This is the accessible text file for GAO report number GAO-06-396 entitled 'Tax Administration: Opportunities to Improve Compliance Decisions and Service to Taxpayers through Enhancements to Appeals' Feedback Project' which was released on April 18, 2006. This text file was formatted by the U.S. Government Accountability Office (GAO) to be accessible to users with visual impairments, as part of a longer term project to improve GAO products' accessibility. Every attempt has been made to maintain the structural and data integrity of the original printed product. Accessibility features, such as text descriptions of tables, consecutively numbered footnotes placed at the end of the file, and the text of agency comment letters, are provided but may not exactly duplicate the presentation or format of the printed version. The portable document format (PDF) file is an exact electronic replica of the printed version. We welcome your feedback. Please E-mail your comments regarding the contents or accessibility features of this document to Webmaster@gao.gov. This is a work of the U.S. government and is not subject to copyright protection in the United States. It may be reproduced and distributed in its entirety without further permission from GAO. Because this work may contain copyrighted images or other material, permission from the copyright holder may be necessary if you wish to reproduce this material separately. Report to the Chairman and Ranking Minority Member, Committee on Finance, U.S. Senate: March 2006: Tax Administration: Opportunities to Improve Compliance Decisions and Service to Taxpayers through Enhancements to Appeals' Feedback Project: GAO-06-396: GAO Highlights: Highlights of GAO-06-396, a report to the Committee on Finance, U.S. Senate. Why GAO Did This Study: Taxpayers disagreeing with Internal Revenue Service (IRS) compliance decisions can request an independent review by IRS’s Appeals Office (Appeals). In 2004 the Commissioner requested that Appeals establish a feedback program to share the results of Appeals’ reviews with the compliance programs. GAO was asked to assess whether (1) information on Appeals results would provide useful feedback to IRS operating divisions to benefit compliance programs, Appeals, and taxpayers through better case resolution and (2) the feedback project is being effectively managed to maximize its potential to improve IRS’s performance and thereby reduce disputes with taxpayers. What GAO Found: Appeals’ case result information has the potential to help compliance programs improve taxpayer service, but realizing improvements requires investments in data collection and analysis that must be considered in light of the likely benefits. Based on a review of 153 Appeals cases, GAO estimates that 41 percent of the 102,623 cases closed in fiscal year 2004 were not fully sustained. Of these, about half were not sustained because Appeals applied a law or regulation differently than the programs. Lacking such information, officials could not assess whether actions like additional guidance were needed. However, identifying specific provisions that were interpreted differently would require data gathering and analysis. Because the differences span a host of laws and regulations, corrective action may only affect a small number of cases. Improved decision making, however, can benefit compliance programs, Appeals, and taxpayers. An initial data analysis, such as identifying programs with high nonsustention rates due to differences in applying laws or regulations, would help to target areas most likely to benefit from feedback. Appeals has taken several initial steps to launch the feedback project. During 2005, for example, Appeals and the compliance programs began to identify additional information needs. In addition, Appeals and the compliance programs could refine the feedback project’s objectives to target the results-oriented improvements that are logical benefits of information sharing. Obtaining agreement between Appeals and the programs on objectives may not be easy because their perspectives differ on the steps needed to improve operations, but is necessary. Also, Appeals’ plans to update its information system to provide additional data on case results will be hindered by inaccurate data. We found that several important data fields had error rates up to 14 percent. Appeals staff cited several reasons for this, including weak data verification procedures. Figure: IRS's Appeals System [See PDF for Image] [End of Figure] What GAO Recommends: GAO recommends that IRS (1) perform an initial analysis of feedback data to identify areas most likely to generate benefits for compliance programs, Appeals, and taxpayers, (2) investigate whether additional actions are needed to improve the consistency of decisions, (3) further develop results-oriented objectives and measures for the feedback project, and (4) build upon current efforts to improve feedback data by establishing internal controls to verify data accuracy on an ongoing basis. In commenting on a draft of this report, IRS agreed with our recommendations. To view the full product, including the scope and methodology, click on the link above. For more information, contact Michael Brostek at (202) 512-9110 or brostekm@gao.gov. [End of Section] Contents: Letter: Results in Brief: Background: Feedback Information Could Be Useful for Improving Performance: Appeals Can Build on Existing Efforts to Improve the Feedback Project: Conclusions: Recommendations: Agency Comments and Our Evaluation: Appendixes: Appendix I: Objectives, Scope, and Methodology: Appendix II: Appeals Workstreams: Appendix III: Appeal Rates: Appendix IV: Comments from the Commissioner of Internal Revenue: Related GAO Products: Tables: Table 1: Reasons Appeals Did Not Fully Sustain Compliance Decisions, Percentage of Nonsustained Cases in GAO's Sample by Reason, and Illustrative Examples, Fiscal Year 2004: Table 2: Comparison of the Composition of Appeals Workload, Rate of Compliance Cases Appealed, Percentage of Appeals Workload, and Cases Not Fully Sustained: Table 3: Reasons that Appeals Did Not Fully Sustain Compliance Program Decisions, Percentage of Cases Not Fully Sustained, Sample of Cases Closed in Fiscal Year 2004: Table 4: Confidence Intervals for Table 1: Table 5: Confidence Intervals for Table 2: Table 6: Confidence Intervals for Table 3: Table 7: Description of Appeals Workstreams: Table 8: Appeal Rates by Workstream, Fiscal Year 2004: Figure: Figure 1: IRS's Appeals System: Abbreviations: ACDS: Appeals Centralized Database System: ACM: Appeals Case Memorandum: CIC: Coordinated Industry Case: CDP: Collection Due Process: DCI: Data Collection Instrument: EITC: Earned Income Tax Credit: IC: Industry Case: IS: Innocent Spouse: IRS: Internal Revenue Service: LMSB: Large and Mid-Sized Business Division: OIC: Offer-In-Compromise: SB/SE: Small Business/Self-Employed Division: TE/GE: Tax-Exempt and Government Entities Division: W&I: Wage and Investment Division: Letter: March 24, 2006: The Honorable Charles E. Grassley: Chairman: The Honorable Max Baucus: Ranking Minority Member: Committee on Finance United States Senate: To help maintain the public's confidence in the tax system, the Internal Revenue Service (IRS) believes it needs to ensure that taxpayers receive a fair, impartial, and uniform resolution of tax disputes. If taxpayers disagree with IRS decisions to assess additional tax or take collection action, they can bring their cases before the IRS Appeals function (Appeals), which independently reviews whether compliance decisions correctly reflect the facts as well as applicable law, regulations, and IRS procedures.[Footnote 1] With a staff of about 1,900 employees and a budget of about $193 million in fiscal year 2005, Appeals annually closes over 100,000 cases where taxpayers disputed IRS's compliance decisions. Our prior reports have found that results-oriented organizations use performance information to continuously identify performance gaps and put that information to work to improve their operations. One possible source of this information in IRS is the results of Appeals cases. Using this information, IRS compliance programs may possibly identify whether Appeals would agree that the case decision represents the best result for IRS and the taxpayer. Then, using Appeals case results information, compliance program managers may possibly be able to identify weaknesses in their programs and improve operations, providing better service to taxpayers, and possibly resolving more cases before an appeal. In line with this concept, the Commissioner recently requested that Appeals establish a feedback project intended to maximize the benefits of sharing Appeals decisions with the compliance programs. Based on your request, this report's objectives are to determine whether (1) information on Appeals results would provide useful feedback to IRS operating divisions to benefit compliance programs, Appeals, and taxpayers through better case resolution and (2) the feedback project is being effectively managed to maximize its potential to improve IRS's performance and thereby reduce disputes with taxpayers. To make determinations in both areas, we reviewed Appeals and compliance program documents and our prior work on performance management, and interviewed Appeals and compliance program officials. To determine whether information on the results of Appeals cases has the potential to provide useful feedback, we reviewed a random sample of Appeals cases closed in fiscal year 2004 to determine whether the cases contained information that could be useful for improving case results. Principally, we collected data on whether Appeals sustained or overturned the compliance program's decision in each case and identified the main reasons for the Appeals decision. The results from that sample have related confidence intervals and precision estimates that are presented in this letter and in appendix I. We conducted our review from October 2004 through October 2005 in accordance with generally accepted government auditing standards. Results in Brief: As an independent reviewer of IRS compliance program's cases, Appeals' case results information has the potential to help those programs improve their service to taxpayers, but realizing improvements requires investments in data collection and analysis that must be considered in light of the likely benefits to the compliance programs, Appeals, and taxpayers. Based on our review of 153 compliance program cases closed in fiscal year 2004, we estimate that 41 percent of the 102,623 cases closed in that year (42,075 cases) were not fully sustained by Appeals.[Footnote 2] Of these cases, Appeals did not sustain 52 percent (21,879 cases) because, at least in part, it applied a law or regulation differently than the compliance programs.[Footnote 3] Because they have not received this type of information, the compliance programs could not assess whether additional guidance, training, or other initiatives were needed. However, identifying more specifically which laws or regulations were interpreted differently by the programs would require an investment to gather and analyze additional data. Because these differing applications span a host of laws and regulations across all of IRS's compliance programs, in many cases the corrective action that might be taken may only affect a relatively small number of cases. Improved decision making, however, can benefit compliance programs, Appeals, and taxpayers. Systematic data analysis, such as identifying programs with high nonsustention rates due to differences in applying laws or regulations, would help identify those programs most likely to benefit from feedback.