This is the accessible text file for GAO report number GAO-13-480 entitled 'Tax Administration: IRS Could Improve Examinations by Adopting Certain Research Program Practices' which was released on May 24, 2013. 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. United States Government Accountability Office: GAO: Report to the Committee on Finance, U.S. Senate: May 2013: Tax Administration: IRS Could Improve Examinations by Adopting Certain Research Program Practices: GAO-13-480: GAO Highlights: Highlights of GAO-13-480, a report to the Committee on Finance, U.S. Senate. Why GAO Did This Study: One way IRS works to understand taxpayer compliance is by operating NRP, which conducts audits or examinations of individual tax returns in a research sample. NRP examinations are fewer in number but more detailed than IRS’s regular, operational examinations, which focus on the tax returns most likely to have substantive noncompliance. GAO was asked to study whether NRP procedures could improve IRS’s operational examinations. In this report, GAO (1) describes the lessons on examination procedures that IRS learned from conducting NRP examinations; and (2) assesses whether additional improvements could be made to operational examinations based on IRS's NRP experience. GAO reviewed documentation on NRP procedures and interviewed relevant officials. GAO did structured interviews with IRS examiners on ideas for improving operational examinations, reviewed IRS documents, and compared the potential impacts of the ideas to criteria, such as whether the changes would make examinations more efficient or accurate. What GAO Found: The Internal Revenue Service’s (IRS) National Research Program (NRP) has helped test or develop practices that have been used in operational examinations of individual tax returns, according to IRS officials. These practices include, for example, helping test IRS’s remote learning system and helping IRS adopt new examination support software. However, examiners GAO interviewed cited four additional ways NRP practices could improve operational examinations. The first two concern using more data in classification—-IRS’s initial review of tax returns to select issues for examination. Operational classifiers do not have access to as much tax return data as NRP classifiers. NRP transcribes more data from paper-filed tax returns. Without the additional data, operational examiners cannot make optimal decisions about what issues to classify, raising the risk of needlessly examining compliant taxpayers. To make data in operational classification better match NRP levels, examiners suggested: * transcribing more data from paper-filed returns, and, * using all data from electronically filed returns. However, additional transcription imposes costs. For example, IRS estimated that it would cost $8 million annually to transcribe more expense data from individual business tax returns. The magnitude of this type of item (for example, an estimated $103 billion was claimed in 2010 for “other expenses” alone) and the high rate of non- compliance (55 percent in the most recent data) make it likely that better targeted examinations could bring in enough added revenue to justify the cost. Additionally, IRS could acquire more data for classification by revising its policy of treating electronically and paper-filed returns the same. However, the change would only affect classification and not examination. Examinations of paper-filed returns and electronically filed returns are based on all the data. Using more data from electronically filed returns in classification would likely reduce the chances IRS would conduct unnecessary examinations of compliant electronic returns. The third way involves clarifying how to save examination case files electronically. Operational examiners have less specific guidance on saving case files than NRP examiners. Clarifying the key files to be saved electronically would be a way to minimize costs and could help make case sharing and other examiner duties more efficient. The fourth way concerns leveraging NRP for examination staff development. Compared to operational examinations, NRP examinations require more detailed documentation and are broader in scope. Examiners told GAO that such NRP experiences help develop the skills of relatively new staff. However, IRS has no official guidance specifically on when NRP examinations could be used to help develop staff. What GAO Recommends: GAO recommends that IRS transcribe additional data from paper-filed returns, use all electronically filed data, clarify guidance on saving examination case files electronically, and develop guidance on when examiners can work NRP cases. IRS generally agreed with the recommendations but did not specifically comment on transcribing more data from paper-filed returns. View GAO-13-480. For more information, contact James R. White at (202) 512-9110 or whitej@gao.gov. [End of section] Contents: Letter: Background: IRS Officials Said that NRP Helped Improve Operational Examinations in Three Ways, Although No Documentation Exists on Effects: Several NRP Examination Procedures and Practices Could Improve Operational Examinations without Undue Cost: Conclusions: Recommendations for Executive Action: Agency Comments: Appendix I: Scope and Methodology: Appendix II: GAO Reports Regarding NRP Examination Procedures and Data Transcription from Paper Returns: Appendix III: How IRS Uses National Research Program Data: Appendix IV: Agency Comments from the Internal Revenue Service: Appendix V: GAO Contacts and Staff Acknowledgments: Table: Table 1: Differences between National Research Program and Small Business/Self-Employed Operational Examination Practices for Individual Tax Returns: Figures: Figure 1: Small Business/Self-Employed Operational Examination Process: Figure 2: National Research Program Process for Measuring Voluntary Reporting Compliance: [End of section] United States Government Accountability Office: GAO: 441 G St. N.W. Washington, DC 20548: May 24, 2013: The Honorable Max Baucus: Chairman: The Honorable Orrin Hatch: Ranking Member: Committee on Finance: United States Senate: Every year, the federal government loses hundreds of billions of dollars of tax revenue through the tax gap--the difference between taxes owed and taxes voluntarily paid on time. The Internal Revenue Service (IRS) estimated that the tax gap for tax year 2006 (the most recent estimate) totaled $450 billion and IRS enforcement programs-- such as examining voluntary compliance on filed income tax returns-- helped recoup about $65 billion of that figure. Resource limitations and interest in minimizing taxpayer burden prevent IRS from examining anything but a small fraction of the total number of individual income tax returns filed for a given tax year. In fiscal year 2012, IRS examined 1.5 million individual tax returns, which was about 1 percent of all individual tax returns filed in 2011. Because such a small percentage of individual returns are examined, IRS must ensure that the examinations are as efficient and effective as possible. One potential source for lessons learned on improving examinations could be IRS's National Research Program (NRP). IRS generates key information about individual taxpayer compliance from NRP, an effort by the IRS to measure compliance for strategic planning and budget purposes. The program involves detailed examinations of samples of individual tax returns that have been drawn annually since tax year 2006.