMIR), which report currency transported into the United States, to track how businesses dispose of cash.[Footnote 20] Agencies also consult CTR data to obtain more detailed information after reviewing SARs. For example, officials from a High Intensity Money Laundering and Related Financial Crimes Areas (HIFCA) noted that they used CTR data for 105 of the 120 reports they filed over a recent 1-year period on investigations initiated after reviewing SARs.[Footnote 21] In this regard, many law enforcement officials, including those from ICE, IRS, and the U.S. Attorney's office, noted that raising the CTR filing threshold of $10,000 would affect adversely their ability to deter money laundering, because the CTR threshold corresponds to those set in other anti-money-laundering provisions. For example, officials from the U.S. Attorney's office indicated that the CTR threshold works in tandem with three other statutorily mandated reporting thresholds, which are also set at $10,000: the CMIR requirement; the Form 8300 requirement; and the bulk cash smuggling statute.[Footnote 22] That no CTRs have been filed on business activities that might be expected to generate them also provides valuable leads to law enforcement. For instance, FinCEN, ICE, and DEA conduct analyses comparing known cash flows documented through other sources with cash flows they would expect CTRs to document. Officials at both FinCEN and ICE reported that a search of BSA data for information on a cash business revealing no CTRs could alert investigators that the business was not using traditional depository institutions and direct their focus to nonbank financial institutions such as money services businesses or to the possibility of currency smuggling. Similarly, the lack of CTRs relating to particular individuals or businesses can provide investigative leads. For example, officials from FinCEN and a Florida law enforcement agency said that the presence of a CMIR, coupled with an absence of related CTRs, could provide intelligence that currency transported into the country was subsequently laundered into the financial mainstream through "structuring" (making a series of cash transactions in amounts less than $10,000). Another law enforcement official from Florida indicated that a lack of CTRs corroborated findings from her agency's surveillance operations that certain laundromats, dry cleaners, and travel agencies had laundered millions of dollars. As shown in figure 1, the state and local law enforcement agencies we surveyed found CTRs to be of most use when developing leads for existing investigations. Officials from law enforcement agencies in California, New York, and Texas--states that were among the highest users of CTR data--indicated that their investigators typically used CTRs to identify a subject's bank account numbers and associates who might be conducting transactions on behalf of the subject. An official from one of these agencies indicated that no other source of information enabled investigators to "map" the financial links between members of a criminal organization as well as the CTR. Figure 1: Extent to Which State and Local Agencies Found That CTRs Assisted Them in Verifying Known or Obtaining Previously Unknown Information: This figure is a bar graph showing extent to which state and local agencies found that CTRs assisted them in verifying known or obtaining previously unknown information. [See PDF for image] Source: GAO. [End of figure] Regarding types of investigations involving CTR data, officials of state and local agencies we surveyed reported that they primarily use CTRs for money-laundering, fraud, and drug investigations (see fig. 2). State law enforcement officials we interviewed told us that they use CTRs for a wide variety of investigations relating to money laundering, drugs, workers compensation fraud, Medicaid fraud, mortgage fraud, and white collar crime. Figure 2: Purposes for Which State and Local Agencies Reported Using CTRs and How They Rated CTRs' Usefulness: This figure is a bar graph showing purposes for which state and local agencies reported using CTRs and how they rated CTRs' usefulness. [See PDF for image] Source: GAO. [End of figure] State and local officials also indicated that the current threshold amount of $10,000 was important to the usefulness of CTRs. Specifically, about 58 percent of the state and local officials we surveyed stated that increasing the CTR filing threshold would result in a "very great" or "great" reduction in the usefulness of the CTR filing requirement to their work. A law enforcement investigator from Illinois indicated that many of the CTRs that his agency's criminal intelligence center reviewed were those that documented total transaction amounts between $10,000 and $20,000. Similarly, an investigator with a Minnesota law enforcement agency indicated that the overwhelming majority of CTRs that he found to be of use to his investigations were for transactions between $10,000 and $20,000. Law Enforcement Agencies Use Aggregated CTR Data to Help Detect Patterns or Trends: Law enforcement agencies noted that CTR information also contributes to pattern and trend analyses. For example, at the request of DEA, FinCEN analyzed CTRs filed by institutions in California by ZIP code, providing a statistical overview of financial activity occurring within those areas that, combined with other law enforcement intelligence, allowed DEA analysts to assess threats on a statewide basis. FinCEN officials indicated that their analysts routinely analyze information from CTRs in conjunction with information from SARs or other BSA reports to develop "big picture" views of suspicious financial activity. In addition, ICE officials noted that their analysts often analyze the "conductor" and account holder information from CTRs to identify individuals moving the largest sums of money on behalf of particular account-holders over time to spot any unusual trends. Further, ICE officials commented that they proactively search CTR information to identify individuals moving large sums of money using the same Social Security number with different personal or business names (according to the officials, name variations is a common technique criminals use to hide their activities and avoid detection). According to FinCEN and IRS officials, analytical tools such as data mining--the application of database technology and techniques, such as statistical analysis and modeling, to uncover hidden patterns and subtle relationships in data--have enhanced their investigative efforts by improving their ability to identify data patterns and trends indicative of money laundering and other financial crimes. CTR Requirements Can Cause Criminals to Act in Ways That Increase Chances of Detection: According to federal law enforcement officials, criminals are forced to undertake more risky and suspicious methods of money laundering than depositing cash into depository institutions because they are well aware of the $10,000 filing threshold for CTRs and the investigative paper trail that it creates. While criminals can use a variety of means to launder their money, law enforcement officials we interviewed pointed to three primary methods that criminals use to avoid the CTR filing requirement: structuring; bulk cash smuggling, or physically moving cash across borders via courier or secreted in cargo; and trade- based money laundering, the process of disguising the proceeds of crime and moving value through the use of trade transactions. Many federal law enforcement officials said that the CTR reporting requirement was critical in supporting the ability of depository institutions, as well as their own investigators, to identify suspicious activity based on the structuring of financial transactions to avoid CTRs. The BSA makes it illegal to structure transactions to avoid triggering otherwise applicable reporting requirements, such as the CTR, allowing federal prosecutors to file charges against individuals who structure their cash transactions.[Footnote 23] The structuring in which criminals engage to avoid CTRs may cause a depository institution to file a SAR. FinCEN analysis of SARs filed by depository institutions from April 1, 1996, through December 31, 2006, showed that 1.5 million SARs, or 48 percent of all SARs filed by depository institutions during this period, were filed based on suspicious activity related to structuring or money laundering.[Footnote 24] Many law enforcement agencies routinely review SARs for evidence of structuring. For example, IRS-Criminal Investigation officials said that their agents are required to review SARs that report structuring. Officials we interviewed at several law enforcement organizations, including three associated with HIFCAs, indicated that they had formed teams to review SARs to generate leads for cases based on structuring and said that they regarded the CTR filing requirement as essential to supporting the ability of depository institutions to identify suspicious activity. Federal law enforcement officials also emphasized that the $10,000 CTR threshold played a key role, by forcing criminals to make many more and smaller transactions than otherwise would be required--thus making them more vulnerable to being reported for structuring. Officials from IRS, ICE, FBI, and U.S. Attorneys Offices indicated that they believed large cash transactions have become more uncommon as consumer access to credit and electronic payment options increased in the 30 years since the threshold was established, making the $10,000 threshold still relevant. The existence of the CTR filing requirement also can force criminals into riskier activities such as bulk cash smuggling.