United States Government Accountability Office GAO Report to Congressional Requesters TAX GAP December 2012 IRS Could Significantly Increase Revenues by Better Targeting Enforcement Resources GAO-13-151 December 2012 TAX GAP IRS Could Significantly Increase Revenues by Better Targeting Enforcement Resources Highlights of GAO-13-151, a report to congressional requesters Why GAO Did This Study What GAO Found Heightened attention to federal deficits The Internal Revenue Service (IRS) spends most of its enforcement resources has increased pressure on IRS to on examinations. Correspondence exams of individual tax returns, which target reduce the tax gap—the difference fewer and simpler compliance issues, are significantly less costly on average between taxes owed and taxes paid on than the broader and more complex field exams. GAO estimated that the time—and better enforce taxpayer average cost (including overhead) of correspondence exams opened in 2007 and compliance. Resource limitations and 2008 was $274, compared to an average of $2,278 for field exams. IRS spent concern over taxpayer burden, almost 20 percent of the $1.6 billion per year that it devoted to exams on returns however, prevent IRS from auditing from taxpayers with positive income of at least $200,000, even though such more than a small fraction of individual returns accounted for only 3 percent of the 136 million individual returns filed per income tax returns filed. How IRS year. (Positive income, a measure that IRS uses to classify returns for exam allocates these limited resources demands careful consideration. planning purposes, disregards losses that may offset this income). As requested, this report (1) describes GAO estimated that, for the 2 years of cases reviewed, correspondence exams how IRS allocates resources across were significantly more productive in terms of direct revenue produced per dollar individual taxpayer compliance of cost than field exams. As shown in the figure below, both types of exams of enforcement programs and across taxpayers with positive incomes of at least $200,000 were significantly more types of taxpayers within each productive than exams of lower-income taxpayers. program; (2) estimates the direct Direct Revenue Return on Investment for Different Types of Exams and Groups of Individual revenue return on investment for the Income Tax Returns Opened in 2007 and 2008 individual taxpayer enforcement programs and the extent of variation across those programs and across types of taxpayers; and (3) determines the potential for gains from shifting resources from lower-yielding programs and types of taxpayers to higher-yielding ones. To accomplish these objectives GAO analyzed IRS data on 2007 and 2008 tax returns, reviewed IRS documentation, and interviewed appropriate IRS officials. What GAO Recommends GAO recommends that IRS review disparities in the ratios of direct Note: EITC is the earned income tax credit revenue yield to costs across different GAO demonstrated how these estimates could be used to inform resource enforcement programs and across allocation decisions. For example, a hypothetical shift of a small share of different groups of cases and consider resources (about $124 million) from exams of tax returns in less productive this evidence as a potential basis for groups shown in the figure to exams in the more productive groups could have adjusting its allocation of enforcement increased direct revenue by $1 billion over the $5.5 billion per year IRS actually resources each year. IRS agreed with collected (as long as the average ratio of direct revenue to cost for each category the recommendations. of returns did not change). These gains would recur annually, relative to the revenue that IRS would collect if it did not change its resource allocation. This particular resource shift would not reduce exam coverage rates significantly and, View GAO-13-151. For more information, therefore, should have little, if any, negative effect on voluntary compliance. contact James White at (202) 512-9110 or firstname.lastname@example.org United States Government Accountability Office Contents Letter 1 Background 3 Examinations of Taxpayers with Less than $200,000 in Positive Income Accounted for Most of the Total Cost of the Four Enforcement Programs 6 Direct Revenue Return on Investment Was Highest for Examinations of Taxpayers with at Least $200,000 in Positive Income 8 Although Some Caution is Warranted, Exam Resource Reallocation Could Produce Significant Direct Revenue Gains 10 Conclusions 15 Recommendations 16 Agency Comments 16 Appendix I Summary Methodology for Data Analysis 18 Appendix II Detailed Tables 21 Appendix III Economic Guidelines for Enforcement Resource Allocation Decisions 24 Appendix IV Comments from the Internal Revenue Service 26 Appendix V GAO Contact and Staff Acknowledgments 28 Tables Table 1: Costs, Direct Revenue, Exam Coverage Rates and No- Change Rates for Different Types and Groups of Exams Opened In Fiscal Years 2007 and 2008 (Weighted Average of the 2 Years) 21 Table 2: Potential Changes in Direct Revenue and Exam Coverage Rates Resulting from a Hypothetical Reallocation of Exam Resources 22 Page i GAO-13-151 Tax Gap Table 3: Ratios of Direct Revenue to Cost for Field Exams Opened in 2007 and 2008 after Accounting for Various Factors 23 Figures Figure 1: Cost Shares and Cost per Case for Four Exam and IRP Programs for Cases Opened in Fiscal Years 2007 and 2008 7 Figure 2: Number of Returns, Exam Costs, and Coverage Rates for Different Categories of Individual Income Taxpayers 8 Figure 3: Ratios of Direct Revenue to Costs for Different Types of Exams and Groups of Individual Income Tax Returns for Cases Opened in Fiscal Years 2007 and 2008 9 Figure 4: Changes in Resources and Direct Revenue for Different Groups of Individual Income Tax Exams under a Hypothetical Reallocation 11 Figure 5: Coverage Rates for Different Types of Exams and Groups of Individual Income Tax Returns Before and After Hypothetical Reallocation 13 Page ii GAO-13-151 Tax Gap Abbreviations AIMS Audit Information Management System ASFR Automated Substitute for Return AUR Automated Underreporter EITC earned income tax credit ERIS Enforcement Revenue Information System IFS Integrated Financial System IRP Information Returns Processing IRS Internal Revenue Service NRP National Research Program SB/SE Small Business/Self-Employed W&I Wage and Investment This is a work of the U.S. government and is not subject to copyright protection in the United States. The published product may be reproduced and distributed in its entirety without further permission from GAO. However, 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. Page iii GAO-13-151 Tax Gap United States Government Accountability Office Washington, DC 20548 December 5, 2012 The Honorable Max Baucus Chairman The Honorable Orrin Hatch Ranking Member Committee on Finance United States Senate The Honorable Charles Grassley Ranking Member Committee on the Judiciary United States Senate Heightened attention to federal deficits has increased pressure on the Internal Revenue Service (IRS) to reduce the tax gap—the difference between taxes owed and taxes paid on time—and better enforce taxpayer compliance. Resource limitations and interest in minimizing taxpayer burden, however, prevent IRS from auditing anything but a small fraction of the total number of individual income tax returns filed for a given