[Footnote 1] While the NRP data provide crucial information for examination planning, IRS could do more analysis on how the procedures and practices that NRP examiners use for NRP examinations might help improve IRS's operational examinations. You asked us to determine whether lessons from NRP examinations could be used to improve operational examinations. In this report, we (1) describe the lessons on examination procedures that IRS learned from conducting NRP examinations and applied to operational examinations; and (2) assess whether additional improvements could be made to operational examinations based on IRS's NRP experience. To describe lessons IRS has already learned, we asked IRS officials what they saw as the main benefits that NRP had already provided operational examinations. Based on those responses, we gathered documentation, to the extent it was available, in an effort to confirm the benefits that the officials cited. We also gathered information from the Small Business/Self-Employed division (SB/SE)--which is largely responsible for field examinations of individual income tax returns--and interviewed relevant SB/SE officials to confirm the cited benefits, as well. To assess whether additional improvements could be made, we conducted a series of individual and group interviews with NRP and operational examiners. We interviewed all but one of the NRP examiners who were scheduled to perform classification of NRP returns in January 2013. [Footnote 2] During these interviews, we gathered feedback on ways that the examiners thought that NRP examination procedures could be applied to operational examinations. We discussed these ideas with operational examiners and with relevant IRS officials to determine their potential benefits, as well as their costs and tradeoffs. We then compared the ideas with criteria developed from our previous report, the Office of Management and Budget, and the Internal Revenue Manual, including whether the idea would meet an IRS need, whether it would have helped correctly identify misreporting, whether it would make examinations more efficient and whether it would reduce taxpayer burden. Where appropriate, we gathered documentation on costs and tradeoffs of the ideas to help determine the plausibility of the ideas that we collected. The data used in the report came from IRS sources or our previous work and therefore reliability had already been reviewed. We re-examined our earlier reviews for applicability to our engagement and determined that these data were also sufficiently reliable for purposes of this report. Appendix I provides additional information on our scope and methodology. We conducted this performance audit from June 2012 through May 2013 in accordance with generally accepted government auditing standards. Those standards require that we plan and perform the audit to obtain sufficient, appropriate evidence to provide a reasonable basis for our findings and conclusions based on our audit objectives. We believe that the evidence obtained provides a reasonable basis for our findings and conclusions based on our audit objectives. Background: Operational Examinations for Tax Returns Filed by Individuals: Part of SB/SE's enforcement responsibilities includes examining individual tax returns to detect misreporting. As shown in figure 1, SB/SE starts the examination process by building an inventory of returns based on a computer scoring system;[Footnote 3] IRS strategic enforcement priorities, such as high income taxpayers and abusive tax avoidance schemes; and examination training purposes. SB/SE examination plans also provide for a balanced approach for return delivery and allocation of staffing. Figure 1: Small Business/Self-Employed Operational Examination Process: [Refer to PDF for image: illustration] Create inventory: IRS uses a computer scoring program called the discriminant function, as well as criteria from special projects and other priorities to create a pool of returns for possible examination. Classification: From the pool of returns created in the previous step, revenue agents and tax compliance officers conduct a first-pass, in a procedure known as classification, to determine whether the return warrants examination. During this stage, issues with the potential for tax change are classified for examination. If the staff classifying the return finds no issues, IRS takes no further action on that return. Return assignment: Examination office and field managers order returns. The inventory created in the previous step is shipped to the groups and assigned to examiners. Managers have the discretion to dismiss cases they do not deem worthwhile. Examination: Examiners review the assigned tax return by interviewing the taxpayers, their representatives, or both. Examiners are to review taxpayers' records and make a determination on whether any additional taxes are owed. Source: GAO analysis of IRS information. [End of figure] Then, in a process called classification, SB/SE examiners identify whether to examine an individual tax return, which tax issues on the return should be examined, and what examination techniques[Footnote 4] should be used. Examiners who classify returns generally use a computer system called the Compliance Data Environment to record which issues (e.g., items that may have been misreported) on that return get passed on to examiners for official examination. The examiners who classify a tax return are not to be the same ones who examine the return. Generally, SB/SE will limit classification to a handful of issues for individual tax returns. After returns have been identified for further review, an inventory is created from which managers assign cases to examiners or dismiss the case as unproductive. IRS officials said that certain examiners have the discretion to suggest declassifying items and the responsibility to examine any additional items that may not have been classified but that appear large, unusual, or questionable. NRP Examinations: IRS carries out detailed research studies of individual taxpayer compliance through its NRP. The data for the studies come from intensive examinations that strive to verify information taxpayers reported or should have reported. In NRP, IRS selects a random sample of individual tax returns each year. IRS uses the NRP data to estimate the tax gap and update return selection formulas used in planning examinations. The data are also intended to be used to design and revise pre-filing programs.[Footnote 5] As shown in figure 2, the NRP process broadly consists of five steps: (1) selection of returns for examination; (2) case building, the process of adding information to the case file, from IRS and other sources;[Footnote 6] (3) classification, which is done using the NRP classification system instead of the Compliance Data Environment; (4) data collection through examinations; and (5) analysis of the examination results. Figure 2: National Research Program Process for Measuring Voluntary Reporting Compliance: [Refer to PDF for image: illustration] Returns selected for NRP sample: Case building: IRS combines the NRP return sent in by the taxpayer with internal and third-party data in a central NRP case file. Classification: Trained NRP classifiers compare IRS and third-party data to the tax return to determine which issues, if any, should be examined. NRP data collection: * Accepted as filed and accepted with adjustments: An average of 984 per year from tax year 2006 to 2010; * Correspondence examinations: An average of 695 per year from tax year 2006 to 2010; * Face-to-face examinations: An average of 12,201 per year from tax year 2006 to 2010. NRP data analysis: Data from NRP cases are transferred to the NRP database. IRS analyzes the data to update workload selection formulas and identify areas of noncompliance. Source: GAO representation of IRS data. [End of figure] NRP primarily uses field examinations but may use correspondence examinations in certain circumstances. Examiners must review all classified items, ensure that the rest of the return is correct, and record examination activity. Information on NRP-specific activity that IRS intends to later analyze is compiled on Microsoft Excel worksheets in a database maintained specifically for NRP information. NRP and SB/SE Operational Examination Differences: NRP and operational examination procedures for individual tax returns share several attributes. For example, examiners use the same examination guidelines for both processes and both have a classification process. However, key differences exist, which are summarized in Table 1. These differences exist partly because of the programs' different objectives, as shown in the table. Table 1: Differences between National Research Program and Small Business/Self-Employed Operational Examination Practices for Individual Tax Returns: Program: Objective; National Research Program Examinations: Gather data about compliance of U.S. taxpayers to assist IRS; Small Business/Self-Employed Operational Examinations: Identify federal tax misreporting and promote voluntary compliance. Program: Case building; National Research Program Examinations: More return data are provided to examiners than is provided for operational classification. Examination case is created for examiner; Small Business/Self-Employed Operational Examinations: Comparatively limited transcription of return data and other IRS and third party information. Examiners must input information needed to begin a case themselves. Program: Classification; National Research Program Examinations: * Many issues are classified; * Examiners must examine all classified issues; * NRP web-based classification system used for processing; * NRP transcribes additional data from paper-filed returns, which are available to classifiers; Small Business/Self-Employed Operational Examinations: * Fewer issues are classified on returns suspected of having significant noncompliance; * Examiners have the discretion to de-classify issues if they believe that the noncompliance is not apparent or worth pursuing in certain circumstances; * Compliance Data Environment used for processing; * Less tax data from paper-filed tax returns are transcribed. Because of IRS policy to treat paper-filed returns the same as electronically filed returns, untranscribed data are not available to classifiers. Program: Computer validation; National Research Program Examinations: NRP-specific lead sheets and case-closing check sheets are computer validated. An examiner cannot close a case without passing validation checks[A]; Small Business/Self-Employed Operational Examinations: No analogous computer validation of lead sheets or case-closing check sheets is done. Program: Workpaper software; National Research Program Examinations: NRP-specific lead sheets in Microsoft Excel allow data recorded by examiners to be extracted for analysis; Small Business/Self-Employed Operational Examinations: Lead sheets in Microsoft Word do not enable automated data extraction but do enable more detailed descriptions of examination activity. Source: GAO analysis of IRS interviews and documentation. [A] Lead sheets identify the steps that should be taken or considered during an examination of a specific issue. NRP lead sheets are more detailed than those used by operational examiners. Case-closing check sheets are a guideline of steps that examiners must take before they can close the case. [End of table] IRS Officials Said that NRP Helped Improve Operational Examinations in Three Ways, Although No Documentation Exists on Effects: IRS officials said that techniques and practices used in the formative years of NRP contributed to various efforts to improve SB/SE operational examinations. IRS was developing NRP at the same time IRS was developing these efforts, which involved changing procedures, applying new technology, and providing better access to taxpayer data held in various parts of the IRS, according to an IRS official. As such, IRS officials said that the program provided a way to test these envisioned improvements to operational examinations in three ways, as discussed below.[Footnote 7] * NRP provided an opportunity to test remote learning for large numbers of IRS employees. NRP was a pioneer in using remote learning software programs that helped convince IRS management that the programs were viable to use.[Footnote 8] * The case building and classification model that NRP developed for the tax year 2001 Form 1040 study contributed to the standard for operational programs. * NRP facilitated the move of the Report Generation Software to a network environment, allowing IRS to move case data electronically across its computer networks.[Footnote 9] NRP piloted the use of batching software in Report Generation Software to open cases and record the initial information available to IRS that goes into the creation of examination case files and the identification of issues on tax returns for possible examination. Although these NRP benefits to operational examinations may have occurred, IRS had no documentation on the effects--positive or negative--of the three cited NRP activities on operational examination practices. IRS officials attributed the lack of corroborating documentary evidence to the fact that the activities cited by NRP officials took place about 10 years ago and were conducted by other IRS staff who were in different positions at the time or who are no longer with the agency. Further, the cited benefits could be viewed as positive side effects of NRP, which had the primary purpose of providing data to identify noncompliance and measure the tax gap, not improving operational examination procedures. Given that any effects on examination were secondary, IRS had not attempted to measure the extent and effects of the cited benefits. Several NRP Examination Procedures and Practices Could Improve Operational Examinations without Undue Cost: Examiners we interviewed identified several NRP practices and procedures that met our criteria for offering potential improvements to operational examinations.[Footnote 10] These practices include adding additional transcribed and electronically filed data that NRP uses to the operational classification process; adopting NRP policies or guidance on saving electronic examination case files; and incorporating NRP returns into examiners' development or training. Examiners raised other meritorious ideas, including requiring Microsoft Excel for all examination workpapers, requiring the use of NRP-style lead sheets and case-closing check sheets, and imaging returns for use in classification. However, based on available evidence and a lack of examiner consensus, these ideas did not meet enough of our criteria to warrant further consideration at this time. Successfully implementing the NRP practices that met our criteria in operational examinations could offer several benefits. To the extent that these changes to operational examinations produce the best tax returns and tax issues for examination, let examiners more quickly access the data they need, and further develop examination expertise, IRS is more likely to achieve benefits, including: * using classification as well as other examination resources more effectively, * identifying more tax noncompliance and more unpaid taxes owed during examinations without using more examination resources, and: * minimizing the burden imposed on taxpayers, particularly those who are compliant. Transcribing and Using More Tax Return Data and Using All Electronically Filed Return Data Could Improve Operational Classification: The NRP practices of transcribing significantly more tax return data from paper submissions than operations and using all available electronically filed data offer benefits for operational examinations and received the most support among examiners. In general, they supported these practices as ways to make operational classification and examination more effective. Because operational classifiers only have access to the types of data transcribed from paper-filed returns, examiners said that they could not immediately access and review all data reported on tax returns, which burdened taxpayers as well as made examiners less efficient in doing their jobs.[Footnote 11] Similarly, IRS officials told us for a previous report that having more tax return information available electronically would reduce burdensome examinations for compliant taxpayers, as well as facilitate enforcement efforts, make case resolution faster, and increase compliance revenue.[Footnote 12] [Side bar] “When non-transcribed items...come to light, [examiners] realize they need to audit these items. This delays the audit as they have to set up a new [document request] and new appointment after they see the full return and after the audit starts. For classifiers, they end up creating audit issues on non-transcribed items that don’t really need to be audited after they see those items on the return.” – IRS examiner interviewed by GAO. [End of side bar] Examiners who classified returns said that the lack of data in operational classification can cause them to classify issues that do not need to be examined. They said additional expense data from NRP on two Form 1040 schedules (Schedule C Profit or Loss from Business and Schedule E Supplemental Income and Loss) are representative of the kind of additional information that could address the problem. [Footnote 13] For example, operational classifiers might unnecessarily flag certain expense items on the schedules as an issue for examination simply because they do not have as many insights as NRP classifiers on the specific details of the items, causing the expenses to appear large, unusual, or questionable. The lack of additional return data also can make the examination process more burdensome for taxpayers being examined, particularly those filing an accurate tax return. Examiners said they sometimes had to examine a tax return issue for which the data were not transcribed. In such cases, examiners ask taxpayers for documents to verify the reporting of a tax issue that would not have been examined if the return data had been transcribed. As one examiner we interviewed said, "When you don't get everything transcribed [for classification], you could end up wasting taxpayers' time and increasing their burden by asking them for more information, particularly for more information that had likely already been provided in the original return but not transcribed." Another examiner said, "It is ridiculous to have to ask a taxpayer for a copy of the return if filed on paper." One method for providing operational classifiers more tax return data is to transcribe more data from paper-filed returns that are not already captured and provide classifiers with access to the additional data electronically.[Footnote 14] Although it may not be worthwhile for IRS to transcribe every piece of data from the millions of individual returns filed on paper each year, IRS could transcribe the tax return schedules that would have the greatest impact on examinations. For example, among IRS's highest priorities for possible transcription are lines from Schedule C. An estimated 75 percent of Schedule Cs were filed electronically in tax year 2010.[Footnote 15] Additional transcription on key expense items could make a significant difference considering the volume and compliance impact of that specific item. For example, the other expenses line on the Schedule C is often used to claim large amounts of deductions that IRS has found in error. Other expenses: * was the most frequently claimed expense (among 23 expense items) and was claimed on an estimated 57 percent of the schedules for tax year 2010, * totaled an estimated $103 billion--the highest expense amount claimed and about 19 percent of all expense claims on the Schedule C for tax year 2010,[Footnote 16] and: * had a noncompliance rate of 55 percent--the second highest of all the expense items on Schedule C, based on the NRP for tax year 2001. These amounts claimed as expenses combined with such a high non- compliance rate mean that it might be possible to find additional misreporting through more precise classification to offset the added transcription costs. At our request, IRS estimated in 2013 that it would cost about $8 million annually to transcribe data about three additional expenses from Schedule C.[Footnote 17] With costs of that level, even a small reduction in expense noncompliance could recoup the costs of transcription. Furthermore, costs could be reduced by transcribing only the portion of Schedules C and E submissions that reported the greatest amounts of income from business operations and that had the greatest chance of being subjected to classification and examination. For example, IRS's 2012 Data Book shows that of more than 16 million business returns with total positive income of less than $1 million, only about 1.4 million have total positive income of at least $200,000 and under $1 million.[Footnote 18] IRS also could apply a similar income cutoff for Schedule E filings. Finally, transcription costs could decline as more returns are filed electronically. According to data in the IRS Oversight Board's Electronic Filing 2012 Annual Report to Congress, the percentage of individual tax returns filed electronically was an estimated 81 percent in 2012, up from 40 percent in 2003. Electronic filing of Schedule C increased from an estimated 38 percent in tax year 2003 to the estimated 75 percent in tax year 2010.[Footnote 19] If costs did not decline, IRS still has the option of making a request for funding a compliance program on transcription in the course of making its annual budget request. For example, for fiscal year 2013, IRS proposed additional funding for activities to address the tax gap. Another option for increasing classification effectiveness is to allow classifiers to use all data reported on electronically filed returns. IRS policy limits the use of electronically filed data for classification to only the types of data that are transcribed from paper-filed returns to assure that paper-filed returns are treated the same as electronic returns. IRS could reverse its policy and allow use of all data filed electronically. IRS officials--including the SB/SE commissioner and deputy commissioner--said they recognize that having more data may help make examination processes more efficient by letting examiners make more- informed decisions. Because of this recognition and feedback from examiners, IRS analyzed whether using electronically filed data led to better examination decisions. The analysis indicated that using more electronically filed data could help classifiers detect more noncompliance. To more rigorously test this, IRS said that it is moving forward with more research to analyze the effects of adding electronically filed data on the amount of additional tax revenue assessed, if any. However, the new studies will not address any changes to taxpayer behavior if IRS reversed its policy on using electronically filed data or the added costs of using the data, such as information technology costs. IRS officials said they remain concerned that altering their policy on using electronic data might adversely affect electronic filing. These studies are still underway and the initial results will not be available until 2014 at the earliest. Regardless of the findings of IRS's upcoming pilots, our findings indicate that using less data to select returns and tax issues for examination increases the likelihood that compliant taxpayers are examined, creating an unnecessary burden for taxpayers filing these returns. In the cases where taxpayers are not compliant, with more data, IRS more likely could identify the noncompliance. Clarifying What Case File Information Should Be Saved Electronically Could Improve Operational Examination Effectiveness: Under NRP's goal of moving toward fully electronic case files, examiners are instructed to save their work electronically. Examiners we spoke with endorsed how NRP communicates the policy, although challenges exist to implementing the end result of complete electronic case files. NRP guidance on electronic case files also could serve as a model to help improve the operational guidance on saving case files electronically, which some observers we interviewed found confusing. IRS's Internal Revenue Manual states that NRP "is moving toward a fully electronic case file,"[Footnote 20] and NRP training states that "all lead sheets and relevant workpapers be completed electronically and included in the electronic case file." Examiners said moving toward requiring electronic case files for operational examinations would provide efficiency advantages. For example, they said electronic case files could save time in processing more than 100,000 audit reconsiderations that IRS works each year by making it easier for examiners to search the old files and identify what previous examiners concluded.[Footnote 21] IRS officials said that NRP conducts training, classification, and reviews remotely, taking advantage of all the electronically available case data. [Side bar] “Electronic workpapers would be more efficient in terms of time and of how much paper examiners have to carry. ... It does take time to scan, but it would be worth it [if] faster equipment than that which IRS currently has does exist.” – IRS examiner interviewed by GAO. [End of side bar] IRS officials previously told us that savings on handling costs was one of the benefits of added electronic documentation.[Footnote 22] IRS must place hundreds of thousands of field examination case files into storage at a federal records center each year. IRS said that it spends about $5 to $9 for each document it recalls from federal record centers. Examiners we spoke with also said that having electronic case files facilitates sharing their work with colleagues and reduces the chances that files will get lost. Constraints to IRS making all operational case files completely electronic mostly deal with the capacity of IRS's computer systems to store electronic case files and to transfer the files over existing bandwidth. IRS officials said they have no estimates on how much it would cost to expand the computer resources to save all files but that it would be considerable. Examiners said that SB/SE's ability to make all case files electronic is constrained because all of its offices do not have sufficient scanning technology to convert paper documents into electronic versions in a timely matter. IRS did not have an estimated cost for providing appropriate scanning technology for all of its offices. If it were too expensive to match NRP's level of electronic case file creation by expanding computer or scanning capacity, another option would be clarifying guidance to focus on the most important examination evidence. For example, NRP training materials and guidance define what constitutes relevant workpapers and list only four exceptions to documents that are permitted to be saved on paper as part of the case file. According to the NRP training, focusing on electronic case files allows subsequent researchers to interpret the examination results. In contrast, operational guidance on saving electronic case files is less clear. On one hand, the Internal Revenue Manual instructs examiners to electronically save any documents provided to the taxpayer's representative or third parties, such as lead sheets, workpapers and other related documents for purposes of an audit trail, as well as workpapers created outside of the Report Generation Software. On the other, SB/SE cautions examiners against saving too much because of space limitations, as the average SB/SE field case file size has grown from about 1 megabyte in 2006 to more than 4 megabytes in 2011. IRS examiner guidance says that examiners should limit the size of documents stored in the Report Generation Software to 5 megabytes; anything larger should be put on DVDs or CDs. According to NRP coordinators we interviewed, the competing operational guidance leaves examiners unsure about the extent that they should save documents electronically. IRS officials also said that they have observed that some operational examiners save more workpapers than necessary to show how they reached their conclusions, and that the NRP guidance helps curb the number of documents saved. IRS Does Not Provide Formal Guidance for Managers on When to Use NRP- Style Examinations as Development Opportunities for Recently Trained Staff: Examiners said that NRP provides a good opportunity for developing examiners who have completed IRS's training but are relatively inexperienced. Compared to operational examinations and to ensure data quality, NRP instructs examiners to document procedures in greater detail than operational examinations and to classify more issues for examinations. These practices could help inexperienced examiners learn about procedures and about more issues than they would on an operational examination. Examiners said that if they were exposed to these NRP practices, they could have more insights to help them make decisions when doing operational examinations. One examiner said working NRP cases "forces agents to look at lots of issues and in depth." Another examiner said that "working on NRPs gives you the big picture. It helps agents become more familiar with the review process." [Side bar] “You learn to know what to ignore and better apply the 80-20 rule.” – IRS examiner interviewed by GAO. The “80-20 rule” refers to how 20 percent of an examiner’s resources will likely yield about 80 percent of the potential tax-change dollars at stake. [End of side bar] To an extent, examiners already may be taking advantage of NRP as a learning tool. Field managers provide employees with the opportunity to perform a limited quantity of higher graded work (like NRP cases) to enhance their career development. One examiner we interviewed said that he had conducted NRP examinations about a year after completing IRS's Form 1040 training module and another said she had "cut her teeth" on NRP examinations, which were among the first she did. However, IRS has no formal guidance requiring or encouraging managers to consider assigning NRP examinations as a development opportunity. Costs of using NRP as a development tool also could be low, given that the returns already have been selected for the NRP sample by the time managers assign the work. Some tradeoffs from using NRP examinations to help develop newer operational examiners warrant consideration. First, for data analysis purposes, the NRP director said he prefers that experienced examiners work NRP cases. Second, he said that IRS adjusts its tax gap figures by an econometric technique that takes examiner skills and experience into account. Managers would have to take care not to assign NRP cases to examiners who were not ready to handle them. Examiners we interviewed said that while NRP examinations are more detailed, the size of an NRP case, which generally may have many classified issues, could overwhelm new examiners. On the other hand, concerns about problems caused by using less experienced examiners in NRP might not apply in all cases. NRP coordinators told us that in their experience, newer but fully trained examiners are conscientious about following the NRP examination procedures as much if not more than examiners with many years of experience. Also, to the extent that using newly trained examiners for NRP cases might affect the econometric technique, IRS could ensure that newly trained examiners who did NRP cases stayed in the pool of examiners working cases in future years. Conclusions: SB/SE's examination program is a key effort in reducing the nation's persistent tax gap for individual returns. Improving examination procedures helps IRS to better detect noncompliance, promote voluntary compliance, and minimize the cost of examinations for IRS and taxpayers. IRS annually makes a considerable investment in NRP, but opportunities exist for IRS to improve operational examinations by applying NRP practices. To the extent that IRS does not pursue these opportunities to maximize its NRP investment, IRS is missing a chance to improve operational examinations by better classifying returns, identifying more noncompliance, and reducing taxpayer burdens. In particular, lessons from NRP suggest that transcribing data filed on paper and using electronically submitted data for Schedules C and E could enhance the efficiency of the operational classification process. Although adding these data would entail costs, the evidence does not necessarily indicate that those costs are great enough to outweigh the benefits or that it would not be beneficial for IRS to request funding for added transcription. Similarly, more consistently saving electronic case files can provide IRS examiners with easier access to a wider swath of information and reduce taxpayer burden by finding critical documents more quickly both during the examination and during follow up work, such as for examination reconsiderations as well as for other IRS staff who need access for analysis or research of examination trends. Given cost concerns about expanding IRS's information system capacity, space constraints could be addressed by having clearer guidance on what files are essential to be electronically saved rather than saving all of them. Clarifying the guidance also could reduce examiners' confusion. Finally, IRS could encourage newly-trained examiners to enhance their development by working selected individual NRP cases, better ensuring that they gain insights that will help them do better operational examinations. Recommendations for Executive Action: To increase the effectiveness of IRS's examinations individual tax returns, GAO recommends that the Commissioner of Internal Revenue take the following four actions. * Transcribe data from paper-filed Form 1040 Schedules C and E that are not currently transcribed and make that data available to SB/SE examiners for classification. If IRS has evidence that the costs related to transcribing all such data on Schedules C and E are prohibitive, IRS could do one or both of the following actions: - transcribe less data by transcribing only the missing data for selected line items, such as certain, large expense line items, or: - develop a budget proposal to fund an initiative for transcribing Schedule C and E. * Make all data collected from electronically submitted Form 1040s available to examiners conducting classification. * Clarify the policy and guidance on which case files should be saved electronically for SB/SE examinations to encourage greater and consistent use of electronic case files. * Develop official guidance to remind managers that NRP cases can be assigned to newly trained examiners, when appropriate, to enhance their career development. Agency Comments: We requested written comments from the Acting Commissioner of Internal Revenue and received comments from the Deputy Commissioner for Services and Enforcement on May 9, 2013. The Deputy Commissioner said that IRS agreed to study the possibility of increasing data transcription but did not comment more specifically. IRS also agreed that making electronically collected data available for classification would be beneficial, saying it intends to look at expanding the use of electronically submitted returns to enhance classification and possibly other areas as well. The agency also agreed to clarify its policy on when case files should be saved electronically and agreed to expand guidance to remind managers that they may assign NRP cases to newly trained examiners when appropriate to enhance their career development. IRS provided technical comments on the draft, which were incorporated as appropriate. We also made other clarifying changes. We have sent copies of this report to the Department of the Treasury, IRS, and other interested parties. This report also will be available at no charge on GAO's web site at [hyperlink, http://www.gao.gov]. If you or your staff have any questions about this report, please contact me at (202) 512-9110 or whitej@gao.gov. 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 are listed in appendix V. Signed by: James R. White: Director, Tax Issues Strategic Issues: [End of section] Appendix I: Scope and Methodology: To describe lessons that IRS has already implemented, we interviewed National Research Program (NRP) on what it saw as the main benefits that NRP had already provided operational examinations. Based on those responses, we gathered documentation, to the extent it was available, to confirm the benefits that the officials cited. We also gathered information from the Small Business/Self-Employed division (SB/SE) by interviewing relevant SB/SE officials to confirm the cited benefits as well. To assess the ways that NRP practices and procedures might improve operational examinations, we used a two step process. First, we identified potential ways that NRP might improve operational examinations by interviewing key IRS officials. Then, in the second step, we compared those ideas against a set of criteria to determine whether the ideas warranted further action. We held exploratory small-group discussions with NRP examiners who also had operational audit experience to obtain preliminary views on NRP practices and procedures that potentially could help improve operational examinations. We conducted these exploratory interviews around the Washington, D.C., area. Based on these discussions and other research on NRP, we identified a set of topics and developed a set of specific, close-ended questions that addressed those topics. We used a structured interview technique with groups of IRS examiners where we asked those questions as well as provided the group participants opportunities to further explain their response choice or to say more about each topic. The structured interview groups comprised either examiners who were NRP classifiers, or operational examiners who had experience with operational classification using the Compliance Data Environment software, which is what operational examiners use to conduct classification. We focused on examiners with classification experience because: a) they had the broadest experience with the examination process, b) were considered to be experts by IRS, and c) our exploratory discussions indicated that some of the NRP benefits to operational audits might center on classification. To choose NRP classifiers, we reviewed a list of examiners who were scheduled to conduct NRP classification in January 2013 that IRS sent us. Operational classifiers were chosen in cooperation with SB/SE liaisons. A criterion we used in requesting participants was that they had experience using the Compliance Data Environment system, so that we could compare the NRP classifiers' experiences with operational classifiers' experiences. In total, we conducted seven interview sessions with groups that ranged in size from three to seven participants at a time, as well as an individual interview. The participants for each interview group were assigned randomly from the list IRS provided. We interviewed all but one of the examiners who were scheduled to participate in the January 2013 classification detail. We discussed the examiners' statements and comments with NRP coordinators who attended the January 2013 NRP classification detail and IRS management to obtain their perspective on these same topics and reaction to views expressed in our group interviews. Because our interviews included all but one of the entire population of NRP classifiers for the January 2013 detail, their responses can be seen as representative of these NRP classifiers' views on the matters we discussed. The views we obtained from operational classifiers are not a representative sample of all operational classifiers and thus reflect only the views of that group. To assess which revenue agent ideas warranted further action, we derived the following criteria. * Would the change be likely to help examiners correctly identify more noncompliance? * Is it likely to make the examination more efficient--i.e., reduce the cost of conducting an examination, or increase benefits for the same cost or increase benefits commensurate with justifiable or no increase in costs? * Is it likely to reduce taxpayer burden? * Would the change meet any other IRS need? These criteria generally are based on our 2005 report, Understanding the Tax Reform Debate (GAO-05-1009SP),[Footnote 23] Office of Management and Budget Circular A-94,[Footnote 24] and the Internal Revenue Manual.[Footnote 25] IRS officials agreed that these criteria were valid. As noted in our 2005 report, the criteria may conflict with each other, and as a result, there often are trade-offs to consider among the criteria when evaluating a particular proposal. In those cases, we used our professional judgment in weighing the relative importance of these criteria in making our recommendations. [End of section] Appendix II: GAO Reports Regarding NRP Examination Procedures and Data Transcription from Paper Returns: The following lists our prior reports regarding National Research Program (NRP) examination procedures. 1. Tax Policy and Administration: Using Data from the Internal Revenue Service's National Research Program to Identify Potential Opportunities to Reduce the Tax Gap [hyperlink, http://www.gao.gov/products/GAO-07-423R]. To ensure that IRS maximizes its return on investment from future NRP studies, we recommended that IRS should develop a plan for capturing complete NRP examination case files that (1) determines the most cost effective means for capturing information electronically and (2) lays out a schedule for when it will begin to capture information electronically. IRS completed a study of capturing electronic information for National Research Program (NRP) examination cases in December 2008. For the tax year 2006 NRP study of individual tax returns, IRS recommended that examiners complete work papers electronically, and mandated electronic work papers for the tax year 2007 study. In a December 2008 document, IRS said that it will require all IRS developed documentation to be in electronic form, but will not scan taxpayer-provided documentation. Because of IRS's actions, we consider these recommendations implemented. 2. Tax Administration: New Compliance Research Effort Is on Track, but Important Work Remains [hyperlink, http://www.gao.gov/products/GAO-02-769]. We recommended that the IRS Commissioner should ensure that testing and modification of NRP case building and classification procedures are complete before IRS begins cadre training, classifying NRP returns, or making any taxpayer contacts. IRS should use some previously audited, non-Earned Income Tax Credit tax returns to evaluate NRP classification procedures and classifier training. We also recommended that the IRS Commissioner should implement plans to select and appropriately train the cadre of examiners and other staff before NRP classification begins. In our June 2003 report on the progress of IRS's NRP implementation,[Footnote 26] we reported that the timing of IRS's NRP staff selection and training satisfied this recommendation. 3. Tax Administration: Alternative Strategies to Obtain Compliance Data [hyperlink, http://www.gao.gov/products/GAO/GGD-96-89]. In our 1996 report on the precursor to NRP, the Taxpayer Compliance Measurement Program (TCMP), we recommended that IRS identify a short- term alternative strategy to minimize the negative effects of the compliance information that is likely to be lost because TCMP was postponed, and develop a cost-effective, long-term strategy to ensure the continued availability of reliable compliance data. We identified several alternatives to the traditional TCMP that would meet some of the data needs that were lost when TCMP was postponed, including (1) conducting multiyear TCMP audits on smaller samples and combining the results; (2) using operational audit data; and (3) conducting a mini TCMP to identify compliance issues, with a more focused TCMP audit on the identified issues. IRS started the NRP program in 2001, using tax returns filed for that tax year. The 2007 NRP study was the first of an ongoing series of annual individual studies using a multi-year rolling methodology. The following reports include discussions of conducting full transcriptions of tax returns. 1. Tax Administration: 2007 Filing Season Continues Trend of Improvement, but Opportunities to Reduce Costs and Increase Tax Compliance Should Be Evaluated [hyperlink, http://www.gao.gov/products/GAO-08-38]. Full transcription could increase the effectiveness of compliance programs. According to IRS officials, transcribing and posting more comprehensive information from individual income tax returns could facilitate the audit process, expedite contacts for faster resolution, reduce handling costs, allow for improved case selection, and potentially better define specific tax gap issues. 2. E-Filing Tax Returns: Penalty Authority and Digitizing More Paper Return Data Could Increase Benefits [hyperlink, http://www.gao.gov/products/GAO-12-33]. In this report, we noted that E-filing provides important benefits to taxpayers, including faster refunds and more accurate returns. It provides a low-cost option for IRS to improve enforcement operations and services to taxpayers. To help increase electronic filing and to better target IRS's efforts, we recommend that IRS determine the relative costs and benefits of transcribing different individual lines of tax return data. IRS should also quantify the benefits of transcribing individual lines and compare them to the individual costs. This analysis could inform budget decisions by allowing IRS to compare the option of additional transcription against any work foregone. As part of this report, we calculated the costs of transcribing data lines from paper-filed returns. [End of section] Appendix III: How IRS Uses National Research Program Data: The National Research Program (NRP) data provide IRS with data for operational and strategic purposes. On an operational level, analysis of NRP data provides insights into how well the IRS meets its day-to- day operating compliance obligations. These concern the enforcement activities associated with compelling non-compliant taxpayers to meet their tax obligations. On a strategic level, compliance measures developed as part of NRP also support IRS's strategic goal of improving taxpayers' voluntary compliance with the tax laws. IRS uses these measures to benchmark the effectiveness of programs to promote voluntary compliance. More specific examples of how IRS uses of NRP data include: * identifying where compliance problems are occurring, so that IRS can improve utilization of limited resources to address those problems; * updating the discriminant function[Footnote 27] and other computerized examination workload selection formulas; * gaining insights into the causes of reporting errors to provide improved taxpayer service;[Footnote 28] * updating estimates of the tax gap;[Footnote 29] and: * studying compliance issues and characteristics of a subset of taxpayers who were examined by NRP. In addition to IRS, we have used NRP data for research purposes. For example, we used the NRP data set as part of our analysis of tax compliance by sole proprietorships study.[Footnote 30] [End of section] Appendix IV: Agency Comments from the Internal Revenue Service: Department of The Treasury: Internal Revenue Service: Deputy Commissioner: Washington, D.C. 20224: May 9, 2013: Mr. James White: Director, Tax Issues: Strategic Issues Team: U.S. Government Accountability Office: Washington, DC 20548: Dear Mr. White: I have reviewed your draft report entitled, "IRS Could Improve Examinations by Adopting Certain Research Program Practices (GAO-13- 480, Job Code 450997). We appreciate your suggestion that we look to lessons learned from National Research Program (NRP) examinations as a source of efficiencies which can be applied to operational examinations. We are always looking for ways to improve our processes and procedures in screening, selecting and examining tax returns. We endeavor to leverage our usage of data wherever possible and we appreciate its value in giving our tax return screeners and examiners more complete information. In our current limited budget situation, we face difficult fiscal choices for technology improvements, but we will continue to look for opportunities to use technology to enhance our existing enforcement activities. Moreover, as the number of electronically filed returns from individuals now exceeds 80% of all returns filed, we intend to review our policy of treating data from electronically and paper filed returns in a similar fashion. We continually work to improve the clarity and consistency of employee guidance. The feedback that improvements can be made regarding guidance on electronic workpapers and case files is a valid one and we will reiterate to our examiners to use electronic workpapers whenever it makes good business sense to do so. Specifically, Small Business/Self Employed Division (SB/SE) Examination function will expand the guidance provided in Internal Revenue Manual (IRM) 4.10.15.1.4, Examination of Returns, Report Generation Software (RGS) and will continue to look for additional areas in our procedural guidance where we can improve clarity. We will continue to provide opportunities to our examiners to participate in NRP examinations, as appropriate based on their experience and skills, while ensuring the data integrity and consistency necessary to support the program. The enclosed response addresses each recommendation separately. If you have questions, please contact me, or a member of your staff may contact Faris Fink, Commissioner, Small Business/Self Employed Division at (202)622-0600. Sincerely, Signed by: Steven T. Miller: Deputy Commissioner for Services and Enforcement: Enclosure: GAO Recommendations and IRS Responses to GAO Draft Report: IRS Could Improve Examinations by Adopting Certain Research Program Practices (GAO-13-480, Job Code 450997): Recommendation 1: Transcribe data from paper-filed Form 1040 Schedules C and E that are not currently transcribed and make that data available to SB/SE examiners for classification. If IRS has evidence that the costs related to transcribing all such data on Schedules C and E exceed the benefits or are otherwise prohibitive, IRS could transcribe less data by transcribing only the missing data for selected line items, such as certain large expense line items. Comment: The IRS agrees with your assessment of the difficulties we experience due to lack of complete transcription of tax return data; however, as you have recognized in your report, the IRS is faced with funding competing priorities in our current fiscal climate. The costs associated with this will include the costs of additional transcription as well as increased data storage and system access. NRP has a separate data repository for the additional data associated with that smaller population of work. Consideration will be given to expanding data transcription and the benefits to be derived given the increase in the numbers of electronically filed returns; and, we agree that we will study the possibility of increasing data transcription of additional return information. Recommendation 2: Make all data collected from electronically submitted Form 1040s available to examiners conducting classification. Comment: As more taxpayers choose to file electronically, the IRS intends to look at expanding the use of data from electronically submitted returns to enhance classification and possibly in other areas as well. We will need to ensure the fairness and integrity of our audit process and balance the potential benefits offered from using e-file data against the potential impact to compliance and participation in the e-file program. Recommendation 3: Clarify the policy and guidance on which case files should be saved electronically for SB/SE examinations to encourage greater and consistent use of electronic case files. Comment: SB/SE field examination will clarify the policy on which case files should be saved electronically for SB/SE examinations. We will also expand the guidance provided in IRM 4.10.15.1.4, Documents Included in RGS, to ensure examiners and managers understand the process of saving documents in the RGS. In addition, the NRP office is currently piloting the electronic case files/workpapers processes in the 1120 study. We will take the lessons learned from that study to improve the guidance in IRM 4.10.15.1.4 and IRM 4.10.9, Workpapers. Recommendation 4: Develop official guidance to remind managers that NRP cases can be assigned to newly trained examiners, when appropriate, to enhance their career development. Comment: Managerial guidance for SB/SE Field Examination's assignment of returns can be found in IRM 1.4.40, Resource Guide for Managers - SB/SE Field and Office Examination Group Manager. While the guidance does not specifically address NRP returns, it does provide information regarding factors that influence the assignment of work. We will expand the guidance to remind group managers that program/project cases, including NRP, may be assigned to newly-trained examiners, when appropriate, to enhance their career development. When making assignment decisions, managers should consider the complexity of the cases plus the experience and training level of the examiners. We will also work with NRP to link their IRM sections related to case assignment to our expanded guidance. [End of section] Appendix V: GAO Contacts and Staff Acknowledgments: GAO Contact: James R. White, (202) 512-9110 or whitej@gao.gov: Acknowledgments: In addition to the contact named above, Tom Short, Assistant Director; Tom Beall; Amy Bowser, Eric Gorman; George Guttman; Sarah McGrath, Robert Gebhart, and Donna Miller made key contributions to this report. [End of section] Footnotes: [1] The annual sample consists of about 12,000 to 15,000 tax returns. NRP originally started with a much larger sample of tax year 2001 returns before switching to the rolling annual sample in 2006. [2] Classification is the process of determining whether a return should be selected for examination, what issues should be examined, and how the examination should be conducted. [3] The scoring system, called the discriminant function, is a mathematical technique used to score income tax returns for examination potential. Generally, the higher the discriminant function score, the greater the probability of significant noncompliance. IRS uses data from NRP to calculate the function. [4] In this context, technique refers to using a correspondence examination, which is done through mail, or field examination, in which IRS meets with taxpayers or the taxpayers' representatives face- to-face. [5] See appendix II for our previous reports on the NRP and appendix III for more detail on how IRS uses NRP data. [6] The NRP case building process compiles a variety of information, such as the taxpayers' past years' returns; information reported to IRS by third parties, such as employers and banks; and information from outside databases, such as property listings. This information is available to the classifier and ultimately the taxpayer. [7] IRS listed other NRP benefits in its FY 2014 Congressional Budget Justification, published in April 2013. When we followed up, IRS officials stood behind the characterization of NRP's benefits in our report. [8] The software programs were CENTRA, which facilitates remote learning, that allowed IRS to provide training at remote field sites and Articulate, a self-paced e-learning platform that allowed IRS to reach a large number of students in an online interactive format. [9] The Report Generation Software is a group of computer programs that support examination activities such as preparing reports, performing complex computations, producing correspondence, and closing cases. [10] Appendix I includes a detailed description of the criteria we used. [11] Since our interviews with the examiners, IRS issued a memo to emphasize the importance of using automated information systems to enhance examination quality. [12] GAO, E-Filing Tax Returns: Penalty Authority and Digitizing More Paper Return Data Could Increase Benefits, [hyperlink, http://www.gao.gov/products/GAO-12-33] (Washington, D.C.: Oct. 5, 2011). [13] Taxpayers use Schedule C to report business income or losses as a sole proprietor. Schedule E is used to report income or losses from endeavors such as real estate rentals, partnerships, estates, trusts, and certain corporations. [14] See appendix II for a list of related GAO reports. [15] Percentage based on Statistics of Income data. The 95 percent confidence interval for the unrounded estimate of 74.6 percent is between 74.4 percent and 74.9 percent. [16] For the $103 billion in "other expenses" estimate, the lower bound is $101 billion and the upper bound is $105 billion at the 95 percent confidence interval. [17] This $8 million includes the cost to keystroke the added expenses, as well as other costs such as for error resolution, quality review, overhead, and employee benefits for the three added expenses. In an earlier calculation for tax year 2008, we calculated that keystroke costs alone for certain, additional Schedule C and E information would be $768,208. [18] The more than 16 million filings do not represent all Schedule C filings. For example, it does not include those that also filed the Earned Income Tax Credit, and therefore does not match the figures used elsewhere in this report. [19] Percentages were derived from Statistics of Income reports. For the 38 percent estimate, the lower bound is 37.6 percent and the upper bound is 38.4 percent at a 95 percent confidence interval. [20] IRS's Internal Revenue Manual, 4.22.6.1, Case Building for Individual Returns (5/20/13). [21] An audit reconsideration is the administrative process the IRS uses to reevaluate prior audits where the IRS assessed additional tax and it remains unpaid, or a tax credit was reversed. If the taxpayer disagrees with the original determination, the taxpayer must provide information not previously considered during the original examination. It is also the process the IRS uses when the taxpayer contests a Substitute for Return (i.e., a return that IRS generates when a taxpayer does not file) determination by filing an original delinquent return. [22] [hyperlink, http://www.gao.gov/products/GAO-12-33]. [23] GAO, Understanding the Tax Reform Debate: Background, Criteria, and Questions, [hyperlink, http://www.gao.gov/products/GAO-05-1009SP] (Washington, D.C.: September 2005). [24] Office of Management and Budget, Circular A-94 Guidelines and Discount Rates for Benefit-Cost Analysis of Federal Programs (Washington, D.C; September 2005). [25] A redacted version of the Internal Revenue Manual is available online at [hyperlink, http://www.irs.gov/irm]. [26] See GAO, Tax Administration: IRS Implementing the National Research Program as Planned, [hyperlink, http://www.gao.gov/products/GAO-03-614] (Washington, D.C.: June 16, 2003). [27] The discriminant function is a computer program IRS uses to score returns for audit potential, to increase the likelihood that noncompliant taxpayers are selected for examination. [28] For example, where NRP examinations turn up systematic compliance errors on particular items for otherwise compliant taxpayers, the IRS may address the source of these errors through redesigned forms, better communications, improved taxpayer education, or proposing legislative changes. NRP data helps IRS assess the effectiveness of compliance programs and treatments it is using. [29] The tax gap estimate is an aggregate of estimates for the three primary types of noncompliance: (1) underreporting of tax liabilities on tax returns; (2) underpayment of taxes due from filed returns; and (3) nonfiling, which refers to the failure to file a required tax return altogether or on time. [30] See GAO, Tax Gap: Limiting Sole Proprietor Loss Deductions Could Improve Compliance but Would Also Limit Some Legitimate Losses, [hyperlink, http://www.gao.gov/products/GAO-09-815] (Washington, D.C.: Sept. 10, 2009). [End of section] GAO’s Mission: The Government Accountability Office, the audit, evaluation, and 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 GAO’s website [hyperlink, http://www.gao.gov]. Each weekday afternoon, GAO posts on its website newly released reports, testimony, and correspondence. To have GAO e-mail you a list of newly posted products, go to [hyperlink, http://www.gao.gov] and select “E-mail Updates.” Order by Phone: The price of each GAO publication reflects GAO’s actual cost of production and distribution and depends on the number of pages in the publication and whether the publication is printed in color or black and white. Pricing and ordering information is posted on GAO’s website, [hyperlink, http://www.gao.gov/ordering.htm]. Place orders by calling (202) 512-6000, toll free (866) 801-7077, or TDD (202) 512-2537. Orders may be paid for using American Express, Discover Card, MasterCard, Visa, check, or money order. Call for additional information. Connect with GAO: Connect with GAO on facebook, flickr, twitter, and YouTube. Subscribe to our RSS Feeds or E mail Updates. Listen to our Podcasts. Visit GAO on the web at [hyperlink, http://www.gao.gov]. To Report Fraud, Waste, and Abuse in Federal Programs: Contact: Website: [hyperlink, http://www.gao.gov/fraudnet/fraudnet.htm]; E-mail: fraudnet@gao.gov; Automated answering system: (800) 424-5454 or (202) 512-7470. Congressional Relations: Katherine Siggerud, Managing Director, siggerudk@gao.gov: (202) 512-4400: U.S. Government Accountability Office: 441 G Street NW, Room 7125: Washington, DC 20548. Public Affairs: Chuck Young, Managing Director, youngc1@gao.gov: (202) 512-4800: U.S. Government Accountability Office: 441 G Street NW, Room 7149: Washington, DC 20548. [End of document]