[Footnote 25] According to an ICE official, smuggling illegal proceeds in bulk cash form makes criminals more vulnerable to detection because it is easier for agents of law enforcement to interdict bulk cash shipments. Similarly, an official associated with the Chicago High Intensity Drug Trafficking Area (HIDTA) reported observing an increase in bulk cash smuggling because criminals would rather take their chances smuggling their proceeds in bulk cash to Mexico.[Footnote 26] In response, the HIDTA has formed a highway patrol to interdict these cash shipments. Officials from ICE, FinCEN, and the Justice Department also reported that the increase in recent years of bulk cash smuggling across the U.S.-Mexican border was an indicator of CTR success in deterring criminals from depositing cash into domestic financial institutions.[Footnote 27] Measuring Usefulness of CTRs Is Difficult: Linking law enforcement's use of CTRs to specific outcome measures is difficult because agencies do not track their use of CTRs, which are typically only one of many sources of information used to support investigations. FinCEN does not routinely publish any information on law enforcement's use of CTR data as it does for other information that financial institutions provide under the BSA. Although concerns about revealing investigatory sources and methods limit dissemination of detailed information on how law enforcement agencies use CTR data, our interviews with depository institution officials suggest that they would better understand the value of meeting their requirements if the institutions were provided with summary information on CTR use. Agencies Have Difficulties Linking CTRs to Specific Investigation Outcomes: While CTRs appear to be valuable for law enforcement purposes, linking their use to specific case outcomes, such as indictments or convictions, is problematic. First, no requirement exists to track the use of CTR data in investigations, and almost all of the officials from the federal, state, and local law enforcement agencies we contacted reported that their agencies did not track their use of CTRs or how the CTRs contributed to case outcomes.[Footnote 28] As a potential indicator of use, we obtained data from FinCEN and IRS on the number of CTR "views"--that is, the number of times that agencies accessed an individual CTR record. IRS tabulates views that occur when agencies access the database through WebCBRS, including its own views through its Intranet access, while FinCEN tabulates views occurring through the Gateway program.[Footnote 29] For example, data show over 1.6 million views of CTRs by federal, state, and local agencies in 2006 (see table 2). Table 2: CTR Views by Agencies, Fiscal Years 2004 through 2006: Agency: IRS; Number of CTR views: 2004: 1,466,518; Number of CTR views: 2005: 1,231,345; Number of CTR views: 2006: 912,405; Number of CTR views: Total: 3,610,268. Agency: ICE; Number of CTR views: 2004: 213,608; Number of CTR views: 2005: 241,692; Number of CTR views: 2006: 207,325; Number of CTR views: Total: 662,625. Agency: FinCEN; Number of CTR views: 2004: 208,609; Number of CTR views: 2005: 105,266; Number of CTR views: 2006: 136,090; Number of CTR views: Total: 449.965. Agency: DEA; Number of CTR views: 2004: 111,294; Number of CTR views: 2005: 108,845; Number of CTR views: 2006: 108,507; Number of CTR views: Total: 328,646. Agency: FBI; Number of CTR views: 2004: 48,364; Number of CTR views: 2005: 62,487; Number of CTR views: 2006: 54,290; Number of CTR views: Total: 165,141. Agency: Federal banking regulators[A]; Number of CTR views: 2004: 31,408; Number of CTR views: 2005: 54,039; Number of CTR views: 2006: 58,006; Number of CTR views: Total: 143,453. Agency: All other agencies[B]; Number of CTR views: 2004: 141,602; Number of CTR views: 2005: 182,181; Number of CTR views: 2006: 171,943; Number of CTR views: Total: 495,726. Agency: Total; Number of CTR views: 2004: 2,221,403; Number of CTR views: 2005: 1,985,855; Number of CTR views: 2006: 1,648,566; Number of CTR views: Total: 5,855,824. Source: GAO analysis of IRS and Department of Homeland Security data. Notes: Number of CTR views does not include those from bulk downloads by the FBI. [A] The Federal Reserve, OCC, OTS, FDIC, and NCUA. [B] Other agencies include the Bureau of Alcohol, Tobacco, Firearms, and Explosives; United States Secret Service; United States Postal Inspection Service; U.S. Attorneys Offices; and the Securities and Exchange Commission; state regulatory agencies; and state and local law enforcement agencies. [End of table] According to FinCEN data, among users who accessed the CBRS database through the Gateway program, state and local law enforcement agencies accounted for about 33 percent and 3 percent of the total CTR views, respectively. These users include state bureaus of investigation and criminal prosecuting offices, such as state offices of attorneys general and county prosecutors' offices. The number of these CTR views was concentrated among a few agencies; for example, 10 agencies from 8 states accounted for almost 64 percent of the total CTR views made by state and local law enforcement agencies, according to FinCEN's data. (For a map showing views by state, see app. II.) However, the numbers of views do not provide any indication of CTRs' usefulness to any specific law enforcement effort or establish a link to any specific outcome of an investigation. Moreover, data on views during this period may not reflect future trends because of the changes in access to the CTR database discussed previously. For example, the numbers of CTRs viewed by state and local users may increase due to the expanding number of users with access and with more understanding of how CTRs can assist their efforts.[Footnote 30] Several officials from state and local law enforcement agencies we interviewed indicated that they believed the use of CTR data through Gateway probably would increase as users better understand the value of BSA data. Also, as noted, FinCEN has recently made bulk downloads of data available to several federal agencies. A second difficulty in measuring the impact specifically of CTR requirements involves the way that CTRs are used--primarily to support investigations that also draw upon many other sources of information. Officials from the federal investigative agencies we interviewed generally stated that outcome measures, such as indictments or convictions, cannot be linked exclusively to CTRs because they are typically one of many leads used to develop an investigation. Similarly, most--about 82 percent--of the state and local agencies we surveyed indicated that the number of investigative leads provided by CTRs was the best outcome measure of their CTR use. Further, officials from several local law enforcement agencies noted that attorneys often negotiate plea agreements with the defendant long before a case goes to trial; thus, no matter how critical the role played by a CTR in the investigation, there would be no trial in which CTRs could be used as evidence. In addition, federal, state, and local law enforcement officials reported that they were more likely to use CTRs as a basis for obtaining subpoenas to access specific bank account records than to use CTRs themselves as evidence in court. Officials from IRS, DEA, and Justice said that by the time a case moves to the trial phase, the prosecution is more likely to use bank records as evidence because those records are generally a more convincing form of evidence of a defendant's transactions. However, investigators would use the CTR to locate the defendant's bank accounts and identify the correct bank records to subpoena. Further, according to an IRS official, CTRs generally were presented in court only when bank records were not available or could not be made available in a timely manner. FinCEN officials reported that the agency does not have outcome measures related to CTR use and analysis because many of the cases FinCEN supported were complex and might not result in tangible success for several years. Officials cited the example of "Operation Cash-Out," where FBI authorities eventually charged persons with attempting to provide funding to al Qaeda. During this investigation, conducted between 2000 and 2006, FinCEN identified more than 14,000 CTRs relating to the investigation. FinCEN Does Not Publish Information on CTR Uses As It Does for Other BSA Data: Although FinCEN has taken some steps to promote awareness of CTRs and their value to the financial community and law enforcement agencies, it does not systematically report information about the numbers of CTRs filed or results of CTR use. Bank officials we interviewed and those responding to our survey strongly questioned whether the CTRs they were filing, especially on customers that they had determined to be at low risk for financial crimes, provided any value to law enforcement. Some officials stated that their resources would be better directed at filing SARs, which they viewed as having greater value to law enforcement. Other institution officials noted that law enforcement agencies had never contacted them. Law enforcement officials have given presentations at banking industry conferences on how BSA