This is the accessible text file for GAO report number GAO-03-356 entitled 'Tax Administration: Federal Payment Levy Program Measures, Performance, and Equity Can Be Improved' which was released on March 06, 2003. This text file was formatted by the U.S. General Accounting Office (GAO) to be accessible to users with visual impairments, as part of a longer term project to improve GAO products’ accessibility. Every attempt has been made to maintain the structural and data integrity of the original printed product. Accessibility features, such as text descriptions of tables, consecutively numbered footnotes placed at the end of the file, and the text of agency comment letters, are provided but may not exactly duplicate the presentation or format of the printed version. The portable document format (PDF) file is an exact electronic replica of the printed version. We welcome your feedback. Please E-mail your comments regarding the contents or accessibility features of this document to Webmaster@gao.gov. Report to Congressional Requesters: United States General Accounting Office: GAO: March 2003: Tax Administration: Federal Payment Levy Program Measures, Performance, and Equity Can Be Improved: GAO-03-356: GAO Highlights: Highlights of GAO-03-356, a report to the Committee on Ways and Means and Subcommittee on Oversight, House of Representatives. March 2003: Tax Administration: Federal Payment Levy Program Measures, Performance, and Equity Can Be Improved: Why GAO Did This Study: According to the Internal Revenue Service (IRS), taxpayers currently owe about $249 billion in delinquent taxes. At the same time, the government pays billions of dollars in Social Security, retirement, and other federal payments to thousands of these individuals. To help IRS administer tax laws fairly and collect delinquent taxes effectively, Congress included a provision authorizing the Federal Payment Levy Program, which allows IRS to continuously levy up to 15 percent of certain federal payments made to delinquent taxpayers. Because of congressional interest about whether the Federal Payment Levy Program is being implemented as intended, GAO was asked to assess how well the program is operating. What GAO Found: The Federal Payment Levy Program enables IRS to continuously levy (take a portion of) federal payments to individuals and businesses owing delinquent taxes. GAO has found the following: * IRS measures only about 27 percent of the revenues that can be attributed to the continuous levy program. IRS does not measure revenues that are received through voluntary payments as taxpayers respond to the notice of intent to levy or certain other results. Understating the program’s impact may hinder IRS in making well-founded decisions on program management and resource allocation. IRS plans to revise its measure of program results but has not yet decided how to do so. * IRS blocks many eligible delinquent accounts from being included in the Federal Payment Levy Program, thereby missing an opportunity to gather information on which debtors are receiving federal payments, that could be used to collect these delinquent taxes more efficiently. IRS recently unblocked some accounts and plans to unblock more, but has not established a time frame to complete these changes. * IRS uses an inaccurate income criterion of ability to pay when determining whether taxpayers receiving Social Security benefits can afford to have their benefits levied under the Federal Payment Levy Program. As a result, fair treatment of taxpayers is compromised because taxpayers with a similar ability to pay their delinquent taxes likely are treated differently. IRS recognizes that the criterion is flawed but continues to use it. Figure: [See PDF for image] [End of figure] What GAO Recommends: To help IRS improve the operation of the levy program, GAO recommends that IRS (1) include more complete data on the range of taxpayers’ actions and tax collections attributable to the program in its new measurement approach, (2) study the feasibility of submitting all delinquent accounts for matching against federal payments, and (3) discontinue use of the income criterion used to determine which Social Security beneficiaries can have their payments levied. IRS agreed to implement the first two recommendations and explore options in regard to the third. www.gao.gov/cgi-bin/getrpt?GAO-03-356. To view the full report, including the scope and methodology, click on the link above.For more information, contact Michael Brostek at (202) 512- 9110 or brostekm@gao.gov. Contents: Letter: Results in Brief: Background: IRS’s FPLP Data Understates Program Results: Blocking Cases from Matching against Federal Payment Records Prevents IRS from Collecting Taxes More Efficiently: IRS’s Criterion for Determining Which Social Security Beneficiaries Can Afford to Have Their Payments Levied Is an Inaccurate Indicator of Ability to Pay: Conclusions: Recommendations for Executive Action: Agency Comments and Our Evaluation: Appendix I: Objectives, Scope, and Methodology: Scope and Methodology: Appendix II: Additional Data on Revenue Collections Attributable to FPLP and Status of Taxpayer Accounts: FPLP Payments by Collection Method: Categorization of Taxpayers: Most Recently Filed Income Tax Returns for Social Security Beneficiaries: Appendix III: Comments from the Internal Revenue Service: Appendix IV: GAO Contacts and Staff Acknowledgments: GAO Contacts: Acknowledgments: Tables: Table 1: Number of Delinquent Taxpayers Receiving Federal Payments out of a Total 1.5 Million Blocked from FPLP: Table 2: FPLP Payments by Collection Method as of August 2002 for Taxpayers Who Received a Notice of Intent to Levy during October- December 2001: Table 3: Categorization of Taxpayers by Delinquent Account Status as of August 2002: Table 4: Distribution of Social Security Payment Recipients by Year of Most Recent Income Tax Return Filed: Figure: Figure 1: Tax Collections Attributable to FPLP as of August 2002 for Taxpayers Receiving a Notice of Intent to Levy during October-December 2001: Abbreviations: ACS: Automated Collection System: FMS: Financial Management Service: FPLP: Federal Payment Levy Program: IRS: Internal Revenue Service: OPM: Office of Personnel Management: TPI: Total Positive Income: 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. Letter: March 6, 2003: The Honorable Bill Thomas Chairman, Committee on Ways and Means House of Representatives: The Honorable Amo Houghton Chairman, Subcommittee on Oversight Committee on Ways and Means House of Representatives: According to the Internal Revenue Service (IRS), taxpayers currently owe about $249 billion in delinquent taxes to the federal government.[Footnote 1] At the same time, the government pays billions of dollars in Social Security, retirement, and other federal payments to thousands of these individuals each year. IRS and federal payment records indicate that nearly one million taxpayers owing about $26 billion in delinquent taxes as of February 2002 were receiving federal payments for federal wages and retirement, Social Security benefits, and goods and services provided to federal agencies. To help IRS collect these delinquent taxes more effectively, Congress included a provision in the Taxpayer Relief Act of 1997 authorizing the establishment of the Federal Payment Levy Program (FPLP), which allows IRS to continuously levy[Footnote 2] up to 15 percent of certain federal payments made to delinquent taxpayers. According to IRS, the program, which began in July 2000, resulted in collecting over $60 million in fiscal year 2002 by directly levying federal payments. Under FPLP, IRS matches its accounts receivable records with federal payment records maintained by the Department of the Treasury’s Financial Management Service (FMS), such as certain Social Security benefit and federal wage records. When the records match, the delinquent taxpayer is sent a notice of intent to levy the payment, which generally gives the taxpayer at least 30 days to either make arrangements to pay the tax debt or provide a reason as to why the payments should not be levied, such as financial hardship. If taxpayers do not respond after 30 days, IRS can instruct FMS to levy their federal payments. Subsequent payments are continuously levied until such time that the tax debt is paid or IRS releases the levy. To determine whether FPLP was being implemented as intended and that the program helps IRS meet its strategic goal of treating all taxpayers fairly, you asked us to assess how well the program is operating. Specifically, our objectives were to determine (1) whether the data IRS uses to manage the program adequately measures program results, (2) how IRS’s decision to block some delinquent accounts from being matched to federal payments under FPLP impacts the agency’s ability to collect taxes efficiently, and (3) whether the criterion IRS uses to include taxpayers receiving Social Security benefit payments into FPLP effectively targets taxpayers who can afford to pay their tax debts. To address our objectives, we: * reviewed documents IRS uses to measure program results and select cases that qualify for FPLP, * discussed program operations with program officials, * analyzed sample cases of delinquent taxpayers who were sent notices of intent to levy during the period October through December 2001, and * analyzed IRS’s accounts receivable and FMS’s payment files on delinquent taxpayer accounts that were not included