tax year. Of the approximately 141 million individual tax returns filed in 2010, only 1.1 percent were formally audited. IRS has developed several compliance enforcement programs with the goal of increasing taxpayer compliance, using resources more efficiently, and minimizing taxpayer burden. These programs, which include computer-based matching systems and streamlined correspondence audits (which do not involve face-to-face meetings with taxpayers) designed for less complex issues, fill different roles in the compliance enforcement strategy. Every year IRS publishes information regarding the coverage rates and additional taxes assessed through these various programs, but relatively little information is available on how much revenue is actually collected as a result of these enforcement activities (called direct revenue). Even less information is available regarding program performance with respect to identifiable subpopulations (varying in terms of income levels, and types of income) covered by the enforcement programs. Given the importance of appropriately allocating limited resources, you asked us to assess the performance of IRS’s enforcement programs. In this report we (1) describe how IRS allocates resources across individual taxpayer compliance enforcement programs and across types of taxpayers within each program; (2) estimate the direct revenue return on investment for the individual taxpayer compliance enforcement programs Page 1 GAO-13-151 Tax Gap and the extent of variation across those programs and across types of taxpayers; and (3) determine the potential for gains from shifting resources from lower-yielding programs and types of taxpayers to higher- yielding ones. To describe how IRS allocates resources across individual taxpayer compliance enforcement programs and across types of taxpayers within each program, we compared the amounts of resources that IRS devotes to each of four broad enforcement programs for individuals who file form 1040 tax returns—Automated Underreporter (AUR), Automated Substitute for Return (ASFR), correspondence examination, and field examination. 1 These programs are described below in the background section. We also report the cost of cases worked and the percentage of total returns covered by correspondence and field exam programs. For the correspondence and field exam programs, we make these comparisons across the principal groups of individual taxpayers (defined in terms of positive income size and characteristics of their returns, such as the presence of the earned income tax credit (EITC)) that IRS uses for exam planning purposes. 2 We use program cost data obtained from IRS’s Integrated Financial System (IFS) and data on the number of cases from IRS’s Enforcement Revenue Information System (ERIS). We also use data on coverage rates from IRS’s Data Book. 3 Our analyses to support all three of our objectives cover enforcement cases opened in fiscal years 2007 and 2008 (the latest years of data available at the time of our analysis in which the large majority of cases have worked their way through the collection process). 4 1 We exclude the Math Error program from our detailed review because it covers 100 percent of tax returns filed and, therefore, does not involve the same types of resource allocation decisions that are relevant to the other programs. 2 In general, total positive income is the sum of all positive amounts shown for the various sources of income reported on the individual tax return and, thus, excludes losses. 3 Internal Revenue Service, Data Book, 2009, Publication 55B, Washington, DC, March 2010, and Data Book, 2008, March 2009. 4 Although IRS typically reports enforcement data according to the fiscal year in which cases are closed, we report our results according to the fiscal year in which the cases were opened because that better reflects the populations of cases that IRS targeted in their annual exam planning process. Page 2 GAO-13-151 Tax Gap To estimate the direct revenue return on investment for the individual taxpayer compliance enforcement programs, we obtained data from ERIS on collections of tax, interest, and penalties directly attributable to specific types of exams for different groups of taxpayers, as well as those attributable to AUR and ASFR cases. We also obtained cost data from IFS for these four programs. For the field exam program, available data allow us to incorporate differences in the length, difficulty, and location of audits into our cost comparisons. Our return on investment measure is computed as the ratio of revenues over costs with the value of the revenues discounted for any delays between the year in which IRS expended the money on the enforcement cases and the year in which the revenues were collected. To determine the potential for gains from shifting resources from lower- yielding programs and types of taxpayers to higher-yielding ones, we use our results from the second objective to demonstrate how modest reallocations of resources would affect total collections, assuming that additional cases examined in a particular group would be approximately as productive as the average case in that group. We also interviewed IRS officials and reviewed the technical literature on tax enforcement to identify factors beyond the direct revenue return on investment that IRS should consider when making adjustments to its resource allocations. We determined for the purposes of this review that the data used were reliable. (See app. I for additional information about our methodology.) We conducted this performance audit from November 2009 through December 2012 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. In January 2012, IRS estimated that the gross tax gap—the difference Background between taxes owed and taxes paid on time—was $450 billion in tax year 2006. IRS estimated that it would eventually recover about $65 billion of this amount through late payments and enforcement actions, leaving a net tax gap of $385 billion. The tax gap has been a persistent problem in spite of extensive congressional and IRS efforts to reduce it. In past work we have said that reducing the tax gap will not likely be achieved through a single solution. Rather, the tax gap must be attacked on multiple fronts and with multiple strategies over a sustained period of time. On the Page 3 GAO-13-151 Tax Gap enforcement front, IRS’s efforts to ensure compliance of individual taxpayers combine several distinct programs that collectively monitor and correct noncompliance with income tax filing, reporting, and payment requirements. These programs fill different roles in the enforcement process and vary in the number of taxpayers covered, the resources used, and their level of automation. Math Error Program: IRS’s Math Error program electronically checks all filed tax returns for obvious math errors as returns are processed. The Math Error program reviews and adjusts items specifically listed in Internal Revenue Code section 6213. The specific issues that the program has authority to review include calculation errors, entries that are inconsistent with or exceed statutory limits, various omissions, inclusions, and entries of information, or incorrect use of an IRS table. Automated Underreporter IRS collects information on taxpayers from employers, financial Program: institutions, and other third parties and compiles these data in the Information Returns Processing (IRP) system. The