data helps them in their investigations. These industry conferences typically include presentations on how law enforcement uses BSA data, but they are not necessarily CTR specific. Officials we interviewed and those responding to our survey stated that they largely did not understand how the CTRs they filed were being used by law enforcement. In contrast to this general lack of information on CTR use, FinCEN routinely publishes information on the numbers of SARs filed and examples of how they have been used by law enforcement agencies. Since October 2000, in conjunction with law enforcement and regulatory agencies, FinCEN has been issuing the SAR Activity Review, which provides information about the preparation, use, and value of SARs that depository institutions, as well as other financial institutions, filed. For example, the October 2007 edition contained expanded descriptions of law enforcement cases to demonstrate the value of BSA data to the law enforcement community, including cases that were proactively initiated as a result of BSA reports, as well as trends in certain crimes identified through SARs. FinCEN also includes some information on the results of requests to financial institutions under section 314(a) of the USA PATRIOT Act. FinCEN officials told us that limited resources currently precluded the agency from routinely analyzing and publishing trend information about CTRs filed by depository institutions. However, the officials noted that the agency recently completed a study of CTR trends and patterns, and they were considering whether to include information on CTRs in the SAR Activity Review. Many officials we interviewed and those responding to our survey indicated that they were genuinely interested in how CTRs were being used. Further, our interviews with depository institution officials suggest that they would better understand the value of meeting their requirements if the institutions were provided with some information on CTR use, similar to that reported on uses of SARs. Financial Institutions Incur Some Costs to Meet Requirements Regardless of the Number of CTRs They File: Our analysis of CTRs filed during calendar years 2004 to 2006 shows that a large proportion was filed by a small number of the largest depository institutions. While these institutions therefore likely incur the greatest expenditure of time and resources, all depository institutions incur some costs to meet CTR requirements, regardless of the number of CTRs they file. This is because institutions must have processes and trained staff in place to identify when and if a CTR is required, including the ability to aggregate same-day cash transactions made by or on behalf of the same person, and to file CTRs correctly. While automation has made these tasks less difficult, many institutions reported that their processes still include "manual" steps. Institutions we contacted were generally unable to quantify their costs for meeting CTR requirements, in large part because they use the same processes and staff for meeting other BSA requirements or for other purposes. While Most Depository Institutions File CTRs, a Small Number of the Largest Institutions Account for the Majority: Our analysis of FinCEN's data on the numbers of CTRs filed annually shows that, from 2004 to 2006, a relatively small number of the nation's approximately 17,000 depository institutions accounted for the large majority of CTRs filed. For example, in 2006, fewer than 30 very large banks (those with assets of $50 billion or more) accounted for over half (55 percent) of new CTRs during this period, while banks with assets between $1 billion and $50 billion accounted for another 30 percent.[Footnote 31] The largest credit unions--those with assets of $100 million or more--accounted for only 1 percent of new CTRs, and credit unions in total accounted for less than 2 percent. (For illustrative purposes, the remainder of this section focuses on CTRs filed in 2006. Details on the numbers of CTRs filed during the 3-year period, including analyses by institutions of different sizes, appear in app. III.) In 2006, nearly two-thirds of depository institutions filed at least one CTR--89 percent of banks and 42 percent of credit unions. However, the CTRs were concentrated among the larger institutions (see fig. 3). For example, the 27 very large banks (representing less than one-half of 1 percent of all banks) filed 55 percent of the CTRs filed. Figure 3: CTRs Filed in 2006 by Banks and Credit Unions, by Asset Size: This figure is a combination of bar graphs showing CTRs filed in 2006 by banks and credit unions, by asset size. [See PDF for image] Source: GAO. Notes: The size categories are based on institutions' assets, as follows: very large banks (greater than $50 billion); large banks (greater than $1 billion - $50 billion); midsize banks (greater than $100 million - $1 billion); small banks ($100 million or less); large credit unions ($100 million or more); midsize credit unions ($10 million to less than $100 million); and small credit unions (less than $10 million). [A] The actual percentage for very large banks was 0.2. [B] The actual percentage for small credit unions was about 0.01. The actual percentage for midsized credit unions was about 0.2 percent. [End of figure] Further analysis of the CTRs filed in 2006 revealed that the 100 largest depository institutions filed 7.8 million CTRs, or 65 percent of the total. One institution--the single largest filer--accounted for 1.7 million CTRs (14 percent of the total). The median number of CTRs filed by banks in each size category was as follows: very large banks, 125,202; large, 1,889; midsize, 154; and small, 17. Regardless of the Number of CTRs Filed, Institutions Incur Costs to Establish and Maintain a CTR Filing Process: Some institutions may rarely file CTRs--for example, our analysis showed that over 5,900 institutions did not file any CTRs in 2006--but nevertheless incur costs to establish a filing process and train their staff to meet CTR requirements. All of our survey and interview respondents reported that they had established processes to file CTRs and thus incurred costs associated with these processes.[Footnote 32] While the following briefly summarizes and depicts a typical process, we found much variation both within institutions--for example, procedures for filing CTRs resulting from aggregated transactions differed from those applicable to a single transaction--and among institutions. At most institutions, the CTR filing process typically involves a number of steps and staff members (see fig. 4). The staff may include tellers, branch supervisors, and compliance officers.[Footnote 33] A teller typically inputs the information needed to fill out the CTR form, during or immediately following a cash transaction greater than $10,000. The teller completes the form, either by hand or through an automated system, and passes it to a branch-level supervisor for review. Once this review is complete, the CTR is sent either electronically or in hard copy to the institution's compliance office for an additional review and compliance check. Once the compliance check is complete, the CTR is signed and sent either electronically to FinCEN or by mail to the IRS. Figure 4: General Process for Filing CTRs: This figure is a flowchart showing the general process for filing CTRs. [See PDF for image] Source: GAO (analysis); Art Explosion (images). [End of figure] To identify cases in which a CTR may be needed if certain individual transactions are aggregated and for other purposes, depository institutions generally keep a daily report of transactions across their branches and service centers and aggregate the transactions by customers' tax identification numbers.[Footnote 34] Typically, compliance office staff review the aggregation report to see if any of the aggregated transactions made by or on behalf of the same person meet the CTR filing threshold. Some depository institution officials we interviewed said that reviewing this aggregation report could take from 1 to 2 hours a day, while others noted that it was a time-consuming process because it was a manual or partly manual process. Further, if information is missing, additional time is required to obtain it. Our analysis of CTR data shows that in 2006, 65 percent of all CTRs filed resulted from aggregated, rather than single, transactions. Many of our survey and interview respondents said that, in general, their CTR process was not complex: Fifty percent said the process was very or somewhat simple for their institution to complete. However, 21 percent of the survey respondents said their process was very or somewhat complex. Officials we interviewed and surveyed also noted that, in general, a number of variations to the basic process outlined in figure 4 exist, depending on circumstances. For example, additional time and effort is needed to fill in any required information that a teller failed to obtain at the time of a transaction. Similarly, if an institution filed a CTR with an error, subsequently filing an amended CTR may involve collecting additional information about the transaction; and "backfiling" (cases in which an institution files a CTR after discovering one needed to be filed) may require time- consuming review of an account's transactions over a period of time. In addition, we found that while most depository institutions generally follow the same process for filing CTRs, significant variations can exist among them, which may be attributed to the quantity of CTRs filed, the number of staff involved, the degree of automation, or institutional preferences in reviewing and processing CTRs. Technology Has Expedited Some Steps in the Filing Process, but Institutions Have Not Automated All Steps: Technology has helped some depository institutions expedite and streamline many or some parts of the CTR process. Overall, 78 percent of institutions responding to our survey reported that at least one part of their CTR filing process was mostly or fully automated. Many of the institutions we spoke with have software systems that prompt the teller when a CTR is necessary for a transaction, and some institutions have systems that allow tellers to electronically access the CTR form at their workstation and enter the necessary information. Additionally, some depository institutions reported that they had software systems that automatically fill in some parts of the form. Also, some banks have invested in software that processes CTRs for final reviews by their compliance office staff. However, the extent of automation varied widely among specific steps in the process (see fig. 5), and no survey respondents reported a completely automated CTR process. For instance, 35 percent of survey respondents said they had a mostly or fully automated process for their tellers to fill out CTR forms, while 48 percent reported this step was largely manual and 16 percent used a mix of manual and automated steps. The step that was most likely to be automated was the aggregating of daily cash transactions: 68 percent of survey respondents reported that their systems automatically generate this aggregation report. The step least likely to be automated was the supervisory review of CTRs; about 10 percent of survey respondents reported that the review processes at the branch level had been automated. While we did not obtain data showing how the extent of automation compares with the volume of CTRs filed, our structured interviews with officials from depository institutions suggest that institutions filing the most CTRs (generally the larger institutions) were more likely to have highly automated processes than smaller institutions filing fewer CTRs. Because of the cost, many smaller banks that do not file as many CTRs may choose not to invest in systems that could provide a greater degree of automation. Figure 5: Extent to Which Steps in the CTR Process Were Automated at Surveyed Institutions: This figure is a combination bar graph showing the extent to which steps in the CTR process were automated at surveyed institutions. [See PDF for image] Source: GAO. [End of figure] The extent of automation could influence the time needed to process CTRs. Overall, survey respondents reported a median time of 25.2 minutes to complete a CTR in 2007 (see fig. 6). In 1998, FinCEN estimated that it took about 24 minutes to complete a CTR.[Footnote 35] Figure 6: Median Time, in Minutes, to Accomplish Each CTR Filing Step: This figure is a bar graph showing median time, in minutes, to accomplish each CTR filing step. [See PDF for image] Source: GAO. Note: The median values for each of the steps have been rounded to the nearest 0.5 minute. The median values for each step do not necessarily sum to the total median time. The 95 percent confidence interval for "teller fills out CTR" is from 4.9 to 8 minutes, the interval for "CTR sent to IRS" is from 2.6 to 4.1 minutes, and the interval for the median time total is from 24.4 to 28.8 minutes. [End of figure] Although FinCEN has taken steps to encourage institutions to file CTRs electronically, 76 percent of our survey respondents said that they filed CTRs by mail, while 14 percent reported that they filed electronically, 6 percent filed by magnetic media, and 4 percent a combination of these methods.[Footnote 36] However, institutions that do not file CTRs electronically may account for a small proportion of all CTRs. According to FinCEN data, 47 percent of all CTRs filed by financial institutions in 2006 were filed electronically, while 22 percent were filed by mail, and 31 percent were filed by magnetic media. Further, the use of electronic filing appears to be growing; in fiscal year 2003, only about 5 percent of CTRs were filed electronically. Some depository institution officials said the ability to e-file has made filing CTRs much easier at their institution. Others stated that they choose not to file electronically because the volume of CTRs they filed did not justify the required time and effort involved. According to FinCEN, electronic filing is best suited for institutions that file a larger volume of CTRs; however, the overall benefits of e-filing to lower-volume filers--for example, the e-filing system provides the institution submitting the CTR with an electronic confirmation of its receipt--in many instances may outweigh development costs. Depository Institutions Could Not Quantify Costs Specific to CTR Requirements: Although they provided some anecdotal estimates, officials of depository institutions we interviewed had difficulty separating costs for meeting CTR requirements from other BSA costs, such as preparing Suspicious Activity Reports.[Footnote 37] In particular, we found that at some banks, some staff and automated systems were used to meet CTR and other BSA filing requirements. While we asked institutions we spoke with to provide estimates of costs based on categories such as personnel, training, and technology, not all institutions were able to do so because they do not typically account for CTR costs in this way. However, officials from institutions we interviewed did describe general types of costs and provided some estimates. In general, personnel costs associated with CTRs may include the cost of training staff on meeting CTR requirements, as well as the cost of labor involved in filing CTRs. They may include salary expenses for tellers, branch managers, and BSA compliance staff. Most institutions said they provide annual training on when and how to file CTRs, and that staff members, including tellers, spent an average of about 1 hour each in CTR training. At smaller institutions, fewer staff may receive training; for example, a compliance officer at one smaller bank told us that 26 staff members received four hours of BSA training annually, a portion of which is dedicated to CTRs. In contrast, the BSA officer at a very large bank said that 1 hour of CTR training is provided annually to 160,000 staff; officials from a very large bank said it registers about 40,000 hours in CTR training each year among its staff. Many depository institutions indicated that training on CTR requirements was part of a larger BSA training course, while a few said they offered training modules focused on CTRs. For example, at a cost of $15,000, one bank has purchased access to a Web-based training program that offers courses on CTRs, as well as other areas of BSA. Furthermore, officials at one very large institution noted that, in particular, they had to conduct training more often for tellers because of high rates of turnover in teller positions. While all institutions incur some level of costs for training their staff, our interviews suggest that typically the higher the number of CTRs an institution filed, the higher the number of associated personnel--and therefore presumably training costs. Similarly, the labor costs associated with actually preparing CTRs would be expected to be larger among the institutions that file the most CTRs--though, as noted, such costs are also affected by the type of process used, including the degree of automation. The highest-volume CTR filers reported having staff solely dedicated to filing CTRs and exemptions; for example, one very large bank employed more than 190 staff at CTR operations centers, and representatives of another very large bank reported 60 staff members who worked exclusively on CTRs. Conversely, representatives of one midsized bank said that they had one and a half full-time equivalent positions in their compliance office dedicated to CTRs. Reflecting the range of numbers of staff that may be involved, officials provided a wide variance of estimated personnel costs. For example, the very large bank with more than 190 dedicated staff (which filed over 1 million CTRs in 2006) estimated the cost to be "several" million dollars. Other large filers also reported high costs for staff salaries that ranged from just less than $1 million to over $5 million. For example, one very large bank that filed almost 1 million CTRs in 2006 estimated the cost at $5.4 million--$3.6 million for the approximately 25,000 tellers involved and $1.8 million in personnel costs for staff dedicated to CTRs. In comparison, officials from one large bank that filed just fewer than 5,000 CTRs in 2006 estimated that personnel costs for tellers and compliance office staff were slightly more than $76,000 for the year. Similarly, the midsize bank that reported one and a half full-time equivalent positions dedicated to CTRs, and filed approximately 2,300 CTRs in 2006, estimated personnel costs of about $31,000 for the compliance office staff but was unable to provide an estimate for the costs associated with the tellers' time. Officials from smaller institutions we spoke with generally estimated lower costs and indicated that CTR filing responsibilities at their institutions were handled by staff that had other responsibilities, as well; one estimated that staff time for filing 65 CTRs in 2006 cost a little less than $2,000. As noted above, institutions have automated processes for meeting CTR requirements to differing extents, and officials cited technology as a significant cost. For example, one large bank was considering adding a CTR module to its current software at a cost of between $60,000 and $70,000; another large bank reported recently spending about $30,000 to purchase a new software component. However, because many of the institutions we spoke with also used these systems for other processes, they were not able to break out the costs exclusively for CTRs. For example, some officials told us their systems cost in the thousands of dollars but that they used the systems for monitoring cash transactions for suspicious activities, as well as for preparing CTRs. As a result, officials we interviewed said that, even if CTR requirements were eliminated, their institutions would still incur both personnel and systems costs because of other BSA compliance activities. An official of a very large bank said that if the CTR requirement were eliminated, the bank would be able to eliminate or reassign 14 staff to other activities but still would need to prepare many of the same reports, such as aggregation reports, because they are used for other purposes, such as identifying suspicious activity. An official of a large bank told us if there were no CTR filing requirement, the bank would realize reductions in some technology costs but would retain staff involved for their expertise and skills in other parts of its BSA program. Uncertainty about Required Documentation and Some Regulatory Requirements May Unnecessarily Discourage Use of Exemptions: FinCEN data show that depository institutions filed about 31,500 Phase I and 39,300 initial Phase II exemptions during 2004-2006.[Footnote 38] However, according to our survey results, many financial institutions with customers considered eligible for exemptions do not actually exempt them but instead continue to file CTRs on the customers' transactions--despite the institutions' recognition that making use of the exemption provisions would enable them to file fewer CTRs. (Complete survey results can be viewed at [hyperlink, http://www.gao.gov/cgi-bin.getrpt?GAO-08-385SP.] Among the reasons cited by institutions was uncertainty about the documentation required to demonstrate that some customers are in fact eligible, accompanied by some concern that examiners from the federal banking regulators would deem the documentation insufficient and cite them for BSA noncompliance. Our discussions with examiners revealed variations in the types of documentation they find acceptable, although our review of data from the banking regulators showed relatively few violations concerning exemptions compared with the number of BSA examinations conducted. Other factors discouraging use of exemptions were the cost and effort involved in meeting FinCEN's regulatory requirements to (1) file an exemption form, and annually review and update the information, particularly for certain customers that are specifically exempted by statute, as appropriate; and (2) biennially file a form to document the continued eligibility of customers that have been exempted under the Phase II regulations--which as a practical matter duplicates the required annual review process for those customers. Factors the institutions indicated might encourage use of exemptions included (1) shortening the waiting period--currently a full year under FinCEN's regulations--before exempting certain customers with a relatively large volume of cash transactions, and (2) making Web-based material available to help train and guide depository institutions' staff in making exemption determinations. Because the transactions of exempt customers are likely to be of little or no value to law enforcement, actions to encourage depository institutions to make greater use of exemptions could avoid the burden of filing some CTRs without harming law enforcement efforts. While Recognizing the Benefits of Exemptions, Depository Institutions Do Not Exempt All Eligible Customers: Exemptions allow institutions to avoid filing CTRs for the exempt customers, but the institutions are not required to exempt eligible customers. According to the results of our survey, institutions that made use of exemptions primarily did so because it allowed them to file fewer CTRs, was cost-effective, and the determinations involved were fairly easy. As shown in figure 7 below, the reasons generally were consistent for both Phase I and Phase II exemptions. Figure 7: Factors That Surveyed Institutions Reported as of Very Great or Great Importance to Their Decision to Exempt Phase I and Phase II Customers: This figure is a combination bar graph showing factors that surveyed institutions reported as of very great or great importance to their decision to exempt Phase I and Phase II customers. [See PDF for image] Source: GAO. Note: For the category "other factors," the 95 percent confidence intervals for the very great/great importance and some/little or no importance estimates are within +/-12 percentage points. [End of figure] The primary reason cited for using the exemption process was that it allowed institutions to file fewer CTRs. While it would be difficult for an institution to track the number of CTRs it "saved" or avoided by exempting a customer, some had; for example, a smaller institution reported that it recently began using exemptions more extensively, and by exempting five more Phase II eligible customers, the institution anticipated filing almost 200 fewer CTRs. (However, the effect of exempting a single customer on the number of CTRs filed cannot be generalized; for example, an exemption might avoid 8 CTRs or 100, depending on the volume of cash transactions in which the customer typically engaged.) Institutions also frequently cited the cost- effectiveness of using exemptions; while they had difficulty estimating the cost of establishing exemptions, just as they did for the costs of filing CTRs, some institutions regarded exempting customers as less costly than filing CTRs. Officials at other institutions we interviewed cited recent advances in commercial software systems that made exemptions easier. For example, software can identify the customers potentially eligible for the Phase II exemption due to the volume of high cash transactions they engaged in during the year. In addition, at least one software vendor makes available for purchase a database of companies listed on stock exchanges that are eligible for the Phase I exemption. Despite the cost-effectiveness of using exemptions, institutions responding to our survey did not exempt all of their eligible customers. For example, while 77 percent of the institutions reported having customers eligible for the Phase I exemption, only 45 percent reported that they always or usually filed Phase I exemptions. Similarly, 83 percent of the institutions reported having customers eligible for the Phase II exemption, but only 49 percent reported that they always or usually filed Phase II exemptions (see fig. 8). Figure 8: Percentage of Depository Institutions with Customers Eligible for Phase I and II Exemptions and Extent to Which They Filed Exemptions in 2006: This figure is a combination pie and bar graph showing percentage of depository institutions with customers eligible for Phase I and II exemptions and extent to which they files exemptions in 2006. [See PDF for image] Source: GAO. Note: The shaded portion of each circle equals the percentage of institutions reporting that they had customers eligible for the exemptions. [End of figure] Some institutions that file large numbers of CTRs--and, therefore, might realize the greatest savings by avoiding CTRs--do not file many exemptions. Some of the reasons for this are discussed in the following sections. (Further details on the results of our survey, including the percentages of institutions that cited specific factors affecting their decisions to exempt or not exempt customers, are presented in app. IV and in GAO-08-385SP.) Uncertainty about Documentation Needed to Demonstrate Eligibility, Accompanied by Concerns of BSA Noncompliance, Deterred Some Exemptions: The leading reason identified by survey respondents that choose not to file Phase II exemptions was difficulty in determining the percentage of a customer's gross revenue derived from lines of business not eligible for exemption. This difficulty--along with other concerns, including that federal banking regulators would deem documentation insufficient--contributed to a reluctance to exempt customers that the institutions considered potentially eligible. The responses of officials of institutions we interviewed were consistent with our survey results. Officials explained that a fair amount of research was required on their part to determine eligibility under the Phase II regulations--for example, examining a business's tax returns or financial statements--and that it was not always clear if the customer qualified for the exemption because it was difficult to determine which part of a business customer's revenue was derived from which activity. The depository institutions that chose to use the Phase II exemption used various methods to document the portion of revenues derived from ineligible activities. Officials of several institutions we interviewed said they arrived at this determination after conducting what they said was exhaustive research, which included analyzing financial statements, searching the Internet, and reviewing available documents if the institution had a lending relationship with the customer, or asking the customer for documents. Officials of other institutions used less labor-intensive methods; for example, an official of one midsize institution indicated that the account officer simply asked customers about the source of their gross revenue and made a notation in the customer's file. Several institutions reported using a letter from the customer to self-certify that no more than 50 percent of their gross revenue came from activities or lines of business ineligible for exemption. Officials from the federal banking regulators generally indicated that they did not have a standardized expectation for what documentation (such as financial statements or tax documents) an institution might use to demonstrate the portion of revenues derived from ineligible activities. They further noted that examiners have some flexibility in determining what level of documentation is required, based on guidance in the: BSA/AML Examination Manual.[Footnote 39] The same manual is used by each of the five federal banking regulators and is available to depository institutions to help guide their BSA compliance activities. However, our interviews with officials and examiners indicated differences among them regarding the type of documentation acceptable. For example, federal regulators and examiners we interviewed had different views about the use of a self-certifying letter and whether depository institutions ought to provide other documentation. While Federal Reserve and FDIC officials said the acceptance of a self- certifying letter would depend on the circumstances, they generally noted that examiners had flexibility in deciding what level of documentation would be acceptable. Officials from OTS and OCC, on the other hand, indicated that a self-certification letter alone would be inadequate to show eligibility. Because, in this instance, the federal banking regulators examine institutions for compliance with FinCEN's regulations, additional guidance from FinCEN could help reduce the difficulties that depository institutions face in making this determination and clarifying, for both the institutions and the regulators, the types of documentation acceptable for demonstrating eligibility. Officials and examiners we interviewed from all of the federal banking regulators indicated that they have found few problems with exemptions, and our review of available violation data for 2005 and 2006 indicated that examiners cited relatively few violations for exemptions. We asked the regulators to disaggregate their data on violations to distinguish those related specifically to exemptions; only the Federal Reserve, FDIC, and OCC were able to provide this level of detail. These three agencies are responsible for examining about 7,800 depository institutions, including the largest banks that likely account for the greatest numbers of CTRs. As shown in table 3, the three agencies collectively found violations associated with exemptions in less than 5 percent of the BSA exams they conducted--a combined total of 227 violations for exemptions in 2005 and 113 violations for exemptions in 2006. Table 3: Exemption Violations Cited in BSA Examinations by FDIC, Federal Reserve, and OCC, 2005 and 2006: 2005; Agency: FDIC; Number of BSA examinations conducted: 3,029; Number of exemption violations issued: 178; Percentage of exemption violations per examination: 5.9. 2005; Agency: Federal Reserve; Number of BSA examinations conducted: 678; Number of exemption violations issued: 10; Percentage of exemption violations per examination: 1.5. 2005; Agency: OCC; Number of BSA examinations conducted: 1,510; Number of exemption violations issued: 39; Percentage of exemption violations per examination: 2.6. 2005; Agency: Total; Number of BSA examinations conducted: 5,217; Number of exemption violations issued: 227; Percentage of exemption violations per examination: 4.4. 2006; Agency: FDIC; Number of BSA examinations conducted: 2,825; Number of exemption violations issued: 80; Percentage of exemption violations per examination: 2.8. 2006; Agency: Federal Reserve; Number of BSA examinations conducted: 815; Number of exemption violations issued: 6; Percentage of exemption violations per examination: .7. 2006; Agency: OCC; Number of BSA examinations conducted: 1,547; Number of exemption violations issued: 27; Percentage of exemption violations per examination: 1.7. 2006; Agency: Total; Number of BSA examinations conducted: 5,187; Number of exemption violations issued: 113; Percentage of exemption violations per examination: 2.2. Source: GAO analysis of Federal Reserve, FDIC, and OCC data. [End of table] Similarly, we asked FinCEN for data on BSA enforcement actions it has taken against depository institutions related to exemptions. (While FinCEN generally coordinates with the federal banking regulators, it may independently take enforcement actions, including imposing penalties and fines, for BSA violations.[Footnote 40]) FinCEN data show that, over the 10-year period 1997 to 2006, it took 110 BSA enforcement actions related to exemptions, 4 of which included fines. (More detailed information on FinCEN's enforcement actions is presented in app. IV.) The fairly low incidence of violations associated with exemptions may reflect depository institutions' decisions to simply not grant exemptions, thus avoiding potential violations. (Some examiners noted that they sometimes encouraged depository institutions to use the exemption process, for example, if the institution was filing many CTRs on customers that were potentially eligible for the exemption.) However, our survey and interviews demonstrate that a lack of clear guidance from FinCEN for documenting eligibility, and the differing interpretations among the federal banking regulators, have the effect of dissuading depository institutions from more frequently using the Phase II exemption. A minority of our survey respondents indicated that they "always" exempt eligible customers--33 percent reported doing so for Phase I-eligible customers and 26 percent for Phase II-eligible customers. Some depository institution officials noted that any compliance deficiency found by BSA examiners was a cause for concern. (About 10 percent of survey respondents reported that they had received a CTR violation or had been fined since 2000.) An official from the very large bank that filed more than 150,000 CTRs in 2006 said it was the bank's official policy not to exempt any new customers that were eligible for the Phase II exemption because, among other things, the bank faced reputation risk if it was cited for a BSA violation, and use of the exemption process opened the bank to examiner criticism and fines. An official from a large community bank said that the bank did not file Phase II exemptions because of concerns about regulatory risk. Officials from several depository institutions we interviewed specifically said it was not clear to them what level of support was needed, and some indicated that they would rather file CTRs than take the risk of not satisfying an examiner. In a 2002 report on the exemption process mandated by section 366 of the USA PATRIOT Act, FinCEN concluded that it should work with the federal banking regulators, as well as banks, to reduce "fear of adverse regulatory consequences from making incorrect exemption determinations."[Footnote 41] Exemptions are addressed in the BSA/AML Examination Manual, which was first published in 2005 and, as noted, is used by the banking regulators and is available to depository institutions. However, 68 percent of our survey respondents said that difficulty in determining whether companies derive more than 50 percent of their revenues from ineligible business activities was a "very great" or "great" factor in their decision not to exempt Phase II- eligible customers. Guidance that could help institutions make greater use of this exemption would help avoid unnecessary CTRs that are of little or no use to law enforcement. Biennial Renewals, Which Duplicate Annual Reviews, Discourage Use of Some Phase II Exemptions: FinCEN's regulations require that depository institutions (1) annually- -at least once a year--review and verify the information supporting any exemptions that they have filed for either Phase I or Phase II customers, and (2) biennially file--on March 15 of the second calendar year following the initial exemption--a renewal form to continue the exemption of Phase II customers. The purpose of the annual review is to ensure that the customers continue to qualify for exemption; according to FinCEN, the biennial renewal provides formal notification to FinCEN that the institution has monitored the customers' transactions as required. About 49 percent of our survey respondents indicated that the time-consuming nature of the biennial renewal was of great or very great importance in contributing to their decision not to exempt customers eligible for Phase II exemptions.[Footnote 42] An official from the very large bank that filed more than 150,000 CTRs in 2006 said it was the bank's official policy not to exempt any new customers that were eligible for the Phase II exemption because of the costs associated with the biennial renewals and the need to keep track of which exemptions had to be renewed in each year. Officials from depository institutions we interviewed, particularly those that did not exempt customers, also said that the need to conduct this review discouraged their use of the exemption. Officials of some depository institutions questioned the value added by biennial renewals, observing that they were already conducting the annual review as well as monitoring all of their customers for suspicious activities, which is part of a strong anti-money-laundering program pursuant to section 352 of the USA PATRIOT Act. Even officials of institutions that nevertheless filed and maintained exemptions considered the requirement to be redundant. For example, officials at one of the very large banks--which had more than 1,900 Phase II exemptions on file--said they filed the "biennial" renewal form every year for every customer, because the bank went through the same steps for the biennial renewals as it did for each required annual review and did not want to risk failing to file a biennial renewal form in the correct year. Further, our analysis of FinCEN data revealed that some institutions file biennial renewal forms on Phase I customers, although they are required only for Phase II customers (in 2006, depository institutions filed 1,382 biennial renewals on Phase I customers). FinCEN established the biennial renewal requirement based on its interpretation of the Money Laundering Suppression Act. Specifically, the act requires the Secretary of the Treasury to prescribe regulations requiring that depository institutions review, at least annually, the qualified business customers that they have exempted and to "resubmit information about such customers" to the Secretary.[Footnote 43] According to FinCEN, the implementing regulations provided for the information to be resubmitted biennially, rather than annually, because the statute does not explicitly set a time frame for the resubmission.[Footnote 44] Further, FinCEN officials believe that the Secretary has general authority to prescribe appropriate exemptions to requirements under the BSA, including revising the regulations to eliminate the biennial renewals.[Footnote 45] FinCEN officials said that the biennial renewal form provides them with evidence that the exempt business remains eligible for the exemption and that the institution has been monitoring the business for suspicious activity. In addition, they reported that FinCEN routinely analyzes biennial renewal forms (along with other information) filed on and by specific depository institutions that are the subjects of compliance or enforcement actions by the federal banking regulators to determine if the institutions properly granted exemptions to eligible customers. However, these activities essentially duplicate those of the bank examiners who, as part of the examination process, ascertain whether institutions properly grant exemptions and monitor their customers for suspicious activity.[Footnote 46] Examiners from a few of the banking regulators indicated that the biennial renewal requirement results in depository institutions collecting the same kinds of information that they collect as part of the annual review of exemptions. Further, all biennial renewals must be filed on March 15, regardless of when the exemption was filed. Officials from the Federal Reserve noted that meeting both requirements can impose significant compliance costs on the institutions, yet the duplication provides no offsetting benefit for supervisory efforts. Eliminating the requirement for biennial renewals could encourage more institutions to make use of Phase II exemptions and reduce the burden associated with filing unnecessary CTRs. Current Regulations Require Institutions to File Exemptions for Customers That Are Statutorily Exempt: Recognizing that the cash transactions of customers that are depository institutions or governmental entities would likely be of little or no use to law enforcement efforts, the Money Laundering Suppression Act specifically directed that the Secretary of the Treasury exempt depository institutions, as appropriate, from filing CTRs on the transactions of these customers. FinCEN did so, but its regulations require depository institutions to file exemption forms if they choose to exempt these types of customers--and to annually review and verify the information supporting the exemption. The statute does not mandate annual reviews for these customers. In essence, the regulations treat these entities like all other customers eligible for Phase I exemptions, including listed companies and majority-owned subsidiaries. Accordingly, if depository institutions choose to exempt these customers, they must perform the same steps and incur costs for annual reviews as they do for other customers they exempt. But depository institutions and governmental entities, in contrast to other Phase I entities such as publicly traded companies, are unlikely to change those characteristics that initially qualified them for exemption.[Footnote 47] For example, a governmental entity is unlikely to become a private company. In any case, a change in the status of a governmental entity or bank would most likely require that the exempted bank account be closed and a new one be opened--triggering a new consideration for exemption. Few institutions we interviewed cited difficulty in determining eligibility for their customers that are other depository institutions or government entities, and many said that, in these cases, they exempt all eligible customers. However, they would have incurred some cost to file the form and to annually review the supporting information. In response to our survey, officials of depository institutions reported that their staff took a median time of about 34 minutes to exempt a Phase I customer, and about 14 minutes for the annual review process.[Footnote 48] Further, some depository institutions do not exempt these customers and continue to incur the cost of filing CTRs. For example, our analysis of FinCEN data shows that, in 2006, almost 87,000 CTRs were filed on over 2,900 depository institutions, and about 45,000 CTRs were filed on some 5,500 government entities.[Footnote 49] These CTRs are unnecessary in that the cash transactions of these entities are not likely to have a high degree of usefulness for law enforcement. According to FinCEN officials, the information provided on the exemption forms for these entities is not required for analytical purposes per se but rather serves as the basis for recording which financial institutions had chosen to exempt specific depository institutions and governmental agencies. However, depository institutions are separately required to keep records of customers' transactions for BSA purposes.[Footnote 50] Federal Reserve officials specifically noted that they believed that the automatic exemption of domestic depository institutions from the CTR filing requirement should be considered and that eliminating the need to file an exemption and keep it current for these entities would make the CTR process more efficient. Continuing to require depository institutions to file forms on these entities--and to incur the cost and effort of annually reviewing the information supporting the exemption--discourages use of the exemption, resulting in CTRs that are likely to be of little or no value to law enforcement. Length of Time Allowed Before Frequent Customers Can Be Exempted May Result in Unnecessary CTRs: FinCEN's Phase II exemption regulations specify that, in order to be eligible for exemption, among other things customers must have held an account for at least 1 year and must have "frequently" engaged in currency transactions in excess of $10,000. In a November 2002 guidance memorandum, FinCEN defined "frequently" as at least eight large currency transactions in a 1-year period (with an exception for seasonal customers).[Footnote 51] Officials of several banks we surveyed said that their use of exemptions for Phase II customers would increase if they were permitted to exempt businesses with frequent cash transactions in less than 12 months. As explained by an official of one institution, a year seems to be an unnecessarily long time if the business is by nature cash-intensive and not suspicious, and the institution regularly files CTRs on the business. Or, as other officials noted, a waiting period of less than 1 year would be appropriate if the ownership of a business changed but the transaction activity remained relatively similar to that under the previous owner, or if known customers chose to form new businesses. We analyzed FinCEN's data to identify the numbers and frequency of CTRs filed on customers that were subsequently exempted. We found that, among customers that were initially exempted in 2006, the median number of CTRs filed in the 12 months preceding the exemption was 14; the median number filed in the 8 months preceding exemption was 11; and in the preceding 6 months, it was 9.[Footnote 52] This analysis demonstrates that many customers that were later exempted engaged in more than the 8 transactions in a 12-month period required by FinCEN-- generating thousands of unnecessary CTRs. FinCEN promulgated its regulations establishing the 12-month requirement before enactment of the USA PATRIOT Act. That law provided for customer identification programs, for which FinCEN regulations require depository institutions to collect sufficient information to verify the identity of customers when they first open an account.[Footnote 53] Thus, depository institutions must require new business customers to provide their name, physical location, and taxpayer identification number, at a minimum, at account opening. Furthermore, as noted above, BSA compliance programs require depository institutions to monitor their customers for suspicious activity. Thus, continuing to require a 12-month period before allowing otherwise nonsuspicious customers with large numbers of cash transactions to be exempted may needlessly cause depository institutions to file CTRs that are not highly useful to law enforcement efforts. Material to Help Train Institutions on Requirements Could Increase Use of Exemptions: FinCEN currently provides material on its Web site, such as answers to frequently asked questions, rulings and guidance, and information on BSA requirements. However, the responses to frequently asked questions and rulings and guidance concerning exemptions are limited and dated.[Footnote 54] Forty-eight percent of respondents to our survey indicated that the availability of Web-based material from FinCEN would greatly or moderately increase their use of the Phase I exemption, and 51 percent of respondents said it would greatly or moderately increase their use of the Phase II exemption. Such material would help train respondents' staff and guide them in interpreting and applying the exemption requirements. Officials of depository institutions we interviewed generally indicated that they currently purchase training modules from vendors, hire trainers, or have their compliance officers develop in-house training. Our work suggests that material to help train staff could assist depository institutions in making some eligibility determinations under both the Phase I and Phase II regulations and help overcome difficulties that often dissuade institutions from greater use of the exemptions. As previously discussed, our survey results and interviews with depository institution officials highlighted difficulties in determining the portion of a customer's gross revenue derived from lines of business not eligible for the exemption, which dissuaded some institutions from using the Phase II exemption. Similarly, about 39 percent of survey respondents indicated that difficulties in determining eligibility was a factor of great or very great importance in their decision not to exempt a customer eligible for the Phase I exemption.[Footnote 55] The difficulties included determining whether customers are "listed" (publicly traded) companies or are majority- owned subsidiaries of such companies. FinCEN's regulations provide that institutions may, among other things, rely on documents filed with the Securities and Exchange Commission or listings of the three stock exchanges published in newspapers or available on Web sites.[Footnote 56] However, officials we interviewed stated that verifying the publicly traded status of customers was not always straightforward, and it is sometimes difficult to determine ownership structures. For example, officials from one bank explained that the bank held a number of accounts for a large publicly traded video rental chain; some of the stores were corporately owned while others were independent franchises. While the company-owned stores would be part of the publicly traded company, the franchise operations were not likely to be publicly traded. The bank did not exempt any of these accounts, however, to avoid the risk of exempting a customer that was not eligible for the Phase I exemption. Further, some survey respondents stated they had difficulty determining eligibility when a customer is a subsidiary of a listed company. One respondent noted that his institution would not attempt to exempt such a customer because it was too difficult to document eligibility in this case. In addition, our analysis of FinCEN's data on exemptions filed from 2004 through 2006 suggests that institutions might benefit from the availability of Web-based material from FinCEN. For example, we found that some institutions were filing biennial renewals for Phase I exemptions, even though FinCEN regulations require renewals only for Phase II exemptions. Also, our interviews with examiners from the federal banking regulators indicated that further training might encourage appropriate use of exemptions. For example, examiners from FDIC and OCC reported that some institutions had difficulty distinguishing between businesses eligible for the Phase I and Phase II exemptions. Some examiners reported that they were educating depository institution staff about the exemption requirements, as well as credit union examiners in particular, on the use of Phase I exemptions for correspondent banks.[Footnote 57] Treasury's 2002 report on the uses of CTRs noted the importance of making the exemption system easier for bank personnel to understand. In preparing that report, FinCEN relied in part on a contractor's survey of depository institutions, including their exemption practices and the reasons underlying them. The contractor concluded that FinCEN should offer a Web-based training module on its Web site to clarify the exemption process. Our work for this report indicates Web-based material that would help train staff could encourage institutions to make greater use of exemptions, thereby avoiding the filing of CTRs that are of little or no use to law enforcement efforts. Providing such material on FinCEN's Web site would be a cost-effective way to help ensure that all institutions have available up-to-date information on how to meet the requirements. Conclusions: Since GAO reported over a decade ago that the large volume of CTR reports had made analysis difficult, expensive, and time consuming, developments in information technology have provided law enforcement with the capacity to simultaneously analyze large quantities of CTR data and link these with other data sets. These technological advancements, as well as the advent of bulk data downloads and expanded access to CTR data by state and local users, have provided law enforcement agencies with greater potential to make use of CTR data in their investigations of a wide variety of financial and related crimes. Further, in addition to supporting specific investigations, CTR requirements aid law enforcement by forcing criminals--who attempt to avoid reportable transactions--to act in ways that increase chances of detection through other methods. Given the multiplicity of sources that federal law enforcement officials may tap in their investigations, and the variety of possible case outcomes, it is understandably difficult to link the use of CTRs with specific outcomes. However, information that law enforcement agencies could provide on how CTRs contribute to their efforts, similar to information they provide on their use of Suspicious Activity Reports, is not systematically provided to depository institutions or shared with state and local law enforcement agencies that have more recently gained access to BSA data. Many depository institutions indicated a desire for some assurance that the information they provide is actually useful to law enforcement efforts. FinCEN routinely collects and makes available information on how Suspicious Activity Reports have contributed to investigations through publication of its SAR Activity Review. A similar approach for collecting and publishing CTR information could provide financial institutions with evidence that their efforts are contributing to detecting and deterring money laundering and other crimes. While recognizing that this effort would entail an investment of resources, we believe it would prove beneficial by providing depository institutions with greater awareness that CTRs are a valuable source of data for law enforcement investigations. With the partial exception of institutions that file the largest numbers of CTRs and have personnel dedicated to that function, most institutions are not able to quantify their costs of complying with CTR requirements, largely because they use the same personnel and automated systems for a variety of purposes. Nevertheless, depository institutions expend what could be considered to be significant amounts of time and resources to meet the requirements. Most depository institutions, based on over 35 years of collective experience in filing CTRs, have established processes that have allowed the filing of most CTRs to become fairly routine. Yet, while technology has helped them meet filing requirements more efficiently, it is clear that most institutions' processes involve steps that cannot be completely automated. These include reviews by compliance officers or other officials to provide assurance that CTRs are correct and will not unduly expose their institutions to risk of being cited for BSA noncompliance by their examiners. Further, because all institutions are subject to compliance with CTR requirements, all incur some costs--for example, in training their staff--regardless of the numbers of CTRs they file. While impacts could therefore vary among institutions depending on the numbers of CTRs they currently file as well as the processes they use, steps to reduce the number of unnecessary CTRs filed could avoid some costs. Increasing use of exemptions would help depository institutions avoid filing unnecessary CTRs, as well as reduce the government's costs to process them. Institutions we surveyed told us t