[Footnote 4] This analysis would also help to focus attention on the costs of developing the feedback data and benefits to be obtained. Currently, IRS does not have plans for these analyses. Appeals has taken several initial steps to launch the feedback project requested by the IRS Commissioner. During 2005, several information- sharing and cooperative projects (e.g., providing copies of case memorandums and basic summary data) were established with some compliance programs on Appeals case results. Appeals and compliance managers can take further steps to build upon those efforts. For example, Appeals plans to update its information system to provide additional information on case results, such as the reasons for Appeal's decisions. In addition, Appeals developed the feedback project's objectives and measures with limited input from the compliance programs. Appeals and the compliance programs can partner on further developing results-oriented objectives that take into account compliance programs' diverse circumstances and clearly articulate the feedback project's expected outcomes and associated performance measures. Obtaining agreement between Appeals and the compliance programs on objectives and performance measures may not be easy because their perspectives differ on the steps needed to improve IRS operations, but it is a necessary step because the programs themselves must play active roles in the project to make any changes that will improve their case results. However, the program will be hindered by inaccurate data. For example, we found that several important data fields related to Appeals case results had error rates up to 14 percent.[Footnote 5] Appeals staff cited several reasons for the errors including a lack of attention to complete and accurate data, confusion about the results of the appeal by those entering the data, and weak data verification procedures. In order to maximize the opportunities for the Appeals feedback project to improve IRS's compliance decisions and service to taxpayers, we are recommending that IRS analyze feedback data to identify areas most likely to benefit from feedback projects; further investigate the most promising areas and assess whether additional actions, such as new guidance or training, are needed to improve the consistency of decisions; further develop results-oriented objectives and associated performance measures for feedback projects; and build upon its current efforts to improve Appeals information for the feedback projects by establishing internal controls to verify, on an ongoing basis, the accuracy of information on case outcomes. In written comments on a draft of this report, the Commissioner of Internal Revenue agreed with our recommendations and said they will help IRS develop a much stronger feedback program. Background: America's tax system is based on taxpayers voluntarily filing tax returns that report the full amount of tax owed and paying any taxes that are due. IRS has four operating divisions: * Wage and Investment Division (W&I) serves the vast number of individual taxpayers including those who file jointly and only have wage and investment income. * Small Business/Self-Employed Division (SB/SE) serves about 45 million small business, individual taxpayers with rental properties and farming businesses, and individuals investing in businesses, such as partnerships. SB/SE also serves corporations and partnerships with less than $10 million in assets and provides field collection services for the other three IRS divisions. * Large and Mid-Size Business Division (LMSB) serves corporations, subchapter S corporations, and partnerships with assets greater than $10 million. These businesses have a large number of employees, have complicated tax and accounting issues, and often conduct business globally. * Tax-Exempt and Government Entities Division (TE/GE) serves three very distinct customer segments. Employee Plans serves private and public retirement plan customers. Exempt Organizations serves customers that are exempt from income taxes, such as charities, civic organizations, and business leagues. Government Entities serves customers from federal, state, and local governments; Indian tribal governments; and tax-exempt bond issuers. These divisions are responsible for providing a full range of services to these taxpayers. Typically, these services would include assisting taxpayers with filing returns, processing those returns and maintaining their accounts, and examining suspected inaccurate returns. Taxpayers who are assessed additional tax and penalties or who have a pending enforcement action to collect delinquent taxes, such as a proposed levy or lien,[Footnote 6] have the right to request a hearing through an administrative appeal before the assessment or collection actions are final.[Footnote 7] IRS notifies the taxpayers in writing of these pending actions and explains their appeal rights. Generally, the taxpayer has 30 days from this notification to request an appeal. The Appeals Process: Appeals' mission is to independently resolve tax disputes prior to litigation on a basis which is fair and impartial to both the government and the taxpayer. To assure their independence, Appeals' staff cannot discuss substantive case issues with compliance staff unless taxpayers or their representative are present. Generally, compliance staff does not directly participate in an appeal or learn about the resulting decision. To identify whether the proposed compliance action should be sustained, Appeals staff review the case file prepared by IRS's compliance program and determine whether that evidence demonstrates that the taxpayer and compliance staff have followed the applicable law, regulation, and IRS procedure. If requested, the taxpayer may meet with Appeals staff and provide additional evidence to support their appeal. To close an examination case, Appeals may (1) agree with the examination program and fully sustain its recommended assessment, (2) disagree and reduce the recommended assessment to partially sustain the assessment, or (3) fully concede to the taxpayer's position and not sustain the assessment. For a collection case, Appeals may (1) agree with and sustain the proposed enforcement action or (2) not sustain the proposed enforcement action by modifying the proposed action (e.g., propose an installment agreement rather than a levy), deferring collection, or fully conceding to the taxpayer's position. If the taxpayer and IRS cannot reach agreement on the outcome of the case through the Appeals process, the taxpayer may have the case reviewed by the U.S. Tax Court, U.S. Court of Federal Claims, or a U.S. district court. In line with its mission to resolve cases prior to litigation, Appeals is also authorized to review the facts of the case in light of the hazards that would exist if the case were litigated.[Footnote 8] Appeals is the only IRS organization authorized to consider hazards of litigation when deciding whether to allow taxes and penalties.[Footnote 9] This means that Appeals may recommend a fair and impartial resolution somewhere between fully sustaining or fully conceding the examiner's proposal that reflects the probable result in the event of litigation. If taxpayers do not reach agreement with IRS examiners on the proposed deficiency, or if they choose not to contact Appeals, IRS will issue a notice of deficiency. This notice describes the deficiency and states that the taxpayer has 90 days to file a petition with the court for a redetermination of the deficiency. However, even though Appeals may be initially bypassed, it still has an opportunity to settle these cases.[Footnote 10] Under IRS procedures designed to encourage resolution of cases at the lowest possible level, the attorney from the local IRS District Counsel's office handling the court case is required to refer the case to Appeals for possible settlement before it is scheduled for trial.[Footnote 11] Figure 1 summarizes IRS's appeals system. Figure 1: IRS's Appeals System: [See PDF for image] [End of figure] Appeals' workload is organized into eight "workstreams" that reflect similarities in the case workload rather than which of IRS's four operating divisions initiated the case. Two of the eight workstreams relate to collection issues and generally originate in two of IRS's four operating divisions responsible for collection issues (Collection Due Process and Offer-in-Compromise workstreams). Three of the eight workstreams include a wide range of generally smaller examination and returns-processing-related penalty cases (Innocent Spouse, Penalty Appeals, and Exam/TEGE). The three other workstreams (Coordinated Industry Case, Industry Case, and Other) cover a small number of complex examinations from IRS's LMSB programs as well as cases that do not fit into other workstreams.[Footnote 12] Appendix II includes definitions of Appeals workstreams, identifies the related operating divisions for the workstream, and the number of cases closed in each workstream during fiscal year 2004. Strategic Approach: Results-oriented organizations consistently strive to improve their performance through strategic planning. As part of this approach, agencies set objectives and measure performance to evaluate whether performance has improved. Specifically, * goals or objectives are the results that a program is expected to achieve, and: * performance measures are selected after goals or objectives are developed, logically related to these goals or objectives, and used to gauge progress toward them. Other federal agencies have previously decided that developing and sharing information on the results of appeals may help them measure performance or at least serve as an indicator of whether their decisions are legally correct. For example, the Merit Systems Protection Board, an independent quasi-judicial agency established to protect merit systems in the federal workplace, has set a performance goal of maintaining or reducing its low percentage of appealed decisions that are reversed or sent back to board judges for a new decision. The board's performance plan for fiscal year 2005 contains an array of case-specific data to measure this performance goal. Feedback Information Could Be Useful for Improving Performance: Appeals overturned about 41 percent of the fiscal year 2004 cases we reviewed and in about half of those cases Appeals disagreed with the way compliance programs applied the law or regulations. This suggests that providing information on Appeals decisions could help compliance program managers improve case results by fostering more proper and consistent case decisions. However, finding the source of possible inconsistencies will require gathering and analyzing additional information and systematic analysis. Improved decision making, however, can benefit compliance programs, Appeals, and taxpayers. Feedback on Appeals Results Could Be Useful: Based on our case review, for cases closed in fiscal year 2004, we estimate that Appeals did not sustain about 41 percent of compliance cases (about 42,075 of the 102,623) that year. We identified six principal reasons for those nonsustentions. As shown in table 1, we estimate that Appeals did not sustain compliance decisions in 52 percent of the cases not sustained (21,879 cases) at least in part because Appeals disagreed with compliance staff's application of tax law or IRS regulations.