in the program. The estimates of the sample cases we took of taxpayers receiving payments from the Social Security Administration have some sampling errors associated with them. Further, estimates about the total group of cases we analyzed (taxpayers receiving retirement payments from the Office of Personnel Management (OPM); vendor payments; and taxpayers receiving Social Security payments) have sampling errors as well. All percentage estimates about the population of taxpayers receiving Social Security payments or the overall population have sampling errors of plus or minus 5 percentage points or less, unless otherwise noted. Our work was done between November 2001 and December 2002 in accordance with generally accepted government auditing standards. (App. I describes our overall objectives, scope, and methodology.): Results in Brief: The data that IRS uses to manage FPLP does not adequately measure a full range of results. The agency’s sole measure of results consists of the amount of revenues collected directly through the continuous levy of federal payments. However, we estimate that direct levy collections account for about 27 percent of the revenues that can be attributed to FPLP, while the remaining 73 percent of FPLP delinquent tax revenue is collected through nonlevy payments, as taxpayers respond to the levy notice by making voluntary payments. IRS is also not measuring the extent that FPLP helps IRS collections to function more efficiently. On the basis of our analysis, we estimate that about 29 percent of the taxpayers who received a notice of intent to levy from FPLP responded by taking action that enabled IRS to remove them from active accounts receivable, thus freeing up IRS resources to pursue other collections. Without full performance management information on collections and efficiency results, IRS can significantly understate the program’s impact and may be hindered in its ability to make well-founded decisions on program management and resource allocation. Acknowledging that FPLP generates indirect results in addition to tax collections made through the continuous levy process, IRS has initiated its own study on how to measure indirect outcomes. This study is expected to be completed in calendar year 2003, at which point IRS will decide how to revise its measurement approach. When FPLP began in July 2000, IRS blocked certain delinquent taxpayers from being identified as receiving federal payments, thereby missing an opportunity to use this information to collect delinquent taxes more efficiently. IRS officials imposed the blocks because of concerns that the potential volume of levies--some 1.4 million taxpayer accounts-- would disrupt ongoing collection activities and likely could not be handled with existing resources. We estimate that only a small fraction of delinquent taxpayers, about 112,000, would actually qualify for levy. However, these 112,000 delinquent taxpayers were collectively receiving about $6.7 billion in federal payments and owed about $1.5 billion in delinquent taxes. If these taxpayer accounts had been matched against federal payment records, IRS would have more information available to determine whether FPLP, or other collection activities, would be more efficient for collection of delinquent taxes for these accounts. In January 2003, IRS unblocked and began matching those delinquent taxpayer accounts identified as receiving either a federal salary or annuity payment. IRS officials plan to unblock a portion of the remaining delinquent accounts sometime in 2005, although they have not yet established a firm time frame for doing so. The FPLP income criterion IRS uses to decide whether taxpayers can afford to have their Social Security benefit payments levied is an inaccurate indicator of ability to pay. The criterion, implemented in January 2002, was intended to identify and exclude from levy the benefit payments of those Social Security beneficiaries who are least able to pay their taxes. However, our analysis of taxpayer behavior for two groups, Social Security benefit cases above and below the income threshold criterion, showed that both responded similarly to a notice of levy by arranging to make some kind of voluntary payment arrangement. In addition, both groups were almost equally likely to rely on Social Security as their sole source of income, suggesting both may experience financial hardship at the same rate. Furthermore, we also found that the income criterion is based on information that is often outdated and incomplete. Because of the inaccuracy of the income criterion, taxpayers with similar abilities to pay their delinquent taxes are likely to be treated differently under FPLP, which conflicts with IRS’s goal of treating taxpayers fairly. Although IRS has also determined that the income criterion is flawed, it continues to use it to identify which Social Security benefit payments will and will not be available for levy under FPLP. We are making several recommendations to help improve FPLP measures of program results, program performance, and to ensure equitable treatment of taxpayers subject to levy. IRS generally agreed with two of our recommendations and agreed to explore options in regard to the third. Background: In the Taxpayer Relief Act of 1997, Congress authorized IRS to collect delinquent tax debt by continuously levying up to 15 percent of certain federal payments made to delinquent taxpayers. In July 2000, IRS first implemented the continuous levy program, now referred to as FPLP. We estimated in prior reviews that once fully operational, the program could potentially recover hundreds of millions of dollars in delinquent tax debt[Footnote 3] annually. In fiscal year 2002 IRS collected more than $60 million in delinquent taxes through continuous levy under FPLP. IRS operates the program with FMS, the agency that receives payment records from and makes payments on behalf of most federal agencies, including federal retirement payments and Social Security benefit and vendor payments.[Footnote 4] With respect to FPLP, FMS compares each taxpayer’s identification number (TIN) and name on agency payment records with the TIN and name control on accounts receivable records provided by IRS. When FMS identifies a delinquent taxpayer scheduled to receive a federal payment, it informs IRS, which then issues a notice of intent to levy to the taxpayer, unless the notice was previously sent. Once taxpayers receive the notice of impending levy, they have several options for action.[Footnote 5] Taxpayers who receive a notice from FPLP have a minimum of 30 days to respond to the notice,[Footnote 6] during which time they may consider several alternatives available to them. Taxpayers may either: * disagree with IRS’s assessment and collection of tax liability, in which case they can appeal the action by requesting a hearing with IRS’s Office of Appeals; * elect to pay off the debt in full; * negotiate with IRS to set up an alternative payment arrangement such as an installment agreement or an offer in compromise;[Footnote 7] or * apply to IRS for a hardship determination, whereby taxpayers demonstrate to IRS that making any payments at all would result in a significant financial hardship. In such cases, IRS may agree to temporarily delay collection action until the taxpayer’s financial condition improves. If taxpayers do not respond to IRS and avail themselves of options within the notification period, IRS will instruct FMS to proceed with the continuous levy by reducing each scheduled payment to the taxpayer by 15 percent--or the exact amount of tax owed if it is less than 15 percent of the payment--until the tax debt is satisfied. The FPLP began with levies of federal employees’ retirement payments and vendor payments issued by FMS. IRS later added additional types of federal payments to the program, including selected Social Security benefits and selected federal salaries. IRS plans to continue expanding the program by adding additional federal employee salaries and other types of federal payments. Not all delinquent taxpayer accounts are eligible for FPLP. IRS has excluded certain accounts from levy; for example, cases where the taxpayer has entered bankruptcy, made alternative arrangements to pay, or demonstrated to IRS that making payments on the outstanding tax debt would result in a financial hardship. With some exceptions, delinquent taxpayer accounts are eligible for FPLP if they are either assigned to IRS’s Automated Collection System (ACS) or to field collections or have been waiting for assignment to one of these areas for at least 1 year.[Footnote 8] Cases in which collection activity has been deferred for at least a year because of low tax liability, as well as cases in which IRS has been unable to either locate or contact the taxpayer are also eligible for FPLP. IRS’s FPLP Data Understates Program Results: IRS tracks FPLP program results by measuring only tax revenue collected through continuous levy, although most of the collections attributable to FPLP result from taxpayers subsequently contacting IRS and either submitting a payment voluntarily,[Footnote 9] or arranging to pay their delinquent taxes through other means such as by entering into an installment agreement. Our analysis of about 98,000 delinquent taxpayers who received a notice of intent to levy during the October through December 2001 period showed that as of August 2002, IRS had collected an estimated $107.1 million from these taxpayers. As shown in figure 1, levies represented an estimated $28.5 million[Footnote 10] of this amount, or 27 percent of the total collections, while other payment methods represented an estimated $78.6 million,[Footnote 11] or 73 percent of the total FPLP-related collections.[Footnote 12] By not measuring the nonlevy payments attributable to FPLP, IRS significantly underestimates the program’s success. Figure 1: Tax Collections Attributable to FPLP as of August 2002 for Taxpayers Receiving a Notice of Intent to Levy during October-December 2001: [See PDF for image] [End of figure] (See app. II for more detailed information on the amount of delinquent taxes collected as a result of FPLP by the type of federal payment taxpayers received (i.e., Social Security benefits, federal retirement payments, and vendor payments) and by the method of collection.): IRS realizes other benefits attributable to FPLP that it does not currently measure. IRS does not measure the extent that FPLP helps IRS function more efficiently by decreasing its accounts receivable inventory. Our analysis of the 98,000 FPLP cases indicated that after receiving a notice of intent to levy, about 29 percent of the taxpayers took action that enabled IRS to remove them from the active accounts receivable inventory or to move them to an inactive status. Specifically, we estimate that subsequent to receiving a levy notice, about 19 percent of the taxpayers resolved their liability while about 10 percent obtained a determination of financial hardship. (For more information on the characterization of account status, see app. II.) By reclassifying some active delinquent accounts to an inactive status and removing others, FPLP helps IRS to more efficiently prioritize its accounts receivable inventory and enables IRS to focus more of its resources on delinquent accounts that have more collection potential. Knowing the full impact of FPLP’s effectiveness would be consistent with IRS’s strategic planning and budgeting process, which emphasizes the importance of assessing the impact of current program operations to efficiently allocate resources. IRS acknowledges that FPLP generates indirect results in addition to revenues collected through the continuous levy process, and has initiated its own study on how to measure the outcomes attributable to the program, which it expects to complete in calendar year 2003. After completing this study, IRS will decide how to revise its measurement approach. Blocking Cases from Matching against Federal Payment Records Prevents IRS from Collecting Taxes More Efficiently: When IRS implemented FPLP in July 2000, it made a decision to temporarily block most delinquent taxpayer accounts placed in ACS and field collections from being matched against federal payment records. ACS accounts were blocked primarily because IRS believed it lacked the resources to issue levy notices and respond to the potential increase in telephone calls from taxpayers responding to the notices and still adequately perform other ACS activities. Specifically, agency officials were concerned that if the 1.4 million delinquent taxpayers with accounts in ACS were added to the program for matching too rapidly, it could disrupt ACS workload processes and likely could not be handled with existing resources. IRS also decided not to match over 55,000 delinquent accounts in field collection because revenue officers believed that this action could interfere with their successfully contacting taxpayers and negotiating a settlement to resolve the delinquent account they had been assigned. We found some of IRS officials’ concerns with respect to matching the nearly 1.5 million ACS and field collection accounts may be unfounded. We matched the nearly 1.5 million accounts to FMS payment records[Footnote 13] and found that only about 112,000 taxpayers were receiving federal payments, as shown in table 1. Table 1: Number of Delinquent Taxpayers Receiving Federal Payments out of a Total 1.5 Million Blocked from FPLP: [See PDF for image] Source: GAO analysis of IRS and FMS data. [End of table] Submitting delinquent taxpayer accounts to FMS for matching against federal payment records does not necessarily mean IRS must levy these accounts. Rather, the matching process performed by FMS would provide IRS with useful additional information to assist in determining what the best collection method may be. For example, of the 108,469 delinquent taxpayers in ACS receiving federal payments at the time of our analysis, IRS had not yet contacted or located more than half-- about 55,900--of these delinquent taxpayers, and could therefore have chosen to levy some of these accounts without disrupting ongoing collections. These 55,900 taxpayers owed about $460 million in delinquent taxes and received $4.1 billion in federal payments during 2002. If IRS had information on matching federal payments for these delinquent accounts, it would have had additional options available to determine how best to pursue collection of the delinquent tax revenue--such as using FPLP. While the ACS and field collection actions may result in the eventual recovery of most of the delinquent taxes associated with these accounts, using the automated FPLP matching process could help IRS collect this revenue in a more timely and efficient manner. Even when ACS or field revenue officers have already contacted the taxpayer, matching--but not necessarily issuing a notice of intent to levy--the account could provide useful information for IRS to consider in collections. For example, revenue officers may not be aware that taxpayers in their case inventory are currently receiving federal payments, and they could use this information to develop a more complete assessment of the taxpayer’s financial situation. In January 2003, IRS unblocked and made available for matching and levy those delinquent accounts identified as receiving federal salary or annuity payments--representing about 20,000 of the 112,000 blocked taxpayers we identified in our analysis. However, other delinquent accounts remain blocked from being matched to FMS payment records. Agency officials said that they plan to unblock a portion of the remaining delinquent accounts sometime in 2005, although they have not yet established a firm time frame for doing so. IRS’s Criterion for Determining Which Social Security Beneficiaries Can Afford to Have Their Payments Levied Is an Inaccurate Indicator of Ability to Pay: IRS established an income threshold to exclude Social Security beneficiaries who cannot afford to pay their taxes from FPLP. However, our analysis of Social Security beneficiaries above and below this income threshold shows that IRS’s income criterion is an inaccurate indicator of a taxpayer’s ability to pay. We found little difference between the two groups in terms of the frequency with which taxpayers either made voluntary payments in response to a levy notice or relied on Social Security as a sole source of income. We also found that the income criterion relies on information that is often outdated and incomplete. IRS Decided to Exclude Selected Social Security Payments from FPLP without Evaluating the Effectiveness of Its Income Level Criterion: In response to concerns raised by the National Taxpayer Advocate regarding potential harm that may be experienced by Social Security beneficiaries who are levied under FPLP, IRS implemented an income criterion for Social Security benefit payments intended to identify and screen from FPLP, those taxpayers who are least able to pay their tax debt. The National Taxpayer Advocate believed that taxpayers who rely solely on Social Security benefits as their income source are most vulnerable to the financial hardship that a continuous levy may cause. In January 2002, IRS implemented an income threshold that excluded Social Security benefit payments from FPLP for 55 percent of the delinquent taxpayers who receive these payments.[Footnote 14] This income level criterion, Total Positive Income (TPI),[Footnote 15] is derived from income information reflected on the most recent income tax return filed by the taxpayer. Under FPLP, the TPI criterion only applies to Social Security benefit payments and not to other forms of payments such as federal annuities or salary payments. We estimate that taxpayers who were receiving Social Security payments and whose TPI was below the income threshold owed approximately $522 million in delinquent taxes.[Footnote 16] Prior to introducing Social Security benefit payments into FPLP, IRS implemented the TPI criterion without doing any tests to indicate whether the TPI criterion was necessary or better than the special procedures IRS had already planned-providing Social Security beneficiaries with a longer notification period relative to recipients of other federal payments. Under the normal notification process, IRS gives delinquent taxpayers 30 days from receipt of a notice of impending levy before it begins to levy the payment. Under FPLP, IRS planned to provide Social Security beneficiaries with a second notice of intent to levy and an additional 30 days to respond. The extended notification period was adopted and is intended to provide Social Security beneficiaries with sufficient opportunity to contact IRS with any questions concerning the levy notice, and, if necessary, demonstrate to IRS that the levy would result in financial hardship. IRS officials said that due to time constraints IRS did not test the extent to which the planned extended two-notice process would have proved sufficient to ensure that Social Security beneficiaries were not subjected to undue financial hardship under FPLP. In addition, IRS did not do any studies to determine what TPI level would best protect financially vulnerable Social Security beneficiaries, or whether the criterion should apply to other federal payments as well. The TPI Criterion Has Several Weaknesses: The TPI criterion has several weaknesses that can impact Social Security beneficiaries both above and below the income threshold. Our analysis has shown that TPI is an inaccurate indicator of those delinquent taxpayers who are the most financially vulnerable. Under current program procedures, IRS does not send either one of the two routine levy notices to Social Security beneficiaries who are below the TPI threshold. However, while phasing Social Security benefit payments into FPLP between October and December of 2001, IRS issued the initial notice of intent to levy to all delinquent Social Security beneficiaries who owed taxes, including those whose TPI was below the threshold and whose benefit payments would therefore be excluded from the program. As a result, we were able to compare data on how taxpayers in both the “able to pay” and “unable to pay” group responded to the levy notice. We found that of the approximately 90,000 Social Security beneficiaries in the below TPI threshold group, an estimated 18 percent made voluntary payments or entered into an installment agreement of their own accord after receiving the one-time notice of intent to levy at the end of 2001. Of the approximately 97,000 Social Security beneficiaries whose TPI was above the threshold, an estimated 12 percent had made voluntary payments or entered into an installment agreement.[Footnote 17] The 6 percent difference between the 12 and 18 percent figures[Footnote 18] is too small to be statistically significant; thus, Social Security beneficiaries above and below the TPI threshold made voluntary payments and entered into installment agreements at comparable rates. While the TPI threshold’s use has categorized certain taxpayers as those whose Social Security payments need to be excluded from FPLP because they are unable to pay, their actions in response to one levy notice demonstrates some ability to pay and, in fact, are similar to the actions of taxpayers above the TPI threshold.[Footnote 19] Our study also showed that IRS granted financial hardship status to Social Security beneficiaries above and below the TPI threshold at similar rates. As part of the regular collections process, any delinquent taxpayer has the right to ask IRS for a hardship determination in which the taxpayer claims that he or she is unable to pay their taxes without incurring undue financial hardship. As part of the determination process, the taxpayer may be required to provide IRS with financial information to substantiate his or her financial condition, and if IRS agrees that paying taxes would constitute a hardship, IRS will temporarily delay collection activity until the taxpayer’s financial condition improves. An estimated 5 percent of Social Security beneficiaries whose TPI is below the threshold and 8 percent of Social Security beneficiaries whose TPI is above the threshold responded to the notice by contacting IRS and obtaining a determination of financial hardship by using IRS’s standard hardship determination process. The difference between the 5 and 8 percent figures[Footnote 20] is too small to be statistically significant; thus, Social Security beneficiaries above and below the TPI threshold entered in hardship status at comparable rates. Furthermore, our analysis also determined that the TPI threshold poorly identifies taxpayers who solely rely on Social Security benefits--the group of taxpayers the National Taxpayer Advocate considered to be most vulnerable to financial hardship from continuous levy. We analyzed information returns data[Footnote 21] for tax year 2001 for a representative sample of Social Security beneficiaries with a TPI below the threshold and reviewed their current income sources. We found that an estimated 46 percent of these Social Security beneficiaries received only Social Security benefits and the remaining 54 percent received income in addition to Social Security benefit payments. These numbers were very close to the numbers we found for Social Security beneficiaries with a TPI above the threshold and, thus, presumably able to afford paying their taxes. About 40 percent of the above TPI group also relied on Social Security benefits as their sole source of income. The difference between the 40 and 46 percent figures[Footnote 22] is not statistically significant, indicating that a Social Security beneficiary under the TPI threshold is almost equally likely to rely solely on Social Security benefit payments as one deemed able to pay. Our analysis also indicated that TPI is frequently outdated because it is based on income information reflected on the most recent income tax return filed by the taxpayer. However, if the taxpayer was not required to file a tax return in recent years,[Footnote 23] his or her TPI information from that last return filed may not be consistent with their current financial situation. For example, a person who had a relatively high TPI according to his or her last income tax return may now be receiving significantly less income, in which case TPI is not an accurate indicator to determine whether they should now be included or excluded from FPLP. We reviewed the currency of the last tax return of Social Security beneficiaries who owed delinquent taxes and estimate that only 17 percent had filed a return for tax year 2001. We found that in total, 53 percent of the Social Security beneficiaries had not filed since tax year 1996 or earlier. (See app. II for more information on the results of this review.): In addition to being frequently outdated, TPI is an incomplete indicator of a person’s full resources. While the TPI calculation does include income earned from assets such as interest and capital gains, it does not include information on a taxpayer’s assets, such as savings account balances, stocks, property ownership, or individual retirement account balances when determining a taxpayer’s ability to pay. Using our review of data for tax year 2001 information returns, we were able to estimate that over 14 percent of Social Security beneficiaries who had a TPI below the threshold made mortgage interest payments, yet were excluded from FPLP. During the regular collections process in which IRS works directly with taxpayers to resolve their delinquent accounts, information on assets is taken into account when assessing the taxpayers’ overall financial condition. Both the National Taxpayer Advocate and the FPLP officials we spoke with acknowledged that the TPI criterion is flawed. FPLP officials, in their own study on TPI implementation, agreed that the TPI criterion had several problems. Similar to our findings, they reported that TPI is often out of date because the majority of last tax returns filed were at least 1 year out of date. IRS also found that TPI does not include consideration of filing status and dependent information, and is not adjusted when changes on the return are made resulting from an audit or amendment. However, IRS has continued to use TPI because it has not identified suitable alternatives, and has not taken time to determine whether the extended notification period effectively meets the needs of the Social Security beneficiary population. Nor has it identified whether Social Security beneficiaries are more vulnerable to financial hardship than other federal payment recipient groups. IRS’s continued use of a criterion that is inadequate in identifying those taxpayers who may be least able to pay, and therefore treats taxpayers with similar abilities to pay differently, is at odds with IRS’s strategic goal emphasizing the importance of treating all taxpayers fairly. Conclusions: IRS’s FPLP is still a relatively new program, and, as such, has not yet realized its full potential. To ensure that it does, IRS needs sufficient information to determine the results the program achieves so that well informed decisions can be made in allocating resources to FPLP. However, when measuring FPLP’s results, IRS only considers the program’s most immediate effect, the tax revenue collected directly by continuous levy, and thereby substantially understates the program’s results. Although IRS recognizes that broader measures of FPLP results are needed, its study to determine how to do so will not be completed until calendar year 2003, at which time IRS will then decide how to revise its measurement approach. FPLP’s full potential also has not been tapped because IRS has not taken full advantage of the information that could be used to determine the most efficient means of collecting delinquent taxes. Rather than analyzing the workload implications of matching additional accounts under the program, IRS blocked certain delinquent accounts from the program on the assumption that including them would create an overwhelming increase in workload. In fact, only a small portion of the blocked delinquent accounts--about 112,000 of the nearly 1.5 million-- matched against federal payments. In January 2003, IRS removed the block on some delinquent accounts but it has not set a firm time frame for unblocking the