Automated Underreporter (AUR) program electronically matches the IRP data against the information that taxpayers report on their forms 1040 as a means of identifying potentially underreported income or unwarranted deductions or tax credits. The matching process takes place months after taxpayers have filed their tax returns. For tax year 2010, AUR identified approximately 23.8 million potential discrepancies between taxpayer income, deduction, and other information reported by third parties and the information supplied by taxpayers on their individual income tax returns. IRS officials said that resource constraints prevent them from contacting taxpayers for all of the cases in which discrepancies are identified. If a mismatch exceeds a certain tax threshold, AUR reviewers decide if it warrants a notice to the taxpayer asking for an explanation of the discrepancy or payment of any additional tax assessed. IRS guidance directs reviewers to consider the reasonableness of the taxpayers’ responses, but reviewers generally do not examine the accuracy of the information in the responses because they do not have examination authority. For certain issues, AUR reviewers may refer cases for a correspondence examination. Automated Substitute for The Automated Substitute for Return (ASFR) program uses data from the Return Program: IRP system to identify persons who did not file returns, construct tax returns for certain nonfilers, and assess tax, interest, and penalties based on those substitute returns. IRS does not pursue all of the constructed returns. Potential cases fall into one of ten priority levels and are worked highest-priority first. ASFR officials said they make budget decisions by Page 4 GAO-13-151 Tax Gap taking into account the resources available to the program and determine the level of new cases that will be worked over the following year. In fiscal year 2011, the ASFR program closed nearly 1.4 million cases. Correspondence Examinations: Correspondence examinations are formal audits of individual taxpayers but do not involve face-to-face meetings with taxpayers. Instead, these examinations target specific issues that are limited in scope and complexity, easily documented, and can be handled quickly and efficiently through correspondence between the taxpayer and the IRS examiner. Tax returns are selected as potential cases through automated business rules that filter or select tax returns according to predetermined criteria. These business rules can detect multiple potential issues, all of which can be worked through a single correspondence exam. Examiners have the authority to review additional issues on a return even if they were not identified by the automatic filters. Field Examinations: Field examinations are conducted in face-to-face meetings between the taxpayer and the IRS examiner. These audits are targeted at individual returns with broader and more complex issues. Unlike correspondence examinations, the field examination program has a classification process where an experienced tax examiner will review a potential case to determine which, if any, issues should be examined. Individual tax returns are selected for field examination in a variety of ways. Some returns are selected in the pursuit of specifically identified compliance issues, such as abusive transactions or offshore compliance. Others are selected on the basis of a statistical formula that attempts to predict the potential for additional tax assessments, and yet others are selected randomly for research purposes. Regardless of why the return was initially selected for audit, an examiner will review the return in its entirety to determine if other issues are present. The responsibility for operating these individual taxpayer enforcement programs largely rests with IRS’s Small Business/Self-Employed (SB/SE) Division, which handles complex individual returns, and Wage and Investment (W&I) Division, which handles simpler returns. SB/SE operates parts of all four IRP and exam programs; W&I operates parts of three programs, excluding field examinations. Page 5 GAO-13-151 Tax Gap Examinations of Taxpayers with Less than $200,000 in Positive Income Accounted for Most of the Total Cost of the Four Enforcement Programs Most IRS Enforcement Correspondence and field exams accounted for more than 80 percent of Resources are Spent on the total administrative costs of the four programs we reviewed over the Examinations 2-year period we examined. (Total costs include direct examination time, training and other offline activities of examiners, supervisory and administrative support, and other overhead costs allocable to each program.) Based on data for hourly costs and time spent on different types of cases that IRS provided, we estimated that the cost per case for field exams, $2,278, was many times greater than those for correspondence exams, $274, AUR, $52, and ASFR, $72. (See fig. 1.) Page 6 GAO-13-151 Tax Gap Figure 1: Cost Shares and Cost per Case for Four Exam and IRP Programs for Cases Opened in Fiscal Years 2007 and 2008 Note: Dollar figures have been adjusted for inflation to 2011 dollars using the GDP deflator. High-Income Taxpayers IRS spent almost 20 percent of the $1.6 billion per year that it devoted to Representing 3 Percent of exams opened in 2007 and 2008 on returns with positive income of at Returns Filed Accounted least $200,000, even though such returns accounted for only 3 percent of the 136 million individual income tax returns filed per year. 5 The share of for Almost 20 Percent of total cost for these returns was greater than their share of total returns Total Exam Costs because they were examined at above average rates and, compared to lower-income returns, field exams were a greater proportion of their examinations. (See fig. 2.) 5 Exams opened in 2007 and 2008 were generally filed in 2006 and 2007. Page 7 GAO-13-151 Tax Gap Figure 2: Number of Returns, Exam Costs, and Coverage Rates for Different Categories of Individual Income Taxpayers Note: Dollar figures have been adjusted for inflation to 2011 dollars using the GDP deflator. For the 2 years of cases we reviewed, exams (both correspondence and Direct Revenue field) of taxpayers with positive incomes of at least $200,000 produced Return on Investment significantly more direct revenue per dollar of cost than exams of lower- income taxpayers. Across income groups, correspondence exams were Was Highest for significantly more productive than field exams in terms of discounted Examinations of direct revenue per dollar of cost. (See fig. 3 and table 1 in app. II.) We estimated that the average direct revenue yield per dollar of cost across Taxpayers with at all correspondence exams of individual taxpayers was $7. In contrast, the Least $200,000 in average direct yield per dollar for field exams of individual taxpayers was $1.8. We also estimated that the direct revenue per dollar of cost was Positive Income about $22 for AUR cases and about $31 for ASFR cases. Page 8 GAO-13-151 Tax Gap Figure 3: Ratios of Direct Revenue to Costs for Different Types of Exams and Groups of Individual Income Tax Returns for Cases Opened in Fiscal Years 2007 and 2008 Note: Dollar figures have been adjusted for inflation to 2011 dollars using the GDP deflator. Direct revenue has been discounted over the gap between the year in which IRS incurred the exam costs and the year in which the revenue was collected. Exams that are more complicated than average are likely to require both more time to complete and more highly skilled examiners, who cost more per hour. In estimating the results for field exams in figure 3, we incorporated differences in the amount of time spent on each field exam, which is recorded in the ERIS database, but we did not account for differences in hourly costs relating to varying skill levels of examiners across cases because the data available for that purpose were limited. 6 Nevertheless, to test the potential sensitivity of our results to this missing 6 According to IRS officials, the ERIS data relating to time spent on correspondence exams are not reliable for our purposes. On their advice, given that neither the time spent on 1040 correspondence exams nor the skill level of examiners typically vary significantly from case to case, we used the same cost estimate for all cases, which IRS provided to us. Page 9 GAO-13-151 Tax Gap factor, we estimated an alternative set of field exam results, using an ERIS data element that reflects the expected difficulty of an exam. We also tested the effect of differences in locality pay for field examiners in different geographic locations. (See app. I for further details.) We found that adjusting for skill levels likely reduces some of the differences in direct revenue per dollar of cost across field exam categories; adjusting for location has a negligible effect. (See table 3 in app. II.) IRS would be able to estimate ratios of direct revenue to cost that better incorporate differences in the hourly costs across examiners with different skill levels if data from IRS’s timekeeping system that records the number of hours that each employee charged to specific exam cases were matched to revenue data for the same cases. Although Some Caution is Warranted, Exam Resource Reallocation Could Produce Significant Direct Revenue Gains Modest Reallocations Our analysis of a hypothetical reallocation of IRS examination resources Might Raise Billions of for this 2-year period indicates that a shift of about $124 million in Dollars in Direct Revenue enforcement resources could have increased direct revenue by $1 billion over the $5.5 billion per year IRS actually collected. This result is based with Little, If Any, Decline on shifting the $124 million from exams of lower-income returns with the in Voluntary Compliance earned income tax credit (EITC) and lower-income business returns without EITC to exams of higher-income returns and lower-income nonbusiness returns without EITC. The result holds true as long as the average ratio of direct revenue to cost for each category of returns remained unchanged. 7 (See fig. 4.) Similar gains would recur annually, 7 The results presented in figure 4 and table 2 in appendix II are based on our estimates of direct revenue to cost from table 1. If, alternatively, we used the ratios for field exams that incorporate the adjustments for difficulty (from table 3) the direct revenue yield of our hypothetical reallocation would be about $900 million. Page 10 GAO-13-151 Tax Gap relative to the revenue that IRS otherwise would collect if it did not change its resource allocation and taxpayer behavior remained substantially the same. Figure 4: Changes in Resources and Direct Revenue for Different Groups of Individual Income Tax Exams under a Hypothetical Reallocation Notes: Negative amounts show resources taken away and corresponding revenue lost while positive amounts show increases. Dollar figures have been adjusted for inflation to 2011 dollars using the GDP deflator. Direct revenue has been discounted over the gap between the year in which IRS incurred the exam costs and the year in which the revenue was collected. We took account of several constraints when designing our hypothetical resource reallocation example. First, we did not want to suggest a large- scale change because some reallocations cannot be made quickly, particularly if they require a different distribution of examiner skills than exists in IRS’s current workforce. The $124 million that we shifted represents less than 8 percent of the $1.6 billion per year that IRS devoted to examinations of individual tax returns for the 2 years we studied and we shifted less than 5 percent of existing field exam Page 11 GAO-13-151 Tax Gap resources ($1.1 billion per year) to correspondence exams. 8 Second, we did not want to end up with extreme coverage rates (either high or low) in any return category. Therefore, we did not reduce the combined coverage rate for any category for which the coverage rate was already close to or below 1 percent, and we kept the highest coverage rate (for returns with positive incomes of $1 million or more) under 11 percent. 9 (Nevertheless, that 11 percent rate is almost twice the current rate for that category.) Finally, given that certain compliance issues can be reviewed effectively only through a field exam, we did not decrease field exam resources in any return category for which we increased correspondence exam resources. Exam resource reallocation can also affect tax collections indirectly by influencing the voluntary compliance of nonexamined taxpayers. These indirect effects are difficult to estimate and IRS has no empirical evidence that would allow it to say whether overall voluntary compliance would increase or decrease as a result of specific resource reallocations. Changes in exam coverage rates are generally believed to affect voluntary compliance by altering taxpayers’ perceived risks of being audited. The higher the risk of being audited, the less inclined taxpayers are to evade taxes. As shown in figure 5, our hypothetical reallocation would have increased combined coverage rates in most of the tax return categories we examined. For those categories in which coverage rates declined, the declines were relatively modest. For these reasons we believe that the direct revenue gains associated with our hypothetical reallocation would not likely be offset by significant indirect revenue losses. However, if larger resource allocations were considered, the lack of empirical evidence on the potential changes in voluntary compliance could leave IRS uncertain of the extent to which direct revenue gains might be offset by negative indirect revenue effects. Although research on this issue is challenging, IRS might be able to leverage its existing efforts to study voluntary compliance through the National Research Program (NRP) to get better information on the influence of enforcement activity on voluntary compliance. 8 See table 1 in appendix II for details of the existing allocation across groups and table 2 for details on the reallocations. 9 The combined coverage rate equals the number of correspondence and field exams for a particular group, divided by the number of returns filed by that group. Page 12 GAO-13-151 Tax Gap Figure 5: Coverage Rates for Different Types of Exams and Groups of Individual Income Tax Returns Before and After Hypothetical Reallocation Consideration of Our analysis focused upon ratios of average direct revenue to average Additional Factors Could cost. We did not incorporate other potentially important considerations Improve Upon Allocation due to data constraints. One such consideration is the extent to which the ratio of direct revenue per dollar of cost may decline for a particular Decisions that are Based category of exams as additional resources are devoted to that category. Solely on Ratios of The revenue yield of each additional return that IRS exams within a Average Direct Revenue to particular return category may be lower than the average revenue- Costs productivity rates we estimated, particularly if IRS’s return selection process for examinations results in returns with the greatest revenue potential being worked first and those with the least potential being worked last. Little is known about the relationship between marginal and average revenue and cost within specific return categories because IRS Page 13 GAO-13-151 Tax Gap currently does not identify the marginal cases worked each year. 10 Until IRS collects some information on marginal cases, such as how the broad characteristics of those returns that would likely be selected (or not selected) in a modest program expansion (or contraction) would differ from the average return actually audited now, planners would have to rely solely upon ratios of average direct revenue to average cost—a less accurate basis for estimating the direct revenue consequences of specific exam resource allocations. An analysis of the marginal revenue yields for specific categories of returns might also enable IRS to reduce the number of audits that result in no direct change in tax liability (although they may have beneficial effects on voluntary compliance). These no-change cases impose burdens on compliant taxpayers. Further, substantial variations across return categories in the percentage of exams that result in no change could be viewed as inequitable because compliant taxpayers in some categories have a greater chance of being burdened than compliant taxpayers in other categories. No-change rates in some higher-income return categories are already relatively high, compared to rates for lower- income categories. For example, the no-change rate for correspondence exams of tax returns with positive income of $1 million or more was about 53 percent for fiscal years 2007 to 2008. (See table 1 in app. II.) However, the highest no-change rates are associated with correspondence exams, which should be less burdensome than field exams. High no-change rates could also be associated with declining revenue yields in marginal cases; however, without a specific study of marginal cases, it is not possible to say whether no-change cases are concentrated among the last cases examined in a particular category or whether they are spread relatively evenly across exams worked throughout the course of the year. 10 Marginal revenue and cost are those associated with the marginal cases worked in each return category. The marginal cases are those that would not have been worked if the resources allocated to a particular category had been slightly less, as well as those additional cases that would have been worked if the resources for the category had been slightly greater. Some of IRS’s exam priorities are related to potential revenue yield; however, IRS also gives high priority to returns randomly selected for the National Research Program. This random selection process is important to the research that identifies significant areas of noncompliance, but it represents one reason to doubt that exam cases worked lower down the list of IRS’s priorities always have lower returns than those worked earlier. Page 14 GAO-13-151 Tax Gap Factors other than revenue yields and IRS budget costs also matter for purposes of an overall cost-benefit evaluation of IRS exam activities. These activities impose compliance costs on taxpayers and economic efficiency costs on society. Return categories with low ratios of direct revenue to IRS budget costs could have offsetting advantages in terms of lower efficiency and compliance costs; however, no empirical evidence of variations in these other effects or costs across the return categories exists, nor would it be easy to obtain. (See app. III for further discussion of these tradeoffs.) The results of our analyses suggest that there is potential for IRS to Conclusions increase the direct revenue yield of selected enforcement programs by hundreds of millions of dollars per year without significant (if any) adverse effect on the indirect effect that examinations have on revenues. However, our results are preliminary and limited in scope. The collection and analysis of additional data would help to both confirm our basic conclusion and assist IRS in more finely adjusting its resource allocation decisions. One priority would be to study the feasibility of estimating the marginal revenue and marginal costs within each program within each taxpayer group. It would be helpful, for example, to estimate at least how the broad characteristics of those returns that would likely be selected (or not selected) in a modest program expansion (or contraction) would differ from the average return actually audited now. Such information would help IRS assess the extent to which revenue productivity would likely decline, if at all, if more exam resources are devoted to a particular group of taxpayers. Another useful project would be to see if some linkage could be made between the amounts of time that specific examiners spend on each case and the revenue collection amounts for each case that are recorded in ERIS. Such a link would enable IRS to estimate ratios of direct revenue to cost that better incorporate differences in the hourly costs across examiners with different skill levels. The collection or estimation of other information that would be useful when allocating resources, such as the influence of enforcement activity on voluntary compliance, is challenging, which is why little is known about those topics to date. Nevertheless, IRS might be able to leverage its existing efforts to study voluntary compliance through the NRP to get better information on the influence of enforcement activity on voluntary compliance. In the absence of the additional data identified above, IRS planners can use the results of an analysis such as ours in combination with their professional judgment to decide whether the potential for direct revenue gains more than offsets the potential for reductions in indirect revenue or Page 15 GAO-13-151 Tax Gap in equity and any increases in compliance or efficiency costs. If the answer is positive, they can adjust their allocation of resources accordingly. Nevertheless, the better empirical basis IRS planners have for making such judgments, the more confident they can be that they are allocating their limited resources to the best effect. To better ensure that IRS’s limited enforcement resources are allocated in Recommendations a manner that maximizes the revenue yield of the income tax, subject to other important objectives of tax administration, such as minimizing compliance costs and ensuring equitable treatment across different groups of taxpayers, the Commissioner of Internal Revenue should: • review disparities in the ratios of direct revenue yield to costs across different enforcement programs and across different groups of cases within programs and determine whether this evidence provides a basis for adjusting IRS’s allocation of enforcement resources each year. As part of this review, IRS should: • develop estimates of the marginal direct revenue and marginal direct cost within each enforcement program and each taxpayer group; • compile data on the amount of time that specific grades of examiners and downstream employees spend on specific categories of exams that can be identified in ERIS; and • explore the potential of estimating the marginal influence of enforcement activity on voluntary compliance, potentially taking advantage of new NRP data. We requested written comments from the Commissioner of Internal Agency Comments Revenue and received a letter from IRS Deputy Commissioner for Services and Enforcement on November 29, 2012, (which is reprinted in app. IV). IRS agreed with our recommendations and agreed that the development of additional key data will require considerable work. In recognition of the time it will take to obtain this information, IRS said it will consider how to apply interim methods, findings, or approximations. As agreed with your offices, unless you publicly announce the contents of this report earlier, we plan no further distribution until 30 days from the report date. At that time, we will send copies to interested congressional committees, the Secretary of the Treasury, the Commissioner of Internal Page 16 GAO-13-151 Tax Gap Revenue, and other interested parties. In addition, the report also will be available at no charge on the GAO website at http://www.gao.gov. If you or your staff have any questions about this report, please contact me at (202) 512-9110 or email@example.com. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. GAO staff who made major contributions to this report are listed in appendix V. James R. White Director, Tax Issues Strategic Issues Page 17 GAO-13-151 Tax Gap Appendix I: Summary Methodology for Data Appendix I: Summary Methodology for Data Analysis Analysis Our principal analysis compares the costs and direct revenues associated with correspondence and field exams that were opened during fiscal years 2007 and 2008 across the principal categories of individual taxpayers that the Internal Revenue Service (IRS) uses for exam planning purposes. These categories, defined in terms of income size and the nature of items reported on the returns, are shown in table 1 of appendix II. Exam Costs IRS provided us with total cost estimates for correspondence exams (with and without the earned income tax credit (EITC)) and field exams (with and without EITC) from its Integrated Financial System. These total cost estimates included all direct examiner costs, training and other off-line activities of examiners, supervisory and administrative support, and other overhead costs allocable to each program. We estimated hourly costs by dividing the total costs by the average examination hours per case, which IRS provided for correspondence exams (with and without the earned income tax credit (EITC)) and field exams (with and without EITC) from its Audit Information Management System (AIMS). 1 These hourly cost estimates are adequate for the relatively high-level comparisons we present in this report. IRS would be able to make more precise estimates for more detailed categories of exams if data from IRS’s timekeeping system that records the number of hours that each employee charged to specific exam cases were matched to revenue data for the same cases. To estimate the cost for each category of field exams we multiplied the hourly cost rates by the number of direct hours reported in the Enforcement Revenue Information System (ERIS) for each category of exam. 2 According to IRS officials the ERIS data relating to time spent on correspondence exams is not reliable for our purposes. 3 On their advice, given that neither the time spent on 1040 correspondence exams nor the 1 IRS provided six separate estimates for the total costs and the average examination hours per case—for Wage and Investment correspondence exams of returns with and without EITC, for Small Business/Self-Employed (SB/SE) correspondence exams of returns with and without EITC, and for SB/SE field exams of returns with and without EITC. 2 ERIS examination hours data originate in AIMS. 3 We determined that ERIS data were sufficiently reliable for other purposes identified in this report. Page 18 GAO-13-151 Tax Gap Appendix I: Summary Methodology for Data Analysis skill level of examiners typically vary significantly from case to case, we used the same cost estimate for all cases, which IRS provided to us. We restated the costs for each case in terms of 2011 dollars by adjusting for inflation. Due to limitations of the ERIS data our cost estimates do not include any downstream costs that IRS’s Collections function may have devoted to these cases or any costs associated with examinations of pass-through entities that could have improved the productivity of some of these 1040 exams. We do not know whether the prevalence of these missing costs varies significantly across our exam categories; however, we do note that the one category we studied that specifically excluded returns with pass-through income had ratios of revenue to costs that were greater or equal to the ratios for the one category identified as likely to include such returns. 4 Direct Revenues We aggregated all of the tax, interest, and penalty collections recorded in ERIS for the same fiscal years, exam types, and taxpayer categories used for the cost side of our analysis. We also included amounts of refunds disallowed due to examinations in our definition of revenues. These amounts represent revenue saved for the government, even though IRS does not have to collect it after the exams. 5 We compiled the revenue data for each fiscal year in which the collections were made (through the end of fiscal year 2011) and restated the revenue in terms of 2011 dollars by adjusting for inflation. 6 Then we discounted the value of collections over the gap between the fiscal year in which IRS incurred the exam costs (which we estimated as being the midpoint of the exam) and the fiscal year in which the revenue was collected. The purpose of this discounting, which is standard practice for cost-benefit analyses, is to 4 Returns without schedules C, E, F, or form 2106 contain no pass-through income; returns with schedule E or form 2106 are the most likely to include such income. 5 One type of revenue savings that we could not include in our analysis due to data limitations was that associated with disallowances of loss or credit carry forwards. Additionally, corrections that IRS examiners make in taxpayers’ favor should also be counted as social benefits arising from the exams because they improve the fairness of the tax system’s application; however, such corrections were not identifiable in IRS’s databases. 6 Additional collections from these 2007 and 2008 cases could continue to come in after 2011, which implies that our ratios of revenue to cost may be slightly understated; however, the vast majority of collections typically are made within two years of exam closures. Page 19 GAO-13-151 Tax Gap Appendix I: Summary Methodology for Data Analysis account for the time-value of money between the time at which the government bears the cost of an activity or investment and the time at which it receives the related benefit. (We used a real discount rate for this discounting because we had already adjusted all of our figures for inflation.) Incorporating Exam The first set of estimates that we present in this report do not reflect Difficulty and Location potential differences in costs across exams due to the degree of experience demanded of the examiner or the location at which the exam was conducted. For a second set of estimates we adjusted the cost of field exams for relative difficulty. To do this we used ERIS data on hours by grade to compute a weighted average pay rate for all exams (for each combination of EITC and non-EITC and field or correspondence for each year). We then adjusted costs for each record by multiplying the cost by the mid-point pay rate for the grade of the record, divided by the weighted average difficulty pay rate for the relevant year and EITC status. For a third set of estimates we made this difficulty adjustment for field exams and then we also adjusted the costs of both correspondence and field exams for location differences by using data on the location of exams, hours, and locality pay for each location to compute a national average locality pay rate (weighted by the number of hours in each location) for each combination of field and correspondence exams of returns with and without EITC and a locality pay rate for each location. We then multiplied the cost estimate for each exam by the ratio of the national rate over the relevant location-specific rate. Effects of a Hypothetical The columns labeled “Change in Resources” in table 2 of appendix II Resource Reallocation show the amounts of IRS budget resources we moved out of or into specified exam categories for our hypothetical reallocation. These shifts were guided by the considerations we noted earlier. We estimated the revenue effect of each shift by multiplying the gain or loss of resources for each category by our estimated ratios of direct revenue to cost for those categories. We estimated the effect on the coverage rate within each category by multiplying the coverage rate prior to the reallocation by the percentage change in each category’s resources caused by the reallocation. Page 20 GAO-13-151 Tax Gap Appendix II: Detailed Tables Appendix II: Detailed Tables Table 1: Costs, Direct Revenue, Exam Coverage Rates and No-Change Rates for Different Types and Groups of Exams Opened In Fiscal Years 2007 and 2008 (Weighted Average of the 2 Years) Correspondence exams Field exams Cost Direct No- Cost Direct No- Combined per year revenue / Coverage change per year revenue / Coverage change coverage Taxpayer group ($ millions) cost rate rate ($ millions) cost rate rate rate Individual income tax returns, total 516.4 7.0 0.8% 15.5% 1,093.7 1.8 0.2% 11.5% 1.0% Returns with total positive income under $200,000: Business and nonbusiness returns with earned income tax credit by size of total gross receipts: Under $25,000 247.5 5.5 1.8% 11.1% 22.9 1.2 0.1% 13.0% 1.9% $25,000 or more 15.6 3.2 4.4% 11.1% 84.2 1.0 2.1% 8.5% 6.5% Nonbusiness returns without earned income tax credit Without Schedules C, E, 114.4 7.9 0.4% 21.0% 89.8 1.6 0.1% 15.5% 0.4% F, or Form 2106 With Schedule E or Form 56.2 6.5 0.9% 10.5% 119.1 1.6 0.3% 7.0% 1.3% 2106 Business returns without earned income tax credit: Nonfarm business returns by size of total gross receipts: Under $25,000 32.9 5.7 0.8% 16.0% 128.3 1.2 0.5% 7.5% 1.2% $25,000 under 13.8 4.4 1.0% 13.0% 98.7 1.0 1.0% 9.5% 1.9% $100,000 $100,000 under 8.2 5.5 2.2% 9.6% 95.6 1.2 2.8% 8.0% 5.0% $200,000 $200,000 or 1.0 2.6 0.3% 45.1% 152.4 1.0 2.2% 15.4% 2.5% more Farm returns 1.0 3.5 0.2% 40.1% 17.6 2.0 0.2% 16.6% 0.5% Returns with total positive income of at least $200,000 and under $1 million: Nonbusiness returns 14.6 25.6 1.7% 34.9% 72.3 2.9 0.6% 15.9% 2.3% Business returns 8.2 13.0 1.2% 34.6% 130.9 1.7 1.6% 17.8% 2.9% Returns with total positive income of 3.1 47.2 2.8% 52.7% 81.8 6.0 3.3% 21.3% 6.2% $1 million or more Source: GAO analysis of IRS data. Page 21 GAO-13-151 Tax Gap Appendix II: Detailed Tables Note: Dollar figures have been adjusted for inflation to 2011 dollars using the GDP deflator. Direct revenue has been discounted over the gap between the year in which IRS incurred the exam costs and the year in which the revenue was collected. Table 2: Potential Changes in Direct Revenue and Exam Coverage Rates Resulting from a Hypothetical Reallocation of Exam Resources Correspondence exams Field exams Change in Change in Change in direct Coverage Change in direct Coverage resources revenue rate Coverage resources revenue rate Coverage Taxpayer group ($ millions) ($ millions) before rate after ($ millions) ($ millions) before rate after Individual income tax returns, 50.0 874.0 -50.0 130.1 total (net) Returns with total positive income under $200,000: Business and nonbusiness returns with earned income tax credit by size of total gross receipts: Under $25,000 -26.4 -145.5 1.8% 1.6% $25,000 or more -3.5 -11.3 4.4% 3.4% -14.9 -14.1 2.1% 1.8% Nonbusiness returns without earned income tax credit: Without Schedules C, E, 41.5 328.4 0.4% 0.5% F, or Form 2106 With Schedule E or 14.7 94.9 0.9% 1.2% Form 2106 Business returns without earned income tax credit: Nonfarm business returns by size of total gross receipts: Under $25,000 $25,000 under -1.7 -7.4 1.0% 0.9% -26.5 -25.4 1.0% 0.7% $100,000 $100,000 under -22.3 -26.8 2.8% 2.1% $200,000 $200,000 or -28.9 -29.4 2.2% 1.8% more Farm returns Returns with total positive income of at least $200,000 and under $1 million: Nonbusiness returns 14.0 356.7 1.7% 3.3% 9.1 26.5 0.6% 0.7% Business returns 8.2 106.8 1.2% 2.5% Page 22 GAO-13-151 Tax Gap Appendix II: Detailed Tables Correspondence exams Field exams Change in Change in Change in direct Coverage Change in direct Coverage resources revenue rate Coverage resources revenue rate Coverage Taxpayer group ($ millions) ($ millions) before rate after ($ millions) ($ millions) before rate after Returns with TPI of $1 million or 3.2 151.4 2.8% 5.7% 33.6 200.2 3.3% 4.7% more Source: GAO analysis of IRS data. Note: Dollar figures have been adjusted for inflation to 2011 dollars using the GDP deflator. Direct revenue has been discounted over the gap between the year in which IRS incurred the exam costs and the year in which the revenue was collected. Table 3: Ratios of Direct Revenue to Cost for Field Exams Opened in 2007 and 2008 after Accounting for Various Factors Adjusted Adjusted for inflation, Adjusted for inflation, for inflation discounting and discounting, difficulty Taxpayer group and discounting difficulty and location Individual income tax returns, total 1.8 1.8 1.8 Returns with total positive income under $200,000: Business and nonbusiness returns with earned income tax credit by size of total gross receipts: Under $25,000 1.2 1.2 1.2 $25,000 or more 1.0 1.0 1.0 Nonbusiness returns without earned income tax credit: Without Schedules C, E, F, or Form 2106 1.6 1.7 1.7 With Schedule E or Form 2106 1.6 1.8 1.8 Business returns without earned income tax credit: Nonfarm business returns by size of total gross receipts: Under $25,000 1.2 1.3 1.3 $25,000 under $100,000 1.0 1.0 1.0 $100,000 under $200,000 1.2 1.2 1.2 $200,000 or more 1.0 1.0 1.0 Farm returns 2.0 1.8 1.9 Returns with total positive income of at least $200,000 and under $1 million: Nonbusiness returns 2.9 2.7 2.7 Business returns 1.7 1.6 1.6 Returns with TPI of $1 million or more 6.0 5.1 5.0 Source: GAO analysis of IRS data. Note: Dollar figures have been adjusted for inflation to 2011 dollars using the GDP deflator. Direct revenue has been discounted over the gap between the year in which IRS incurred the exam costs and the year in which the revenue was collected. Page 23 GAO-13-151 Tax Gap Appendix III: Economic Guidelines for Appendix III: Economic Guidelines for Enforcement Resource Allocation Decisions Enforcement Resource Allocation Decisions An economic cost-benefit evaluation of IRS’s overall activities would involve a comparison of the social costs and social benefits associated with those activities. 1 IRS’s function is to collect tax revenue that the federal government transfers among citizens as cash payments or in the form of goods and services. The collection process imposes costs on society but produces no direct benefit itself. The government’s use of the collected revenue may ultimately produce a net benefit for society if the social value of that use exceeds the social cost of raising the revenue. IRS has no influence over how tax revenue is used; it can only contribute to increasing the net social benefit by increasing the amount of revenue collected for a given amount of social cost (or decreasing the social cost of raising a given amount of revenue). Specific resource allocation choices can be compared on the basis of the amount of revenue they produce for a given amount of total social costs. Components of Total The social costs of tax collection comprise the following: Social Costs • IRS budget costs. • Tax burden. This is the actual money collected from taxpayers. Amounts collected as a result of IRS enforcement activities from taxpayers who, otherwise, would have been noncompliant may have a zero social cost. The cost that those amounts represent can be attributed to the tax law, rather than to IRS enforcement efforts. If those additional amounts of taxes due are not collected, the tax burdens evaded by noncompliant individuals are offset by the additional taxes that compliant taxpayers must pay in order to support a given government budget. • Compliance burden. IRS’s enforcement activities can affect the costs that taxpayers incur when complying with the tax law by increasing the time and money that they spend preparing their returns and interacting with IRS. 1 Social costs are the sum of all private costs and any external costs to society arising from the production of a good or service. Private costs are incurred by producers of goods and services and are reflected in the prices charged to consumers for those goods and services. External costs are costs arising from production, such as pollution, that are borne by third parties (other than the producers or consumers). Similarly, social benefits are the sum of all private benefits and any external benefits arising from the production of a good or service. Page 24 GAO-13-151 Tax Gap Appendix III: Economic Guidelines for Enforcement Resource Allocation Decisions • Efficiency costs. IRS’s enforcement activities can alter the tax avoidance and evasion behavior of individuals, which affects the efficiency of resource allocation in the economy. If an enforcement activity increases the aggregate costs of tax avoidance and evasion, economic efficiency and the average standard of living is reduced. Conversely, if the activity reduces such aggregate costs, economic efficiency would improve. • Equity costs. IRS’s resource allocation can affect how exam-related compliance burdens are distributed across different groups of taxpayers and also how the risk of noncompliant taxpayers getting penalized for evasion varies across groups. It is difficult to know what society as a whole would view as an equitable distribution of these burdens and risks; therefore it is difficult to assess the equity effects of any particular reallocation of resources. The only component of social costs that can be reliably measured is the IRS budget cost, and it is difficult to attribute even that cost to very specific enforcement activities (such as specific audits). Consequently, IRS planners cannot consider all types of social costs in a rigorously quantitative manner when making their resource allocation decisions. Marginal Analysis Economists use the term “margin” when referring to the scopes of the various types of decisions that individuals make. For example, if IRS examination planners were deciding how to allocate the last million dollars of their budget between different types of audits, the marginal social cost of the choice they made would be a million dollars, plus the sum of all other social costs resulting from the IRS activities supported by that million dollars. The marginal revenue would be the amount of additional tax collections attributable (both directly and indirectly) to those activities. The most economically efficient choice would be the one that produced the highest ratio of marginal revenue to marginal social cost. The ratio of marginal revenue to marginal social cost provides a basis for comparing the cost of collecting taxes by different approaches. Such comparisons can be made across broadly defined approaches (e.g., increasing taxpayer services to promote higher voluntary compliance versus increasing enforcement efforts to reduce noncompliance). Alternatively, as in this study, comparisons could be made across more narrowly defined alternatives (e.g., devoting more resources to audits of taxpayers with incomes below a certain amount versus devoting those resources to audits of taxpayers with incomes above that amount). Page 25 GAO-13-151 Tax Gap Appendix IV: Comments from the Internal Appendix IV: Comments from the Internal Revenue Service Revenue Service Page 26 GAO-13-151 Tax Gap Appendix IV: Comments from the Internal Revenue Service Page 27 GAO-13-151 Tax Gap Appendix V: GAO Contact and Staff Appendix V: GAO Contact and Staff Acknowledgments Acknowledgments James R. White, (202) 512-9110 or firstname.lastname@example.org GAO Contact In addition to the contact named above, James Wozny (Assistant Staff Director), Kevin Daly (Assistant Director), Michael Brostek, Ethan Acknowledgments Wozniak, Suzanne Heimbach, Sara Daleski, Lois Hanshaw, Karen O’Conor, Ray Bush, Elizabeth Fan, and Robert MacKay made key contributions to this report. (450795) Page 28 GAO-13-151 Tax Gap The Government Accountability Office, the audit, evaluation, and GAO’s Mission 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. The fastest and easiest way to obtain copies of GAO documents at no Obtaining Copies of cost is through GAO’s website (http://www.gao.gov). Each weekday GAO Reports and afternoon, GAO posts on its website newly released reports, testimony, and correspondence. To have GAO e-mail you a list of newly posted Testimony products, go to 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, 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 on Facebook, Flickr, Twitter, and YouTube. Connect with GAO Subscribe to our RSS Feeds or E-mail Updates. Listen to our Podcasts. Visit GAO on the web at www.gao.gov. Contact: To Report Fraud, Website: http://www.gao.gov/fraudnet/fraudnet.htm Waste, and Abuse in E-mail: email@example.com Federal Programs Automated answering system: (800) 424-5454 or (202) 512-7470 Katherine Siggerud, Managing Director, firstname.lastname@example.org, (202) 512- Congressional 4400, U.S. Government Accountability Office, 441 G Street NW, Room Relations 7125, Washington, DC 20548 Chuck Young, Managing Director, email@example.com, (202) 512-4800 Public Affairs U.S. Government Accountability Office, 441 G Street NW, Room 7149 Washington, DC 20548 Please Print on Recycled Paper.
Tax Gap: IRS Could Significantly Increase Revenues by Better Targeting Enforcement Resources
Published by the Government Accountability Office on 2012-12-05.
Below is a raw (and likely hideous) rendition of the original report. (PDF)