[Footnote 13] Providing feedback on such disagreements could help compliance managers improve case results by taking action to foster the proper and consistent application of tax laws and regulations. For example, compliance managers could assess whether guidance or manuals, supervision, quality control, or other management tools should be revised to ensure that cases are properly closed. Identifying more specifically which laws or regulations were applied differently by the compliance programs would require an investment to gather and analyze additional data. For instance, in table 1 we identified the handling of a state tax refund as an example of differing applications of tax laws and regulations. To determine whether this is a common problem or an isolated instance, officials would have to investigate the issue by, for example, drawing a random sample of cases or questioning first-line managers and staff. Because these differing applications span a host of laws and regulations across IRS's compliance programs, the corrective action that might be taken may only affect a relatively small number of cases. In complex cases, Appeals and the compliance managers may need to work together to develop a mutual understanding of how laws and regulations should be applied. Table 1: Reasons Appeals Did Not Fully Sustain Compliance Decisions, Percentage of Nonsustained Cases in GAO's Sample by Reason, and Illustrative Examples, Fiscal Year 2004: Reason Appeals did not fully sustain compliance decision: Application of laws or regulations; Percentage of cases not fully sustained based on the reason [A[B]]: 52; Illustrative case examples: Examination revised taxable income to include the state refund for the prior year. Appeals determined that the taxpayer had not itemized on his or her federal return for the prior year and received no tax benefit. Therefore, examination should not have included the refund in taxable income. Reason Appeals did not fully sustain compliance decision: Additional information provided by the taxpayer or not accepted by compliance; Percentage of cases not fully sustained based on the reason [A[B]]: 44; Illustrative case examples: Examination disallowed a taxpayer's dependency exemptions and Earned Income Tax Credit (EITC) in a correspondence examination of the tax return. On appeal, the taxpayer presented supporting documentation, including birth certificates and school records, to substantiate the claim. Reason Appeals did not fully sustain compliance decision: Original audit work or significant rework by Appeals; Percentage of cases not fully sustained based on the reason [A[B]]: 14; Illustrative case examples: The taxpayer contended that he or she was not liable for a tax assessment since the taxpayer had lost his or her wallet and someone else used the taxpayer's Social Security Number to file the return. The taxpayer filed in Tax Court. Since the case had not been reviewed by Appeals, Chief Counsel forwarded the case to Appeals for review. Appeals determined that the taxpayer was not liable for the deficiency since the taxpayer was the victim of identity theft. Reason Appeals did not fully sustain compliance decision: Taxpayer not responsive to compliance; Percentage of cases not fully sustained based on the reason [A[B]]: 13; Illustrative case examples: Examination determined that a taxpayer had not filed a tax return and thus incurred a tax deficiency. Both compliance and Appeals requested the tax return. The taxpayer did not submit the tax return until the case was sent to Chief Counsel to prepare for a Tax Court hearing. Once the tax return was processed, the taxpayer was due a refund. Reason Appeals did not fully sustain compliance decision: Hazards of litigation; Percentage of cases not fully sustained based on the reason [A[B]]: 13; Illustrative case examples: Failure to file and related penalties imposed against the taxpayer were abated when the taxpayer presented unique and sympathetic facts (e.g., the corporation was a nonprofit organization primarily using volunteer workers) that would be a hazard if the case proceeded to litigation. Reason Appeals did not fully sustain compliance decision: Appeals changed collection alternative; Percentage of cases not fully sustained based on the reason [A[B]]: 11; Illustrative case examples: Taxpayer requested a Collection Due Process hearing because he or she wanted an alternative method of collection other than the levy proposed by compliance. During the appeal, the taxpayer requested and entered into an installment agreement that allowed the taxpayer to pay the delinquent tax over time through periodic payments. Source: GAO sample of Appeals cases. [A] These percentages are based on a subset of 63 nonsustained cases. The margin of error for these estimates is larger than for the sample as a whole. Confidence intervals are shown in appendix I. [B] Since a case can have multiple reasons why it is not sustained, the percentage of cases not fully sustained does not total to 100 percent. [End of table] As also shown in table 1, we estimate that Appeals did not sustain compliance decisions in 44 percent of the cases because the taxpayer provided additional information to Appeals. For cases in this category, officials would need to investigate whether compliance staff could have done more to obtain the information needed to resolve the tax before the case was appealed. For example, compliance managers might assess whether staff clearly articulated the type and extent of information needed, gave the taxpayer sufficient time to respond, or received the information but did not use it appropriately to resolve the case. Similar data gathering and analysis would be needed for the other reasons we identified for Appeals not sustaining cases in order for the information to be useful in improving compliance's decision making. For example, for cases where Appeals had to perform original audit work or significant rework, compliance managers would need to identify why their staff did not perform the necessary work while the case was still their responsibility. For cases where Appeals accepted a collection alternative, compliance managers might assess whether it was because the taxpayer had not requested an alternative, the taxpayer's financial circumstances had changed since compliance worked on the case, or a request for an alternative was inappropriately rejected by compliance staff. For cases where taxpayers did not respond to compliance, compliance managers might assess whether staff had made sufficient attempts to contact the taxpayer. Improved Decision Making Would Benefit Compliance Programs, Appeals, and Taxpayers: As shown in table 2, the appeal rate--the percentage of cases appealed- -varies across Appeals' workstreams from 29 percent for LMSB's Coordinated Industry Case program (CIC) cases to one-tenth of 1 percent for cases in the Penalty Appeals and Other workstreams.[Footnote 14] Managers of programs with high appeal rates told us that they would benefit from Appeals feedback information in improving decision making. For example, with relatively high appeals rates and complex tax issues frequently considered in both the Industry Case (IC) and CIC programs, LMSB managers believe that Appeals case result information is important for managing their programs to update policies and procedures, modify or assess new training needs, or identify needed changes in the tax law.[Footnote 15] Similarly, Offer-in-Compromise program managers say they have benefited from working with Appeals staff on studies analyzing why cases were not sustained by Appeals. One study indicated that compliance managers and Appeals needed to reevaluate or reinforce some of their policies as well as be more consistent in following established procedures for assessing financial information, such as calculating transportation expenses, establishing the value of cars, and estimating future income. Table 2: Comparison of the Composition of Appeals Workload, Rate of Compliance Cases Appealed, Percentage of Appeals Workload, and Cases Not Fully Sustained: Appeals workstreams: Coordinated Industry Case; Number of appeals cases closed in fiscal year 2004: 653; Percentage of Appeals workload: 0.6; Percentage of cases compliance closed that are appealed: 29; Number of cases in GAO's sample: N/A[A]; GAO sample percentage of cases not fully sustained: N/A[A]. Appeals workstreams: Offer-in-Compromise (OIC); Number of appeals cases closed in fiscal year 2004: 17,887; Percentage of Appeals workload: 17.2; Percentage of cases compliance closed that are appealed: 13; Number of cases in GAO's sample: 28; GAO sample percentage of cases not fully sustained: 14[B]. Appeals workstreams: Industry Case (IC); Number of appeals cases closed in fiscal year 2004: 670; Percentage of Appeals workload: 0.6; Percentage of cases compliance closed that are appealed: 14; Number of cases in GAO's sample: N/A[A]; GAO sample percentage of cases not fully sustained: N/A[A]. Appeals workstreams: Innocent Spouse (IS); Number of appeals cases closed in fiscal year 2004: 4,713; Percentage of Appeals workload: 4.5; Percentage of cases compliance closed that are appealed: 8; Number of cases in GAO's sample: 7; GAO sample percentage of cases not fully sustained: 57[C]. Appeals workstreams: Collection Due Process (CDP); Number of appeals cases closed in fiscal year 2004: 32,226; Percentage of Appeals workload: 31.0; Percentage of cases compliance closed that are appealed: 1; Number of cases in GAO's sample: 53; GAO sample percentage of cases not fully sustained: 19[A]. Appeals workstreams: Exam/TEGE; Number of appeals cases closed in fiscal year 2004: 28,592; Percentage of Appeals workload: 27.5; Percentage of cases compliance closed that are appealed: 0.4; Number of cases in GAO's sample: 40; GAO sample percentage of cases not fully sustained: 73[B]. Appeals workstreams: Penalty Appeals; Number of appeals cases closed in fiscal year 2004: 14,647; Percentage of Appeals workload: 14.1; Percentage of cases compliance closed that are appealed: 0.1; Number of cases in GAO's sample: 16; GAO sample percentage of cases not fully sustained: 69[C]. Appeals workstreams: Other; Number of appeals cases closed in fiscal year 2004: 4,558; Percentage of Appeals workload: 4.4; Percentage of cases compliance closed that are appealed: 0.1; Number of cases in GAO's sample: 9; GAO sample percentage of cases not fully sustained: 56[C]. Appeals workstreams: Total; Number of appeals cases closed in fiscal year 2004: 103,946; Percentage of Appeals workload: 100; Percentage of cases compliance closed that are appealed: 0.3; Number of cases in GAO's sample: 153; GAO sample percentage of cases not fully sustained: 41. Source: GAO sample of Appeals cases closed during fiscal year 2004, IRS Data Book 2003, and IRS data. [A] N/A means not applicable. IC and CIC workstreams are not included in our sample because these cases are very complex and the supporting documentation is voluminous. See appendix I. [B] GAO's sample results for three workstreams--Collection Due Process, Exam/TEGE, and Offer-in-Compromise--are statistically generalizable to the specific workstream population. The rate of cases not fully sustained for Exam/TEGE differs from Collection Due Process and from Offer-in-Compromise cases. The margin of error for these estimates is larger than for the sample as a whole due to the smaller sample sizes when looking at each workstream separately. Confidence intervals are shown in appendix I. [C] GAO's sample results for Innocent Spouse, Penalty Appeals, and Other are not statistically generalizable because the sample sizes were not large enough. Consequently, these workstream results apply only to cases in our sample. [End of table] Managers of programs with low appeal rates may not see as much benefit in obtaining feedback from Appeals. With an appeal rate of less than one-half of 1 percent, managers in W&I, the source of many cases in the Exam/TEGE workstream, explained that they had limited interest in devoting resources to analyzing Appeals feedback information, although they would review any analysis Appeals provided to them. Managers told us that, given the challenges facing W&I, they needed to focus resources on other issues. However, although managers might not see much direct benefit for their programs, reducing the appeal rate for compliance programs could benefit Appeals. As shown in table 2, the Exam/TEGE, Penalty Appeals, and Other workstreams have appeal rates of less than 1 percent, but the cases from these three workstreams make up about half of Appeals' workload. In addition, the cases that are appealed are generally not fully sustained. The percentage of cases in our sample that were not sustained ranged from 73 percent for Exam/TEGE to 56 percent for the Other workstream.[Footnote 16] From Appeals' perspective, improving case results in these workstreams could represent a target of opportunity for reducing its case load and increasing its efficiency. Analysis of our sample found that across all workstreams, Appeals cases that are fully sustained require about half of the staff hours of cases that are not fully sustained. If providing feedback to compliance programs improved their decision making, taxpayers would benefit as well. For example, if compliance programs used feedback information to improve their understanding of how to apply tax laws and regulations, they could reduce the number of taxpayers requesting an appeal and therefore resolve cases more quickly and with more uniform decisions. Further, since Appeals managers said some taxpayers decide not to pursue an appeal even though they disagree with a compliance decision, more consistent application of the tax laws or regulations could also improve the fairness and accuracy of their outcomes. The Exam/TEGE workstream can be used as a hypothetical example of the potential effect of these benefits. If the quality of compliance case decisions were to improve and as a result the percentage of cases fully sustained in Appeals were to increase from 28 percent to 38 percent, Appeals would save an estimated 7 staff years.