remaining accounts. Removing the block on the remaining accounts does not mean IRS would necessarily have to levy the accounts. Rather, it can use the information gained from matching the accounts to federal payment records to help identify the most efficient means of resolving the accounts, which may include levying some portion of them. Finally, use of an income based criterion to identify whether Social Security benefit payments should be excluded from the program has likely resulted in unequal treatment of similarly situated taxpayers, which conflicts with IRS’s strategic goal of treating all taxpayers fairly. IRS adopted this criterion (1) without testing its effectiveness and (2) without determining if excluding any beneficiaries from FPLP was even a necessary step, given measures that IRS had already taken to address possible hardship cases. While FPLP officials concur that it has many weaknesses, IRS continues using the criterion to exclude selected Social Security benefit payments from FPLP. Recommendations for Executive Action: To help ensure that IRS is operating FPLP in a manner that achieves the program’s full potential and ensures equitable treatment of taxpayers, we recommend that the Acting Commissioner of Internal Revenue: * Include in IRS’s planned new approach to measuring FPLP results data on the full range of taxpayers’ actions and tax collections attributable to FPLP, including nonlevy collections and account resolutions. * Study the feasibility of submitting all eligible delinquent accounts to FMS on an ongoing basis for matching against federal payment records under FPLP, and use information from any matches to assist IRS in determining the most efficient method of collecting delinquent taxes, including whether to use FPLP. * Discontinue using the TPI criterion as an indicator of Social Security beneficiaries’ ability to pay delinquent taxes and rely on the extended two-notice process to identify beneficiaries for whom a levy would be a hardship. Determine whether sending a second notice that explains the financial hardship exception to all Social Security beneficiaries subject to levy is sufficient to identify hardship situations. If not, develop and test a criterion that reliably identifies those Social Security beneficiaries for whom a levy would represent an undue hardship. Agency Comments and Our Evaluation: We received written comments on a draft of this report from the Acting Commissioner of Internal Revenue (see app. III). The Acting Commissioner generally agreed with our recommendations and provided technical comments and clarifications that we have incorporated throughout this report where appropriate. To enhance IRS’s ability to measure the full range of direct and indirect results of FPLP operation, the Acting Commissioner agreed to include in the agency’s planned new approach to measuring program results, data on nonlevy collections and account resolutions. The Acting Commissioner said that, as IRS gets closer to implementing its new measurement approach, it would like to share its methodology with us. However, the Acting Commissioner raised concerns with the part of our recommendation calling for IRS to consider account resolutions when measuring FPLP results. He said that contrary to our draft report’s statement that FPLP frees up resources through account resolutions that can be used to pursue accounts with more collection potential, FPLP has generated an increased workload for IRS staff that diverts them from more productive uses. He concluded that it is difficult to assess the total costs and/or indirect benefits of the program in terms of resources freed up to pursue other collection activity. We agree that doing such an assessment is challenging. However, IRS managers’ ability to accurately gauge FPLP effectiveness, as well as validate resource allocation decisions among various IRS collection activities should be based, to the extent practical, on data that accurately and completely reflect program results. While we recognize that IRS does have workload that comes from FPLP cases, and that workload may be lower priority in some cases than other alternative casework, FPLP also leads to many case closures that require little IRS employee time. In addition, since all cases subject to the FPLP have already received notices from IRS about their delinquent accounts, IRS had previously judged the accounts as meriting collection action even if they might involve increased collection staff workload. In our draft report, we had recommended that IRS submit all delinquent accounts to FMS on an ongoing basis for matching against federal payment records under FPLP and use information from any matches to assist IRS in determining the most efficient method of collecting delinquent taxes, including whether to use the FPLP to do so. While reviewing the draft report, IRS officials raised concerns, which they did not express in our previous discussions with them, that computer programming costs associated with implementing this draft recommendation might make the recommendation infeasible for certain types of taxpayer accounts. In light of those concerns, we agreed to modify our recommendation to instead recommend that IRS study the feasibility of submitting all delinquent accounts for matching. In his letter, the Acting Commissioner agrees with this revised recommendation and indicates that efforts to do so are underway. In reference to our recommendation that IRS discontinue use of the TPI criterion, the Acting Commissioner stated IRS’s commitment to ensuring all taxpayers are treated fairly and that IRS is concerned with the issues we raised regarding the TPI criterion. The Acting Commissioner did not agree to discontinue the use of the TPI immediately, but said that IRS would take the next 120 days to work with the National Taxpayer Advocate and program administrators to assess deficiencies in the current process and to develop a suitable solution. This action generally is responsive to our recommendation. However, if developing and implementing a suitable solution extends beyond the 120-day milestone set by the Acting Commissioner, we believe IRS should, as we recommended, discontinue use of the TPI criterion and rely on the two- notice procedure already in place while IRS continues to work on a solution it believes would be more suitable. As our report notes, IRS has been aware of some limitations with TPI for some time and has yet to identify a suitable solution. We will send copies to the Ranking Minority Member, House Committee on Ways and Means; Ranking Minority Member, Subcommittee on Oversight, House Committee on Ways and Means; and the Chairman and Ranking Minority Member, Senate Committee on Finance. We will also send copies to the Acting Commissioner of Internal Revenue and other interested parties. Copies of this report will also be made available to others upon request. The report will also be available on GAO’s Web site at http://www.gao.gov. If you have any questions concerning this report, please contact me at (202) 512-9110 or Ralph Block at (415) 904-2150. Key contributors to this work are listed in appendix IV. Michael Brostek Director, Tax Issues: Signed by Michael Brostek: [End of section] Appendix I: Objectives, Scope, and Methodology: Our objectives were to determine (1) whether the data the Internal Revenue Service (IRS) uses to manage the Federal Payment Levy Program (FPLP) adequately measures program results, (2) how IRS’s decision to block some delinquent accounts from being matched to federal payments under FPLP impacts the agency’s ability to collect taxes efficiently, and (3) whether the criterion IRS uses to include taxpayers receiving Social Security Administration benefit payments in FPLP effectively targets taxpayers who can afford to pay their tax debts. Scope and Methodology: We obtained and reviewed the data used by IRS to track program results in an effort to determine whether IRS’s data on FPLP operations adequately measures program results, including direct levy collections and other nonlevy collections such as those made through voluntary payments, installment agreements, and offers in compromise that occur in response to FPLP collection notices. We also selected and analyzed three groups of delinquent taxpayers that received a notice of intent to levy between October and December 2001. These groups were (1) all taxpayers receiving federal retirement payments from the Office of Personnel Management (OPM), (2) all taxpayers receiving vendor payments from federal agencies whose payments are processed by the Financial Management Service (FMS), and (3) a random sample of taxpayers receiving Social Security benefit payments. We analyzed a total of 1,540 taxpayers that were comprised of (1) 699 delinquent taxpayers receiving OPM payments, (2) 484 delinquent vendors receiving federal payments, and (3) a random sample of 357 delinquent taxpayers who were receiving Social Security payments. Since OPM and vendor payments have been eligible for FPLP matching since program inception, we analyzed only the new matches that occurred during the selected time period for these populations. However, because IRS first started matching all Social Security payments during the October-December 2001 time period, we randomly selected a sample of 357 taxpayers whose total positive income (TPI) was above the income threshold and who were receiving Social Security benefit payments. We weighted our observations on those 357 sampled cases in order to project to the total population of 97,133 taxpayers in this category.[Footnote 24] Using electronic taxpayer information files provided by IRS, we analyzed subsequent transactional activity that occurred on the taxpayer accounts reviewed between the initial match in the fourth quarter of 2001 and as of August 26, 2002. We also analyzed the level and type of account activity that occurred prior to inclusion in FPLP, including the elapsed time since the last significant action initiated by either IRS or the taxpayer, to determine whether the account activity that occurred after IRS issued a notice of intent to levy could be attributed to the continuous levy program. We attributed voluntary tax payments to inclusion in FPLP in only those cases where the delinquent account was not in any other collection status, that is, field collection or in the Automated Collection System (ACS) inventory. We discussed the range of impacts that may result after a taxpayer receives a notice of intent to levy under FPLP with program officials, which they agreed go beyond a continuous levy payment. We interviewed IRS officials in FPLP, ACS, and field collection areas to determine whether some of the taxpayer accounts that are currently blocked could be subject to FPLP, including determining the reason for blocking these accounts as well as the likelihood of unblocking them in the future. To estimate the number of blocked delinquent accounts that would match federal payment records if IRS were to introduce accounts that are blocked from being included in the program, we obtained and matched IRS’s accounts receivable records as of February 2002 with agency payment records obtained from FMS.[Footnote 25] To determine whether the TPI criterion IRS uses to levy Social Security payments is effectively targeting taxpayers who can afford to pay their tax debts, we performed additional analysis on the random sample of 357 delinquent taxpayers receiving Social Security payments included in the first of our study objectives. In addition, we analyzed a random sample of 405 delinquent taxpayers who received Social Security payments and who were under the TPI threshold. We weighted our observations on the 405 cases to project to the population of 90,307 taxpayers in this category. We reviewed income data from IRS’s taxpayer information files and information returns database to determine (1) each taxpayer’s reliance on Social Security benefits as a sole source of income, (2) whether taxpayers below and above the TPI threshold differed in their reaction to receiving a notice of intent to levy under FPLP, and (3) whether the TPI criterion accurately reflected each taxpayer’s current income level. We also met with the National Taxpayer Advocate as well as IRS program officials to discuss why the TPI criterion is used to screen Social Security payments that are to be included in FPLP. We also met with officials from the Social Security Administration to get their views on phasing benefit payments into FPLP. We did our work at IRS and FMS headquarters in Washington, D.C.; the Social Security Administration headquarters in Baltimore, Maryland; the IRS area offices in Oakland, California and Sacramento, California; and the IRS Return Processing Center in Fresno, California. [End of section] Appendix II: Additional Data on Revenue Collections Attributable to FPLP and Status of Taxpayer Accounts: This appendix provides additional details on the results of our analysis, specifically (1) the estimated $107.1 million in delinquent tax collections that can be attributed to the FPLP, (2) the categorization of taxpayer action after entering the FPLP, and (3) the distribution of Social Security payment recipients by year of the most recent income tax return filed. FPLP Payments by Collection Method: Our analysis included a group of taxpayers in FPLP receiving Social Security benefit payments, federal government retirement payments from OPM, and vendor payments. For detailed information on the distribution of delinquent tax dollars collected by various methods, including continuous levy under FPLP, see table 2. Table 2: FPLP Payments by Collection Method as of August 2002 for Taxpayers Who Received a Notice of Intent to Levy during October- December 2001: Dollars in millions: SSA: FPLP nonlevy payments; Dollars collected: $73.4; 95 percent confidence: interval of dollars: collected[A]: 54.6 to 123.4; Percentage of dollars: collected[B]: 73. Installment agreement payments; Dollars collected: 2.8; 95 percent confidence: interval of dollars: collected[A]: 2.4 to 3.5; Percentage of dollars: collected[B]: 3. Litigation, claim, and offer in compromise payments; Dollars collected: 1.9; 95 percent confidence: interval of dollars: collected[A]: 1.1 to 1.9; Percentage of dollars: collected[B]: 2. Voluntary payments; Dollars collected: 68.7; 95 percent confidence: interval of dollars: collected[A]: 52.7 to 119.0; Percentage of dollars: collected[B]: 69. FPLP levy payments; Dollars collected: 26.9; 95 percent confidence: interval of dollars: collected[A]: 23.8 to 30.1; Percentage of dollars: collected[B]: 27. Total; Dollars collected: $100.3; 95 percent confidence: interval of dollars: collected[A]: [Empty]; Percentage of dollars: collected[B]: 100. OPM; FPLP nonlevy payments; Dollars collected: 0.3; 95 percent confidence: interval of dollars: collected[A]: [Empty]; Percentage of dollars: collected[B]: 49. Installment agreement payments; Dollars collected: 0.02; 95 percent confidence: interval of dollars: collected[A]: [Empty]; Percentage of dollars: collected[B]: 4. Litigation, claim, and offer in compromise payments; Dollars collected: 0.0007; 95 percent confidence: interval of dollars: collected[A]: [Empty]; Percentage of dollars: collected[B]: 0. Voluntary payments; Dollars collected: 0.3; 95 percent confidence: interval of dollars: collected[A]: [Empty]; Percentage of dollars: collected[B]: 45. FPLP levy payments; Dollars collected: 0.3; 95 percent confidence: interval of dollars: collected[A]: [Empty]; Percentage of dollars: collected[B]: 51. Total; Dollars collected: $0.6; 95 percent confidence: interval of dollars: collected[A]: [Empty]; Percentage of dollars: collected[B]: 100. Vendor; FPLP nonlevy payments; Dollars collected: 4.9; 95 percent confidence: interval of dollars: collected[A]: [Empty]; Percentage of dollars: collected[B]: 79. Installment agreement payments; Dollars collected: 0.9; 95 percent confidence: interval of dollars: collected[A]: [Empty]; Percentage of dollars: collected[B]: 14. Litigation, claim, and offer in compromise payments; Dollars collected: 0.2; 95 percent confidence: interval of dollars: collected[A]: [Empty]; Percentage of dollars: collected[B]: 4. Voluntary payments; Dollars collected: 3.8; 95 percent confidence: interval of dollars: collected[A]: [Empty]; Percentage of dollars: collected[B]: 62. FPLP levy payments; Dollars collected: 1.3; 95 percent confidence: interval of dollars: collected[A]: [Empty]; Percentage of dollars: collected[B]: 21. Total; Dollars collected: $6.2; 95 percent confidence: interval of dollars: collected[A]: [Empty]; Percentage of dollars: collected[B]: 100. Overall; FPLP nonlevy payments; Dollars collected: 78.6; 95 percent confidence: interval of dollars: collected[A]: 59.8 to 128.6; Percentage of dollars: collected[B]: 73. Installment agreement payments; Dollars collected: 3.7; 95 percent confidence: interval of dollars: collected[A]: 3.3 to 4.4; Percentage of dollars: collected[B]: 3. Litigation, claim, and offer in compromise payments; Dollars collected: 2.1; 95 percent confidence: interval of dollars: collected[A]: 1.3 to 2.1; Percentage of dollars: collected[B]: 2. Voluntary payments; Dollars collected: 72.8; 95 percent confidence: interval of dollars: collected[A]: 56.8 to 123.1; Percentage of dollars: collected[B]: 68. FPLP levy payments; Dollars collected: 28.5; 95 percent confidence: interval of dollars: collected[A]: 25.4 to 31.7; Percentage of dollars: collected[B]: 27. Total; Dollars collected: $107.1; 95 percent confidence: interval of dollars: collected[A]: [Empty]; Percentage of dollars: collected[B]: 100. Source: GAO analysis of IRS data. [A] Data on Social Security cases sampled has been projected to represent the total population of 97,133 taxpayers receiving Social Security payments and whose TPI was above the criterion threshold. [B] Percentages may not add due to rounding. [End of table] Categorization of Taxpayers: To assess the full range of FPLP results, we reviewed the status of accounts for a group of taxpayers after they had been in the program for an extended period of time.[Footnote 26] Based on our examination of the available IRS transactions data for each delinquent account, we assigned these sampled taxpayers to categories representative of the most recent activity that had occurred on their account as of August 2002. In general, we found that the taxpayer status fell into one of eight categories: (1) the account had been resolved through payment and/or abatement of the tax liability; (2) the account was still being actively matched and/or levied under FPLP; (3) the taxpayer had made at least one voluntary payment on the delinquent debt; (4) the taxpayer had made installment agreement payments on the debt or had initiated the installment agreement process; (5) the taxpayer had either initiated litigation, a claim for a refund, or the offer in compromise process; (6) the delinquent account had been transferred to either the ACS or the field for manual collection; (7) the delinquent account was removed from FPLP for various reasons, for example, the taxpayer was deceased or for other unknown reasons; and (8) IRS made a determination that the taxpayer had a financial hardship and was currently unable to make any payments on the debt. Table 3 provides more detailed information on the categorization of taxpayer activity we observed for the accounts reviewed. Table 3: Categorization of Taxpayers by Delinquent Account Status as of August 2002: Resolved account; Taxpayers: SSA: 17,685; Taxpayers: OPM: 257; Taxpayers: Vendor: 295; Taxpayers: Overall: 18,237; [Empty]; Percentage of taxpayers: SSA: 18; Percentage of taxpayers: OPM: 37; Percentage of taxpayers: Vendor: 61; Percentage of taxpayers: Overall: 19. Account currently in FPLP; Taxpayers: SSA: 42,445; Taxpayers: OPM: 212; Taxpayers: Vendor: 91; Taxpayers: Overall: 42,748; [Empty]; Percentage of taxpayers: SSA: 44; Percentage of taxpayers: OPM: 30; Percentage of taxpayers: Vendor: 19; Percentage of taxpayers: Overall: 43. Voluntary payment made; Taxpayers: SSA: 2,993; Taxpayers: OPM: 49; Taxpayers: Vendor: 15; Taxpayers: Overall: 3,057; [Empty]; Percentage of taxpayers: SSA: 3; Percentage of taxpayers: OPM: 7; Percentage of taxpayers: Vendor: 3; Percentage of taxpayers: Overall: 3. Installment agreement; Taxpayers: SSA: 5,442; Taxpayers: OPM: 59; Taxpayers: Vendor: 20; Taxpayers: Overall: 5,521; [Empty]; Percentage of taxpayers: SSA: 6; Percentage of taxpayers: OPM: 8; Percentage of taxpayers: Vendor: 4; Percentage of taxpayers: Overall: 6. Litigation, claim, and offer in compromise; Taxpayers: SSA: 2,993; Taxpayers: OPM: 30; Taxpayers: Vendor: 16; Taxpayers: Overall: 3,039; [Empty]; Percentage of taxpayers: SSA: 3; Percentage of taxpayers: OPM: 4; Percentage of taxpayers: Vendor: 3; Percentage of taxpayers: Overall: 3. Transferred to manual collection process; Taxpayers: SSA: 6,802; Taxpayers: OPM: 29; Taxpayers: Vendor: 29; Taxpayers: Overall: 6,860; [Empty]; Percentage of taxpayers: SSA: 7; Percentage of taxpayers: OPM: 4; Percentage of taxpayers: Vendor: 6; Percentage of taxpayers: Overall: 7. Out of FPLP; Taxpayers: SSA: 9,251; Taxpayers: OPM: 38; Taxpayers: Vendor: 18; Taxpayers: Overall: 9,307; [Empty]; Percentage of taxpayers: SSA: 10; Percentage of taxpayers: OPM: 5; Percentage of taxpayers: Vendor: 4; Percentage of taxpayers: Overall: 9. Determined unable to pay; Taxpayers: SSA: 9,523; Taxpayers: OPM: 25; Taxpayers: Vendor: 0; Taxpayers: Overall: 9,548; [Empty]; Percentage of taxpayers: SSA: 10; Percentage of taxpayers: OPM: 4; Percentage of taxpayers: Vendor: 0; Percentage of taxpayers: Overall: 10. Total; Taxpayers: SSA: 97,133; Taxpayers: OPM: 699; Taxpayers: Vendor: 484; Taxpayers: Overall: 98,316; [Empty]; Percentage of taxpayers: SSA: 100; Percentage of taxpayers: OPM: 100; Percentage of taxpayers: Vendor: 100; Percentage of taxpayers: Overall: 100. Source: GAO analysis of IRS data. Note: Data on the sampled Social Security cases has been projected to represent the total population of 97,133 taxpayers receiving Social Security payments and whose TPI was above the criterion threshold. Totals may not add due to rounding. The 95 percent confidence interval on the overall percentage estimates ranges from plus or minus 1.5 to 5 percentage points. [End of table] Most Recently Filed Income Tax Returns for Social Security Beneficiaries: We analyzed the yearly distribution of the most recent income tax return filed for Social Security payment recipients who owed delinquent taxes. Table 4 shows that only 17 percent of these taxpayers had filed for tax year 2001, and that 53 percent had not filed since 1996 or earlier. Table 4: Distribution of Social Security Payment Recipients by Year of Most Recent Income Tax Return Filed: Tax year of last return filed: 2001; Number of taxpayers: 31,865; Percentage of taxpayers who filed: 17; Cumulative percentage of taxpayers: 100. Tax year of last return filed: 2000; Number of taxpayers: 11,246; Percentage of taxpayers who filed: 6; Cumulative percentage of taxpayers: 83. Tax year of last return filed: 1999; Number of taxpayers: 13,121; Percentage of taxpayers who filed: 7; Cumulative percentage of taxpayers: 77. Tax year of last return filed: 1998; Number of taxpayers: 14,995; Percentage of taxpayers who filed: 8; Cumulative percentage of taxpayers: 70. Tax year of last return filed: 1997; Number of taxpayers: 16,870; Percentage of taxpayers who filed: 9; Cumulative percentage of taxpayers: 62. Tax year of last return filed: 1996 and earlier; Number of taxpayers: 99,343; Percentage of taxpayers who filed: 53; Cumulative percentage of taxpayers: 53. Tax year of last return filed: Total; Number of taxpayers: 187,440; Percentage of taxpayers who filed: 100; Cumulative percentage of taxpayers: [Empty]. Source: GAO analysis of IRS data. Note: Data on the combined Social Security samples of taxpayers above and below the TPI criterion have been projected to represent the total population of 187,440 taxpayers receiving Social Security payments. The 95 percent confidence interval on the percentage estimates ranges from plus or minus 1.6 to 3.4 percentage points. [End of table] [End of section] Appendix III: Comments from the Internal Revenue Service: DEPARTMENT OF THE TREASURY INTERNAL REVENUE SERVICE WASHINGTON, D.C. 20224: COMMISSIONER: February 27, 2003: Mr. Michael Brostek Director, Tax Issues U.S. General Accounting Office 441 G Street, N.W. Washington, D.C. 20548: Dear Mr. Brostek: This is in response to your letter and draft report entitled “Federal Payment Levy Program Measures, Performance, and Equity Can Be Improved.” The Federal Payment Levy Program (FPLP) has proven to be a valuable, cost-effective method of securing tax revenues from federal payment sources. As we expand FPLP to additional federal employee salaries and other types of federal payments, we will continue our efforts to fully capture the program’s business results, and improve performance in a manner that ensures fair and equitable treatment for all taxpayers. Your report suggests that FPLP frees up resources, enabling us to pursue accounts with more collection potential. We agree that FPLP automated processes enable us to close accounts. However, we have experienced a workload increase due to taxpayer inquiries made as a result of FPLP and the actions necessary to close these cases. Many of these taxpayer contacts are on cases resulting in less productive use of our resources. With the plans for continued expansion of FPLP, it is difficult to assess the total costs and/or indirect benefits of the program in terms of resources freed up to pursue other collection activity. Our response to the Recommendations for Executive Action follows: Recommendation: Include in IRS’s planned new approach to measuring FPLP results data on the full range of taxpayers’ actions and tax collections attributable to FPLP, including non-levy collections and account resolutions. Response: We agree with the report’s finding that FPLP measures only the direct revenue collected as a result of the levy and that we should also measure indirect revenue through non-levy payments. The program staff will work with our Information Technology Services (ITS) to determine programming requirements needed to include: indirect revenue as a measurement of FPLP. We anticipate developing these requirements and including them in a Request for Information Services in early May. Based on the requirements, ITS will provide an estimated implementation date. We would like to share our proposed methodology with you as we develop it. We will also explore the feasibility of tracking other account resolutions, such as “Determined Unable to Pay,” as a result of FPLP. Recommendation: Study the feasibility of submitting all eligible delinquent accounts to FMS on an ongoing basis for matching against federal payment records under FPLP, and use information from any matches to assist IRS in determining the most efficient method of collecting delinquent taxes, including whether to use FPLP. Response: We are currently conducting a one-time test match with FMS on Automated Collection System (ACS) cases that are eligible for FPLP. We will study the results of that test to determine the cost-effectiveness of implementing matching as an ongoing process. In addition, we have expanded our universe of accounts subject to FPLP. As you noted in your report, we will include delinquent accounts of taxpayers who received either a federal salary or annuity payment in Fiscal Year (FY) 2003, and we have also requested programming changes for FY 2005 to include additional cases currently in ACS. Recommendation: Discontinue using the Total Positive Income (TPI) criterion as an indicator of Social Security beneficiaries’ ability to pay delinquent taxes and rely on the extended two-notice process to identify beneficiaries for whom a levy would be a hardship. Determine whether sending a second notice that explains the financial hardship exception to all Social Security beneficiaries subject to levy is sufficient to identify hardship situations. If not, develop and test a criterion that reliably identifies those Social Security beneficiaries for whom a levy would represent an undue hardship. Response: We are committed to ensuring all taxpayers are treated fairly. We are concerned with the issues you outlined in your report regarding the TPI criterion we use for Social Security beneficiaries. We implemented this process in response to concerns the National Taxpayer Advocate (NTA) raised about the potential adverse effect of levies on this particular group of taxpayers. We believe it is critical that we take the next 120 days to work with the NTA and program administrators to assess the deficiencies in the current process and to develop a suitable solution. If you have any questions, please contact me or have a member of your staff contact Pamela G. Watson, Director, Filing and Payment Compliance at (404) 338-8686. Sincerely, Bob Wenzel Acting Commissioner: Signed by Bob Wenzel [End of section] Appendix IV: GAO Contacts and Staff Acknowledgments: GAO Contacts: Michael Brostek (202) 512-9110 Ralph T. Block (415) 904-2150: Acknowledgments: In addition to those named above, Tom N. Bloom, Allen T. Chan, Jeanine Lavender, Ellen Rominger, Amy Rosewarne, Samuel Scrutchins, Wendy Turenne, James J. Ungvarsky, Elwood D. White, and Thomas Venezia made key contributions to this report. FOOTNOTES [1] This represents total unpaid assessments as of September 30, 2002. Federal accounting standards identify unpaid assessments as (1) taxes due from taxpayers for which IRS can support the existence of a receivable through taxpayer agreement or a favorable court ruling (federal taxes receivable); (2) assessments IRS has made of additional taxes owed in which neither the taxpayer nor the court has affirmed that the amounts are owed; and (3) write-offs, for which IRS expects no collection due to factors such as the taxpayer’s death, bankruptcy, or insolvency. [2] Levy is the legal process by which IRS orders a third party to turn over property in its possession that belongs to the delinquent taxpayer named in a notice of levy. A continuous levy remains in effect from the date it is first made until the tax debt is fully paid or IRS releases the levy. [3] U.S. General Accounting Office, Tax Administration: IRS’ Levy of Federal Payments Could Generate Millions of Dollars, GAO/GGD-00-65 (Washington, D.C.: April 2000) and Tax Administration: Millions of Dollars Could Be Collected If IRS Levied More Federal Payments, GAO-01-711 (Washington, D.C.: July 2001). [4] The Office of Personnel Management issues federal retirement payments. Social Security benefit payments outlined in Title II, Federal Old-Age, Survivors, and Disability Insurance Benefits, of the Social Security Act, are subject to continuous levy under FPLP. Social Security benefit payments such as lump-sum death benefits, benefits paid to children, and special benefits for persons aged 72 and over by 1971 are not included in FPLP. In addition, Supplemental Security Income payments under Title XVI and payments with partial withholding to repay a debt owed to Social Security will not be levied through FPLP. Vendor payments are issued to businesses or individuals that provide goods or services to the federal government. [5] Taxpayers notified of an impending FPLP levy have typically already received several previous balance due notices as part of IRS’s standard notification process. [6] Taxpayers who have been matched on a scheduled benefit payment from the Social Security Administration receive a second notice, with an additional 30 days to respond, if they take no action after receiving the first notice of intent to levy. [7] Installment agreements allow the full payment of the debt in smaller, more manageable amounts. An offer in compromise approved by IRS allows taxpayers to settle their unpaid debt for less than the full amount of the balance due. [8] The ACS is a telephone collection system that uses a computerized inventory system containing information on balance due accounts and investigations of delinquent tax returns. Delinquent accounts assigned to the field collection inventory system are assigned to a revenue officer in the field who pursues the account. [9] We refer to payments made by taxpayers after receiving a levy notice as voluntary payments because the taxpayers subsequently remitted payments to IRS without further action on IRS’s part. [10] The 95 percent confidence interval ranges from $25.4 million to $31.7 million. [11] The 95 percent confidence interval ranges from $59.8 million to $ 128.6 million. [12] We attributed payments made on delinquent accounts to FPLP if taxpayers had taken no significant actions to resolve their delinquency prior to entering the program and receiving a notice of intent to levy. Our analysis showed that prior to their inclusion into FPLP, an average of 16 months had elapsed since the vendors had initiated any action on their delinquent accounts; an average of 21 months had elapsed since taxpayers receiving federal retirement payments had initiated any action on their accounts; and an average of 57 months had elapsed since Social Security payment recipients had initiated any significant action on their delinquent accounts. The 95 percent confidence interval on the Social Security estimate ranges from 54.1 to 60.7 months. [13] We analyzed IRS’s accounts receivable files data as of February 2002, which showed that IRS blocked the delinquent accounts of nearly 1.5 million individuals and businesses owing about $14.2 billion in tax debt from FPLP. We matched the blocked delinquent taxpayers against federal payments for Social Security benefits, federal retirement, and federal salary that were made in February 2002, and for vendor payments that were made during the second quarter of fiscal year 2002. [14] IRS data as of July 2002. [15] TPI is calculated by summing the positive values from the following income fields from a taxpayer’s most recently filed individual income tax return: wages; interest; dividends; distributions from partnerships, small business corporations, estates, or trusts; Schedule C net profits; Schedule F net profits; and other income such as Schedule D profits and capital gains distributions. Losses reported for any of these values are treated as a zero. The TPI threshold is sensitive information, and therefore, available for official use only. [16] Data as of August 2002. The 95 percent confidence interval ranges from $383 million to $660 million. [17] As of August 2002. [18] The 95 percent confidence intervals for the 18 percent and 12 percent overlap. The interval for the 18 percent figure ranges from 14 percent to 22 percent, while the interval for the 12 percent figure ranges from 9 percent to 16 percent. [19] Although the taxpayers’ actions demonstrate some ability to pay, we recognize that some portion of those who voluntarily settled their delinquent accounts after receiving a levy notice may have done so despite being in a financial hardship position. A case-by-case review of the taxpayers’ circumstances would be needed to determine to what extent, if at all, taxpayers above and below the TPI who made voluntary payments were nevertheless in a hardship situation. [20] The 95 percent confidence intervals for the 5 percent and 8 percent overlap. The interval for the 5 percent figure ranges from 3 percent to 8 percent, while the interval for the 8 percent figure ranges from 5 percent to 11 percent. [21] An information return is a tax document businesses are required to file to report certain business transactions to IRS. For example, these transactions include (but are not limited to) wages paid to employees; interest and dividend payments; pension distributions; and mortgage interest paid. [22] The 95 percent confidence intervals for the 40 percent and 46 percent overlap. The interval for the 40 percent figure ranges from 35 percent to 45 percent while the interval for the 46 percent figure ranges from 41 percent to 51 percent. [23] The determination of whether an individual is required to file an income tax return is based on their filing status, age, and the amount of annual gross income. For example, in 2002, couples that filed jointly and were over 65 years of age with a gross income of less than $15,650 would not be required to file an income tax return. [24] Because we followed a probability procedure based on random selection for the samples we selected, each of these samples is only one of a large number of samples that we might have drawn. Since each sample could have provided different estimates, we express our confidence in the precision of our particular sample’s results as a 95 percent confidence interval. This is the interval that would contain the actual population value for 95 percent of the samples we could have drawn. As a result, we are 95 percent confident that each of the confidence intervals in this report will include the true values in the study population. [25] The payment records obtained cover various periods of time. Vendor payments are for the second quarter of fiscal year 2002, salary payments represent one biweekly pay period in February 2002, and all other payments are for the month of February 2002. [26] Roughly 8 to 11 months. 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 GAO Reports and Testimony: 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-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 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 daily E-mail alert for newly released products” under the GAO Reports 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, Waste, and Abuse in Federal Programs: Contact: Web site: www.gao.gov/fraudnet/fraudnet.htm E-mail: fraudnet@gao.gov 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: