United States General Accounting Office GAO Report to Chairman and Ranking Minority Member, Committee on Finance, U.S. Senate January 2003 IRS LOCKBOX BANKS More Effective Oversight, Stronger Controls, and Further Study of Costs and Benefits Are Needed GAO-03-299 a January 2003 IRS LOCKBOX BANKS More Effective Oversight, Stronger Highlights of GAO-03-299, a report to the Controls, and Further Study of Costs and Committee on Finance, U.S. Senate Benefits Are Needed Lockbox banks are commercial FMS has contractual agreements with four lockbox banks, which operate banks that process certain taxpayer 11 lockbox sites at nine locations on IRS’s behalf. Of the more than receipts on behalf of the Internal $2 trillion in tax receipts that IRS collected in fiscal year 2002, lockbox Revenue Service (IRS). Following banks processed approximately $268 billion. an incident at a lockbox site during 2001, which involved the loss and destruction of about 78,000 tax The findings of GAO’s study include the following: receipts totaling more than • Nothing inherent in the lockbox contractual agreements would $1.2 billion, the Committee asked necessarily contribute to mishandling of tax receipts. Although a desire GAO to examine whether to avoid negative consequences, such as financial or other penalties (1) provisions of the contracts allowed for by the agreements, could motivate bank employees to make under which lockbox banks operate address previously poor decisions, penalty provisions are necessary to help the government identified problems or might address inadequate performance. The results of an ongoing investigation contribute to mishandling of tax of the 2001 incident may help IRS and FMS determine whether new receipts, (2) oversight of lockbox provisions or modifications to existing provisions are needed. banks is adequate, (3) internal • Although IRS and FMS have significantly increased their presence at controls are sufficient, and (4) IRS lockbox sites, oversight of lockbox banks during fiscal year 2002 was not and Treasury’s Financial fully effective to ensure that taxpayer data and receipts were adequately Management Service (FMS) had considered the costs and benefits safeguarded and properly processed. Inadequate oversight resulted of contracting out the functions mainly from (1) a lack of clear oversight directives and policies, performed by lockbox banks. (2) failure to perform key oversight functions, and (3) conflicting roles and responsibilities of IRS personnel responsible for day-to-day oversight of lockbox banks. • Internal controls, including physical security controls, need to be GAO is making recommendations strengthened at IRS lockbox locations (see table below). In addition, the to improve oversight and internal controls at IRS lockbox banks. In processing guidelines under which IRS lockbox banks operate need to addition, GAO is recommending be revised to improve receipt-processing controls, employment that a study of the benefits and screening, and courier security. costs, including opportunity costs, • IRS and FMS have not performed a comprehensive study of the costs of using lockbox banks to process and benefits of using lockbox banks. The most recent study, in 1999, tax receipts be completed before omitted some costs that may have affected the result. For example, the the current lockbox bank contracts study did not consider opportunity costs—benefits forgone that might expire in 2007. have resulted from alternative uses of the money. Because of these In commenting on a draft of this omissions and several changes that have affected costs and benefits, a report, FMS and IRS agreed with new study will be needed before lockbox contracts expire in 2007. our recommendations and have initiated or plan to initiate actions Internal Control Issues Found at Lockbox Locations in Calendar Year 2002 to implement them. Lockbox bank location 1 2 3 4 5 6 7 8 9 Physical security issues X X X X X X X X X Processing control issues X X X X X X X X www.gao.gov/cgi-bin/getrpt?GAO-03-299. Courier security issues X X X Employee screening issues X X X X X To view the full report, including the scope and methodology, click on the link above. Source: Findings based on reviews at the nine lockbox locations. For more information, contact Steven J. Sebastian (202-512-3406). Contents Letter 1 Results in Brief 3 Background 6 Scope and Methodology 10 It Is Not Known Whether Contract Provisions Could Contribute to Improper Handling of Taxpayer Receipts 12 Oversight of Lockbox Banks Was Not Fully Effective 15 Lockbox Banks’ Internal Control Deficiencies Expose the Federal Government to Theft and Loss 25 A Comprehensive Study Evaluating Costs for IRS Processing Versus Using Lockbox Banks for All Types of Tax Receipts Has Not Been Performed 30 Conclusions 37 Recommendations for Executive Action 38 Agency Comments and Our Evaluation 42 Appendixes Appendix I: Internal Control Weaknesses 44 Appendix II: GAO Analysis of IRS’s and FMS’s August 1999 1040 Tax Payment Comparative Cost Benefit Study 63 Appendix III: Comments from the Department of the Treasury 69 GAO Comments 75 Appendix IV: GAO Contacts and Staff Acknowledgments 76 GAO Contacts 76 Acknowledgments 76 Tables Table 1: Internal Control Issues Found at Lockbox Locations in Calendar Year 2002 26 Table 2: Comparison of the Cost to the Federal Government of Alternative Receipt Processing Approaches, Fiscal Years 2001-2007 32 Table 3: IRS and Lockbox Banks Cost and Saving Estimates for the Federal Government, Fiscal Years 2001 through 2007 63 Table 4: IRS and Lockbox Banks Cost and Saving Estimates Under IRS Scenario I for the Federal Government, Fiscal Years 2001 through 2007 64 Table 5: IRS’s Fiscal Year 2001 Labor Cost 66 Page i GAO-03-299 IRS Lockbox Banks Contents Figures Figure 1: Flow of lockbox collections and dataFMS formalized the lockbox 7 Figure 2: Fiscal Year 2002 Dollar Value and Volume of Collections for Three Major Types of Tax Receipt Collection Mechanisms 9 Abbreviations AES automated entry system AWSS Agency-Wide Shared Services EFTPS Electronic Federal Tax Payment System FAR Federal Acquisition Regulation FMS Financial Management Service FSD Financial Services Division IRM Internal Revenue Manual IRS Internal Revenue Service LPG Lockbox Processing Guidelines NPV net present value OIG Office of Inspector General PCD program completion date TFM Treasury Financial Manual TIGTA Treasury Inspector General for Tax Administration TTB Treasury Time Balances 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. It may contain copyrighted graphics, images or other materials. Permission from the copyright holder may be necessary should you wish to reproduce copyrighted materials separately from GAO’s product. Page ii GAO-03-299 IRS Lockbox Banks A United States General Accounting Office Washington, D.C. 20548 January 15, 2003 Leter The Honorable Max Baucus Chairman, Committee on Finance United States Senate The Honorable Charles E. Grassley Chairman (designate), Committee on Finance United States Senate In its role as the nation’s tax collector, the Internal Revenue Service (IRS) collected over $2 trillion during fiscal year 2002. Of that amount, $1.5 trillion was collected through electronic means,1 financial institutions serving as lockbox banks collected more than $268 billion, and IRS offices collected $86 billion. A lockbox bank is a commercial bank with a designated post office box to which taxpayers are instructed to mail their payments and related tax documents. Lockbox banks process the documents, deposit the receipts, and then forward the documents and data to IRS’s Submission Processing Centers, which update taxpayers’ accounts. The Department of the Treasury’s Financial Management Service (FMS) has contractual agreements2 with four lockbox banks, which operate 11 lockbox sites at nine locations on IRS’s behalf. The intent of the lockbox program is to accelerate the deposit of tax receipts and increase interest savings, thus enhancing the efficiency of government cash management. Lockbox banks have been used since 1984 for this purpose. 1 Business and individual taxpayers use the Electronic Federal Tax Payment System (EFTPS) to pay taxes. EFTPS is an electronic system managed by two financial agent banks chosen by the Department of the Treasury. 2 Pursuant to 12 U.S.C. 265 and other authorities, the Secretary of the Treasury has authority to designate financial institutions as depositaries and financial agents of the U.S. government to perform certain financial services, including lockbox services. These agreements are legally binding but are not considered procurements subject to the provisions of the Federal Property and Administrative Services Act (41 U.S.C. §§ 251-260) or the Federal Acquisition Regulation (FAR). In this report, we refer to them as contractual agreements and use the terms “contracts” and “contractors” for ease of reference. Page 1 GAO-03-299 IRS Lockbox Banks Instances of fraud, waste, and abuse have occurred at IRS lockbox banks over the years. In 2001, approximately 78,000 federal tax receipts, valued at more than $1.2 billion, were lost or destroyed at a lockbox bank operated by the Mellon Financial Corporation in Pittsburgh, Pennsylvania. Instances of employees stealing and cashing taxpayer receipts have also occurred at lockbox banks. Although FMS and IRS have taken steps to increase monitoring of and internal controls at lockbox banks as a result of the loss and destruction of tax receipts at the Pittsburgh lockbox site, we and others, including the Treasury Inspector General for Tax Administration (TIGTA) and offices within IRS and FMS, continue to find internal control weaknesses at lockbox banks.3 The purpose of this report is to respond to your request that we review existing and planned security and internal control measures at all lockbox banks that have agreements with FMS to provide tax receipt processing services for IRS. Specifically, you asked that we • determine whether the new lockbox agreements in effect beginning in 2002 address previously identified problems and whether contract provisions may contribute to improper handling of taxpayer returns, • determine the adequacy of FMS’s and IRS’s oversight in monitoring lockbox banks’ adherence to the contractual agreements and lockbox processing guidelines, • determine whether controls over processing and safeguarding of taxpayer receipts and data are in place and are working effectively, and • assess the extent to which IRS and FMS considered the costs and benefits of contracting out the functions performed by the lockbox banks instead of using IRS employees. To meet these objectives, we (1) reviewed reports relevant to oversight and management of the lockbox program, (2) reviewed laws, regulations, and guidance related to federal cash management activities, (3) interviewed FMS and IRS officials, (4) compared the 1993 and 2002 lockbox bank 3 See U.S. General Accounting Office, Internal Revenue Service: Progress Made, but Further Actions Needed to Improve Financial Management, GAO-02-35 (Washington, D.C.: Oct. 19, 2001); U.S. Department of the Treasury, Treasury Inspector General for Tax Administration, Nationwide Guidelines and Controls for Lockbox Banks Need Further Improvement, 2002- 30-180 (Washington, D.C.: Sept. 18, 2002). Page 2 GAO-03-299 IRS Lockbox Banks contractual agreements, and (5) reviewed FMS and IRS policies, guidelines, checklists, and reports from oversight reviews. We also visited two IRS Submission Processing Centers and all nine lockbox locations, and reviewed and discussed with FMS and IRS studies on the costs and benefits of processing tax receipts by lockbox banks. As agreed with your office, we did not specifically review the mishandling incident at the Pittsburgh lockbox site since an ongoing investigation was in process. Results in Brief We found nothing inherent in the new 2002 lockbox bank contractual agreements or the prior agreements that would necessarily contribute to mishandling of taxpayer receipts. The agreements do contain penalty provisions that can lead to negative consequences for banks if their work does not meet quality standards or is not performed within required time frames. The consequences range from financial penalties to termination of the lockbox agreement. Although a desire to avoid negative consequences could motivate lockbox bank employees to make poor decisions in handling taxpayer receipts, penalty and termination provisions are necessary to help the federal government address inadequate contractor performance. As with any federal contract, effective oversight of contractor performance is critical. FMS and IRS made some enhancements to the 2002 agreements, such as the addition of a new performance penalty and clarification of other provisions. Because TIGTA’s investigation of the incident involving the loss and destruction of tax receipts by employees at the Pittsburgh lockbox site during the 2001 April peak processing period is still ongoing, it is unclear whether any provisions in the lockbox agreements may have contributed to the mishandling incident. When the results of the investigation are known, FMS and IRS should determine whether contract provisions need to be modified or whether additional controls need to be implemented. Although both IRS and FMS have significantly increased their presence at lockbox banks and made other improvements, oversight of lockbox banks was not fully effective for fiscal year 2002 to ensure that taxpayer data and receipts were adequately safeguarded and properly processed. The weaknesses in oversight resulted largely from (1) a lack of clear directives and documented policies and procedures for various oversight functions, (2) key oversight functions not being performed, and (3) conflicting roles and responsibilities for IRS lockbox coordinators. The lack of defined oversight roles and responsibilities resulted in (1) IRS and FMS failing to take official action on bank requests for waivers from required processing Page 3 GAO-03-299 IRS Lockbox Banks guidelines, thus permitting banks to operate for months with internal control deficiencies, (2) FMS not always obtaining responses from lockbox bank management to findings and recommendations from FMS on-site reviews, (3) key IRS officials not always participating in security reviews in which they had agreed to participate and for which they had the expertise, and (4) instances of incomplete and inadequate on-site reviews. In addition, the IRS personnel who perform compliance reviews at the lockbox sites have conflicting responsibilities—to monitor bank compliance with the processing guidelines and to assist the banks with required production goals—that could affect their ability to objectively oversee lockbox banks for which they have responsibility. These deficiencies weaken the effectiveness of FMS’s and IRS’s oversight of lockbox banks. Also, without clearly defined and documented oversight requirements, there is less assurance that the oversight improvements made during 2002 will be sustained. We found during our reviews at the nine lockbox bank locations that internal controls, including physical security controls, need to be strengthened and that lockbox processing guidelines need to be revised to ensure that taxpayer data and receipts are adequately safeguarded. For example, we found required door alarms at several banks that were barely audible or did not elicit the expected responses from security guards, and we found prohibited types of personal belongings, such as bulky coats or carrying bags that could be used to conceal items, in the tax receipt processing areas of several locations. These weaknesses increase the risk that taxpayer data or receipts could be stolen from the processing area. We also found that improvements are needed in the lockbox processing guidelines with respect to processing controls, employment screening, and courier security. For example, IRS lockbox processing guidelines do not require lockbox bank couriers and permanent employees to undergo the same type of background investigation that IRS couriers and contractors undergo, even though lockbox bank couriers and permanent employees have the same access to taxpayer data and receipts as IRS couriers and contractors. Ensuring that effective lockbox processing guidelines are in place and being followed decreases the risk of loss, theft, and mishandling of taxpayer receipts and data. IRS and FMS have not performed a comprehensive study that evaluates the cost of IRS processing tax receipts and the cost of lockbox banks processing tax receipts for all types of tax receipts currently processed by the banks. The most recent study, jointly performed by IRS and FMS in 1999, considered the costs of processing individual tax receipts only and Page 4 GAO-03-299 IRS Lockbox Banks did not consider all of the costs and benefits of using lockbox banks rather than IRS to process individual tax receipts. In addition, the study focused exclusively on the costs and benefits to the federal government resulting from speedier deposit of tax receipts. Having adopted this approach using Treasury guidance, IRS and FMS did not consider certain costs, most notably opportunity costs, or the benefits forgone that might have resulted from alternative uses of the money spent to achieve speedier deposits. Agencies receive budgets that they are expected to use to achieve their missions economically and efficiently. IRS and FMS did not consider whether forgoing the speedier deposit of tax receipts and using IRS’s funds elsewhere, such as within other high-yield compliance activities, might result in financial benefits to the government that would be greater than those generated by accelerating the deposit of tax receipts. FMS and IRS also did not clearly define in the 1999 study the type of analysis that had been conducted. Clearly defining the type of analysis undertaken is important because different types of analyses take into account different types of costs and benefits. The specific types of costs and benefits considered can, in turn, affect a study’s conclusions. Since the completion of the 1999 study, changes at IRS and the lockbox network could affect processing and the related costs. Because of these changes, the 1999 study will not be useful for determining whether IRS and FMS should continue using lockbox banks to process tax receipts when the current lockbox agreements expire in 2007. We are making recommendations to improve both oversight of and internal controls at lockbox banks. We are also recommending that FMS and IRS thoroughly review the results of the ongoing TIGTA investigation of the 2001 incident at the Pittsburgh lockbox site and that they implement additional controls and revise the agreements as necessary to decrease the likelihood that such an incident will occur again. Because IRS and FMS will need to decide before 2007 whether to continue using lockbox banks to process tax receipts, we are recommending that a study be done in time to support this decision, that the type of study be clearly defined, and that all appropriate costs be considered. FMS and IRS agreed with our recommendations and have initiated or plan to initiate actions to implement them. Page 5 GAO-03-299 IRS Lockbox Banks Background In an effort to expedite receipt processing,4 the IRS conducted its first pilot project to obtain lockbox services from a commercial bank in 1984. The receipts processed were limited to tax receipts for estimated tax payments, which are typically paid by taxpayers on a quarterly basis. The bank was compensated from the interest it earned on a compensating balance— funds placed in the bank’s account by FMS. Since that time, the IRS lockbox program has expanded to cover taxpayers in all states and receipts for individual income tax returns, employment tax returns, and other miscellaneous types of taxes. Most of the returns are received during the April peak processing period and the smaller peak periods during January, June, and September. Certain taxpayers who owe money and are making payments are instructed to mail returns and payments to post office boxes maintained by the lockbox banks. The lockbox sites deposit the receipts to an account with Treasury and send processed documents (tax return forms and payment vouchers), computer tapes containing taxpayer data, and unprocessable receipts5 to IRS Submission Processing Centers for further processing and recording in the taxpayer accounts (see fig. 1). 4 Lockbox collection services are recognized by 31 U.S.C. 3720(a) as a procedure executive agencies may use to comply with requirements for the collection and timely deposit of funds. 5 Unprocessable receipts are tax receipts that lack adequate information, such as taxpayer identification numbers, to be processed by the lockbox bank and are forwarded to IRS Submission Processing Centers for further research. Page 6 GAO-03-299 IRS Lockbox Banks Figure 1: Flow of lockbox collections and dataFMS formalized the lockbox Taxpayer Post Office Box Lockbox Bank Federal Reserve Bank Money Money Money Data Data Data Data Data Financial Management Internal Revenue Update taxpayer Service Service accounts Source: GAO, based on Department of the Treasury, Financial Management Service, Cash Management Made Easy (Washington, D.C.: April 2002). Some images copyright 1998 Nova Development Corporation. FMS formalized the lockbox processing arrangements in 1993 by establishing contractual agreements with commercial banks to process tax receipts on behalf of IRS. In 2002, new, but similar, agreements were established. Like the 1993 agreements, the 2002 agreements are 5-year agreements with two 1-year extension options.6 The current lockbox network consists of four banks, three of which operate multiple sites that 6 The lockbox agreements entered into in 1993 took effect in calendar year 1994. The agreements had 5-year terms and allowed for two 1-year extensions. After both 1-year extensions were exercised, the agreements were extended by FMS for an additional 8 months. Page 7 GAO-03-299 IRS Lockbox Banks support the 10 IRS Submission Processing Centers across the country.7 The agreements with FMS require that these banks operate their sites according to IRS’s Lockbox Processing Guidelines (LPG). The LPGs provide the detailed procedures the banks are required to follow in providing lockbox services for IRS and are updated as needed. Both FMS and IRS monitor lockbox bank compliance with the agreements and the LPGs, and each lockbox site has an IRS employee who serves as a lockbox coordinator.8 For fiscal year 2002, IRS lockbox banks processed more than 66 million receipts, totaling about $268 billion, which accounted for approximately 13 percent of total tax receipts in dollars and 32 percent in volume. 7 The four banks operate at a total of nine locations. Because two of the locations have 2 sites each—1 for individual returns and 1 for business returns—there are a total of 11 lockbox sites. 8 Each lockbox site is assigned a lockbox coordinator from the IRS Submission Processing Center for which it processes tax payments. The coordinator is primarily responsible for ensuring that his or her designated lockbox promptly deposits tax payments. The coordinator is also responsible for performing compliance and quality reviews and for coordinating lockbox oversight and operating activities between lockbox management, FMS, and IRS National Office and the Submission Processing Center. Page 8 GAO-03-299 IRS Lockbox Banks Figure 2: Fiscal Year 2002 Dollar Value and Volume of Collections for Three Major Types of Tax Receipt Collection Mechanisms Dollars Volume $86 billion 32.6 million $268 billion 66.8 million $1.57 trillion 66.0 million IRS Submission Processing Center collection Lockbox collection Electronic Federal Tax Payment System Source: IRS data. Note: These pie charts exclude 46.1 million and almost $80 billion of receipts related to Federal Tax Deposits (Coupons) in fiscal year 2002. The intent of the lockbox program is to enhance federal cash management by accelerating the deposit of tax receipts, which would increase interest float savings (or interest cost avoidance) to the government and reduce the amount Treasury would have to borrow to pay government obligations. The estimate of interest float savings resulting from IRS’s use of lockbox banks has varied throughout the years. In calculating interest float savings, IRS and FMS assumed, based on a 1988 joint IRS/FMS study,9 that lockbox banks could process receipts and deposit the funds to a Treasury account 3 days faster than IRS. However, in a 1998 report, the Treasury Office of Inspector General (OIG) questioned the validity of this assumption and recommended that IRS acquire relevant and reliable comparative cost data 9 U.S. Department of the Treasury, Internal Revenue Service, Evaluation of Alternative Systems for Collecting Form 1040 Estimated Tax Payments (Washington, D.C.: May 1988). Page 9 GAO-03-299 IRS Lockbox Banks on all aspects of the lockbox program to identify the most cost-effective option to use for processing and depositing tax receipts.10 In response to the Treasury OIG report, IRS and FMS hired a contractor to compare lockbox bank processing to IRS processing for individual tax (Form 1040) receipts. The contractor reported in July 1998 that lockbox banks could make funds available to the federal government an average of 2 days faster than IRS. In 1999, IRS and FMS formed a taskforce to study the costs and benefits of continuing to use lockbox banks for processing Form 1040 receipts. Based on its study, the IRS/FMS taskforce recommended that such processing remain with the lockbox banks rather than be returned to IRS for fiscal years 2001-2007.11 Scope and To accomplish our work, we reviewed reports relevant to oversight and management of the lockbox program, including reports prepared by FMS, Methodology IRS, GAO, the Treasury OIG, TIGTA, and internal and external auditors from several lockbox banks. We reviewed laws, regulations, and guidance related to the cash management activities of the federal government. We also interviewed FMS and IRS officials from several headquarters divisions. As agreed with your office, we did not review the incident involving the loss and destruction of taxpayer receipts and data at the Pittsburgh lockbox, since an ongoing investigation was in process. The following procedures were also performed for each of the objectives: • To determine if new contractual agreements address previously identified problems and correct provisions that could contribute to improper handling of taxpayer returns, we reviewed and analyzed the provisions in the 1993 and 2002 lockbox bank contractual agreements. We also compared the agreements to determine whether changes had been made. • To determine the adequacy of FMS’s and IRS’s oversight of lockbox banks, we reviewed FMS and IRS policies, guidelines, checklists, 10 U.S. Department of the Treasury, Office of Inspector General, Review of the Effectiveness of Using Commercial Bank Lockboxes for Federal Income Tax Payments, OIG-98-097 (Washington, D.C.: Aug. 20, 1998). 11 IRS/FMS Joint Taskforce, 1040 Tax Payment Comparative Cost Benefit Study (Washington, D.C.: August 1999). Page 10 GAO-03-299 IRS Lockbox Banks schedules of site visits, and reports from oversight reviews. We performed site visits at two IRS Submission Processing Centers to observe IRS reviews of documentation received from the lockbox banks. At each Submission Processing Center, we interviewed relevant management and staff concerning lockbox bank oversight policies, procedures, and practices. • To determine if lockbox banks’ security and internal controls to safeguard taxpayer receipts and returns are sound and properly implemented, we observed physical security and internal controls and interviewed lockbox personnel at all nine lockbox locations during the April 2002 peak processing period and at two lockbox sites during the June 2002 peak processing period. At each site, we also reviewed lockbox bank employee personnel records for a nonrepresentative selection of permanent and temporary lockbox employees. In addition, we compared the 2001 and 2002 LPGs for changes related to safeguarding tax receipts and data, receipt processing, employee screening, and courier requirements. We also compared IRS’s Internal Revenue Manual (IRM)12 and other directives, which are IRS’s detailed policies, with the 2002 LPGs. • To determine whether the costs and benefits of processing tax receipts through lockbox banks instead of processing them at IRS were considered, we reviewed federal guidance and economic literature on cost-benefit and cost-effectiveness analyses of federal programs and policies. In addition, we analyzed prior FMS and IRS studies and the support for the data, assumptions, and methodology used in the 1999 report to estimate the costs and benefits of processing tax receipts through lockbox banks versus processing them at IRS. We performed our work from April 2002 through November 2002 in accordance with U.S. generally accepted government auditing standards. We requested written comments on a draft of this report from the Secretary of the Department of the Treasury or his designee. These comments are discussed in the Agency Comments and Our Evaluation section of this report, incorporated in the report as applicable, and reprinted in appendix III. 12 The IRM is IRS’s internal operating manual that sets forth the agency’s various operating policies and procedures. Page 11 GAO-03-299 IRS Lockbox Banks It Is Not Known We found nothing inherent in the new 2002 lockbox bank contractual agreements or the prior agreements that would necessarily contribute to Whether Contract mishandling of taxpayer receipts. The agreements contain penalty Provisions Could provisions that can result in negative consequences if banks do not perform work that meets quality standards or do not perform work within required Contribute to Improper time frames. The consequences range from financial penalties to Handling of Taxpayer termination of the lockbox agreement. Although a desire to avoid negative Receipts consequences could motivate lockbox bank employees to make poor decisions in their handling of taxpayer receipts, the penalty and termination provisions are necessary to help the federal government address inadequate contractor performance on the two performance dimensions—quality and timeliness—deemed critical by IRS and FMS. FMS and IRS made some enhancements to the 2002 agreements, such as the addition of a new performance penalty and clarification of other provisions. Because TIGTA’s investigation of the incident involving the loss and destruction of tax receipts by employees at the Pittsburgh lockbox site during the 2001 April peak processing period is still ongoing, it is unclear whether any provisions in the lockbox agreements may have contributed to the mishandling incident. When the results of the investigation are known, FMS and IRS should determine whether contract provisions need to be modified or whether additional controls need to be implemented. Lockbox Bank Agreements Factors used to assess contractor performance are cost, timeliness, and Contain Timeliness and quality of service provided. Contracts often contain specific standards for acceptable performance as well as provisions for rewarding or penalizing Quality Factors to Assess contractors according to their performance in these areas. Such provisions Performance may inadvertently encourage contractors to focus their efforts on one area to the detriment of their performance in other areas. For example, if a contract’s provisions reward timely performance more than they reward high-quality performance, contractors may be encouraged to take shortcuts that improve timeliness but detract from quality of work. As a result, it is important that contract incentives be appropriately designed and balanced to obtain acceptable levels of performance in all relevant areas. To ensure that contract provisions are operating as intended, effective oversight of contractors is essential. The lockbox banks are paid a fixed price for each item they process, and their total compensation depends on the number of each type of receipt they process. Lockbox banks have no direct influence over the volume of Page 12 GAO-03-299 IRS Lockbox Banks receipts. The compensation paid is therefore not a factor in measuring IRS lockbox performance, and the lockbox contractual agreements contain no direct incentives, positive or negative, related to cost. Lockbox bank performance is measured on timeliness and quality factors, specifically focusing on expediting the flow of funds to Treasury and ensuring that receipts are accurately processed. Because the basis for using lockbox banks was IRS’s and FMS’s belief that the banks could process tax receipts faster than IRS could, timeliness is a key measure of lockbox bank performance. Except during peak processing periods, the agreements require that lockbox banks deposit receipts within 24 hours of their receipt at the lockbox bank. During peak processing periods, all receipts must be processed and ready for deposit by the assigned program completion date (PCD). 13 According to the terms of the lockbox contractual agreements, FMS may assess banks penalties or terminate their contractual agreements if they fail to meet these deadlines. The agreements also provide that the amount of any penalty assessed for late deposit of tax receipts shall be based on the amount of interest Treasury lost because of the delay. FMS imposed financial penalties on two lockbox banks for not meeting the PCD during the April 1997 peak processing period. Meeting quality standards is another critical aspect of IRS lockbox banks’ performance. Processing errors can place unnecessary burdens on taxpayers and delay processing of tax receipts. For example, a processing error might cause a lockbox bank to withdraw more funds from a taxpayer’s account than the amount actually written on the check. If the taxpayer’s account contained inadequate funds to cover the incorrect withdrawal amount, the taxpayer’s bank could assess penalties for insufficient funds. If the error caused the bank to withdraw less than the amount owed, IRS might erroneously assess the taxpayer interest and penalties for an incomplete payment. The lockbox agreement provisions allow FMS to assess lockbox banks financial penalties or to terminate an agreement for poor performance. FMS added a new provision to the 2002 lockbox agreements that is designed to facilitate reimbursement to IRS and FMS for costs they incur due to specific failures in performance, such as 13 The program completion date is the date by which the banks must finish processing all tax receipts associated with the April peak processing period and when they return to the nonpeak workload requirement to process each receipt within 24 hours of receiving it. Page 13 GAO-03-299 IRS Lockbox Banks costs resulting from lockbox banks’ errors in processing tax receipts. In addition, the 2002 agreements clarified certain existing penalty provisions. We found nothing inherent in the 1993 or 2002 lockbox contractual agreements that would necessarily contribute to mishandling of taxpayer receipts. Although a desire to avoid negative consequences, such as financial penalties or contract termination, could cause lockbox bank employees to make poor decisions, penalty and termination provisions are necessary to help the federal government address inadequate contractor performance. To help ensure that contractors are adhering to contract terms and to reduce the risk that lockbox banks might compromise taxpayer data, effective oversight of lockbox sites is essential. Cause of the 2001 Incident The exact cause of the 2001 incident involving the loss and destruction of Remains Under taxpayer data and receipts at the Pittsburgh lockbox site has yet to be officially reported. The site had a history of performance problems for Investigation which the bank had been assessed financial and other penalties. In 1997, FMS assessed the bank that operated the Pittsburgh site more than $1.4 million in penalties for failing to meet the assigned PCD and therefore delaying availability of funds to the Treasury. In September 2000, FMS placed the site on probation because of numerous uncorrected security violations, including commingling of corporate and other government agency processing with IRS processing. FMS, with IRS’s concurrence, removed the site from probationary status 2 ½ months later, after a site review conducted during the probationary period indicated that bank management had corrected all but one of the security violations.14 A subsequent, unannounced review by FMS and IRS 3 ½ months after the site was taken off probation also found that past violations of security requirements had not recurred and that the site was, for the most part, in compliance with security requirements. Nevertheless, approximately 2 months after this review, the Pittsburgh site was found to have lost or destroyed tens of thousands of tax returns, and, as a result, FMS terminated its contractual agreement for the site. As of September 30, 2002, IRS had spent over $4 million to obtain duplicate receipts and returns from 14 One violation remained, which was to be corrected when the bank moved into a new facility in December 2000. Page 14 GAO-03-299 IRS Lockbox Banks the affected taxpayers, and the federal government had lost an estimated $13.5 million in interest as a result of the incident. In October 2002, Mellon Bank agreed to pay the government $18.1 million to cover administrative costs and expenses associated with the incident. However, TIGTA is still investigating the incident to determine whether criminal charges should be filed against any of the bank employees. As of November 2002, TIGTA and the applicable U.S. Attorney’s Office were unable to discuss this investigation with us. Until the investigation is completed, we cannot determine whether the site’s interpretation of contract provisions was a contributing factor to the 2001 incident or whether provisions need to be added or revised to help prevent a similar incident from occurring. Oversight of Lockbox FMS and IRS oversight of lockbox bank operations is a key control for ensuring that funds collected through the lockbox banks are protected Banks Was Not Fully against fraud and mismanagement. The oversight functions performed by Effective FMS and IRS include various on-site and off-site reviews to ensure compliance with LPGs and contract terms, evaluation of requests for waivers from LPG requirements and proposed compensating controls, and enforcement of penalties against banks that fail to meet LPG and contract terms. In calendar year 2002, the agencies made significant improvements to their oversight of lockbox operations, mostly in response to the Pittsburgh lockbox incident. However, we found that the oversight of lockbox banks was not fully effective in protecting the government’s interests due to (1) a lack of clear directives and documented policies and procedures for various oversight functions, (2) key oversight functions not being performed, and (3) conflicting roles and responsibilities for IRS lockbox coordinators. These issues reduce the overall effectiveness of IRS and FMS oversight of lockbox banks. Additionally, the lack of clearly defined oversight requirements increases the risk that the oversight improvements made during 2002 may not continue in the future. According to IRS officials, they are in varying stages of completing several memoranda of understanding to identify and document oversight roles and responsibilities. Until these roles and responsibilities are agreed to and documented, however, oversight weaknesses are likely to continue to exist. Page 15 GAO-03-299 IRS Lockbox Banks FMS and IRS Improved FMS and IRS made significant improvements to the oversight of lockbox Oversight in 2002 but Have banks in 2002 compared with prior years. These improvements include (1) enhanced monitoring of peak processing operations, (2) involvement of Not Incorporated key IRS officials in security reviews, (3) establishment of a new office with Improvements in Agency a full-time FMS official responsible for oversight of lockbox security, and Policies and Procedures (4) establishment of a new performance penalty provision to reimburse the government for poor quality performance. However, many of the improvements have not yet been institutionalized in the form of agency policies and procedures. As such, there is less assurance that this increased oversight will continue in the future. Monitoring of Peak Period Prior to 2002, FMS’s on-site presence during peak processing periods was Operations Was Enhanced limited to only a few days each year and occurred near the end of the peak periods to ensure that production goals were achieved. IRS’s on-site presence at lockbox banks was generally limited to a lockbox coordinator, who was present for the duration of the peak period. However, these coordinators face competing demands of ensuring that lockbox sites promptly deposit tax payments, performing quality and compliance reviews, and assisting with other processing issues, such as training lockbox staff. In reaction to the Pittsburgh incident, FMS and IRS concluded that they needed more on-site presence at lockbox banks during peak operations. In April 2002, each lockbox site had at least one FMS official, one IRS headquarters official, and the site’s designated lockbox coordinator present for the entire April peak processing period. This increased on-site presence provided IRS and FMS with more comprehensive coverage of April peak operations. During 2002, FMS and IRS also placed a heavier emphasis on monitoring lockbox sites’ daily production status. According to IRS and FMS officials, their focus historically has been on monitoring a lockbox site’s ability to meet the overall April peak processing period’s PCD. IRS and FMS found that this approach presented problems because lockbox banks tended to address production problems during the peak processing period only when meeting the PCD was questionable. The solution employed by the banks was to bring in additional temporary staff near the end of the peak period to be able to meet the PCD. FMS and IRS officials indicated that their limited on-site presence affected the agencies’ ability to detect production problems on a real-time basis. Only after the peak period did the lockbox banks bring production issues to the attention of IRS and FMS. Page 16 GAO-03-299 IRS Lockbox Banks In 2002, FMS and IRS officials focused on monitoring lockbox banks’ daily production by reviewing production reports, observing production activity on the processing floor, and reporting production issues to IRS and FMS headquarters as soon as issues arose. During April 2002 and subsequent smaller peak periods in June and September 2002, each lockbox site also submitted, on a daily basis, an “FMS Daily Status Report.” These reports noted the daily status of critical production issues, such as staffing shortages and equipment problems, that could cause delays in the timely deposit of tax receipts or could affect performance. FMS and IRS headquarters officials reviewed these reports for potential problems and contacted lockbox management and on-site agency officials for follow-up and to facilitate timely resolutions of the problems. While these changes enhanced monitoring of peak operating periods, these improvements have not yet been incorporated into agency policies. Currently, only the lockbox coordinator is required by IRS policies to be on-site during the peak processing period. Personnel from FMS and IRS headquarters are not mandated by agency policies to be at lockbox sites during this time. Additionally, IRS officials indicated that they might not have adequate headquarters staff to assist with future on-site reviews. As a result, there is less assurance that increased on-site presence of staff at lockbox sites during peak processing periods will continue. IRS Personnel with Physical After the 2001 Pittsburgh incident, IRS concluded that its participation in Security Expertise Now Perform joint IRS/FMS annual unannounced security reviews should be performed Security Reviews by IRS staff with security expertise. Prior to 2002, personnel from IRS’s Wage and Investment Division, who have primary responsibility to manage the lockbox program, performed physical security reviews but did not have physical security expertise. During 2002, IRS’s Agency-Wide Shared Services (AWSS), which has staff with physical security expertise and performs physical security reviews at IRS Submission Processing Centers, began participating in the joint IRS/FMS unannounced security reviews. However, IRS policies only require the Wage and Investment Division to perform unannounced security reviews. AWSS’s participation in lockbox reviews is based only on an oral agreement to perform such reviews. As such, there is less assurance that AWSS will continue to perform lockbox security reviews in the future. FMS Established the Bank Prior to 2002, FMS’s Financial Services Division (FSD), which had the Review Office administrative responsibility for negotiating and entering into lockbox contracts with financial institutions, also had the responsibility to oversee lockbox banks’ compliance with contract terms, such as security Page 17 GAO-03-299 IRS Lockbox Banks requirements. To effectively perform its oversight responsibilities, FMS recognized a need to establish a full-time position responsible for oversight of lockbox security. In August 2002, FMS formally established the Bank Review Office. The director and staff of this office are now responsible for FMS’s oversight of the security of federal government lockbox banks. Their responsibilities now include performing on-site reviews, following up on corrective actions to address review findings, and reviewing the adequacy of security requirements for lockbox banks. The director of this office now serves as the main FMS contact point on most oversight issues for IRS and lockbox banks. New Performance Penalty A new provision was added to the 2002 contractual agreements to assist the Provision Established government in obtaining reimbursement from lockbox banks for direct costs incurred by IRS to correct performance failures on the part of lockbox banks. This provision enhances the government’s ability to penalize lockbox banks for poor quality work and be reimbursed for the additional costs IRS incurs to rework transactions erroneously processed by lockbox banks. Effective use of this provision requires additional guidelines and procedures to help management decide whether certain situations merit pursuing the penalty provision. Such guidelines and procedures should include (1) IRS’s expectations for unacceptable error rates, (2) procedures to identify and document lockbox errors, and (3) procedures to track IRS costs incurred as a result of rework. During 2002, IRS had not established guidelines and procedures for effective use of the new performance penalty provision. IRS cited three reasons why guidelines and procedures had not yet been established. First, IRS has not yet determined thresholds for unacceptable error rates. Second, IRS officials indicated that the agency had spent significant resources to document and build a legally defensible case to obtain reimbursement of costs related to the 2001 incident. Based on lessons learned from this case, IRS concluded that it needed to establish a more cost-effective means to accurately identify and document lockbox errors, as opposed to errors caused by taxpayers or IRS. During 2001, IRS tested a process to identify and document lockbox errors but concluded that the process was too labor-intensive and might not provide accurate and legally defensible data. Therefore, IRS is still exploring other methods to obtain these data. Finally, IRS officials explained that by law, reimbursements from lockbox banks would have to be remitted to the U.S. Treasury general fund and that they had explored legal options to keep the reimbursed funds for IRS’s own use. Page 18 GAO-03-299 IRS Lockbox Banks The reimbursement provision is a critical tool available to the government as a means to recover from lockbox banks costs incurred as a result of poor quality work and as an enforcement tool to encourage banks to implement effective controls and procedures to accurately process receipts. However, until the necessary guidelines and procedures are established, IRS and FMS cannot effectively use this oversight tool. As a result, the government may be paying for poor quality work and incurring additional costs to correct errors. FMS and IRS Did Not Despite significant improvements, we found several instances where key Perform Certain Key oversight functions were not performed, which resulted in an increased risk of loss to the government and taxpayers. Specifically, (1) IRS and FMS Oversight Functions did not take timely action on lockbox sites’ requests for waivers from LPG requirements, (2) IRS did not always participate in unannounced security reviews, (3) FMS did not always obtain formal responses from lockbox bank management to unannounced security reviews, and (4) IRS lockbox coordinators did not always complete reviews for peak processing periods. Additionally, the guidance used for these reviews needs to be strengthened. IRS and FMS Did Not Take The LPGs set forth security and processing requirements for lockbox sites. Timely Action on Formal They also allow deviations from these requirements if bank management Requests for Waivers submits a written waiver request to IRS and FMS. The bank must demonstrate its site’s legitimate inability to meet a requirement and must implement an alternate procedure or compensating control. In practice, after a bank submits a written waiver request stating the reason for its site’s inability to meet the LPG requirement and explaining its compensating control to mitigate risks of loss to the government, IRS and FMS allow the site to operate with its compensating control while they review the waiver request. However, because some waiver requests will eventually be denied after FMS and IRS conclude that the compensating controls are not adequate, some sites with inadequate controls are, in effect, allowed to operate in noncompliance with the LPGs until their waiver is officially denied. Therefore, to protect the government from losses resulting from a site’s noncompliance, FMS and IRS have a fiduciary duty to approve or disapprove waiver requests and effectively and promptly communicate final decisions. Prior to April 2002, IRS and FMS approved and disapproved waiver requests orally. To better coordinate their efforts, the agencies decided to develop a written joint process to assess waiver requests in April 2002. However, this process was never formalized in agency policies. IRS Page 19 GAO-03-299 IRS Lockbox Banks provided us a draft document of the joint waiver assessment process. According to the draft document, lockbox banks would submit a waiver request form to FMS. FMS would assess the waiver request and forward the request with FMS’s recommendations to an IRS waiver coordinator. The IRS waiver coordinator would disseminate the waiver requests to various units within IRS responsible for assessing the waiver requests. Once appropriate IRS officials had made their decisions and signed off on the waiver forms, the IRS waiver coordinator would forward the waiver form with IRS’s decision to FMS. FMS would notify the bank of the joint decision by returning the waiver request form with both agencies’ responses. The draft document indicates that the whole process should take about 6 days from FMS’s receipt of the waiver request form. According to our review of waiver requests made from April 16 through April 22, 2002, FMS forwarded to IRS its recommendations on most of the requests within 5 days of receiving them from lockbox banks. However, IRS took 5 months to officially sign half of the waivers. According to IRS officials, this delay resulted from a misunderstanding on the part of some IRS officials about whether IRS had to complete the waiver forms if IRS had already orally informed FMS of its decision. FMS postponed notifying the banks of a decision on their requests because it was waiting to receive a formal written decision from IRS for each request. However, 4 months after FMS received the requests for waivers and with no official decision from IRS, FMS decided to inform the lockbox sites of its unilateral disapproval of the banks’ waiver requests to mitigate the risk of loss to the government from lockbox sites not operating in compliance with the LPGs. Although most of the waiver requests were not related to critical requirements, one lockbox waiver pertained to a critical LPG security requirement. Under this requirement, temporary employees must provide photo identifications to the guards in exchange for badges allowing them access to the processing area. This control procedure was designed to validate the identity of individuals claiming to be employees before they enter the processing area. Bank management at one lockbox bank believed that its automated entry system provided adequate compensating controls against unauthorized access to tax data and receipts and that the manual verification of each employee’s identity was unnecessary. The bank therefore did not perform the procedure. FMS and IRS eventually denied the request, stating that the lockbox bank’s compensating controls did not provide sufficient protection against unauthorized entry. However, because of the breakdown in the joint waiver assessment process and the resulting delay in notifying the bank of the agencies’ decision, tax receipts and data Page 20 GAO-03-299 IRS Lockbox Banks were unnecessarily exposed to an increased risk of theft. Bank management had informally requested a waiver for the same issue in August 2001, before FMS and IRS established their formal waiver process in April 2002. IRS officials indicated that they had already orally informed the bank that its initial request for a waiver was denied and had also informed FMS during a meeting that the subsequent request submitted on the official waiver request form in April 2002 was denied. Therefore, IRS officials concluded that it was not necessary to review the April 2002 formal waiver request. FMS postponed notifying the banks of the decision because it was waiting for IRS’s decision in writing. After FMS decided to inform banks of its unilateral decisions on their formal waiver requests, it notified this site of its decision to deny the waiver request in mid-August 2002. As a result, the site operated in noncompliance with this security requirement for over 7 months, from January through August 2002. This period included the April 2002 peak period, when the bank operated in three shifts with as many as 300 employees per shift. IRS and FMS officials indicated that they are developing several memoranda of understanding between the two agencies to better coordinate oversight efforts, including the joint waiver assessment process. IRS Did Not Always Participate In response to a Treasury OIG recommendation,15 IRS and FMS began in Unannounced Security performing unannounced security reviews of lockbox bank sites. IRS’s IRM Reviews requires its Wage and Investment Division to conduct joint unannounced security reviews with FMS. However, IRS did not participate in the first three unannounced security reviews in 2002. The IRS official who would have participated in these security reviews had an extensive technical background in physical security, which would have been helpful in detecting physical security deficiencies. According to IRS officials, IRS did not participate in the reviews because the units responsible for performing various security reviews for both agencies were reorganizing during early 2002 and the agencies had failed to effectively communicate who would be the responsible parties to perform the unannounced security reviews. IRS officials indicated that they plan to include coordination of security reviews between the two agencies in a memorandum of understanding currently being developed. 15 OIG 98-097. Page 21 GAO-03-299 IRS Lockbox Banks FMS Did Not Always Obtain Effective oversight of lockbox banks requires appropriate follow-up on Responses from Lockbox Bank corrective actions taken to address deficiencies found during on-site Management to Unannounced reviews. As part of its follow-up procedures, FMS requires lockbox bank Reviews management to provide an official management response to reports on its unannounced reviews. In their response, bank managers are to indicate whether corrective actions have been taken or propose corrective actions and the dates by which the bank intends to implement them. Personnel from FMS’s FSD are responsible for obtaining management responses. However, FSD failed to obtain formal management responses for three unannounced security review reports issued in 2001 because FSD staff were preoccupied with the Pittsburgh lockbox incident and with implementing the 2002 lockbox network. Additionally, the tracking procedure in place to remind staff to obtain management responses was ineffective. As a result, FMS did not have sufficient information to assess the adequacy or timeliness of proposed corrective actions or corrective actions already taken. FMS and IRS found the same security problems identified by these reviews several months later. Although two of the three locations had closed after 2001, bank managers from one of the closed locations were transferred to manage a new lockbox location in the 2002 lockbox network. In a subsequent visit to this new location, FMS found the same security violation that it had found in 2001 at the closed site regarding the lack of a seeding program.16 Additionally, a subsequent visit by FMS and IRS to the location that continued to operate in 2002 revealed problems with malfunctioning perimeter door alarms that were similar to problems identified during the previous unannounced security review. The lack of effective controls to ensure that bank managers take corrective actions increases the risk that identified security weaknesses will not be corrected. According to FMS officials, the recently created Bank Review Office, whose staff have full-time responsibility for security oversight, is now responsible for following up on management responses to security reviews. Additionally, FMS officials indicated that they plan to create an automated tracking system to better track the status of management responses. 16 A seeding program is a control procedure used by bank management to discourage and detect employee theft. Bank managers or TIGTA place cash or checks with blank payee names among IRS tax payments without the employees’ knowledge and closely monitor employees to determine whether they steal the cash or checks or promptly report their discovery to bank managers. Page 22 GAO-03-299 IRS Lockbox Banks Lockbox Coordinators’ Reviews The IRM specifically directs lockbox coordinators to perform on-site Were Not Adequately Performed reviews of lockbox banks during peak processing periods. The scope of these on-site reviews includes assessments of the lockbox site’s compliance with critical processing and security requirements. Lockbox coordinators record the results of their reviews on forms specifically designed to show results of observations and sampling of transactions to determine the lockbox site’s compliance with specific security and processing requirements. However, for the January and April 2002 peak processing periods, lockbox coordinators did not adequately perform these reviews. Specifically, coordinators did not perform reviews for four lockbox sites for the January 2002 peak processing period. In addition, some of the information to be recorded on the forms used by the coordinators to document their reviews was not provided. For the April 2002 peak period, all the coordinators submitted the results of their reviews, but some of the reviews and the forms used to document them were only partially completed. IRS officials explained that the reviews could not be completed because several lockbox sites were new to the lockbox network during 2002 and therefore required full-time support from their lockbox coordinators. Because the reviews were not always completed, IRS may not have detected instances of noncompliance and therefore would not have been able to take immediate corrective actions. The guidelines for the reviews conducted by lockbox coordinators also need to be strengthened. To provide adequate oversight of lockboxes, lockbox coordinator reviews should include steps to assess critical controls and procedures for lockbox security and processing. In general, the lockbox coordinators’ reviews provide coverage for most of these areas. However, we found other critical controls and procedures not sufficiently covered by lockbox coordinator reviews because the guidelines and review procedures were either unclear or did not require these controls and procedures to be subject to review. These deficiencies could lead to IRS’s failure to detect significant instances of noncompliance. For example, while TIGTA found that couriers with criminal records at one lockbox location were given access to taxpayer data before their FBI fingerprint checks were completed, the lockbox coordinator’s review found the same location to be in compliance with courier requirements. We found that there is no specific step in the lockbox coordinator review procedures to determine whether all couriers given access to taxpayer data have completed favorable FBI fingerprint checks, as required by the LPGs. Page 23 GAO-03-299 IRS Lockbox Banks The review procedures only broadly ask whether the courier service is in compliance with the LPGs. Additionally, while the review procedures require coordinators to sample employee files to determine lockbox compliance with FBI fingerprinting requirements, there is no similar review step to determine lockbox compliance with FBI fingerprint checks for couriers and contractors. Lockbox coordinators we interviewed stated that during their on-site reviews, they did not verify whether couriers had completed favorable FBI fingerprint checks before they were given access to taxpayer data. As discussed later in this report, we also found problems with some lockbox sites’ controls to account for cash payments received from taxpayers and items found during candling.17 For example, the internal logs used by some lockbox sites to record cash received and items found during candling did not reconcile with the logs they submitted to IRS, raising questions about possible missing payments. Lockbox coordinators did not detect this problem because they are not required to and do not review internal lockbox logs for cash payments and candled items. Lockbox Coordinators’ IRS lockbox coordinators are responsible for the day-to-day administration Conflicting Roles and of the lockbox program, including providing guidance and training to lockbox staff and ensuring that the lockbox banks promptly address Responsibilities Could production problems that may delay the deposit of funds to the Treasury. Impair Their Objectivity Lockbox coordinators are also responsible for performing on-site reviews to assess work quality and the bank’s compliance with security requirements. This creates a situation in which lockbox coordinators have multiple and conflicting responsibilities. When faced with competing demands, the likelihood increases that the day- to-day pressures of administering the lockbox program will take precedence over oversight responsibilities. For example, according to IRS officials, the lockbox coordinators’ failure to complete their peak period reviews resulted from competing demands. Lockbox coordinators are responsible for performing quality reviews and for assisting lockbox banks with processing issues. IRS Wage and Investment officials indicated that some of the lockbox coordinators could not complete their reviews 17 Candling is a process that uses a light source to determine if any contents remain in envelopes before their destruction. Items found during candling must be recorded on a log. Page 24 GAO-03-299 IRS Lockbox Banks because more pressing matters concerning lockbox operations, particularly at new lockbox sites, required most of their attention. The close working relationship with bank management that lockbox coordinators develop while helping bank management meet processing goals could also impair coordinators’ objectivity and independence when they are evaluating the bank’s compliance with LPG requirements. The Treasury OIG raised this issue in 1998 when it recommended that coordinators periodically rotate out of their coordinator positions to help maintain independence. However, IRS has not yet addressed this issue, citing its need to retain experienced lockbox coordinators to assist with the implementation of the 2002 lockbox network. IRS officials indicated that there are plans to reorganize positions of lockbox coordinators to address the independence issue. Until this issue is addressed, IRS has no independent overseers with sole responsibility for monitoring lockbox bank compliance during peak processing periods. Lockbox Banks’ The Comptroller General’s Standards for Internal Control in the Federal Government18 require that access to resources and records, such as IRS Internal Control receipts and taxpayer data, be limited to authorized individuals to reduce Deficiencies Expose the risk of unauthorized use or loss to the government. Lockbox banks, working as financial agents of the federal government, have a responsibility the Federal to protect taxpayer receipts and data entrusted to them. Tax receipts, such Government to Theft as cash and checks, are highly susceptible to theft, and unauthorized use of and Loss taxpayer data could result in identity theft and financial fraud.19 For example, from October 1, 2000, to April 30, 2002, TIGTA initiated investigations of theft of receipts valued at almost $2 million from the IRS lockbox network.20 It is therefore critical that lockbox banks implement a strong system of internal controls for the lockbox sites they operate. Prior audits by GAO and TIGTA have noted internal control weaknesses at lockbox sites. For example, in fiscal year 2000, we found that background 18 U.S. General Accounting Office, Standards for Internal Control in the Federal Government, GAO/AIMD-00-21.3.1 (Washington, D.C.: November 1999). 19 Taxpayer data on tax forms could include taxpayer name, social security number, and address. 20 This is in addition to the approximately $1.2 billion in receipts that were lost or destroyed at the Pittsburgh site. Page 25 GAO-03-299 IRS Lockbox Banks screening for a temporary lockbox employee’s criminal history was limited to a local police records check and that some employees were given access to taxpayer data and receipts before completion of their background screening. We also found that lockbox couriers were subject to less stringent standards than IRS couriers. For example, lockbox couriers were not required to travel in pairs when transporting taxpayer data. We reported that these control weaknesses could lead to thefts of taxpayer receipts and data at lockbox banks.21 IRS and lockbox banks have implemented several additional safeguarding controls in response to audit findings and the 2001 incident at the Pittsburgh site. For example, IRS now requires lockbox employees to have obtained favorable results on their FBI fingerprint checks, which are nationwide checks of criminal records, before they receive access to tax data. IRS has also enhanced lockbox courier security guidelines. Nevertheless, during our recent visits to all nine lockbox locations, we found internal control deficiencies in the areas of (1) physical security, (2) processing controls, (3) courier security, and (4) employment screening. These control weaknesses, if uncorrected, could lead to significant losses to the government and taxpayers because they increase the risk of loss, theft, or mishandling of taxpayer receipts and data. Table 1 demonstrates the pervasiveness of the internal control issues found during calendar year 2002. These issues are discussed briefly below and are discussed in more detail in appendix I. Table 1: Internal Control Issues Found at Lockbox Locations in Calendar Year 2002 Lockbox bank location 1 2 3 4 5 6 7 8 9 Physical security issues X X X X X X X X X Processing control issues X X X X X X X X Courier security issues X X X Employee screening issues X X X X X Source: Findings based on reviews at the nine lockbox locations. 21 GAO-02-35. Page 26 GAO-03-299 IRS Lockbox Banks Physical Security Given the large volume and assembly-line nature of tax receipt processing, Weaknesses Could Allow lockbox operations generally occur in areas with open floor plans, where taxpayer data and receipts are easily accessible to individuals on the Unauthorized Access to processing floor. Thus, the vulnerability of sensitive tax data and receipts Taxpayer Data and Receipts to theft or misuse is heightened. This vulnerability underscores the need for effective controls to deter and detect unauthorized access to taxpayer data and receipts. However, during our visit to lockbox locations, we observed numerous physical security breaches. Of the nine lockbox locations we visited, we found the following: • At four locations, perimeter doors leading into processing areas were unlocked or door alarms did not function effectively. • At two locations, guards were not responsive to alarms. • At one location, employee identification was not adequately verified. • At two locations, employment status of temporary employees was not adequately verified. • At two locations, visitor access to and activity in processing areas were not adequately controlled. • At six locations, guards did not closely monitor items brought into or out of the processing areas or prevent unauthorized items, such as personal belongings, in the processing area. • At seven locations, camera surveillance needed to be improved. These security weaknesses increase the risk of theft and mishandling of taxpayer data and receipts and reduce the possibility of the timely detection of such incidents. Processing Controls Need In addition to physical security controls, lockbox banks are required to Improvement to Better implement processing controls to maintain accountability for and security over transactions as they are processed in the normal course of operations. Account for and Protect During our visits, we found deficiencies in processing controls designed to Taxpayer Data and Receipts account for or protect tax data and receipts at most of the lockbox locations. Specifically, of the nine locations we visited, we found the following: Page 27 GAO-03-299 IRS Lockbox Banks • At four locations, candling procedures were not adequate to minimize the risk of accidental destruction of tax data and receipts. • At four locations, lockbox bank management did not reconcile candling logs to properly account for all items found during candling. • At six locations, lockbox bank managers did not perform required reviews of the candling process or did not adequately document results of their reviews. • At three locations, controls were not in place to adequately safeguard and account for cash receipts. • At seven locations, returned refund checks were not adequately protected against theft. • At one location, timely payments were incorrectly processed as delinquent. Inadequate accounting and safeguarding of tax payments and related vouchers or returns could lead to taxpayer burden, such as penalties and interest for failure to pay tax liabilities, if these items are accidentally destroyed, stolen, or incorrectly processed. Additionally, inadequate processing controls could result in errors not being detected promptly or additional work for IRS employees who must correct taxpayer accounts as a result of errors. Courier Security Lockbox banks employ courier services to deliver (1) mail from post Weaknesses Expose Tax offices to lockbox processing sites, (2) processed checks to depository banks, and (3) tax data and unprocessable receipts to IRS Submission Data and Receipts to Theft Processing Centers. During peak processing periods, couriers are entrusted and Loss with transporting thousands of pieces of mail each day that contain tax data, cash, checks, and deposits worth millions of dollars. Given the susceptibility of these items to theft and loss, it is important that they be carefully safeguarded while in transit to and from lockbox locations. However, our review and reviews conducted by TIGTA and IRS found several problems with courier security during calendar year 2002. These problems relate to failure to comply with courier security requirements as well as deficiencies in the current requirements. Specifically, at the nine locations, we and other reviewers found the following: Page 28 GAO-03-299 IRS Lockbox Banks • At one location, couriers with criminal records were given access to tax data before bank management received results of their FBI fingerprint checks. • At two locations, couriers were not properly identified. • At two locations, courier vehicles containing tax data and receipts were not locked. • At all locations, background screening requirements for lockbox couriers were less stringent than screening requirements for IRS couriers. • At all locations, courier requirements did not prohibit unauthorized individuals in courier vehicles or require lockbox staff to inspect courier vehicles for unauthorized passengers. Until these courier security weaknesses are addressed, tax data and receipts in transit to and from lockbox locations are exposed to higher risks of theft and loss. Background Screening of Because lockbox employees are entrusted with handling sensitive taxpayer Lockbox Employees Needs information and billions of dollars in receipts annually, ensuring worker integrity through a carefully managed recruiting and hiring process is an Further Improvement area that demands special attention from IRS, FMS, and lockbox management. However, during our review and those performed by TIGTA and IRS, we found that lockbox banks did not always comply with the FBI screening requirements and that further refinements to background investigation requirements for lockbox employees are needed. Specifically, for the nine locations, we and other reviewers found the following: • At five locations, employees were given access to taxpayer data and receipts before bank management received results of their FBI fingerprint checks. • At all locations, requirements for background investigations of lockbox permanent employees were unclear and resulted in variations in the scope and documentation of background investigations conducted on them. Page 29 GAO-03-299 IRS Lockbox Banks • At all locations, FBI fingerprint checks for lockbox employees who are not U.S. citizens with lawful permanent residence status may not be adequate to disclose criminal histories for individuals who have only recently established residence in the United States. These weaknesses increase the risk of theft of tax data and receipts by individuals who may be unsuitable to work at IRS lockbox sites. A Comprehensive IRS and FMS have not performed a comprehensive study evaluating the full range of costs and benefits of IRS processing tax receipts versus the Study Evaluating Costs lockbox banks processing the receipts for all types of tax receipts for IRS Processing processed by the banks. Over the years, several studies have been performed evaluating the degree of interest float savings resulting from the Versus Using Lockbox use of lockbox banks. IRS and FMS jointly performed the most recent Banks for All Types of study in 1999 in response to a Treasury OIG recommendation. In the 1999 Tax Receipts Has Not joint study, IRS and FMS considered the costs and benefits to the federal government of using lockbox banks rather than IRS to process Form 1040 Been Performed tax receipts more quickly. Having adopted this perspective in accordance with Treasury regulations, IRS and FMS did not consider some costs, such as opportunity costs, or the results from alternative uses of the money spent to achieve speedier deposits. Foregoing speedier deposit of tax receipts and using the funds elsewhere could, according to recent IRS data, result in financial benefits to the government that could be greater, depending on the study assumptions used. In effect, IRS and FMS did not consider the implications of such alternatives available to IRS. In addition, the study did not clearly define the type of analysis done. Differing types of analysis would require consideration of differing costs and benefits and could result in different decisions than were made on the basis of the 1999 study. Due to changes in IRS, the lockbox bank network, and other factors, the 1999 study will not be useful to support IRS and FMS officials’ future decision about whether to continue the lockbox arrangement when the current agreements expire in 2007. IRS and FMS officials initially stated that they had not planned to conduct a new study before the lockbox agreements expire in 2007, but that a new study might be appropriate given many changes and the passage of time since 1999. Page 30 GAO-03-299 IRS Lockbox Banks The 1999 Study Results Did In keeping with instructions in the Treasury Financial Manual (TFM), the Not Consider All Costs and purpose of the IRS/FMS study was to determine whether using lockbox banks or IRS to process most individual income tax receipts would Benefits minimize the total cost to the federal government. The TFM defines total cost to include agency costs, the cost of purchased services, and any loss of interest earnings to the government due to delays in depositing receipts.22 The study described three scenarios (described in app. II) in comparing estimated lockbox bank and IRS processing costs for fiscal years 2001 through 2007 based on different assumptions. The study used scenario I, the most conservative scenario, to support the decision to continue using lockbox banks. This scenario showed that IRS could spend about $56 million to internally process tax receipts, or could join with FMS to spend a total of $144.9 million to use lockbox banks to process tax receipts more quickly.23 The study concluded that the lockbox processing would be more costly but would save the government $100.5 million in interest that otherwise would be lost due to slower deposits if IRS processed the receipts. These interest savings more than offset the additional processing costs, producing net savings over 7 years of $11.6 million to the federal government from using lockbox banks rather than IRS to process tax receipts. Scenario I is used throughout this report in the cost comparisons shown, unless otherwise noted. Table 2 shows these cost comparisons. 22 Volume 1, Part 6, Section 8025.30 Collection Mechanisms. 23 IRS would pay for $60.3 million of this cost out of its budget; FMS would use Treasury Time Balances (TTB) to compensate the banks for the remaining $84.5 million. TTBs are deposited in lockbox banks and the interest that banks earn on those deposits is their compensation. We did not analyze these funding sources. The estimated federal funds rate used for this study was 5 percent. Page 31 GAO-03-299 IRS Lockbox Banks Table 2: Comparison of the Cost to the Federal Government of Alternative Receipt Processing Approaches, Fiscal Years 2001-2007 Dollars in millions Costs for Lockbox costs lockbox Costs for IRS minus IRS alternative alternative costsa IRS expenditures $60.3 $56.0 $4.4 FMS compensation to banks 84.5 0 84.5 Interest lost by government due to slower deposits 0.0 100.5 (100.5) Costs (savings) to federal government $144.9 $156.5 ($11.6) Source: GAO analysis of 1999 IRS/FMS study. a Differences between IRS and lockbox processing costs may be affected by rounding. However, IRS and FMS did not consider various costs in these estimates. Most notably, they did not consider opportunity costs related to alternate uses of IRS funds to accelerate tax deposits. Each year, the Congress approves a budget for an agency and provides discretion, within certain constraints, on agency spending. Given resource constraints, IRS must effectively choose how to spend its fixed budget. IRS and FMS decided that it would be worth spending $4.4 million more out of IRS’s budget to use lockbox banks to process tax receipts compared to IRS’s slower process because net savings to the government would reach $11.6 million over the 7 years. IRS and FMS did not consider whether greater financial benefits could accrue to the government if IRS processed receipts and used the estimated extra $4.4 million from IRS’s budget to generate higher revenues through other programs or activities.24 Recent estimates by IRS’s Research Division have pointed to other activities—such as tax enforcement—in which spending the extra $4.4 million would have generated from at least $44 million to well over $100 million. For example, in some enforcement 24 The $84.5 million of funding provided by FMS through TTBs could not be reallocated to other IRS activities because Treasury permits TTBs to be used solely to pay for depository services that expedite deposits to the Treasury. Page 32 GAO-03-299 IRS Lockbox Banks programs for just individual taxpayers, the ratio of estimated marginal tax revenue gained to additional spending was • 13:1 to pursue known tax debts through phone calls, • 32:1 to identify unreported income by computer matching tax returns and information returns, such as Forms W-2 and 1099, and • 11:1 to audit tax returns through correspondence. Our reference to these alternate uses and ratios is for illustrative purposes. We did not analyze the basis for IRS’s estimates that produced these ratios. However, we have found over the years that IRS has had difficulty readily accumulating the costs of various activities because IRS lacks a cost accounting system.25 Despite these caveats, if these estimated ratios and scenario I estimates are approximately accurate, IRS and FMS might have made a different decision by considering opportunity costs.26 Although opportunity costs may be the most significant costs not considered, the study also excluded certain direct IRS and FMS costs, as discussed below. • IRS and FMS costs to oversee lockbox bank processing: As discussed earlier in this report, these agencies use staff to oversee and review operations of lockbox banks. • Costs to reduce the risk in processing tax receipts: This risk became evident during the 2001 filing season, with the incident at the Pittsburgh lockbox affecting about 78,000 taxpayer receipts. As of September 30, 2002, the government had lost an estimated $13.5 million in interest from missing tax receipts, and IRS has spent about $4.3 million to resolve problems with taxpayer accounts. In October 2002, Mellon Bank agreed to pay the government $18.1 million to cover administrative costs 25 U.S. General Accounting Office, Financial Audit: IRS’s Fiscal Years 2002 and 2001 Financial Statements, GAO-03-243 (Washington, D.C.: Nov. 15, 2002). 26 If IRS and FMS chose the internal processing alternative, the federal government would have lost the estimated $100.5 million in interest earnings but would have saved the $84.5 million paid through the TTBs and an additional $4.4 million from IRS’s budget. This choice would have been better from a federal budget standpoint as long as IRS generated at least $16 million (the difference between $100.5 million and $84.5 million) by investing the $4.4 million elsewhere. Page 33 GAO-03-299 IRS Lockbox Banks and expenses associated with the incident; however, these costs continue to grow as IRS is still resolving some issues. • Costs to process different types of receipts: The study only considered Form 1040 tax receipts in response to a Treasury OIG report that questioned having lockbox banks rather than IRS process such tax receipts. Lockbox banks are also paid to process other types of tax receipts related to more than 10 other tax forms, such as Forms 1041 and 941.27 It is not known if these costs would be significant enough to change the study’s conclusions. It is also not known to what extent such costs would be offset by additional direct costs to IRS associated with returning the receipt processing function to IRS. Alternate Study Approaches IRS and FMS characterized the study at various times as a cost-benefit Could Affect the Costs and analysis and as a cost-effectiveness analysis. These types of studies would include different costs and benefits from those included in the 1999 study. Benefits Considered and the If IRS and FMS had done a cost-benefit or cost-effectiveness study, the Study Results resulting conclusions may have differed. Being clear about the type of analyses being done and the limitations and uses of the related results helps decision makers to interpret and use those results. The 1999 study was neither a cost-benefit nor a cost-effectiveness analysis in economic terms. Cost-benefit and cost-effectiveness analyses are two tools that are commonly used to determine whether government investments or programs can be justified on economic principles. These tools also help to identify the best alternative from a range of competing investment alternatives. Economists commonly agree that, when either of these two analyses is used to evaluate federal programs that affect 27 The study did not consider whether lockbox banks instead of IRS should process certain types of forms, or whether lockbox banks should process all or none of the forms. Page 34 GAO-03-299 IRS Lockbox Banks private citizens or other entities, the analysis should include costs and benefits to individuals and entities outside of the federal government.28 The IRS/FMS study was not a cost-benefit analysis because it did not consider costs or benefits to individuals or entities outside the federal government, such as taxpayers.29 By considering costs and benefits beyond the government, the study should address whether the benefits gained by the federal government (on behalf of society as a whole) outweigh the costs imposed on certain members of society, such as taxpayers. The lockbox program affects taxpayers and/or their banks by reducing their interest float benefits through quicker depositing of their tax receipts. The additional interest gained by the federal government through accelerated tax deposits comes with a similar loss of interest to taxpayers who mail in payments.30 The acceleration of deposits largely shifts who benefits. This shifting of interest benefits may have some value to society in terms of a more equitable sharing of the costs of government; however, it is difficult to put a dollar value on improved equity and that value is not necessarily equal to the dollar amount of interest that is transferred from taxpayers and/or banks to the government. Since these interest gains by the federal government made the use of lockbox banks beneficial, a different valuation of these gains could have resulted in a different decision about whether to contract with the banks. 28 See OMB, Circular A-94 Revised, Guidelines and Discount Rates for Benefit-Cost Analysis of Federal Programs (Washington, D.C.: Oct. 29, 1992), section 6; and OMB, Circular A-130, Management of Federal Information Resources (Washington, D.C.: Nov. 30, 2000), section 7. 29 A cost-benefit analysis estimates the discounted values of the expected benefits and costs of an investment or program and then subtracts the costs from the benefits to determine the net present value (NPV). The costs and benefits associated with an investment or program are typically spread over many years. Discounting (the computation of present values) is necessary to reflect the fact that a dollar of benefit or cost in a future year is worth less than a dollar of benefit or cost in the current year. Programs with larger positive NPVs are generally preferred over those with smaller or negative NPVs. 30 Some or all of the interest loss actually may be incurred by the taxpayers’ banks. Page 35 GAO-03-299 IRS Lockbox Banks The study also is not a cost-effectiveness analysis because it did not compare program alternatives that had the same objectives.31 Instead, the alternatives of using lockbox banks or IRS staff to deposit tax receipts assumed that lockbox banks would achieve a faster speed of depositing receipts than IRS. In order to determine the cost-effectiveness of the lockbox program, the analysts would have had to compare the costs of that program to the costs of one or more IRS alternatives that would achieve the same speed of depositing. It is not clear that a cost-effectiveness study comparing like processing speeds would have yielded the same result as the 1999 study. The specific type of study that should be done to support a decision about whether to accelerate the deposit of tax receipts through lockbox banks is a matter of judgment. However, because the type of analysis that flows from the decision about the type of study to do can affect the study results, it is important that study leaders clearly define the type of analysis to be undertaken and why. 1999 Study Will Not Support Since 1999, many changes have occurred at IRS and the lockbox bank IRS’s and FMS’s Future network that could affect processing and future cost comparisons. For example, IRS began moving to a new organizational structure in October Decision about Continuing 2000, which has changed where and how IRS processes certain types of tax the Lockbox Bank returns and receipts. In addition, the network of lockbox banks (in terms of Arrangement beyond 2007 how many and which banks are involved) has changed. Starting in 2002, processing also included new security requirements, such as having FBI fingerprint checks done for each employee or contractor, to reduce the risks of thefts. Finally, changes in the 1999 study assumptions would be likely by 2007. Such assumptions involve • the number of days of interest float; • the number of tax receipts and number of Forms 1040 filed electronically; 31 A cost-effectiveness analysis is used to determine the least-cost approach for achieving a specified objective. This analysis does not provide a justification for pursuing the objective but simply identifies the best approach after deciding on the objective. A fundamental characteristic of a cost-effectiveness analysis is that the specified objective does not vary across alternatives being compared. Page 36 GAO-03-299 IRS Lockbox Banks • IRS labor, computer, and space costs; • lockbox bank charges; and • interest rates. For example, IRS and FMS assumed that lockbox banks would process tax receipts 1.384 days faster than IRS under scenario I based on a 1998 study by a contractor. 32 However, IRS and FMS cannot be certain that any advantage that lockbox banks might have had in 1999 still exists due to the changes to IRS’s structure and the lockbox bank network. IRS and FMS officials stated that they had not planned to conduct a new study before the lockbox agreements expire in 2007. However, they have indicated that such a study may be warranted given many changes and the passage of time since 1999. Conclusions Approximately $268 billion in tax receipts IRS collected in fiscal year 2002 were processed through lockbox banks. Given the significance of tax receipts processed by lockbox banks, effective oversight and sound internal controls at lockbox sites are critical to protect taxpayer data and receipts. The loss and destruction of tax receipts at the Pittsburgh lockbox site highlighted the need for increased scrutiny and oversight of these banks. Our review of the 1993 and 2002 lockbox contractual agreements revealed nothing inherent in their provisions that would necessarily encourage lockbox employees to mishandle taxpayer receipts. It is possible that in an effort to avoid penalties allowed under the agreements, such as financial penalties or contract termination, lockbox bank employees might make poor decisions about handling taxpayer receipts; however, these are important provisions designed to help the government address inadequate contractor performance. 32 The contractor concluded that lockbox banks processed tax receipts on average 2 days faster than IRS. However, IRS and FMS revised the 2-day lockbox advantage in each of the three scenarios in the 1999 study on the basis of assumptions, such as a promise by the lockbox banks to compress processing by 1 day. See appendix II for further explanation of the three scenarios that IRS and FMS used in the 1999 study. Page 37 GAO-03-299 IRS Lockbox Banks While FMS and IRS significantly improved their oversight of lockbox banks during 2002, oversight and internal control deficiencies impeded the effectiveness of this oversight in minimizing the risk to the federal government and taxpayers. These deficiencies need to be addressed to provide increased assurance that taxpayer data and receipts are properly safeguarded. IRS’s and FMS’s oversight of lockbox banks has not been fully effective primarily because their oversight roles and responsibilities have not been sufficiently defined and documented. Additionally, numerous internal control weaknesses need to be corrected and certain provisions of the lockbox processing guidelines need to be revised. Until these oversight and internal control deficiencies are addressed, the security of taxpayer data and receipts may be compromised. The most recent study done by IRS and FMS in 1999 followed Treasury regulations in considering only the costs and benefits to the federal government of achieving speedier tax deposits by using lockbox banks to process individual tax receipts (Form 1040). In doing so, the study did not consider other types of receipts processed by the lockbox banks or some relevant direct costs in comparing lockbox and IRS processing costs nor did it consider opportunity costs. Accounting for opportunity costs would be consistent with agencies’ responsibility to use budget funds economically and efficiently and should be considered regardless of the type of analysis that IRS and FMS undertake. However, the type of analysis affects other types of cost and benefit data that should be considered and therefore may affect the study results. IRS and FMS were not clear on the type of analysis done in 1999. Because certain data and assumptions from the 1999 study are now obsolete, IRS and FMS will need to conduct another study to determine whether to continue using lockbox banks when the agreement expires in 2007. Recommendations for To decrease the likelihood that further incidents involving the loss and destruction of taxpayer receipts and data will occur, we recommend that Executive Action the Commissioner of FMS and Acting Commissioner of IRS thoroughly review the results of TIGTA’s investigation of the 2001 incident at the Pittsburgh lockbox site when it is completed and, if the results warrant, implement additional controls and modify the lockbox contractual agreements as appropriate. Page 38 GAO-03-299 IRS Lockbox Banks To improve the effectiveness of government oversight of lockbox banks, we recommend that the Commissioner of FMS and the Acting Commissioner of IRS • document IRS’s and FMS’s oversight roles and responsibilities in agency policy and procedure manuals and determine the appropriate level of IRS and FMS oversight of lockbox sites throughout the year, particularly during peak processing periods; • establish and document guidelines and procedures in IRS and FMS policy and procedure manuals for implementing the new penalty provision for lockbox banks to reimburse the government for direct costs incurred in correcting errors made by lockbox banks; • finalize and document the recently developed waiver process in IRS and FMS policy and procedure manuals and ensure that decisions on requests for waivers are formally and promptly communicated to lockbox management; and • establish and document a process in IRS and FMS policy and procedure manuals to ensure that lockbox bank management formally responds to IRS and FMS oversight findings and recommendations promptly and that corrective actions taken by lockbox bank management are appropriate. To improve the effectiveness of government oversight of lockbox banks, we also recommend that the Acting Commissioner of IRS • establish and document a process in IRS policy and procedure manuals to ensure that IRS officials with the appropriate levels of expertise continue to participate in announced and unannounced security reviews of lockbox banks; • ensure that the results of on-site compliance reviews are completed and promptly submitted to IRS’s National Office; • revise the guidance used for compliance reviews so it requires reviewers to (1) determine whether lockbox contractors, such as couriers, have completed and obtained favorable results on IRS fingerprint checks and (2) obtain and review all relevant logs for cash payments and candled items to ensure that all payments are accounted for; and Page 39 GAO-03-299 IRS Lockbox Banks • assign individuals, other than the lockbox coordinators, responsibility for completing on-site performance reviews. To improve internal controls at lockbox banks, we recommend that the Commissioner of FMS and the Acting Commissioner of IRS • require that internal control deficiencies are corrected by lockbox bank management and that IRS and FMS take steps through ongoing monitoring to ensure that the following LPG requirements are routinely adhered to: • perimeter doors are locked and alarms on perimeter doors are functioning, • guards are responsive to alarms, • employees’ identity and employment status are verified prior to granting access to the processing floor, • visitor access to and activity in the processing area are adequately controlled, • employee access and items brought into and out of processing areas are closely monitored by guards, • surveillance cameras and monitors are installed in ways that allow for effective, real-time monitoring of lockbox operations, • envelopes are properly candled, • lockbox bank management perform and adequately document candling reviews, • returned refund checks are restrictively endorsed immediately upon extraction, • lockbox couriers are properly identified prior to granting them access to taxpayer data and receipts, and • employees have received favorable results on fingerprint checks before they are granted access to taxpayer data and receipts. Page 40 GAO-03-299 IRS Lockbox Banks • revise the lockbox processing guidelines to require that • before lockbox bank couriers receive access to taxpayer data and receipts they undergo and receive favorable results on background investigations that are deemed appropriate by IRS and are consistent across lockbox banks, • before permanent lockbox bank employees receive access to taxpayer data and receipts they undergo and receive favorable results on background investigations that are deemed appropriate by IRS and are consistent across lockbox banks, • lockbox bank guards inspect courier vehicles for unauthorized passengers and unlocked doors, • candling procedures for the various types of extraction methods be clarified, • during candling, lockbox bank employees record which machines and which extraction clerks missed items, • lockbox bank management reconcile items found during candling to the candling records, • lockbox bank management reconcile cash payments to internal cash logs and the cash logs they provide to IRS, and • lockbox employees immediately seek processing guidance from the lockbox coordinator if envelopes with timely postmark dates are received after the postmark review period has ended. Because IRS and FMS must decide before 2007 whether to continue using lockbox banks to process tax receipts or to return that function to IRS, we recommend to the Secretary of the Treasury that a study be done in time (1) for its findings to be considered in the decision-making process and (2) to make any improvements to lockbox processing that the study indicates are necessary or to return the processing to IRS. Regardless of the type of analysis chosen, we recommend that the Secretary of the Treasury • clearly define the type of analysis being done and why, and follow through to identify and analyze costs and benefits relevant to the type of analysis, Page 41 GAO-03-299 IRS Lockbox Banks • consider the opportunity costs associated with the proposed investment in using lockbox banks to accelerate the deposit of tax receipts, and • include the direct costs associated with oversight, risk reduction, and non-Form 1040 tax receipts. Agency Comments and Treasury’s response to a draft of this report was jointly prepared by IRS and FMS. In responding to the report, IRS and FMS agreed with our findings Our Evaluation and recommendations and stated that they have initiated or plan to initiate actions to implement our recommendations. IRS and FMS agreed to continually monitor lockbox banks’ adherence to internal controls and to modify the LPGs to improve consistency standards and clarify instructions. IRS and FMS also agreed to complete an analysis of lockbox processing prior to the expiration of the current lockbox agreements to determine how best to satisfy IRS’s remittance processing needs. In their response, IRS and FMS indicated many actions they have taken since October 2001 to improve lockbox operations. We identified many of these improvements during our review and documented them in our report, such as the establishment of the Bank Review Office within FMS and the development of memoranda of understanding concerning oversight roles and responsibilities. These actions and agreements will need to be promptly completed and thoroughly documented to satisfactorily address many of the issues raised in this report. The complete text of Treasury’s response is reprinted in appendix III. As agreed with your office, unless you announce its contents earlier, we plan no further distribution of this report until 30 days after its issuance date. At that time, we will send copies to the Chairmen and Ranking Minority Members of the Senate Committee on Appropriations; Senate Committee on Governmental Affairs; Senate Committee on the Budget; Subcommittee on Treasury, General Government, and Civil Service, Senate Committee on Appropriations; Subcommittee on Oversight of Government Management, Restructuring, and the District of Columbia, Senate Committee on Governmental Affairs; House Committee on Appropriations; House Committee on Ways and Means; House Committee on Government Reform; House Committee on the Budget; Subcommittee on Government Efficiency, Financial Management, and Intergovernmental Relations, House Committee on Government Reform; and Subcommittee on Oversight, Page 42 GAO-03-299 IRS Lockbox Banks House Committee on Ways and Means. We will also provide copies to the Chairman and Vice-Chairman of the Joint Committee on Taxation, the Secretary of the Treasury, the Commissioner of FMS, the Acting Commissioner of IRS, the Director of the Office of Management and Budget, the Chairman of the IRS Oversight Board, and other interested parties. Copies will be made available to others upon request. In addition, the report will be made available to others at no charge on GAO’s Web site at http://www.gao.gov. If you have any questions about this report, please contact Steve Sebastian at (202) 512-3406 (SebastianS@gao.gov) or Mike Brostek at (202) 512-9110 (BrostekM@gao.gov). Additional key contributors to this assignment are listed in appendix IV. Steven J. Sebastian Director Financial Management and Assurance Michael Brostek Director Tax Administration and Justice Page 43 GAO-03-299 IRS Lockbox Banks Appendix I Internal Control Weaknesses Ap A pp ep ned nx ide x Iis During calendar year 2002, we visited all nine lockbox locations to review their internal controls designed to protect taxpayer data and receipts. TIGTA auditors and IRS reviewers also performed reviews of lockbox controls in 2002. Below are the specific internal control issues with (1) physical security, (2) processing controls, (3) courier security, and (4) employee screening, that we and other reviewers found at lockbox sites. Physical Security The Lockbox Processing Guidelines (LPG) establish physical security and Weaknesses Could Allow other processing requirements. Lockbox banks are required by their agreements with FMS to abide by these guidelines. However, during our Unauthorized Access to visit to lockbox locations, we observed numerous physical security Taxpayer Data and Receipts breaches in violation of the LPGs. We also identified areas where the LPGs could be strengthened. These matters are discussed in the following sections. Perimeter Doors Were Unlocked To detect attempted breaches into secured space, lockbox guidelines and Alarms Did Not Function require all perimeter doors leading into IRS lockbox space to be equipped with alarms. The guidelines also require guards to ensure that such doors are locked. However, at four lockbox locations, we noted problems with perimeter door security. At one location, we found a perimeter door unlocked. At another location, a perimeter door was not equipped with an audible alarm during operating hours. Bank management did not think an audible alarm was necessary because an additional interior door that was locked during the day controlled access to the processing area. However, we observed that this interior door was propped open during our visit. At this location, we also found that the alarm for another door was barely audible over the noise from the production floor and immediately ceased when the door was closed, limiting the opportunity for security personnel to determine which door opened and to investigate any possible unauthorized entrance or exit. At a third location, the alarms for one set of doors were not turned on during operating hours, and the alarm for another perimeter door failed to activate because, according to a bank official, the alarm had not been properly set. At a fourth location, the door alarms were not audible at the guard post because the guards had turned down the volume. These security weaknesses could result in unauthorized entry to and exit from the lockbox processing areas, increasing the risk of theft of tax data and receipts. Page 44 GAO-03-299 IRS Lockbox Banks Appendix I Internal Control Weaknesses Guards Were Not Responsive to Door alarms serve to alert lockbox staff to a possible breach of security. To Alarms be an effective physical security control, alarms require quick response time by a security force. However, at two of the lockbox locations where we found problems with perimeter door security, we also noted that security guards were not responsive to alarms. For example, at one location, we tested the alarm twice. On the first test, no guard responded to the alarm. On the second test, it took 1 minute after the door was opened before we saw a guard approaching the door. Bank managers believed that their guards should have responded faster. At another location, the perimeter door alarm sounded twice, and both times the guards did not respond. According to the bank manager, the guards’ post orders did not instruct guards to respond to alarms. He also added that it was difficult for the guards to hear the alarms from the guard station. The presence of security guards serves to detect unusual activities and to deter crime. An ineffective security force may not only limit the overall effectiveness of a security system but also may inadvertently encourage security breaches from individuals who decide to take advantage of this weakness. Employee Identification Was Not Lockbox locations are set up with a main entrance where guards can Adequately Verified observe and control the traffic into and out of the processing area. The LPGs require that temporary employees provide photo identification to the guards in exchange for a badge allowing access to the processing floor. This ensures the identity of temporary employees is validated prior to entering the processing area. We found that at one location, guards did not routinely verify temporary employee identification when they entered the processing area. Bank management believed that manual, daily verification of temporary employee identification was not necessary because the location’s automated entry system (AES) provided sufficient controls to limit access to authorized individuals. The AES allowed entry into the building and processing floor to individuals with AES cards, which control the door and turnstile locks. AES cards are issued to temporary employees after the guards have verified the employees’ identities with valid photo identification cards on their first day of work. Once these employees are issued AES cards, guards no longer verify the employees’ identification before they enter the building and processing floor. Temporary employees wear handwritten name tags with no photo identification to easily verify their identity. As a result, if an unauthorized individual obtains an AES card, the lack of routine verification of employee identification increases Page 45 GAO-03-299 IRS Lockbox Banks Appendix I Internal Control Weaknesses the likelihood that an individual could gain unauthorized access to the building and the processing floor, thereby increasing the risk of theft of tax data and receipts. Employment Status of Effectively limiting access at lockbox locations to authorized individuals Temporary Employees Was Not requires controls that not only verify the identity of employees, but also Adequately Verified determine whether individuals who present themselves as employees are currently authorized to have access to the site. We found controls over verification of employment status to be ineffective at two locations. At one location, the controls implemented to validate an individual’s current employment were ineffective.33 The two-step manual process to identify temporary employees and check their current employment status before issuing access badges could have allowed unauthorized individuals to enter the processing area. The first step required temporary employees to obtain name tags from one station. Staff at this station checked the employees’ identification cards against a roster of current employees and issued them handwritten name tags. The second step required temporary employees to obtain access badges from the guard station. The guards issued access badges to anyone with a valid photo identification and handwritten name tag, making an assumption that those with name tags were current employees. As a result, an unauthorized individual could have circumvented these controls and gained access to the processing floor by making a name tag and presenting it to the guards. At another location, guards did not compare temporary employees’ identification cards to the temporary agency’s list of current employees until after the employees were given access to the processing area. This increases the risk that unauthorized individuals could gain access to taxpayer data and receipts and not be promptly detected. Visitor Access to and Activity in Anyone entering the lockbox processing area must wear an identification Processing Area Were Not badge. In addition, individuals who have not had an FBI fingerprint check, Adequately Controlled but require access to the processing floor, must be escorted by a guard. However, at one lockbox location, we observed a copy machine repairman with no identification badge who was unescorted in the processing area. The guards indicated they were unaware whether the contractor had received an FBI fingerprint and therefore one of the guards had escorted 33 Unlike the previous example, this location did not use an automated entry system to control access to the processing area. Instead, temporary employees were manually issued access badges and name tags each day. Page 46 GAO-03-299 IRS Lockbox Banks Appendix I Internal Control Weaknesses the repairman into the processing area. However, during our observation, the guard was across the room, too far away to effectively observe what the repairman was doing. We also observed that none of the employees near the repairman challenged his presence. Bank managers later showed us proof that the repairman had successfully completed an FBI fingerprint check and explained that the guards should have given him a visitor badge, but did not need to escort him. Although in this case the contractor turned out to have an adequate security clearance, our observation indicated a need for guards to be better trained on procedures for granting access to contractors and for guards and employees to be more alert to activities on the processing floor. We also found malfunctioning AESs designed to control entrance into and exit from processing areas. Two lockbox locations use an AES to control access to the processing area. According to bank management for these two locations, the AES must register an exit for a specific badge before it can be used again to enter the processing area. In addition, the AES should not allow an individual to exit without a registered entry. In other words, in order for anyone to use an access badge to exit the processing area through the controlled access points, the system must first have a record of that badge as having been used to enter, otherwise exit from the area would be denied. However, due to a programming error, we found that the AES did not function as intended. Specifically, we found that a single visitor badge could be used repeatedly by different individuals, one after another, to gain access to the processing area, because the AES did not require a registered exit between registered entries. Moreover, we found that the badges allowed individuals to exit the processing area even though the AES did not initially record their entrance. This AES deficiency compromised the lockbox banks’ ability to control and monitor visitors’ entrance into and exit from the processing floors. At one location, managers corrected the error before we left; at the other location, managers agreed to correct the error. As a result of these weaknesses, tax receipts and data are exposed to an increased risk of theft from visitors who are not adequately monitored. Items Brought Into and Out of The LPGs prohibit individuals from bringing personal belongings, such as Processing Areas Were Not purses and shopping bags, into processing areas. The guidelines also Closely Monitored prohibit individuals from wearing baggy clothing or bulky outerwear inside processing areas. Guards stationed at the main entrance of processing areas enforce this rule. However, at five lockbox locations, we found that the guards did not effectively monitor individuals entering the processing floors to enforce Page 47 GAO-03-299 IRS Lockbox Banks Appendix I Internal Control Weaknesses this requirement. At three of these locations, we observed employees bring personal belongings or wear bulky outerwear inside the processing area. In one instance, a guard brought a purse inside the processing area. At four of these locations, we were able to bring personal belongings, such as purses, into the processing floor and were not challenged by the guards. The guards we interviewed informed us that they either failed to observe personal belongings brought into the processing area or did not know the requirements of the lockbox guidelines concerning personal belongings. Guards are also required to question individuals who attempt to remove property from the lockbox locations. However, at one location, the guards failed to search papers we carried out of the processing area. At another location, we found that on the day of our visit there were no guards present to observe employees leaving the processing area because bank management had not informed the guards that several employees would be working beyond their normal work hours. The guards finished their shift and left before the employees were dismissed. As a result of these security breaches, individuals could have used personal belongings and bulky clothing to conceal and remove tax data and receipts from lockbox locations undetected. Camera Surveillance Needs The LPGs require surveillance cameras to be installed at lockbox locations Improvement and security guards to monitor the cameras to observe critical areas, such as the loading docks, secure storage areas, mailroom, and extraction area. However, we found that the camera surveillance at seven lockbox locations we visited could be improved. At two locations, the cameras were stationary and did not have zoom capability to effectively monitor critical areas. At another three locations, camera monitors to survey activities on the processing floor were located in the managers’ offices; however, because of other duties, the managers were frequently on the processing floor and were not able to observe the monitors.34 FMS visited one of these three locations in early 2002 to perform an unannounced security review and also reported this as a finding. Additionally, at another of these three locations, there were no surveillance cameras in the former administrative offices located within 34 One of these locations also maintained a monitor at the guard station. However, the monitor only surveyed activity at the parking lot and docking area. The monitor used to survey the processing floor was maintained at the site manager’s office. Page 48 GAO-03-299 IRS Lockbox Banks Appendix I Internal Control Weaknesses the processing area. Lockbox managers recently vacated this area and did not consider installing surveillance cameras because no processing activity occurred there. However, this area housed unused file cabinets and desks with drawers where tax data and checks could be hidden. One of the offices was also used by a computer contractor while on site. At the two other locations, we found that the monitors used to display the images from surveillance cameras were ineffective. Specifically, the monitors displayed up to 16 images on one screen making each individual image barely visible to effectively monitor or detect illegal activities. Additionally, since these monitors were visible to employees and visitors, they were ineffective deterrent controls to those who noticed that the surveillance could not effectively monitor activities on the processing floor. The images would be more effective in deterring criminal activities if they were more visible. As a result of these weaknesses, lockbox camera surveillance was not capable of consistently and effectively detecting unusual activity or unsafe practices and providing early warning of possible security compromises. Processing Controls Need In addition to physical security controls, the lockbox guidelines also Improvement to Better provide requirements for processing controls to maintain accountability for, and security over, transactions as they are processed through the Account for and Protect normal course of operations. During our visits to lockbox locations, we Taxpayer Data and Receipts found deficiencies in processing controls designed to account for or protect tax data and receipts at several of the lockbox sites. Specifically, taxpayer information and receipts particularly vulnerable to theft, such as cash, were not carefully processed and safeguarded. Moreover, the records to account for items found during candling, a process to detect overlooked items remaining in envelopes, and cash payments were inadequate to allow lockbox managers to easily and promptly detect lost or stolen items. Candling Procedures and Review To prevent the accidental destruction of taxpayer data and receipts, Were Not Adequate lockbox guidelines require envelopes to undergo a candling process. The LPGs also require lockbox site managers to periodically review the effectiveness of their site’s candling procedures. During our review of lockbox operations, we found that at some lockbox locations (1) deficiencies in the candling processes may lead to the accidental destruction of tax data and receipts, (2) accounting for found items was insufficient to detect missing checks within a reasonable time, and (3) management review of candling processes was either lacking or not clearly documented. Page 49 GAO-03-299 IRS Lockbox Banks Appendix I Internal Control Weaknesses Inadequate Candling Procedures Could Lead to Accidental Destruction of Tax Data and Receipts Lockbox staff open envelopes manually by hand or with the assistance of a mail extraction machine. Some mail extraction machines slit envelopes on one side, allow employees to extract their contents, and have the capability to electronically detect overlooked items remaining in the envelopes. More advanced machines slit envelopes on three sides and extract the contents. The method used to open the envelopes determines how the envelopes should be candled. The LPGs provide guidance on candling procedures and require all envelopes to be candled twice before destruction. Envelopes opened with the assistance of a mail extraction machine are considered to have gone through the first candling.35 After processing, employees are required to review each envelope through a source of light, such as a candling table, to determine if any contents remain in the envelope. This is considered the second candling. For manually opened envelopes, envelopes that are slit on three sides and flattened sufficiently meet candling requirements without further light source viewing. Envelopes that are manually slit on only one side are reviewed twice on the candling table. During our visits to lockbox locations, we found that at four locations, envelopes were not sufficiently candled to prevent the accidental destruction of tax data and receipts because of malfunctioning machines, careless candling practices, and ineffective candling guidelines. • At one location, we found two mail extraction machines that malfunctioned and failed to detect checks remaining in the envelopes. Additionally, some envelopes were only candled once because employees often used the mail extraction machines as desks. Employees manually opened the envelopes, completely bypassing the first candling step to be performed by the machines. These envelopes were then candled only once on the candling table. We also observed that employees were inattentive when candling envelopes on the candling table, allowing envelopes to overlap and making it difficult to fully illuminate each envelope or all parts of an envelope. 35 Based on the LPGs, the first candling occurs when envelopes are processed through the mail extraction machines. The second required candling occurs on the lighted table used to perform candling. Page 50 GAO-03-299 IRS Lockbox Banks Appendix I Internal Control Weaknesses • At a second location, envelopes that were manually opened were not slit on three sides or candled twice. • Two other locations did not properly candle all envelopes because the candling requirements in the LPGs do not specify procedures to be used with more advanced extraction machines that slit the envelopes on three sides and extract the contents. Management at these locations believed that because the envelopes were opened on three sides, they met the candling guidelines for manually opened envelopes. Therefore, they concluded that no further candling was required. However, an IRS official subsequently explained that envelopes opened by these types of machines should be subject to a second candling until IRS performs a study to determine their effectiveness. The fact that we found a $3,300 check that had not been detected by one of these advanced machines, located in another site,36 indicates that they can also malfunction and result in the destruction of taxpayer data and receipts. As a result, candling procedures did not effectively reduce the risk of accidental destruction of tax receipts and data. Accounting for Items Found During Candling Was Insufficient Lockbox guidelines require lockbox employees to account for each item found during candling. Some lockbox locations use two forms to record items found--an internal candling log and a Form 953537 or equivalent, which is required to be submitted to the lockbox coordinator each month for IRS review. Employees prepare the internal log while candling and later transfer the data onto the Form 9535. Since all items found during candling must be reported to IRS, the Form 9535 should record the same number and amounts of checks found as noted on the internal log being maintained as processing is occurring. However, the lockbox guidelines do not require a reconciliation of one set of records to the other, nor do they require a reconciliation of items found during candling to their candling records. During our review, we found that lockbox banks did not always have procedures to ensure that all items found were accurately recorded on both 36 This location candled their envelopes twice even though they used the advanced mail extraction machine. 37 Form 9535 is the Record of Lockbox Discovered Remittance and Correspondence. A discovered remittance is a tax receipt discovered outside the normal check processing operation, such as during candling. Page 51 GAO-03-299 IRS Lockbox Banks Appendix I Internal Control Weaknesses sets of logs and that bank managers could not properly account for all items found. At one location, the two sets of logs that bank management provided to us could not be reconciled. For example, a check was recorded on the internal log but not on the Form 9535. Management could not provide an adequate explanation or documentation to explain the discrepancy. At this location, we also noted delays of up to 6 days in transferring records of items found from the internal log to the Form 9535. Lockbox management explained that because of the volume of work during the April peak, the Form 9535 could not always be completed or the checks could not always be processed the same day they were found during candling. Another location’s internal log only recorded tallies of the quantity of items found, which did not match the number of items listed on the log provided to IRS. Because there was not enough information on the internal log, we could not determine whether items, such as checks, recorded on the internal log but not on the Form 9535, were ever processed and credited to the taxpayers’ accounts. We identified this same issue during a visit to this location in October 2001 as part of our audit of IRS’s fiscal year 2001 financial statements.38 After this visit, lockbox management and IRS agreed to address this problem. However, as of April 2002, this problem had not been corrected. IRS indicated that it plans to address this issue in the January 2003 revisions to LPGs. At two other locations, we also found that no reconciliation occurred between items found during candling and the candling log. As a result of these weaknesses, lockbox management and IRS may not promptly detect missing checks. We also noted that the lockbox guidelines do not specifically require the banks to determine which machine or which extraction clerk39 missed items that were subsequently found during candling. Such information would help lockbox bank managers promptly track and repair malfunctioning machines and provide performance feedback to extraction clerks. 38 U.S. General Accounting Office, Management Report: Improvements Needed in IRS’s Accounting Procedures and Internal Controls, GAO-02-746R (Washington D.C.: July 18, 2002). 39 An extraction clerk is an individual responsible for manually extracting contents from envelopes. Due to the nature of the work, extraction clerks are among the first individuals who handle tax data and receipts at lockbox sites. Page 52 GAO-03-299 IRS Lockbox Banks Appendix I Internal Control Weaknesses Management Review of Candling Process Was Insufficient or Not Adequately Documented Because IRS officials are not at lockbox locations daily, IRS relies on lockbox managers to ensure that their staff complies with lockbox guidelines. In the case of candling requirements, IRS guidelines require lockbox managers to sample candled items to determine the effectiveness of the candling process and to report results of the reviews to IRS. However, at four locations, we found that lockbox managers, who indicated they had performed these reviews, failed to document or clearly document the results. At two additional locations, managers stated that they did not sample candled envelopes. The manager at one of these locations believed that her frequent observation of the candling process was sufficient to ensure that the envelopes were properly candled. Given the problems we observed with the candling process at several of the locations we visited, management reviews and proper documentation of such reviews are critical to help ensure that problems are promptly identified and corrected. Controls Were Not in Place to Lockbox sites receive tax receipts in the form of cash as well as checks. Adequately Safeguard and Cash receipts are highly susceptible to theft, and lockbox guidelines have Account for Cash Receipts specific requirements for safeguarding and accounting for cash receipts. The guidelines require cash to be stored in locked containers to prevent theft. The keys to these containers must not be left in unsecured places, such as desk drawers. The guidelines also require cash receipts to be recorded on a log. At three lockbox locations, we noted weaknesses in the safeguarding and accounting for cash. At one location, the locked cash box was stored in a locked drawer. However, the keys to both the cash box and the locked drawer were stored in an open drawer because bank management wanted supervisors to have quick access to the cash box to expedite the deposit of cash. At another location, cash for the business and individual tax payments were stored in two separate safes. Three staff members had keys to the safes, which could be opened by individuals acting alone. Additionally, the safe used for business payments was small and moveable. The individual access to the safes and the ease of mobility of one of the safes compromised the security of the stored cash. At two locations, we noted discrepancies between the bank’s safe log and the log that the lockbox managers were required to provide to IRS. A safe log is an internal lockbox document that employees complete before they Page 53 GAO-03-299 IRS Lockbox Banks Appendix I Internal Control Weaknesses place cash in the safe or cash box. The safe log should identify the taxpayer, the amount paid, and date cash was discovered. When we reviewed the two sets of logs we found that the dates, amounts, item counts, and taxpayer identification information in the logs did not agree, and bank managers could not reconcile some of the discrepancies. For example, at one location, five items on the safe log were not on the IRS log. One of the items was a $140 cash receipt. Because the safe log did not record taxpayer information for any of the receipts, we could not verify whether the receipt was ever posted to the taxpayer’s account. There is currently no requirement in the LPGs for bank managers to reconcile their internal cash logs to the cash log sent to IRS. Such reconciliation could have quickly detected the discrepancies we noted. Bank management attributed some of these discrepancies to human error, inexperienced staff, and staff failure to take the accounting for cash seriously because the amounts found are typically small. However, individual taxpayers have made cash payments totaling hundreds of dollars. Failure to properly secure and account for cash receipts could result in theft, the untimely detection of theft, inaccurate posting of tax receipts, and unnecessary burden on taxpayers whose cash receipts are lost or incorrectly posted. Therefore, it is critical for lockbox banks to diligently safeguard and account for cash receipts. Returned Refund Checks Were Lockbox banks sometimes receive federal tax refund checks that have Not Adequately Protected been returned by taxpayers as payment against other tax liabilities. Some Against Theft of these checks have been endorsed by taxpayers and are therefore negotiable. As a safeguard against theft and misuse of returned refund checks, lockbox guidelines require lockbox extraction clerks to promptly stamp a restrictive endorsement on returned refund checks. These checks are subsequently forwarded to IRS for further processing. At seven locations, however, we observed that returned refund checks were not always stamped upon extraction and at some locations were set aside, unsecured, to be stamped later by a different individual. At two locations, we found the returned refund checks without the required stamps ready to be shipped to IRS. Lockbox management attributed many of these exceptions to staff oversight or inadequate training. Several years ago, IRS investigators discovered the theft of seven returned refund checks totaling $300,000 at one IRS Submission Processing Center. Such thefts can also occur at lockbox sites. Thus, it is critical that restrictive endorsements be placed on returned refund checks as soon as they are extracted. Page 54 GAO-03-299 IRS Lockbox Banks Appendix I Internal Control Weaknesses Timely Payments Were The LPGs, as currently written, have resulted in timely tax payments being Incorrectly Processed As processed as delinquent. To determine timeliness, lockbox employees are Delinquent required to examine the postmarks on envelopes in which tax receipts were received. If the postmark indicates that the receipt is delinquent (i.e., the postmark is subsequent to the return or payment due date), the receipt should be processed as delinquent and the envelope should be attached to the corresponding document and forwarded to IRS. If lockbox employees determine that the receipt is timely (i.e., the postmark is prior to the return or payment due date), the envelope need not be retained. However, beginning with the first day of the month following the month the payment is due, lockbox guidelines require lockbox employees to use the date the mail is received as the transaction date to be recorded in the taxpayer’s account as the payment date. Furthermore, the lockbox guidelines do not require lockbox employees to review and retain the envelopes in which the tax receipts were enclosed. At one lockbox location we visited, we observed that on May 1, 2002, lockbox employees received and extracted tax receipts from two envelopes postmarked prior to or on the April 15 payment due date. Because it was already past the period during which lockbox employees were required to examine postmarks and retain envelopes, lockbox employees processed the payments as delinquent and would have discarded the envelopes even though they were aware that the envelopes were postmarked on or before the payment due date. When we brought these two transactions to the lockbox coordinator’s attention, the lockbox coordinator concluded that, in these instances, the taxpayers had made timely payments. The lockbox coordinator subsequently adjusted the taxpayers’ accounts to reflect these as timely payments. We recognize that careful examination of postmarks on envelopes for all receipts received after the payment due dates slows down the payment processing and we recognize the need to establish reasonable cut-off dates for this review process. However, lockbox employees should immediately seek processing guidance from the lockbox coordinator when incidents such as the ones noted above come to their attention to avoid burdening taxpayers with erroneous penalties and interest for late payments. Additionally, as written, the LPGs could lead to potential taxpayer burden and unnecessary costs to IRS associated with correcting the status of taxpayer’s accounts. Page 55 GAO-03-299 IRS Lockbox Banks Appendix I Internal Control Weaknesses Courier Security Is Lockbox banks entrust courier employees with transporting thousands of Inadequate to Protect Tax pieces of mail containing tax data, cash, checks, and deposits worth millions of dollars per day. Given the susceptibility of these items to theft Data and Receipts from and loss, it is extremely important that they be carefully safeguarded while Theft or Loss in transit to and from lockbox sites. We previously reported on weaknesses in lockbox courier security and noted that lockbox courier guidelines were not as stringent as IRS courier guidelines.40 For example, unlike IRS courier requirements, the LPGs did not require courier employees to pass a limited background investigation, thus increasing the risk of theft of tax data and receipts by couriers with past criminal histories. The LPGs also did not require lockbox couriers to be insured for $1 million and to travel in pairs while transporting IRS data. In fact, in past audits, we found that lockbox banks used only one courier employee to transport IRS data. These weaknesses in courier security increased the risk of theft and loss of taxpayer data and receipts. In response to our audit findings, IRS enhanced the lockbox courier requirements. The 2002 LPGs now require that couriers used by lockboxes pass favorable FBI fingerprint checks, be bonded for $1 million, travel in pairs, transport IRS data from the lockbox facility to its destination with no stops in between, provide dedicated service to IRS, and lock and attend courier vehicles while IRS data are contained within the vehicles. Despite these enhancements, we and other auditors41 continue to find weaknesses in lockbox courier security because of the lockbox sites’ failure to consistently comply with the revised guidelines. In addition, lockbox courier guidelines could be further refined to improve the security of tax data and receipts. Couriers Given Access to Tax IRS recently revised the background screening requirements for lockbox Data Before Fingerprint Check couriers. The revised LPGs, effective January 2002, prohibit couriers from Results Received having access to IRS data until lockbox managers have received results of their FBI fingerprint checks and resolved any questionable fingerprint 40 GAO-02-35. 41 TIGTA 2002-30-127. Page 56 GAO-03-299 IRS Lockbox Banks Appendix I Internal Control Weaknesses results. However, during a recent TIGTA audit42 of one lockbox site, auditors found that three couriers were allowed access to taxpayer information before the lockbox received the results of their fingerprint checks. Lockbox managers subsequently received results of the FBI fingerprint checks, which indicated that two of these couriers had criminal histories. Nevertheless, TIGTA found that lockbox management continued to allow the two couriers and an additional courier, whose FBI fingerprint check also indicated a criminal history, access to taxpayer data while follow-up investigations, which subsequently cleared them, were underway. TIGTA auditors attributed this weakness to lockbox management’s failure to develop and implement procedures to ensure that couriers are granted proper clearance before they receive access to IRS data. Lockbox Couriers Not Properly Lockbox courier standards require courier employees to wear Identified identification badges and lockbox banks to implement procedures to properly identify couriers. However, at two lockbox locations, we found that couriers did not wear their identification badges. At one of these locations, lockbox employees did not verify the courier’s identification before entrusting him with taxpayer data because the employees indicated that they were familiar with the courier. Additionally, at this location, the courier access list, which lists couriers authorized to access tax data and photo identification of couriers, was maintained at the guard station and not easily accessible to employees who must verify couriers’ identities daily. Although not a requirement in the LPGs, some locations have posted the courier access lists by loading dock doors to facilitate the identification of couriers. Unless lockbox employees diligently perform their duties to properly identify couriers, tax receipts and data are exposed to higher risk of theft from former couriers who have recently been terminated or unauthorized individuals posing as couriers. Courier Vehicle Containing Tax The LPGs require courier vehicles to be locked whenever IRS data are Data and Receipts Was Not contained in the vehicle until it reaches its destination. Additionally, the Locked vehicle must always be under the supervision of one of the couriers and never left unattended. At one lockbox site, we observed that couriers did not lock the courier vehicle containing tax data. The couriers stated that they generally did not lock the doors because they never left the IRS packages unattended. The lockbox guard did not check to see if the vehicle 42 TIGTA 2002-30-127. Page 57 GAO-03-299 IRS Lockbox Banks Appendix I Internal Control Weaknesses was locked because there is no requirement to do so. IRS reviewers also observed couriers failing to lock courier vehicles during their review of another lockbox location in April 2002. Failure to ensure that courier vehicles are locked while in possession of taxpayer data and receipts increases the risk of loss of such items. Background Screening of We also found that background screening requirements continue to be less Lockbox Couriers Are Less stringent for lockbox couriers than for IRS couriers. IRS couriers are Stringent Than Requirements for subject to both an FBI fingerprint check and a basic background IRS Couriers investigation for contractors before they are given access to tax receipts. This background investigation includes a check of other federal agency and defense clearance investigation databases for results of previous background investigations and a check for any outstanding tax liabilities. In contrast, lockbox couriers are only required to favorably clear an FBI fingerprint check and are subject to no further background investigations. As a result, IRS may not discover information, such as outstanding tax liabilities, that might cause IRS to deny them access to taxpayer data. IRS is aware of the risks associated with lockbox couriers and is considering enhancing the lockbox guidelines to require lockbox couriers to undergo basic background investigations similar to those required of IRS couriers. Unauthorized Individuals Not IRS courier standards specifically prohibit the presence of unauthorized Prohibited From Courier individuals in courier vehicles and require IRS personnel to inspect courier Vehicles vehicles daily to ensure that no unauthorized passengers are in the vehicle. The LPGs, on the other hand, contain no prohibition of unauthorized individuals in courier vehicles and do not require lockbox staff or guards to inspect courier vehicles for unauthorized passengers. The guidelines state only that IRS reserves the right to inspect courier vehicles and drivers. Because IRS representatives are not on-site every day and there is no requirement for lockbox employees to inspect vehicles, unauthorized individuals could ride in courier vehicles and have access to taxpayer data and receipts without lockbox management’s knowledge. Further Improvements Because lockbox employees are entrusted with handling sensitive taxpayer Needed on Background information and billions of dollars in receipts annually, ensuring worker integrity through a carefully managed recruiting and hiring process is an Screening of Lockbox area that demands special attention from IRS, FMS, and lockbox Employees management. We previously reported that the screening of permanent and Page 58 GAO-03-299 IRS Lockbox Banks Appendix I Internal Control Weaknesses temporary lockbox employees was inadequate and untimely.43 Specifically, instead of referring to a national database to check for criminal records, lockbox banks limited the screening of criminal background investigations for temporary employees to police records checks in counties that individuals voluntarily disclosed as prior residences. Therefore, the police records checks may be incomplete for some individuals who chose not to disclose counties in which they committed a crime and have criminal records. In addition, lockbox permanent employees were allowed to handle cash, checks, and taxpayer data before their fingerprint checks were completed. IRS management has been responsive to our recommendations and has enhanced its policy on screening of permanent and temporary lockbox employees. The LPGs now require permanent and temporary employees to undergo FBI fingerprint checks. In contrast to the previous police records checks performed by county, FBI fingerprint checks are national in scope. An individual’s fingerprints are matched against fingerprints maintained in the FBI’s national database of criminal records. As a result, criminal records checks performed for lockbox applicants are no longer dependent on the applicant to accurately and completely disclose prior residences. The guidelines have also been updated to prohibit access to taxpayer data and receipts until lockbox management receives the results of an individual’s fingerprint checks. Results that show a possible criminal history must be resolved before the individual in question is allowed access to the lockbox site. The guidelines also require permanent lockbox employees to undergo an appropriate background investigation, as determined by an IRS security officer, in addition to an FBI fingerprint check. Despite these policy improvements, we found that lockbox banks did not always comply with the FBI fingerprint requirements and that further refinements are needed regarding background investigation requirements for lockbox employees. Employees Given Access to Based on our review of lockbox personnel files, we found that lockbox Taxpayer Data and Receipts banks are generally complying with the new guidelines. However, we and Before Fingerprint Check IRS reviewers nonetheless found instances of noncompliance at lockbox Results Received banks. This shows a need for IRS and FMS to ensure that lockbox management clearly understand the screening requirements and have implemented effective controls to prevent permanent and temporary 43 GAO-02-35. Page 59 GAO-03-299 IRS Lockbox Banks Appendix I Internal Control Weaknesses employees’ access to tax data until they have favorably completed an FBI fingerprint check. Specifically at three lockbox locations, we noted noncompliance related to the screening of permanent staff. At one location, we found that 4 out of a nonrepresentative selection of 25 permanent employees whose personnel files we reviewed began working at the lockbox location before bank management received their fingerprint check results. Bank management and IRS personnel explained that this situation occurred because IRS had verbally waived compliance with the screening requirements due to the fact that the bank was experiencing delays in obtaining timely responses on fingerprint checks at the beginning of 2002. At the second location, weak controls to ensure that all employees successfully complete FBI fingerprint checks allowed a permanent bank employee to process receipts and taxpayer data for 3 months before lockbox managers discovered that the employee had not undergone a fingerprint check. The employee was removed from the processing floor until the fingerprint check was completed and approved. At the third location, a permanent employee was allowed to work for several days before the FBI fingerprint check was completed because bank management misunderstood the fingerprint check requirement for lockbox employees. During its April 2002 peak review, IRS officials found similar problems at two other lockbox locations. One location allowed a temporary employee access to tax data before the completion of the employee’s fingerprint check. At the other location, the temporary agencies listed 12 employees as eligible to work, of which 6 were already working, even though they had not received their FBI clearance checks or were denied clearance to access tax data. As discussed earlier, taxpayer information and receipts are easily accessible to anyone on the processing floor. Therefore, it is critical for lockbox banks to ensure that these items are properly safeguarded by diligently complying with all aspects of the LPGs, which include screening lockbox employees. Late screening of lockbox employees could result in theft or loss in instances where bank management unknowingly allows individuals with criminal backgrounds access to IRS data and receipts. Guidelines for Background The current LPGs require permanent lockbox employees to favorably Investigations of Lockbox complete an FBI fingerprint check and an appropriate background Permanent Employees Need investigation, as determined by an IRS personnel security officer. However, Improvement the LPGs do not define what is considered an appropriate background investigation for permanent lockbox employees and what information regarding the results of the background investigation should be provided to Page 60 GAO-03-299 IRS Lockbox Banks Appendix I Internal Control Weaknesses IRS. As a result, the scope of and documentation for background investigations performed on permanent lockbox employees varies greatly among lockbox banks and is not consistent with background investigations required of other IRS contractors. According to IRS officials, while some lockbox banks subject their permanent employees to the required FBI fingerprint check and additional background investigation, such as county criminal records and credit history checks, other lockbox banks subject their permanent employees to FBI fingerprint checks with no further background screening. Additionally, IRS officials found that investigation results they receive from banks do not provide adequate information to determine whether the individual should be allowed access to taxpayer data. For example, background investigation results may indicate that a criminal records check was completed but not whether any arrests or convictions were found. Other results of background investigations may indicate that an arrest or conviction was found during a criminal records check but not the basis for the arrest or how recently or frequently the offense occurred. IRS officials also explained that some lockbox banks could not provide documentation of results of background investigations performed on their permanent employees because, as a result of recent bank mergers, lockbox management did not have access to those records. According to IRS officials, IRS did not foresee the problems with background investigations for permanent lockbox employees. As the variance in the scope of background investigations and in the adequacy of documentation of their results became evident, IRS made a decision to accept favorable results of FBI fingerprint checks as the minimum criterion for allowing permanent lockbox employees access to taxpayer data. As a result, the level of background screening performed on permanent employees is inconsistent among lockbox banks and with requirements for other IRS contractors, such as IRS contracted couriers whose backgrounds are checked against other investigation databases and for tax liabilities, as previously discussed. Additionally, permanent employees who were granted access to tax data based only on the results of favorable FBI fingerprint checks, in effect, received the same level of background screening as temporary employees and less than that of IRS contracted couriers, even though permanent employees have more influence and authority over lockbox operations, and are granted more access rights to various sections of the lockbox sites. Because some banks subject their permanent employees to less scrutiny when performing background investigations than other banks, IRS may not be aware of critical Page 61 GAO-03-299 IRS Lockbox Banks Appendix I Internal Control Weaknesses information that could have been uncovered by a more thorough background investigation, such as recent criminal records not yet reported to the FBI, which might cause IRS to deny them access to taxpayer data. FBI Checks May Be Inadequate TIGTA auditors recently reported that lockbox banks often employ non- for Lockbox Employees Who U.S. citizens with lawful permanent residence to process IRS tax Have Recently Established payments.44 Although the IRS hires only U.S. citizens, IRS and FMS have Residency in the United States allowed lockbox banks to hire non-U.S. citizens. TIGTA auditors found this policy to be consistent with guidelines from the Department of the Treasury regarding the hiring of contract employees. However, this hiring practice may pose unnecessary risks to IRS materials because the FBI fingerprint check, which is national in scope, may have very little information to disclose if these individuals lived in this country for only a short period of time. The Department of State and the Immigration and Naturalization Service perform some background checks before issuing visas to nonresidents or upgrading visas that may allow individuals to achieve lawful permanent resident status. However, neither the TIGTA auditors nor the IRS know the extent of these background checks. In response to this finding, IRS agreed to form a task group to review the current standards. If IRS determines that the standards do not provide adequate protection or the risk is not reduced by other security measures, IRS will incorporate more stringent requirements into the LPGs after coordinating with FMS and the Department of the Treasury. The uncertainty of criminal histories of non-U.S. citizens hired by lockbox banks may lead to hiring of individuals with criminal histories which, in turn, increases the risk of theft of receipts or misuse of tax data. For instance, TIGTA auditors reported that evidence regarding the theft of checks from one lockbox site indicates the involvement of a crime ring from a foreign country in the negotiation of and possibly in the actual theft of taxpayer checks. 44 TIGTA 2002-30-180. Page 62 GAO-03-299 IRS Lockbox Banks Appendix II GAO Analysis of IRS’s and FMS’s August 1999 1040 Tax Payment Comparative Cost Benefit Study Appendx iI In August 1999, an IRS/FMS taskforce issued a study entitled 1040 Tax Payment Comparative Cost Benefit Study. The study estimated the costs and interest savings from processing Form 1040 tax receipts at IRS compared to lockbox banks using three different IRS scenarios. Table 3 shows the IRS/FMS taskforce results for all three IRS scenarios for each of 7 fiscal years (fiscal years 2001 through 2007) and overall. Table 3: IRS and Lockbox Banks Cost and Saving Estimates for the Federal Government, Fiscal Years 2001 through 2007 Dollars in thousands IRS scenarios Scenario I Scenario II Scenario III IRS processing costs $55,958 $59,597 $52,408 Lockbox bank processing costs (144,879) (144,879) (144,879) Interest float saving from lockbox banks 100,523 217,897 978,637 Net savings from using lockbox banks $11,603 $132,615 $886,166 Source: GAO analysis of 1999 IRS/FMS study. Note: Totals for net savings may not add due to rounding. All three IRS scenarios used the same estimated lockbox bank processing costs of $144.9 million. IRS interest float savings and processing costs varied because assumptions differed across scenarios. Scenario I estimated a 1.384 interest float savings while scenarios II and III used 3 days and 10.6 days, respectively. Processing costs for scenario I assumed additional processing equipment, additional staff, additional space, and a 10 percent increase in processing productivity. Scenario II assumptions were the same as I except for assuming no increase in processing productivity. Scenario III assumptions were the same as II except for assuming no additional processing equipment, staff, or space. We focused on IRS scenario I for further analysis because it was the one used to justify the decision to continue using lockbox banks to process tax receipts. Table 4 shows the IRS/FMS taskforce results for scenario I in more detail for each of 7 fiscal years (2001 through 2007). Page 63 GAO-03-299 IRS Lockbox Banks Appendix II GAO Analysis of IRS’s and FMS’s August 1999 1040 Tax Payment Comparative Cost Benefit Study Table 4: IRS and Lockbox Banks Cost and Saving Estimates Under IRS Scenario I for the Federal Government, Fiscal Years 2001 through 2007 Dollars in thousands Fiscal Year Cost category 2001 2002 2003 2004 2005 2006 2007 Total IRS costs Labora $5,289 $5,676 $6,210 $6,625 $7,056 $7,507 $7,978 $46,341 Basic supportb 292 308 331 348 365 383 400 $2,427 Equipment depreciation 1,318 1,318 1,318 1,318 1,318 0 0 $6,590 Site preparation depreciation 120 120 120 120 120 0 0 $600 Total IRS processing cost $7,018 $7,422 $7,980 $8,411 $8,859 $7,890 $8,378 $55,958 Lockbox bank costs Basic and standardc $10,671 $10,916 $11,584 $12,077 $12,579 $13,092 $13,616 $84,535 Ancillaryd 7,617 7,792 8,269 8,621 8,979 9,345 9,720 $60,344 Total lockbox bank cost $18,288 $18,709 $19,853 $20,698 $21,558 $22,437 $23,336 $144,879 Cost difference: Lockbox bank compared to IRS ($11,270) ($11,286) ($11,874) ($12,287) ($12,699) ($14,548) ($14,957) ($88,921) Interest saving difference: Lockbox bank compared to IRS $12,518 $13,074 $13,846 $14,407 $14,976 $15,556 $16,146 $100,523 Net saving $1,248 $1,787 $1,973 $2,120 $2,277 $1,008 $1,188 $11,603 Source: GAO analysis of 1999 IRS/FMS study. Note: Totals may not add due to rounding. a Labor includes costs to sort, process, and deposit tax receipts plus benefits, overhead, and inflation. b Basic support includes costs for service and supplies, equipment, and printing. c Basic and standard includes costs for depositing tax receipts in the bank. d Ancillary includes costs for other services, such as sorting and shipping tax returns to IRS. We analyzed the documented support for the data used to develop estimates in the study. The support often came from historical data on lockbox banks and IRS’s processing. We generally found some documented support on the methodology and assumptions used for the costs and revenue estimates. We could not compare support for the specific cost estimates, however, because the lockbox banks only had a basic charge for processing tax receipts and an additional charge to sort tax returns and ship them to IRS. Page 64 GAO-03-299 IRS Lockbox Banks Appendix II GAO Analysis of IRS’s and FMS’s August 1999 1040 Tax Payment Comparative Cost Benefit Study The estimation methodology and assumptions used in the study were the same for each year. To illustrate the methodology and assumptions, we reviewed how the estimates were developed for the first year—fiscal year 2001. For example, as shown in table 4, the net saving of $1.2 million is the difference between IRS and lockbox bank costs and IRS and lockbox bank interest savings. For costs, the study estimated that processing the tax receipts through lockbox banks would cost $11.3 million more than processing them through IRS. For interest savings, the study estimated that using lockbox banks would save $12.5 million more than using IRS. For cost estimates, a key factor was the estimated number of tax receipts, which was based on the actual number of 1998 tax receipts and projected for future years using expected growth rates. To understand IRS’s costs for fiscal year 2001, we analyzed its four components—labor costs, basic support costs, equipment depreciation, and site preparation depreciation. A discussion of each of the four cost components follows. IRS labor costs were often estimated from IRS’s Cost Estimate Reference guide that provides estimated costs for particular activities at IRS.45 The guide includes estimated labor cost and staff hours for processing tax returns.46 We traced each cost estimate in the study to the IRS cost guide. We also discussed the cost estimates with the IRS analyst who made and documented the computations for the study. Table 5 breaks down the IRS labor cost estimate for fiscal year 2001. 45 Internal Revenue Service, Cost Estimate Reference, Document 6746, revision November 1998. 46 We made no attempt to analyze IRS’s data collection systems that support the data in the cost guide. Page 65 GAO-03-299 IRS Lockbox Banks Appendix II GAO Analysis of IRS’s and FMS’s August 1999 1040 Tax Payment Comparative Cost Benefit Study Table 5: IRS’s Fiscal Year 2001 Labor Cost Rate per Staff Hourly Activity Volume hour hours staff pay Cost Sorting returns 10,093,903 103.9 97,150 $9.24 $897,666 Processing receipts 7,831,622a 171.4 45,692 $10.24 $467,886 Depositing receipts 10,093,903 267.9 37,678 $10.17 $383,185 Quality assuranceb $13.94 $80,531 c Overhead $13.81 $1,862,679 Benefitsd $886,067 e Labor inflation $710,651 Total labor cost $5,288,665 Source: GAO analysis of the 1999 study estimates. a Processing tax receipts involves a lower volume because the study estimated IRS could process 2,262,281 more receipts without a cost impact, assuming an estimated 10 percent productivity increase from using new processing equipment. b Quality assurance cost is computed by multiplying the staff hours for sorting, processing, and depositing (180,520 hours) by 3.2 percent and by $13.94 hourly staff pay (factors from IRS cost guide). c Overhead cost is computed by multiplying staff hours for sorting, processing, depositing, and quality assurance (186,297 hours) by 72.4 percent and by $13.81 hourly staff pay (factors from IRS cost guide). d Benefits are computed by adding the costs for the first five activities and multiplying by 24 percent (benefit rate from IRS cost guide). e Labor inflation is estimated from federal pay increases, with a compounded rate of 15.5 percent through fiscal year 2001 applied to the cost of the first six activities ($4,578,014). Explanations of the other three components in IRS’s processing cost estimates follow. • Basic support cost: $291,679 Consists of service and supplies, equipment, and printing on the basis of rates listed in the IRS cost guide. • Equipment depreciation: $1,317,796 IRS would need to spend an estimated $6,588,980 on hardware, furniture, and software if IRS processed Form 1040 tax receipts instead of lockbox banks. This cost was depreciated over a 5-year period in equal annual amounts. Page 66 GAO-03-299 IRS Lockbox Banks Appendix II GAO Analysis of IRS’s and FMS’s August 1999 1040 Tax Payment Comparative Cost Benefit Study • Site preparation depreciation: $120,000 IRS would need to spend $600,000—$300,000 at each of two IRS locations—to prepare space to accommodate new equipment required to process the increased volume of tax receipts. This cost was depreciated over a 5-year period in equal annual amounts. We also analyzed the added interest savings if lockbox banks processed the tax receipts instead of IRS. The IRS/FMS taskforce study followed a formula in Treasury regulations to compute this estimate. For fiscal year 2001, the factors in that formula included • total tax receipts = $45,224,421,259 divided by • total deposit days of 250 multiplied by • interest float of 1.384 days multiplied by • an estimated federal funds rate of 5 percent. The number of deposit days was specified in the Treasury regulations. The interest float represents how much faster lockbox banks could process tax receipts compared to IRS in three areas, totaling 1.384 days: • Mail float 0.035 days • Availability float 0.349 days • Compressing the program completion date (PCD) 1.000 day Mail float is measured from the time a taxpayer mails a payment until it arrives at a lockbox bank or IRS. Availability float is measured from the time a receipt is deposited until the funds are credited to the Treasury. The PCD is the day when lockbox banks must finish processing during peak workload periods and return to a schedule of depositing receipts within 24 hours. We examined the basis for each of these three estimates. Mail and availability float figures were taken from a July 1998 interest float study done by a contractor for FMS. The PCD figure came from an agreement by lockbox banks to compress PCD by 1 day while the study concluded that IRS could not match the compression for a number of reasons. A new interest float study would have to be done to know the actual float Page 67 GAO-03-299 IRS Lockbox Banks Appendix II GAO Analysis of IRS’s and FMS’s August 1999 1040 Tax Payment Comparative Cost Benefit Study advantage, if any, from using lockbox banks rather than IRS to process the tax receipts. Page 68 GAO-03-299 IRS Lockbox Banks Appendix III Comments from the Department of the Treasury Appendx Ii Note: GAO comments supplementing those in the report text appear at the end of this appendix. See comment 1. Page 69 GAO-03-299 IRS Lockbox Banks Appendix III Comments from the Department of the Treasury See comment 2. See comment 1. Page 70 GAO-03-299 IRS Lockbox Banks Appendix III Comments from the Department of the Treasury Page 71 GAO-03-299 IRS Lockbox Banks Appendix III Comments from the Department of the Treasury Page 72 GAO-03-299 IRS Lockbox Banks Appendix III Comments from the Department of the Treasury Page 73 GAO-03-299 IRS Lockbox Banks Appendix III Comments from the Department of the Treasury Page 74 GAO-03-299 IRS Lockbox Banks Appendix III Comments from the Department of the Treasury The following are GAO’s comments on the Department of the Treasury’s letter dated December 20, 2002. GAO Comments 1. See “Agency Comments and Our Evaluation” section. 2. IRS and FMS indicated the need for one technical clarification regarding our use of the terms “contracts,” “contractor,” and “contractual agreements” with respect to lockbox banks and recommended that we delete all references to “contracts” and “contractors.” IRS and FMS stated that when lockbox banks perform services for IRS, they act in a financial agent capacity on behalf of Treasury and that this function does not constitute a procurement or contract within the meaning of the Federal Property and Administrative Services Act or the Federal Acquisition Regulation (FAR). We recognize that the lockbox agreements are not procurements for purposes of the act or the FAR, but we did not change the language used in the report for ease of reference. It should be noted that Treasury also uses contract terminology in discussing lockbox agreements. Specifically, the Treasury Financial Manual gives FMS “the exclusive authority to contract for lockbox services with the selected bank and the agency” and further states that “an agency is prohibited from entering into new contractual agreements … without the prior approval of FMS.” In addition, in the IRM, IRS defines a lockbox depositary agreement as a “contractual agreement signed by IRS, FMS and the Lockbox that provides the requirements of the activities performed as the commercial depositories.” Our use of contract terminology in this report is consistent with Treasury’s use of such terminology in the TFM and the IRM. We did add a footnote (see footnote 2) to clarify that while the agreements with the lockbox banks are legally binding, they are not procurements subject to the provisions of the Federal Property and Administrative Services Act or the FAR, and to indicate that we use the terms “contracts” and “contractors” in the report for ease of reference. Page 75 GAO-03-299 IRS Lockbox Banks Appendix IV GAO Contacts and Staff Acknowledgments Appendx i IV GAO Contacts Steve Sebastian (202) 512-3406 Mike Brostek (202) 512-9110 Acknowledgments In addition to those named above, Larry Dandridge, Marshall Hamlett, Aaron Holling, Jeffrey Jacobson, Casey Keplinger, Laurie King, Delores Lee, Yola Lewis, Larry Malenich, Julia Matta, Tom Short, and Esther Tepper made key contributions to this report. (191024) Page 76 GAO-03-299 IRS Lockbox Banks GAO’s Mission The General Accounting Office, the investigative arm of Congress, exists to support Congress in meeting its constitutional responsibilities and to help improve the performance and accountability of the federal government for the American people. GAO examines the use of public funds; evaluates federal programs and policies; and provides analyses, recommendations, and other assistance to help Congress make informed oversight, policy, and funding decisions. GAO’s commitment to good government is reflected in its core values of accountability, integrity, and reliability. Obtaining Copies of The fastest and easiest way to obtain copies of GAO documents at no cost is through the Internet. GAO’s Web site (www.gao.gov) contains abstracts and full- GAO Reports and text files of current reports and testimony and an expanding archive of older products. The Web site features a search engine to help you locate documents Testimony using key words and phrases. You can print these documents in their entirety, including charts and other graphics. Each day, GAO issues a list of newly released reports, testimony, and correspondence. GAO posts this list, known as “Today’s Reports,” on its Web site daily. The list contains links to the full-text document files. To have GAO e-mail this list to you every afternoon, go to www.gao.gov and select “Subscribe to GAO Mailing Lists” under “Order GAO Products” heading. Order by Mail or Phone The first copy of each printed report is free. Additional copies are $2 each. A check or money order should be made out to the Superintendent of Documents. GAO also accepts VISA and Mastercard. Orders for 100 or more copies mailed to a single address are discounted 25 percent. Orders should be sent to: U.S. General Accounting Office 441 G Street NW, Room LM Washington, D.C. 20548 To order by Phone: Voice: (202) 512-6000 TDD: (202) 512-2537 Fax: (202) 512-6061 To Report Fraud, Contact: Web site: www.gao.gov/fraudnet/fraudnet.htm Waste, and Abuse in E-mail: email@example.com Federal Programs Automated answering system: (800) 424-5454 or (202) 512-7470 Public Affairs Jeff Nelligan, Managing Director, NelliganJ@gao.gov (202) 512-4800 U.S. General Accounting Office, 441 G Street NW, Room 7149 Washington, D.C. 20548 United States Presorted Standard General Accounting Office Postage & Fees Paid Washington, D.C. 20548-0001 GAO Permit No. GI00 Official Business Penalty for Private Use $300 Address Service Requested
IRS Lockbox Banks: More Effective Oversight, Stronger Controls, and Further Study of Costs and Benefits are Needed
Published by the Government Accountability Office on 2003-01-15.
Below is a raw (and likely hideous) rendition of the original report. (PDF)