[Footnote 17] Another potential cost saving would result if fewer taxpayers appealed because the quality of compliance case decisions improved. For example, if the number of cases from the Exam/ TEGE workstream that are appealed fell by 10 percent, Appeals would save an estimated 17 staff years.[Footnote 18] More Systematic Data Analysis Would Help to Identify Useful Feedback: Identifying which compliance programs would benefit most from feedback is important given that Appeals hears a wide variety of cases, the cases are spread across the operating divisions, and Appeals does not fully sustain cases for a variety of reasons. This dispersion means that in some situations the costs IRS would incur to analyze Appeals data and devise and implement improvements in operations may not be justified given how few cases could be affected. When analyzing our case sample, we found that overall (1) about half of all not fully sustained cases cited either the application of laws and regulations or additional information as the reason for nonsustention and (2) certain workstreams have significantly higher nonsustention rates than others. As shown in table 3, by considering these two facts in combination, we found that two workstreams-Penalty Appeal and Exam/ TEGE-had a large percentage of cases that were not sustained for these two reasons. Other information already available might also be used to identify the most promising areas in which to conduct feedback projects. For example, those cases that are most costly to Appeals to work on, measured for instance by staff hours per case, may yield the most savings to Appeals if the cases could be resolved in the compliance programs without an appeal being made. Appeals and compliance programs have been selecting their projects more on the basis of manager judgment than through data analysis, such as those cases described above. Table 3: Reasons that Appeals Did Not Fully Sustain Compliance Program Decisions, Percentage of Cases Not Fully Sustained, Sample of Cases Closed in Fiscal Year 2004: Reason Appeals did not fully sustain compliance decision[B]: Application of laws or regulations; Percentage of issues cited in cases not fully sustained[A]: CDP: 3.2; Percentage of issues cited in cases not fully sustained[A]: OIC: 2.2; Percentage of issues cited in cases not fully sustained[A]: Innocent Spouse: 4.3; Percentage of issues cited in cases not fully sustained[A]: Penalty Appeals: 9.7; Percentage of issues cited in cases not fully sustained[A]: Exam/TEGE: 12.9; Percentage of issues cited in cases not fully sustained[A]: Other: 3.2; Percentage of issues cited in cases not fully sustained[A]: Total: 35.5. Reason Appeals did not fully sustain compliance decision[B]: Additional information provided by the taxpayer or not accepted by compliance; Percentage of issues cited in cases not fully sustained[A]: CDP: 4.3; Percentage of issues cited in cases not fully sustained[A]: OIC: 1.1; Percentage of issues cited in cases not fully sustained[A]: Innocent Spouse: 0.0; Percentage of issues cited in cases not fully sustained[A]: Penalty Appeals: 1.1; Percentage of issues cited in cases not fully sustained[A]: Exam/TEGE: 20.4; Percentage of issues cited in cases not fully sustained[A]: Other: 3.2; Percentage of issues cited in cases not fully sustained[A]: Total: 30.1. Reason Appeals did not fully sustain compliance decision[B]: Original audit work or significant rework by Appeals; Percentage of issues cited in cases not fully sustained[A]: CDP: 2.2; Percentage of issues cited in cases not fully sustained[A]: OIC: 0.0; Percentage of issues cited in cases not fully sustained[A]: Innocent Spouse: 0.0; Percentage of issues cited in cases not fully sustained[A]: Penalty Appeals: 1.1; Percentage of issues cited in cases not fully sustained[A]: Exam/TEGE: 5.4; Percentage of issues cited in cases not fully sustained[A]: Other: 1.1; Percentage of issues cited in cases not fully sustained[A]: Total: 9.7. Reason Appeals did not fully sustain compliance decision[B]: Taxpayer not responsive to compliance; Percentage of issues cited in cases not fully sustained[A]: CDP: 3.2; Percentage of issues cited in cases not fully sustained[A]: OIC: 0.0; Percentage of issues cited in cases not fully sustained[A]: Innocent Spouse: 0.0; Percentage of issues cited in cases not fully sustained[A]: Penalty Appeals: 0.0; Percentage of issues cited in cases not fully sustained[A]: Exam/TEGE: 5.4; Percentage of issues cited in cases not fully sustained[A]: Other: 0.0; Percentage of issues cited in cases not fully sustained[A]: Total: 8.6. Reason Appeals did not fully sustain compliance decision[B]: Hazards of litigation; Percentage of issues cited in cases not fully sustained[A]: CDP: 0.0; Percentage of issues cited in cases not fully sustained[A]: OIC: 0.0; Percentage of issues cited in cases not fully sustained[A]: Innocent Spouse: 1.1; Percentage of issues cited in cases not fully sustained[A]: Penalty Appeals: 3.2; Percentage of issues cited in cases not fully sustained[A]: Exam/TEGE: 3.2; Percentage of issues cited in cases not fully sustained[A]: Other: 1.1; Percentage of issues cited in cases not fully sustained[A]: Total: 8.6. Reason Appeals did not fully sustain compliance decision[B]: Appeals changed collection alternative; Percentage of issues cited in cases not fully sustained[A]: CDP: 5.4; Percentage of issues cited in cases not fully sustained[A]: OIC: 1.1; Percentage of issues cited in cases not fully sustained[A]: Innocent Spouse: N/A[C]; Percentage of issues cited in cases not fully sustained[A]: Penalty Appeals: N/A[C]; Percentage of issues cited in cases not fully sustained[A]: Exam/TEGE: N/A[C]; Percentage of issues cited in cases not fully sustained[A]: Other: 1.1; Percentage of issues cited in cases not fully sustained[A]: Total: 7.5. Reason Appeals did not fully sustain compliance decision[B]: Total; Percentage of issues cited in cases not fully sustained[A]: Total: 100. Source: GAO sample of Appeals cases. [A] The 63 cases not fully sustained from the sample of 153 cases cited 93 reasons for nonsustained cases. Totals many not add due to rounding. [B] Confidence intervals for the Penalty Appeals and Exam/TEGE application of laws and regulations and additional information provided by the taxpayer or not accepted by compliance for Exam/TEGE are presented in appendix I. Confidence intervals for the other cells are not provided because the sample size is not large enough to project results. [C] N/A means not applicable. These cells are not applicable because Innocent Spouse, Penalty Appeals, and Exam/TEGE are Appeals workstreams related to tax assessments, and the alternatives for the collection of delinquent tax do not apply. [End of table] As discussed earlier, Appeals and compliance program managers will need to sort through possible reasons why some areas appear to have high levels of nonsustained cases. This may require several iterations of data analysis, discussion, and manager judgment. Once officials have identified the areas with the greatest potential for improvement, Appeals and compliance programs can explore low-cost avenues for using feedback information. For example, in two of the three workstreams with the highest percentage of appealed cases, Appeals and compliance programs have completed some projects based on Appeals case results. The joint study on Offer-in-Compromise cases not sustained by Appeals was conducted by a small team of compliance and Appeals staff and involved the review of 113 cases in 1 week. In contrast, Appeals and LMSB concluded that jointly reviewing fully conceded issues in the CIC program was too expensive because these cases can involve numerous complex issues. Rather, Appeals has started to provide to LMSB all Appeals Case Memorandums (ACM) as a low-cost solution for providing the information to target possible areas needing improvement.[Footnote 19] However, Appeals has not yet explored potential avenues for using feedback information in other large workstreams, such as Penalty Appeals. Appeals Can Build on Existing Efforts to Improve the Feedback Project: Appeals has taken several steps to launch and begin expanding the feedback project. As Appeals and compliance managers gain experience in analyzing and using feedback information, Appeals, in partnership with compliance managers, can build upon those efforts by identifying the additional feedback information that needs to be shared and further developing results-oriented objectives. In addition, potentially useful feedback data contain errors that undermine their usefulness. Appeals Is Providing Some Feedback Based on Officials' Initial Judgments of Its Needs: Appeals has taken several steps to launch and expand the feedback project. For example, officials are sending ACMs to certain compliance programs. During 2005, Appeals started to send ACMs to: * LMSB Industry Case and Coordinated Industry Case programs, * W&I for the Innocent Spouse and EITC programs, * SB/SE for the Collection Due Process, Offer-in-Compromise, and International Examination programs,[Footnote 20] and: * TE/GE for some Exempt Organizations cases. The Collection Due Process program also receives some summary-level information on whether the taxpayer and Appeals agreed on the outcome of an appeal as well as the Appeals inventory level. Appeals and the compliance programs are working together to determine which additional programs should receive specific feedback information. In addition, Appeals and compliance programs' staff meet regularly through advisory board meetings. The advisory boards were created to focus on important cross-functional issues, solve problems, identify new issues arising in the compliance programs, and generally maintain close working relationships. For example, as previously discussed, Appeals and the SB/SE collection staff jointly worked on a review of Offer-In-Compromise cases to determine why Appeals accepted some offers that the compliance program rejected and are considering similar efforts with other compliance programs. Generally, decisions on what information to initially share with the compliance programs have grown out of discussions between Appeals and compliance staff and reflect their best judgment about the information that likely will help the compliance programs improve case results. Appeals and the compliance programs are still determining what additional feedback information should be shared. Appeals, in coordination with the compliance programs, is revising its case closing documents to provide additional information describing the basis for the resolution of a case. For example, Appeals is working with the compliance programs to provide information on whether additional information was considered by Appeals, cases were closed based on hazards of litigation, or taxpayers did not respond or delayed their responses. Compliance program managers told us that providing more detailed information tailored to their needs will help them to improve their results. Appeals plans to implement a revised Appeals case closing system during 2006. Appeals managers believe, and we agree, that the compliance programs will likely identify additional information needs in the future as they begin to analyze and use the information. For example, some compliance managers have told us that information on sustention rates would be useful. However, Appeals' has no plans to develop this information. Partnership with Compliance Programs Could Help Ensure That Feedback Information Will Be Used to Achieve Desired Program Results: Although Appeals has worked with the compliance programs on many aspects of the feedback project, Appeals developed the objectives and performance measures for the feedback project with relatively little input from the compliance programs. After initially developing the objectives and measures, Appeals distributed them to compliance program representatives for comment but received little response. Appeals officials therefore concluded the compliance programs agreed with the objectives and measures. Best practices in strategic planning, of which setting objectives is a part, call for the involvement of stakeholders.[Footnote 21] In the case of the feedback project, involving the compliance programs in establishing the project objectives is particularly important because the programs themselves must play active roles in the project to make any changes that will improve their case results. Appeals officials acknowledge that mutually agreed upon objectives and measures would increase the likelihood that compliance programs would use the feedback information provided. However, gaining consensus may not be easy. The following illustrates the importance of involving the compliance programs in these decisions and of the potential difficulties that could arise. Appeals could adopt an objective for the feedback project of improving the sustention rate for compliance program cases that go to Appeals. That is, if the quality of compliance decisions is improved through feedback of Appeals case results, more appealed cases should be upheld. However, some of the compliance program officials we interviewed do not want improvement in the sustention rate to be an objective for a variety of reasons. For example, officials note that Appeals can change a case result for reasons that are out of their control. As discussed before, Appeals is authorized to close cases based on hazards of litigation and compliance programs are not. As a result, compliance managers are concerned that including cases closed based on hazards of litigation as part of a sustention rate would be unfair. Other managers do not agree that Appeals always makes the correct decisions on compliance cases. Thus, the active involvement of compliance program officials in the selection of objectives would be important to determining the strengths and weaknesses of potential objectives. Further, the involvement of compliance programs in establishing objectives and project measures may better ensure that the feedback project is focusing on desired results. As defined by Appeals in April 2005, the objectives of the case feedback project are to: * build strong relationships between Appeals and operating divisions and functions, * capture and share trend data, * analyze trend data-and provide meaningful commentary to the operating divisions and functions, and: * influence operating division policy and procedure. Although these objectives indicate some of the activities that are integral to feedback sharing and a desired outcome--influence on operating divisions' policies and procedures--involving the operating divisions in considering program objectives would provide an opportunity to build on these objectives to more fully define the results intended for the feedback project. The Commissioner did not specify the benefits that he thought should result from sharing of Appeals case information with the operating divisions and their compliance programs. However, as discussed earlier, sharing this information has the potential to improve the operations of the divisions and, consequently, the quality of their case decisions, potentially increasing the case sustention rate and taxpayer satisfaction with the Appeals process while also decreasing the time to complete an appeal. In addition, sharing information may also improve Appeals' decision making by, for example, clarifying IRS's interpretation of new or particularly complex tax laws so that both Appeals and compliance managers apply them consistently. By working with the compliance programs, Appeals would have the opportunity to further refine the project objectives to more specifically identify which of these possible results-oriented improvements are being sought by the project. For example, as mentioned earlier, the Merit Systems Protection Board has set a performance goal of maintaining or reducing its low percentage of appealed decisions that are reversed or sent back to the board. To the extent that new objectives are identified, Appeals and the compliance programs would need to ensure that appropriate performance measures are developed to track progress toward those objectives. Project Is Correcting Flaws in Data, but Sufficient Internal Controls to Monitor and Verify Data Not Yet Developed: When we compared data in Appeals' Centralized Database System (ACDS)[Footnote 22] to documentation in closed Appeal case files, we found significant error rates related to data that would be used for a case feedback project. The highest error rates in the fields were related to the results of an appeal, such as in the revised tax, revised penalty amounts, or case-closing code field.[Footnote 23] For example, 14 percent of the cases contained errors in the revised tax field. These errors related to the outcome data that likely would be included as part of any feedback information provided to the compliance programs and would diminish the information's usefulness to compliance program managers. Further, 12 of 165 cases (7 percent) could not be analyzed because the files could not be located or essential Appeals documents were not available.[Footnote 24] On the basis of error rates identified, we reviewed internal controls[Footnote 25] for processing case results data and identified several internal control weaknesses that may have contributed to inaccurate data in ACDS. For example, we were informed by Appeals that some appeals officers, who are responsible for working the taxpayers' case, did not verify that ACDS data, such as the amount of tax or penalty owed by the taxpayer, was entered into ACDS accurately. Appeals policy requires that the appeals officers verify the key data in ACDS, such as the statute of limitations date, when a case is received. When a case is completed, Appeals procedures require the case manager, who supervises the appeals officer, to review and sign case-closing documents, which include data such as the amount of proposed tax or penalty and case-closing code. The closed-case data are then entered into the ACDS information system by the Appeals Processing Services staff. According to Appeals, once the case is sent to Processing Services for data entry, the appeals officer and case manager generally do not see the case again and do not know whether the closing data have been entered into ACDS accurately. Appeals guidance does not require that the appeals officer or Processing Services staff verify whether the data were accurately entered. According to Processing Services staff, appeals officers may not ensure that case-closing documents are complete. For example, data, such as the amount of revised tax or penalty (the amount of tax or penalty as determined by Appeals) or the closing code, may not be entered on the closing document by the appeals officers. Processing Services staff said that in these cases, they must review the case file to determine the correct closing data and enter that data into ACDS. The staff stated that identifying the correct data may be difficult in complex cases. Other internal controls only partly compensate for the lack of data entry verification. Appeals performs an annual Inventory Validation Listing process for open cases where critical fields in ACDS are verified and errors identified are corrected in ACDS. Since only open cases are reviewed, fields with closing data, such as revised tax, revised penalty, and closing code, are not reviewed. These closing fields are critical to the feedback loop process and without verification inaccurate data could be sent to the compliance programs. Appeals is making efforts to improve the accuracy of the data in ACDS. Appeals, for the first time, completed a data reliability study of ACDS in 2005. This study consisted of a random probability sample of 1,568 Appeals cases where data fields that were considered critical or were used daily were tested. From the study, Appeals identified data accuracy and internal control issues that were consistent with our findings. Appeals found that some fields in ACDS had lower than expected accuracy rates. For example, the revised tax field for the Innocent Spouse workstream had an accuracy rate of 71.9 percent, while the revised penalty field for the Other workstream was 78.1 percent.[Footnote 26] Appeals also identified that improvements were needed in (1) internal controls including training of Processing Services staff on ACDS input procedures, (2) ACDS data fields with lower than expected accuracy rates, and (3) Appeals' section of the Internal Revenue Manual, which includes guidelines for standard data accuracy reviews. Appeals has been revising its database and related data entry procedures to improve the accuracy of the data in ACDS. Case-closing documents are being redesigned in a computer-based format so that only data which are appropriate to the case under appeal could be selected, thus reducing the potential for errors. Although Appeals is making efforts to improve the accuracy of the data in ACDS, it has not completed plans to address all of the identified data accuracy issues. Appeals will likely continue to experience data accuracy issues unless it improves its internal controls to verify, on an ongoing basis, the accuracy of case data entered into ACDS. Conclusions: Using the results of Appeals case outcomes has the potential to improve compliance programs' case results and service to taxpayers with benefits that could accrue to the divisions, Appeals, and taxpayers. Nevertheless, given the scope of Appeals' work, careful targeting of investments to use Appeals information is needed to ensure that the benefits will be significant enough to justify the costs IRS incurs to collect and analyze Appeals data and make changes in policies, procedures, or practices based on those analyses. Because relatively few compliance program cases may be affected by the use of some Appeals feedback information, officials need to be judicious in selecting topical areas to study. Opportunities exist to move beyond professional judgment in selecting these areas to a more data-driven approach. However, to maximize the benefits of sharing Appeals information, as intended by the Commissioner, the officials need to better define what the program is intended to achieve and how results will be measured. Appeals and the compliance programs need to enter into an active partnership to develop results-oriented objectives and associated performance measures for the feedback project. Finally, the feedback project must be built on reliable data, which requires that better internal controls be instituted to drive down the error rates in key data that will be provided to the compliance programs. Recommendations: We are making recommendations to the Commissioner of Internal Revenue to ensure that the feedback project reaches its maximum potential in improving case results. Specifically, we recommend that the Commissioner direct Appeals: * in partnership with the compliance programs, to analyze Appeals case- results data, such as the workstream sustention rates, reasons for nonsustention, or staff hours spent per case, to identify areas in which improvements are likely to generate the greatest benefits to the compliance programs, Appeals, and taxpayers; * in partnership with the compliance programs, to further investigate the most promising areas and assess whether actions, such as additional guidance or training, are needed to improve the quality of compliance programs' case decisions; * in partnership with the compliance programs, to further develop results-oriented objectives and associated performance measures for the feedback project; and: * to build upon its current efforts to improve the quality of Appeals information for the feedback project by establishing internal controls to verify, on an ongoing basis, the accuracy of the data entered into Appeals information systems on case results. Agency Comments and Our Evaluation: The Commissioner of Internal Revenue provided written comments on a draft of this report in a March 6, 2006, letter which is reprinted in appendix V. The Commissioner agreed with our recommendations and said they will help IRS develop a much stronger feedback program. With regard to the first recommendation, the Commissioner said IRS would continue quarterly meetings between the operating divisions and Appeals and report national-level feedback data at least annually to identify specific compliance programs where shared benefits would be realized. We agree that these actions would be a first step toward implementing the recommendation. However, as discussed in the report, more systematic data analysis would help Appeals and the compliance programs identify areas more likely to realize the benefits of using feedback data. As discussed in the Commissioner's comments, this may involve reviewing external data, such as the National Taxpayer Advocate's reports, as well as other data to identify areas that may yield the most savings to IRS if the cases were resolved in the compliance programs without appeals. The analysis of existing data is a necessary step toward tailoring analyses to each of IRS's compliance programs. As agreed with your offices unless you publicly announce its contents earlier, we plan no further distribution of this report until 30 days after its issue date. At that time, we will send copies of this report to the Secretary of the Treasury, The Commissioner of Internal Revenue, and other interested parties. Copies will be made available to others upon request. This report is available at no charge on GAO's Web site at [Hyperlink=http://www.gao.gov]. If you or your staff have any questions, please contact me at (202) 512- 9110 or Jonda Van Pelt, Assistant Director, at (415) 904-2186. We can also be reached by e-mail at brostekm@gao.gov or vanpeltj@gao.gov, respectively. Contact points for our offices of Congressional Relations and Public Affairs may be found on the last page of this report. Key contributors to this report were Carl Barden, Evan Gilman, Leon Green, Shirley Jones, Laurie King, Ellen Rominger, and Michael Rose. Signed by: Michael Brostek: Director, Tax Issues Strategic Issues Team: [End of section] Appendix I: Objectives, Scope, and Methodology: Our objectives were to determine whether (1) information on Appeals results has the potential to provide useful feedback to the Internal Revenue Service (IRS) operating divisions to benefit compliance programs, Appeals, and taxpayers through better case resolution and (2) the feedback project was being effectively managed to maximize its potential to improve IRS's performance and thereby reduce disputes with taxpayers. To determine whether information on the results of Appeals cases has the potential to provide useful feedback and whether the feedback project is being effectively managed, we interviewed 24 Appeals executives, managers, and staff who work with compliance program staff on feedback issues, coordination, or information systems issues. We also interviewed 58 compliance program executives, managers, and staff selected by operating division liaisons to represent their compliance programs because of their familiarity with Appeals issues. We discussed with these officials the type of feedback data that are being collected by Appeals and sent to the compliance program officials as well as the type of feedback data compliance program officials would like to receive from Appeals. We reviewed documents provided by Appeals on the feedback project. We reviewed the Appeals Centralized Database System (ACDS) to determine whether it contained sufficient case results information. We found that it did not contain sufficient information for our analyses, such as whether Appeals agreed with the compliance decision. Therefore, to develop this information, we selected a random probability sample of case files to review. The sample was drawn from an initial population of 103,946 Appeals cases closed in ACDS for fiscal year 2004. However, since Industry Case (IC) and Coordinated Industry Case (CIC) cases, which originate from IRS's Large and Mid-Size Business Division (LSMB), are complex and the supporting documentation is voluminous, we excluded these 1,323 cases from the population. Therefore, the final population size was 102,623 cases. Of the 165 cases selected in our sample, we reviewed 153 cases to determine the results of the cases. The remaining 12 cases could not be analyzed because the files could not be located or essential Appeals documents were not available. We assessed the known characteristics of the 12 cases not received against those of the 153 received for potential systematic differences. Based on this nonresponse bias analysis, we concluded that it was acceptable to treat the 12 cases as missing at random. We reviewed documents, such as the Appeals Case Memorandum and the Case Activity Record, and determined whether the cases were fully sustained, partially sustained, or not sustained. In determining the extent to which a case was sustained, we based our decision on the determination made by Appeals in the Appeals Case Memorandum using the following scale: "fully sustained" indicated that in our judgment Appeals agreed with compliance on all issues appealed by the taxpayer; "partially sustained" indicated that Appeals agreed with at least one but not all of the issues; and "not sustained" indicated that Appeals did not agree with any of the issues. We also reviewed the cases to determine the reasons the cases were not sustained by Appeals. Since cases could include several compliance issues, there may have been multiple reasons why a case was not sustained. We recorded each decision and the reason for the decision cited in the Appeals case file for a case not being sustained on a data collection instrument (DCI) that we developed. The analysts who participated in reviewing the case files and recording the information on the data collection instrument were knowledgeable about the appeals process and how to interpret the information in the case files. To ensure that the data entered on the DCIs conformed to GAO's data quality standards, each completed DCI was reviewed by at least one other GAO analyst. The reviewer compared the data recorded on the DCI to the data in the case files to determine whether he or she concurred with the interpretation of the case files and the way the data were recorded on the DCI. When there were differing perspectives, the analysts met and reconciled them. Tabulations of the DCI items were automatically generated using a statistical software package to develop case outcome information. For these analyses, the computer programs were checked by a second, independent analyst. We developed case outcome information for each of the Appeals workstreams except IC and CIC. For the CDP, Exam/TEGE, and OIC workstreams, our sample sizes were large enough to generalize the results separately for each workstream, or to have a margin of error small enough to produce meaningful workstream estimates. Because we followed a probability procedure based on random selection, our sample 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, plus or minus 8 percentage points. This is the interval that would contain the actual population value for 95 percent of the samples we could have drawn. For example, Appeals did not sustain 41 percent of the cases in the sample, which has a 95 percent confidence interval of 33 percent as a lower bound and 49 percent as an upper bound. Workstream estimates come from subsets of the sample. Thus workstream-specific estimates have larger confidence intervals due to the smaller sample size. Tables 4, 5, and 6 present the confidence intervals for sample data presented in the report. Table 4: Confidence Intervals for Table 1: Reason Appeals did not fully sustain compliance decision: Application of laws or regulations; Percentage of cases not fully sustained: 52; Confidence intervals: Lower bound: 40.0; Confidence intervals: Upper bound: 64.8. Reason Appeals did not fully sustain compliance decision: Additional information provided by the taxpayer or not accepted by compliance; Percentage of cases not fully sustained: 44; Confidence intervals: Lower bound: 32.1; Confidence intervals: Upper bound: 56.8. Reason Appeals did not fully sustain compliance decision: Original audit work or significant rework by Appeals; Percentage of cases not fully sustained: 14; Confidence intervals: Lower bound: 6.7; Confidence intervals: Upper bound: 25.5. Reason Appeals did not fully sustain compliance decision: Taxpayer not responsive to compliance; Percentage of cases not fully sustained: 13; Confidence intervals: Lower bound: 5.6; Confidence intervals: Upper bound: 23.6. Reason Appeals did not fully sustain compliance decision: Hazards of litigation; Percentage of cases not fully sustained: 13; Confidence intervals: Lower bound: 5.6; Confidence intervals: Upper bound: 23.6. Reason Appeals did not fully sustain compliance decision: Appeals changed collection alternative; Percentage of cases not fully sustained: 11; Confidence intervals: Lower bound: 4.5; Confidence intervals: Upper bound: 21.7. Source: GAO. Note: Based on a subset of 63 cases not sustained by Appeals. [End of table] Table 5: Confidence Intervals for Table 2: Appeals workstreams: Offer-in-Compromise; GAO sample: percentage of Appeals cases not fully sustained: 14; Number of cases in the sample for this workstream: 28; Confidence intervals: Lower bound: 4; Confidence intervals: Upper bound: 33. Appeals workstreams: Innocent Spouse; GAO sample: percentage of Appeals cases not fully sustained: 57; Number of cases in the sample for this workstream: 7; Confidence intervals: Lower bound: 18; Confidence intervals: Upper bound: 90. Appeals workstreams: Collection Due Process; GAO sample: percentage of Appeals cases not fully sustained: 19; Number of cases in the sample for this workstream: 53; Confidence intervals: Lower bound: 9; Confidence intervals: Upper bound: 32. Appeals workstreams: Exam/TEGE; GAO sample: percentage of Appeals cases not fully sustained: 73; Number of cases in the sample for this workstream: 40; Confidence intervals: Lower bound: 56; Confidence intervals: Upper bound: 85. Appeals workstreams: Penalty Appeals; GAO sample: percentage of Appeals cases not fully sustained: 69; Number of cases in the sample for this workstream: 16; Confidence intervals: Lower bound: 41; Confidence intervals: Upper bound: 89. Appeals workstreams: Other; GAO sample: percentage of Appeals cases not fully sustained: 56; Number of cases in the sample for this workstream: 9; Confidence intervals: Lower bound: 21; Confidence intervals: Upper bound: 86. Appeals workstreams: Total; GAO sample: percentage of Appeals cases not fully sustained: 41; Number of cases in the sample for this workstream: 153; Confidence intervals: Lower bound: 33; Confidence intervals: Upper bound: 49. Source: GAO. [End of table] Table 6: Confidence Intervals for Table 3: Reasons Appeals did not fully sustain compliance decision: Application of laws and regulations; Penalty Appeals: Percentage not sustained: 9.7; Penalty Appeals: Confidence interval: Lower bound: 4.5; Penalty Appeals: Confidence interval: Upper bound: 17.6; Exam/TEGE: Percentage not sustained: 12.9; Exam/TEGE: Confidence interval: Lower bound: 6.8; Exam/TEGE: Confidence interval: Upper bound: 21.5. Reasons Appeals did not fully sustain compliance decision: Additional information provided by the taxpayer or not accepted by compliance; Penalty Appeals: Percentage not sustained: 1.1; Penalty Appeals: Confidence interval: Lower bound: N/A; Penalty Appeals: Confidence interval: Upper bound: N/A; Exam/TEGE: Percentage not sustained: 20.4; Exam/TEGE: Confidence interval: Lower bound: 12.8; Exam/ TEGE: Confidence interval: Upper bound: 30.1. Source: GAO. Note: N/A means not applicable because the confidence interval was not computed. [End of table] To determine how effectively the feedback project was being managed, we reviewed documents supplied by Appeals and compliance program officials, such as meeting minutes for the advisory boards and strategic planning documents. We also interviewed these officials and reviewed our prior work on best practices for developing information that can be used to improve agency performance. To compute appeal rates, we compared compliance cases closed in fiscal year 2003 by workstream to the Appeals cases closed in fiscal year 2004. Since Appeals typically required about a year to complete a case, the 2004 Appeals closings were cases that were most likely closed by compliance programs during 2003. Further, IRS uses a similar approach to compute audit rates. To identify the number of compliance cases by workstream, we used data published in IRS's fiscal year 2003 Databook. Data on cases closed for the Innocent Spouse and Offer-in-Compromise workstreams was not available in the Databook and was provided by IRS staff. Data Reliability: We assessed whether the case results data contained in ACDS were sufficiently reliable for our use. We selected the first 100 cases from our random sample of 165 cases to make this determination. We interviewed knowledgeable Appeals officials about the data, performed electronic testing of relevant data fields for obvious errors in accuracy and completeness, and collected and reviewed documentation about the data and the system. We also reviewed prior Treasury Inspector General for Tax Administration reports. Of the 100 cases selected for our sample, we reviewed 92 cases from all of the Appeals workstreams except as mentioned earlier, the IC and CIC workstreams. The remaining 8 cases could not be analyzed because essential Appeal documents were not available. We compared documents in closed Appeals cases, such as the Appeals Case Memorandum, to data in ACDS. However, Appeals did not always provide documentation for the basis of the compliance determination; therefore, in some cases, we were unable to determine if data, such as the amount of tax proposed by compliance, were accurate. We had Appeals verify data errors in fields that were specific to case results information, such as the amount of revised tax and penalty, as well as the closing code. Due to the high error rate of some data fields in our sample, we reviewed internal controls used in the processing of case results data at one Appeals area office. This review consisted of observation and inquiry of Appeals officials on Appeals' case processing procedures and review of Appeals documentation. We also spoke to officials in Appeals headquarters concerning weaknesses identified in Appeals' internal controls. On the basis of our data reliability review of ACDS, we determined that data in ACDS were not sufficiently reliable for our use. Instead of relying on that data, we used data developed from our sample of Appeals cases and continued our analyses of Appeals' internal controls. We conducted our review at Appeals headquarters in Washington, D.C., and one Appeals area office from October 2004 through October 2005 in accordance with generally accepted government auditing standards. [End of section] Appendix II: Appeals Workstreams: Appeals' workload is organized into eight workstreams. These workstreams include cases that have similar characteristics rather than reflecting the IRS operating division where they originated. For example, cases in the Collection Due Process workstream include only appeals by taxpayers under provisions of the IRS Restructuring and Reform Act 1998, which authorizes an independent review by Appeals of proposed levies and filed liens. These cases could originate in either the Wage and Investment Division or the Small Business and Self- Employed Division, since either division could propose a levy or file a lien. Other workstreams include a wide range of cases from across IRS operating divisions. The Exam/TEGE workstream includes appeals for compliance actions, including recommended assessments and proposed penalties originating from much of IRS's reporting and filing compliance program, with the exclusion of LMSB cases. These appeals can include diverse issues, such as recommended assessments related to the Earned Income Tax Credit or large charitable organizations, such as universities or hospitals. During fiscal year 2004, Appeals completed nearly 104,000 cases. Table 7 describes these workstreams and the IRS operating divisions where these cases where proposed. Table 7: Description of Appeals Workstreams: Appeals workstream: Collection Due Process (CDP); Definition: The IRS Restructuring and Reform Act of 1998 authorized an independent review by Appeals and by the courts of proposed levies and filed liens. The taxpayers, under some circumstances, could have the original assessment reviewed and can request an appeal for liens and levies that are past the 30-day period; Originating IRS division (s): SB/SE; W&I; Appeals cases closed in fiscal year 2004: 32,226; Percentage of total cases: 31.0%. Appeals workstream: Exam/TEGE; Definition: This workstream includes appeals for tax disputes originating with three of four IRS operating divisions' examination functions. The workstream does not include cases originating in LMSB's Industry Case and Coordinated Industry Case programs. This workstream includes cases from a wide range of major IRS compliance programs, including the campus compliance, field examination, exempt organization examinations, employee plans examinations, international issues examinations, tax shelters, and examinations of excise, employment, and estate and gift tax returns; Originating IRS division (s): W&I; SB/SE; TE/GE; Appeals cases closed in fiscal year 2004: 28,592; Percentage of total cases: 27.5%. Appeals workstream: Offer-in-Compromise (OIC); Definition: The Internal Revenue Code authorizes the Secretary of the Treasury to compromise tax debts prior to reference to the Department of Justice for prosecution or defense. Under current regulations, debts may be compromised if there is doubt as to liability, doubt as to collectibility, and for effective tax administration. Appeals cases in this workstream are primarily reviewed by SBSE's centralized OIC collection functions at the Brookhaven and Memphis campuses since offer cases may include taxpayers from each of IRS's four operating divisions; Originating IRS division (s): SB/SE; LMSB; W&I; TE/GE; Appeals cases closed in fiscal year 2004: 17,887; Percentage of total cases: 17.2%. Appeals workstream: Penalty Appeals; Definition: Most of the requests in this workstream relate to requests for abatement of penalties for a taxpayer's failure to file, failure to pay, and failure to deposit. Typically, these are penalties generated through automated compliance tests; Originating IRS division (s): W&I; SB/SE; TE/GE; LMSB; Appeals cases closed in fiscal year 2004: 14,647; Percentage of total cases: 14.1%. Appeals workstream: Innocent Spouse; Definition: Under the Innocent Spouse program, one spouse requests relief from a joint tax liability after assessment because (1) a joint return had an understatement of tax due to erroneous items of the nonrequesting spouse, (2) the requesting spouse did not know and had no reason to know there was a tax understatement at the time the return was signed, and (3) taking into account all facts and circumstances, holding the requester for relief liable for the tax would be unfair. Appeals reviews determinations after rejection by W&I; Originating IRS division (s): W&I; Appeals cases closed in fiscal year 2004: 4,713; Percentage of total cases: 4.5%. Appeals workstream: Other; Definition: These include appeals for several other tax disputes that are not included in other Appeals workstreams. These cases include the Collection Appeals program cases, trust fund recovery penalty cases, requests for abatement of interest, and tax disclosure cases; Originating IRS division (s): W&I; SB/SE; LMSB; TE/GE; Disclosure; Appeals cases closed in fiscal year 2004: 4,558; Percentage of total cases: 4.4%. Appeals workstream: Coordinated Industry Case; Definition: These are appeals related to the recommended assessment of additional tax and penalties for the largest corporate taxpayers in the United States. These taxpayers have been audited by a team of revenue agents rather than a single agent.[A]; Originating IRS division (s): LMSB; Appeals cases closed in fiscal year 2004: 653; Percentage of total cases: .6%. Appeals workstream: Industry Case; Definition: These are appeals related to the recommended assessment of additional tax and penalties for corporations with $10 million or more in assets that are not in the Coordinated Industry Case program.[A]; Originating IRS division (s): LMSB; Appeals cases closed in fiscal year 2004: 670; Percentage of total cases: .6% Appeals workstream: Total cases closed by Appeals; Appeals cases closed in fiscal year 2004: 103,946; Percentage of total cases: 100%. Source: GAO. [A] Appeals cases from the Coordinated Industry Case and Industry Case program were not included in our sample. [End of table] [End of section] Appendix III: Appeal Rates: Taxpayers in each workstream requested Appeals of recommended assessments or other compliance actions, such as proposed levies and filed liens, at widely differing rates. To compute the appeal rate for each workstream, we compared the number of compliance cases closed for each workstream to the number of cases Appeals closed. We compared fiscal year 2003 compliance case closings to fiscal year 2004 Appeals case closings because Appeals averaged 260 calendar days during fiscal year 2004 to complete its work on a case.[Footnote 27] For example, as reported in table 16 of the IRS Databook for 2003, during fiscal year 2003, IRS filed 548,683 notices of federal tax liens, served 1,680,844 notices of levy, and made 399 seizures for a total of 2,229,926 compliance actions. Each of these actions could be the basis for a CDP appeal. During fiscal year 2004, Appeals completed work on 32,226 CDP cases, for an Appeal rate of 1.445 percent or 1 percent. Table 8: Appeal Rates by Workstream, Fiscal Year 2004: Appeals workstream: CIC; Number of compliance cases closed in fiscal year 2003: 2,287; Number of cases closed by Appeals in fiscal year 2004: 653; Appeal rate: 29%. Appeals workstream: IC; Number of compliance cases closed in fiscal year 2003: 4,769; Number of cases closed by Appeals in fiscal year 2004: 670; Appeal rate: 14%. Appeals workstream: OIC[A]; Number of compliance cases closed in fiscal year 2003: 136,822; Number of cases closed by Appeals in fiscal year 2004: 17,887; Appeal rate: 13%. Appeals workstream: IS[B]; Number of compliance cases closed in fiscal year 2003: 57,606; Number of cases closed by Appeals in fiscal year 2004: 4,713; Appeal rate: 8%. Appeals workstream: CDP[C]; Number of compliance cases closed in fiscal year 2003: 2,229,926; Number of cases closed by Appeals in fiscal year 2004: 32,226; Appeal rate: 1%. Appeals workstream: Exam/TEGE[D]; Number of compliance cases closed in fiscal year 2003: 6,926,956; Number of cases closed by Appeals in fiscal year 2004: 28,592; Appeal rate: .4%. Appeals workstream: Penalty Appeals[E]; Number of compliance cases closed in fiscal year 2003: 17,827,467; Number of cases closed by Appeals in fiscal year 2004: 14,647; Appeal rate: .1%. Appeals workstream: Other[F]; Number of compliance cases closed in fiscal year 2003: 6,078,884; Number of cases closed by Appeals in fiscal year 2004: 4,558; Appeal rate: .1%. Appeals workstream: Total; Number of compliance cases closed in fiscal year 2003: 33,264,717; Number of cases closed by Appeals in fiscal year 2004: 103,946; Appeal rate: .3%. Sources: GAO analysis of IRS data. Note: Compliance cases closed in fiscal year 2003 are most likely to have been closed by Appeals during fiscal year 2004. [A] Offers received in 2003; offer closing data not available. [B] Case closing data provided by IRS. [C] Number of liens filed and levy notices issued, rather than cases closed. [D] Total examinations, less IC and CIC examinations, plus Information Returns program cases, EP/EO examinations, and EITC notices. [E] Excluding employment tax penalties. [F] Total employment tax penalties. Total compliance cases closed in fiscal year 2003 does not include disclosure appeals and abatement of interest. Data on these compliance actions would not have a material effect on the computation of the appeal rate. [End of table] Other approaches could be used to compute appeal rates. Our analysis used compliance cases closed as the basis for measuring appeal rates because (1) a uniform, published source of data was available and provided data on six of the eight Appeals workstreams and (2) it broadly compares IRS compliance programs to the Appeals program. Another approach for measuring appeal rates, for example, could use only cases closed where IRS recommended an additional tax assessment and not include cases where no tax was proposed, because taxpayers would not have a basis for requesting an appeal. In some programs this difference may be substantial. For example, in the Offer-in-Compromise program, according to unpublished data provided by IRS, 58 percent of fiscal year 2003 Offer-in-Compromise cases were closed by Compliance because the offer was not processable or was returned to the taxpayer. Accordingly, the taxpayer did not have a basis for an appeal. Eliminating these cases from the Offer-In-Compromise cases closed in 2003 would more than double the appeal rate from 13 percent to 31 percent. However, limited data are available to use other approaches for computing appeal rates. For example, about 30 percent of the cases closed in Appeals' second largest workstream, Exam/TEGE, originated from the Earned Income Tax Credit and the Automated Underreporter programs. Data were not published on the proportion of these cases that were closed with recommended assessments. [End of section] Appendix IV: Comments from the Commissioner of Internal Revenue: Department Of The Treasury: Internal Revenue Service: Washington, D.C. 20224: Commissioner: March 6, 2006: Mr. Michael Brostek: Director, Tax Issues: Strategic Issues Team: United States Government Accountability Office: Washington, D.C. 20548: Dear Mr. Brostek: Thank you for the opportunity to respond to your draft audit report entitled Tax Administration: Opportunities to Improve Compliance Decisions to Taxpayers through Enhancements to Appeals' Feedback Project (GAO-06-396). Your audit report is very timely since we are in the early stages of program development for the Appeals feedback loop process. Your recommendations will help us develop a stronger feedback process. Your report provides an assessment of two issues: 1) whether information on Appeals results would provide useful feedback to IRS operating divisions and would benefit its compliance programs and 2) whether the feedback process is being effectively managed to maximize its potential to improve IRS's performance and reduce disputes with taxpayers. We agree with the four audit recommendations. Our feedback process is built around on-going interactions, meetings and discussions between Appeals and each operating division. Adopting your recommendations will allow us to strengthen the process by which we provide both individual case resolution information as well as program level-data. I am pleased your report reflects that no one approach for feedback will be possible since the issues and cases vary so significantly by operating division. This supports our plan to tailor feedback to each operating division. We will work and resolve issues through an advisory board process. Our comments on the draft report's specific recommendations are enclosed. If you have any questions, please contact Karen Ammons, Deputy Chief, Appeals at (202) 435-5600. Sincerely, Signed by: Mark W. Everson: Enclosure: Our comments on the report's specific recommendations follow: Recommendation 1: Perform an initial analysis of feedback data to identify areas most likely to generate benefits for compliance programs, Appeals, and taxpayers. To address your recommendation we will continue quarterly Operating Division/ Appeals Board meetings and report national level feedback loop data at least annually to identify specific compliance programs where shared benefits would be realized. Recommendation 2: Investigate whether additional actions are needed to improve the consistency of decisions. The Advisory Board is the forum where we explore additional actions needed to improve consistency of decisions. Through quarterly board meetings we will identify areas where joint Appeals/ operating division program reviews and/or data analysis will serve as tools to identify areas for improvement. The advisory boards will also review external data sources, such as the National Taxpayer Advocates' prior year reports to the Congress and customer satisfaction survey data, to help identify programs which may benefit from more extensive review. In addition, Appeals is in the process of automating feedback data and reports. Once implemented the Advisory Boards will review the data for each operating division and use it to guide and adjust program operations. Recommendation 3: Further develop results-oriented objectives and measures for the feedback project. We agree a shared cross-functional set of objectives and performance measures are necessary for the success of the Feedback Project. From analysis of the feedback data, the cross-functional Advisory Boards will set specific Feedback Project objectives for the feedback project for the coming year. The board will target the most critical types of cases having benefits to both the operating divisions and Appeals and identify the performance measures for the following year's improvement activities. Recommendation 4: Build upon current efforts to improve feedback data by establishing internal controls to verify data accuracy on an ongoing basis. Appeals has just completed an extensive data accuracy review of its Appeals Centralized Database System (ACDS) resulting in a comprehensive report. This review was conducted in conjunction with the Office of Program Evaluation and Risk Analysis (OPERA) using valid sampling methodology. This review and report produced a number of recommendations which Appeals is implementing and will enhance data accuracy and improve the data reliability. Some of the most important recommendations Appeals is adopting are: * Create standard data accuracy reviews by technical and managerial employees as well as headquarters analysts during operational reviews. * Create specific ACDS data validation programs to prevent data entry errors. * Establish clear guidelines on data accuracy procedures. [End of section] Related GAO Products: Tax Administration, Planning for IRS's Enforcement Process Changes Included Many Key Steps but Can Be Improved. G [Hyperlink, http:// www.gao.gov/cgi-bin/getrpt?GAO-04-287] AO-04-287. Washington, D.C.: January 20, 2004. Tax Administration, IRS Needs to Further Refine Its Tax Filing Season Performance Measures. GAO-03-143. Washington, D.C.: November 22, 2002. IRS Modernization: IRS Should Enhance It's Performance Management System. GAO-01-234. Washington, D.C.: February 23, 2001. Standards for Internal Control in the Federal Government. GAO/AIMD-00- 21.3.1. Washington, D.C.: November 1999. Executive Guide: Measuring Performance and Demonstrating Results of Information Technology Investments. GAO/AIMD-98-89. Washington, D.C.: March 1998. Executive Guide: Effectively Implementing the Government Performance and Results Act. GAO/GGD-96-118. Washington, D.C.: June 1996. (450368): FOOTNOTES [1] Compliance programs include numerous IRS examination programs, which may assess additional taxes, and its collection programs, which may take enforcement actions (e.g., liens, levies, seizures) to collect delinquent taxes. Each IRS operating division manages multiple compliance programs. [2] This estimate is based on our random sample. We are 95 percent confident that the actual proportion is between 33 percent and 49 percent. [3] Fifty-two percent of the cases in our sample were not fully sustained for this reason. We are 95 percent confident that the actual proportion is between 40.0 percent and 64.8 percent. [4] Sustention rate is the proportion of cases sent to Appeals that are not changed following review. [5] Of the cases in our sample, 14.1 percent identified errors in the revised tax field (Appeals' determination of the tax liability). We are 95 percent confident that the actual proportion is between 7.7 percent and 23.0 percent. [6] Under the Internal Revenue Code, "levy" is defined as the seizure of a taxpayer's assets to satisfy a tax delinquency. A "lien" is a legal claim, filed in accordance with state property law that attaches to property to secure payment of a debt. [7] 26 CFR 601.106 and Internal Revenue Code sections 6320 and 6330. [8] Hazards of litigation are a substantial uncertainty (1) as to how the courts would interpret and apply the law, (2) about the court's likely factual findings, or (3) about the admissibility or weight that would be given to a specific item of evidence. 26 CFR 601.106(f)2 and Internal Revenue Manual section 8.6.1.3.1. Hazards of litigation are not considered for collection cases. [9] Appeals is the only IRS administrative function with the authority to consider settlements of tax controversies and as such has the primary responsibility to resolve these disputes without litigation to the maximum extent possible. Internal Revenue Manual section 1.2.1.8.4. [10] Collection Due Process cases must be considered by Appeals before review by a court. [11] 26 CFR 601.106(d) 3(iii) and Internal Revenue Manual section 8.4.1.2. [12] Generally these include requests for abatement of interest, Trust Fund Recovery Penalty cases, the Collection Appeals program, and Disclosure issues. [13] Cases may have multiple reasons for closing because a case may include more than one possible instance of noncompliance. Generally, each issue is analyzed separately on its merits and the appeals determination may be based on separate reasons for each issue. [14] The CIC program examines the largest corporate taxpayers in the United States using a team of revenue agents rather than a single agent. [15] The IC program examines corporations with $10 million or more in assets that are not in the Coordinated Industry program. [16] Our sample results for Innocent Spouse, Penalty Appeals, and Other are not statistically generalizable because the sample sizes were not large enough. Consequently, these workstream results apply only to cases in our sample. [17] Estimate is based on our sample of Appeals cases closed in fiscal year 2004. We estimate the savings would be 6.7 staff years and are 95 percent confident that the savings would be between 1.1 and 12.3 staff years. [18] Estimate is based on our sample of Appeals cases closed in fiscal year 2004. We estimate the savings would be 16.7 staff years and are 95 percent confident that the savings are between 12.4 and 21.1 staff years. [19] ACMs document Appeals' decision in a case and explain why Appeals decided to either sustain or overturn compliance decisions. ACMs typically include an analysis of the facts and the applicable tax laws and regulations for the case. [20] The EITC is a refundable tax credit originally intended to offset the burden of Social Security taxes and provide a work incentive for low-income taxpayers. The IC, CIC, Innocent Spouse, Collection Due Process, and Offer-in-Compromise programs are described in appendix II. [21] GAO, Executive Guide: Effectively Implementing the Government Performance and Results Act, GAO/GGD-96-118 (Washington, D.C.: June 1996), p. 13. [22] ACDS is a computer-based system used to control and track cases throughout the appeal process. It generates Appeals management statistics and reports. [23] The case-closing field primarily describes whether the taxpayer agreed or disagreed with Appeals' determination and whether the case was appealed to court. For some workstreams, closing codes also describe whether Appeals sustained the case. [24] Another ongoing assignment also has not located a significant percentage of Appeals files. A subsequent study will more fully analyze the issues related to unavailable case files. [25] Agencies establish internal controls to prevent and detect errors and fraud. [26] Appeals is 95 percent confident that for the revised tax field, the actual proportion is between 65.0 and 78.7 percent, and for the revised penalty field, the actual proportion is between 71.9 and 84.3 percent. [27] This computation also follows the approach that IRS uses to compute audit rates. For that computation, IRS compares the number of returns filed to the number of audits closed in the following year. GAO's Mission: The Government Accountability 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 e-mail alerts" under the "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. Government Accountability 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. Government Accountability Office, 441 G Street NW, Room 7149 Washington, D.C. 20548: