This is the accessible text file for GAO report number GAO-09-769 entitled 'Home Mortgage Interest Deduction: Despite Challenges Presented by Complex Tax Rules, IRS Could Enhance Enforcement and Guidance' which was released on August 31, 2009. This text file was formatted by the U.S. Government Accountability Office (GAO) to be accessible to users with visual impairments, as part of a longer term project to improve GAO products' accessibility. Every attempt has been made to maintain the structural and data integrity of the original printed product. Accessibility features, such as text descriptions of tables, consecutively numbered footnotes placed at the end of the file, and the text of agency comment letters, are provided but may not exactly duplicate the presentation or format of the printed version. The portable document format (PDF) file is an exact electronic replica of the printed version. We welcome your feedback. Please E-mail your comments regarding the contents or accessibility features of this document to Webmaster@gao.gov. This is a work of the U.S. government and is not subject to copyright protection in the United States. It may be reproduced and distributed in its entirety without further permission from GAO. Because this work may contain copyrighted images or other material, permission from the copyright holder may be necessary if you wish to reproduce this material separately. Report to the Joint Committee on Taxation: United States Government Accountability Office: GAO: July 2009: Home Mortgage Interest Deduction: Despite Challenges Presented by Complex Tax Rules, IRS Could Enhance Enforcement and Guidance: GAO-09-769: Contents: Letter: Background: IRS's Enforcement and Research Programs Do Not Provide an Overall Picture of Mortgage Interest Deduction Noncompliance: Taxpayers Face a Variety of Problems Trying to Comply with Mortgage Interest Deduction Rules, and IRS Faces Challenges in Detecting Noncompliance: Options Exist for Expanding Mortgage Information Reporting and Using Private Sector Data to Enhance Enforcement: IRS's Guidance to Taxpayers and Training Materials for Examiners Did Not Indicate the Range of Possible Mortgage Interest Deduction Situations: Tax Software Differed in the Ease of Navigating Mortgage Interest Deduction Instructions: Conclusions: Recommendations for Executive Action: Agency Comments and Our Evaluation: Appendix I: Scope and Methodology: Appendix II: Schedule A, Itemized Deductions: Appendix III: Examination Results Involving Mortgage Interest Deductions: Appendix IV: Examples Illustrating Mortgage Interest Deduction's Complexity: Appendix V: Descriptive Data on Home Mortgage Interest Deduction Claims and Form 1098 Filings: Appendix VI: Margins of Error Tables for Estimates in Appendix V: Appendix VII: SMR Research's Data on Home Equity Debt and How IRS Might Use Similar Data for Enforcement: Appendix VIII: Comments from the Internal Revenue Service: Appendix IX: GAO Contact and Staff Acknowledgments: Tables: Table 1: Limitations on Mortgage Interest Deductibility by Type of Debt: Table 2: Amounts of Mortgage Interest Deductions, Home Sales Prices, and Home Equity Debt in 2001 and 2006: Table 3: What IRS's Main Research and Enforcement Programs Do and Do Not Show about Mortgage Interest Deduction Compliance: Table 4: Results of IRS's Fiscal Year 2008 Examinations of Schedule A, Line 10 for CIP and non-CIP Examinations: Table 5: IRS Worksheet to Figure Home Mortgage Interest: Table 6: Number and Dollar Total of Schedule A, Line 10 Deductions per Number of Form 1098 Filings, 2004 through 2006: Table 7: Distribution of Schedule A, Line 10 Deductions in Select Percentiles per Number of Form 1098 Filings, 2004 through 2006: Table 8: Results of Comparing Total Schedule A, Line 10 Deduction Amounts to Amounts Reported on Form 1098, 2004 through 2006: Table 9: Percentile Distributions for Schedule A, Line 10 Amounts Compared with Form 1098 Amounts and the Difference between Schedule A, Line 10 Amounts and Form 1098 Amounts, 2004 through 2006: Table 10: Year-to-Year Changes in Amount Deducted on Schedule A, Line 10, 2004 through 2006: Table 11: Year-to-Year Changes in the Number of Form 1098 Filings, with Distribution of Change Amounts on Schedule A, Line 10, 2004 through 2006: Table 12: Numbers of Matching Loan Account Numbers on Forms 1098, 2004 through 2006: Table 13: Number of Individual Tax Returns with Schedule A Mortgage Interest Deductions and Non-Schedule A Mortgage Interest Deductions, 2004 through 2006: Table 14: Margins of Error for Number and Dollar Total of Schedule A, Line 10 Deductions per Number of Form 1098 Filings, 2004 through 2006: Table 15: Margins of Error for Distribution of Schedule A, Line 10 Deductions in Select Percentiles per Number of Form 1098 Filings, 2004 through 2006: Table 16: Margins of Error for Results of Comparing Total Schedule A, Line 10 Deduction Amounts to Amounts Reported on Form 1098, 2004 through 2006: Table 17: Margins of Error for Percentile Distributions for Schedule A, Line 10 Amounts Compared with Form 1098 Amounts and the Difference between Schedule A, Line 10 Amounts and Form 1098 Amounts, 2004 through 2006: Table 18: Margins of Error for Year-to-Year Changes in Amount Deducted on Schedule A, Line 10, 2004 through 2006: Table 19: Margins of Error for Year-to-Year Changes in the Number of Form 1098 Filings, with Distribution of Change Amounts on Schedule A, Line 10, 2004 through 2006: Table 20: Margins of Error for Number of Matching Loan Account Numbers on Forms 1098, 2004 through 2006: Table 21: Margins of Error for Numbers of Individual Income Tax Returns with Schedule A Mortgage Interest Deductions and Non-Schedule A Mortgage Interest Deductions, 2004 through 2006: Table 22: Types of Transactions Included in SMR Research's Home Equity Debt Calculations: Figures: Figure 1: IRS Chart Illustrating How to Determine Whether Mortgage Interest Is Fully Deductible as Shown in Publication 936: Figure 2: Current Form 1098 Information Compared with Options for Additional Mortgage Debt Information: Figure 3: Internal Revenue Service Schedule A on Which the Home Mortgage Interest Deduction Is Taken: Abbreviations: AIMS: Audit Information Management System: AMT: Alternative Minimum Tax: CDW: Compliance Data Warehouse: CIP: Compliance Initiative Project: EOAD: Examination Operational Automation Database: IRS: Internal Revenue Service: NRP: National Research Program: SOI: Statistics of Income: [End of section] United States Government Accountability Office: Washington, DC 20548: July 29, 2009: The Honorable Charles B. Rangel: Chairman: Joint Committee on Taxation: House of Representatives: The Honorable Max Baucus: Vice Chairman: Joint Committee on Taxation: United States Senate: The home mortgage interest deduction is the third most expensive federal income tax expenditure, with the government expected to forgo about $80 billion of revenue for the deduction in 2009.[Footnote 1] Subject to various limitations, taxpayers may deduct interest on home- secured loans, such as mortgages, mortgage refinancings, and home equity loans, including those taken as lump sum amounts and home equity lines of credit. The rules that taxpayers must follow in determining the proper amount of mortgage interest to deduct can be complex. For example, there are limitations on the amount of debt for which interest can be deducted, special rules for refinancing, situations where alternative minimum tax (AMT) considerations apply, and rules on the deductibility of prepaid interest amounts called points. In general, complex tax rules increase the potential for noncompliance. You asked us to study the home mortgage interest deduction to determine if there are administrative issues that need to be addressed to improve taxpayer compliance and Internal Revenue Service (IRS) enforcement. For this report, we (1) provide information on how IRS detects taxpayers' noncompliance with the home mortgage interest deduction rules and what it knows about the extent of noncompliance; (2) identify the problems, if any, taxpayers face in attempting to comply with the deduction and describe IRS's challenges in detecting mortgage interest deduction noncompliance; (3) assess options to give IRS more information to enforce compliance with the rules; (4) determine whether IRS's guidance to taxpayers and its examiners' guidance and training on the deduction provide enough information to properly calculate the taxpayers' allowable mortgage interest deduction; and (5) describe how tax-return preparation software programs handle the deduction. You also asked us to provide descriptive information on taxpayers' mortgage interest deductions and mortgage interest payments reported on Form 1098, Mortgage Interest Statement. Appendix V provides this information. Consideration of statutory changes was beyond the scope of our report. To address our objectives, we analyzed IRS's efforts to determine taxpayer compliance with the mortgage interest deduction. This included examining information on IRS's compliance initiative projects that focused on the deduction and on IRS's other routine examination activities.[Footnote 2] We also analyzed information from IRS's Statistics of Income (SOI) Division and Compliance Data Warehouse (CDW) databases to compile information about individuals' year-to-year changes in claiming the deduction and obtained statistical information from the private sector on home equity loans, home equity lines of credit, and refinancings. We reviewed relevant studies and publications and interviewed cognizant IRS officials and representatives from the tax preparation and mortgage banking industries to obtain information on the issues surrounding administration of the deduction. We determined whether IRS's guidance to taxpayers prominently mentioned the limitations on the amount of debt for which interest can be deducted. To determine whether IRS examiners' training and guidance provided enough information on properly calculating the mortgage interest deduction, we compared IRS's examination training and guidance with tax law and instructions. We also analyzed how several tax preparation software packages that are widely used by individual taxpayers or paid preparers handled the calculation of the deduction by reviewing appropriate parts of the software to determine how they treated the debt limitations and by interviewing company officials. We performed tests that determined the data used in this report were reliable for our purposes. Appendix I provides a more detailed discussion of our scope and methodology. We conducted this performance audit from May 2008 through July 2009 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: Taxpayers have been allowed to deduct mortgage interest payments on their federal tax returns since Congress enacted the federal income tax in 1913. At that time, the deduction for home mortgage interest was part of the deduction allowed for any interest paid. The Tax Reform Act of 1986 limited deductions for nonbusiness interest.[Footnote 3] The 1986 act specifically disallowed deductions by individuals of personal interest but included a limited exception for qualified mortgage interest. The 1986 act was amended the next year to include dollar limits on the amount of home mortgage debt for which interest payments can be deducted.[Footnote 4] Taxpayers may deduct the interest they pay on loans secured by qualified homes--either their main home or their main home and a second home. These loans include first or second mortgages, home equity loans, and home equity lines of credit. Boats and recreational vehicles may qualify as homes if they have sleeping, cooking, and toilet facilities. As table 1 shows, two kinds of debt may qualify for the mortgage interest deduction. The first kind is acquisition debt, which is debt incurred in acquiring, constructing, or substantially improving a qualified home and that is secured by the home. Taxpayers may deduct all of the interest paid on acquisition debt incurred on or before October 13, 1987, known as grandfathered debt. Taxpayers' interest deductions for debt on qualified homes purchased after October 13, 1987, is limited to the interest on $1 million of acquisition debt, reduced by any grandfathered debt. Table 1: Limitations on Mortgage Interest Deductibility by Type of Debt: Category of limitation or deductibility: Dollar limitation on debt on which interest is deductible; Acquisition debt: Limited to $1 million[A] reduced by grandfathered debt; Home equity debt: Limited to lesser of (1) $100,000[A] or (2) home's fair market value minus acquisition debt and grandfathered debt. Category of limitation or deductibility: Limitation on use of loan proceeds on which interest is deductible; Acquisition debt: Limited to acquiring, constructing, or substantially improving a home; Home equity debt: May be used for anything, including credit card payments, cars, tuition, and vacations. Category of limitation or deductibility: Interest deductibility for alternative minimum tax (AMT) purposes; Acquisition debt: Generally deductible; Home equity debt: Generally not deductible unless proceeds are used to buy, build, or improve a home. Category of limitation or deductibility: Limitation on the refinancing amount on which interest may be deducted; Acquisition debt: Limited to the balance of the old mortgage before refinancing; refinancing cannot increase acquisition debt except to substantially improve the home; Home equity debt: Refinancing amount above the acquisition debt balance is home equity debt, up to the home equity debt limitation. Category of limitation or deductibility: Deductibility of points; Acquisition debt: Deductible in the year paid if various conditions are met and the loan is used to buy, build, or improve taxpayer's main home; if certain other tests are met, points are deductible over the life of the loan; when refinancing, except in states covered by the Eighth Circuit Court of Appeals or when used to improve taxpayer's main home and certain conditions are met, points are deductible over the life of the loan; Home equity debt: Deductible over the life of the loan. Source: GAO analysis of mortgage interest deduction rules. Note: Deductibility assumes itemization of deductions, legal liability for the loan, and a mortgage secured by a qualified home in which the taxpayer has an ownership interest. [A] Acquisition debt and home equity debt are limited to $500,000 and $50,000, respectively, for married people filing separately. [End of table] The second kind of qualified debt is home equity debt, which is any non- acquisition debt secured by the home. Its proceeds may be used for anything other than to buy, build, or substantially improve the home. For example, the proceeds may be used to pay off credit cards or finance cars or vacations. Home equity debt is limited to the home's fair market value minus the acquisition debt and grandfathered debt, or to $100,000, whichever is less. Home equity debt for tax purposes should not be confused with what is commonly called a home equity loan. The latter may qualify as either acquisition debt or home equity debt depending on how the proceeds are used.[Footnote 5] Table 1 also summarizes the deductibility of points. A point is a term used to describe certain charges paid in connection with obtaining a mortgage and is calculated at 1 percent of a mortgage's principal. Generally, points are considered prepaid interest, and, as such, the deduction is usually allocated over the life of the loan. However, if various conditions are met, the taxpayer may deduct the points in the year paid.[Footnote 6] Determining whether points are fully deductible in the year paid may require that taxpayers go through as many as 10 decision steps. Lending institutions and other entities engaged in a trade or business that receive $600 or more during the year in mortgage interest and certain points from an individual taxpayer are required to file a Form 1098, Mortgage Interest Statement. Form 1098 reports are sent both to the taxpayer and IRS to show how much mortgage interest was paid to the lender that year. Taxpayers report their deductible mortgage interest information to IRS on lines 10 through 13 of Schedule A, the schedule itemizing deductions for the Form 1040 individual income tax return (see appendix II). Typically, the larger deductible amounts are reported on line 10, which shows home mortgage interest and points reported to taxpayers on Form 1098. Lines 11 and 12 of Schedule A are for claiming mortgage interest and points, respectively, not reported on Form 1098,[Footnote 7] and, starting in 2007, line 13 is used for deductible qualified mortgage insurance premiums. Some interest payments reported on Form 1098 may be deductible on other Form 1040 schedules, such as Schedule C for sole proprietorship businesses or Schedule F for farming. Changes in the housing market over the years may have affected the overall amount of home mortgage interest that taxpayers claimed. Table 2 shows the change in various indicators related to the mortgage interest deduction from 2001 to 2006. The rise in home prices and thus mortgage amounts, partly account for the large increases in mortgage interest deductions claimed by taxpayers in 2006 compared to 2001. In 2006, the median sales price for existing U.S. single-family homes was up from 2001. Median sales prices varied widely across metropolitan areas, with some locations exceeding $700,000. The home equity loan totals reported by the Federal Reserve Board also increased significantly from 2001 to 2006. According to the U.S. Census Bureau's 2006 American Community Survey, about a fifth of owner-occupied housing units with a mortgage had a home equity loan. Also, according to the Joint Center for Housing Studies of Harvard University, 85 percent of homeowners who refinanced their mortgages in 2006 took cash out. Table 2: Amounts of Mortgage Interest Deductions, Home Sales Prices, and Home Equity Debt in 2001 and 2006: Indicator related to the mortgage interest deduction: Taxpayers' Schedule A, line 10 deductions; 2001: $323 billion; 2006: $437 billion. Indicator related to the mortgage interest deduction: Number of returns with amounts claimed on line 10; 2001: 36 million; 2006: 39 million. Indicator related to the mortgage interest deduction: Average line 10 deduction; 2001: $9,000; 2006: $11,200. Indicator related to the mortgage interest deduction: Median home sales price; 2001: $156,600; 2006: $221,900. Indicator related to the mortgage interest deduction: Home equity loan amounts reported by the Federal Reserve Board; 2001: $439 billion; 2006: $1,019 billion. Indicator related to the mortgage interest deduction: The part of home equity taken out as cash above the amount needed for the refinancing of prime, first-lien, conventional mortgages, as reported by Freddie Mac; 2001: $83 billion; 2006: $318 billion. Sources: IRS, Statistical Abstract of the U.S., Federal Reserve Board, Freddie Mac, and GAO calculations. Note: Numbers are rounded and dollars are not adjusted for inflation. Mortgage interest deduction limitations have not changed with inflation. [End of table] IRS's Enforcement and Research Programs Do Not Provide an Overall Picture of Mortgage Interest Deduction Noncompliance: Although IRS's enforcement and research programs found some mortgage interest deduction compliance problems, the methods leave gaps in what is known about the extent and specific nature of noncompliance. The four main programs that IRS uses to enforce or research mortgage interest deduction compliance include the following. * A computer matching program designed to identify taxpayers whose mortgage interest deduction exceeds amounts reported on Form 1098: Of the approximately 1.7 million mismatches the program identified above a certain threshold in 2005, the latest year available, IRS followed up on about 135,000 and changed tax assessments or refunds for about 53,000.[Footnote 8] Tax assessments totaled about $216 million, much more than in previous years, and averaged about $4,300. * IRS's National Research Program (NRP), the main ongoing study of taxpayer compliance based on examinations of a random sample of tax returns: IRS found that between 12 and 14 percent of individual taxpayers who claimed home mortgage interest on line 10 of Schedule A misreported the amount. The misreporting was split about evenly between taxpayers underreporting the deduction and taxpayers overreporting it. * Routine examinations done by correspondence, in an IRS office, or on site: IRS does not maintain separate examination results solely on line 10 issues. Most of the approximately 33,000 examinations done in fiscal year 2008 by IRS revenue agents and tax compliance officers that included a review of line 10 resulted in increases to the line 10 deduction (i.e., the deduction increased) or in no recommended change. For cases involving line 10 where examiners recommended greater tax amounts, the median tax owed for all issues examined including line 10 was $4,612, according to our analysis of IRS records. For cases closed in fiscal year 2008 in which examiners decreased the line 10 amounts and lowered taxpayers' deductions, the median decrease in the line 10 mortgage interest deduction was $6,430. Other results are listed in appendix III. * Two special compliance initiative projects (CIP) primarily examining acquisition debt for a limited segment of individual taxpayers with high mortgage interest deductions or high adjusted gross incomes: About 9,000 examinations were closed in fiscal year 2008 for both projects. IRS recommended about $95 million in tax changes for cases closed in fiscal years 2006 through 2008. Examiners decreased line 10 deductions for most cases closed in fiscal year 2008. The median decreases were about $29,000 for the first project and $28,000 for the second. For cases in which examiners increased the tax owed, the median increases in tax were about $8,000 in both projects. IRS examiners attributed the changes to taxpayers and paid tax return preparers not knowing or ignoring the mortgage interest deduction rules and to problematic tax preparation software.[Footnote 9] Appendix III also presents results from CIP exams. Table 3 shows, for these four main programs, that gaps exist in what IRS can measure or enforce concerning taxpayers' mortgage interest deductions. Table 3: What IRS's Main Research and Enforcement Programs Do and Do Not Show about Mortgage Interest Deduction Compliance: Program: Matching program; The program shows: * Situations in which taxpayers deduct more interest than what was reported on Form 1098; The program does not show: * Noncompliance in which taxpayers should have deducted less interest than the amount shown on Form 1098; * Discrepancies that fall under IRS thresholds for follow-up. Program: National Research Program; The program shows: * Population estimates for misreported mortgage interest for 2001; The program does not show: * Recent noncompliance information; complete results from IRS's ongoing compliance studies are not expected to be available for more than 2 years; * Noncompliance in which taxpayers' deductions and Form 1098 amounts match. NRP automatically excluded returns from further consideration if the Form 1098 amounts matched the Schedule A amounts, unless another problem was found. Program: Routine examinations; The program shows: * Adjustments to taxpayers' line 10 deductions and other tax changes for each tax year; The program does not show: * Noncompliance that examiners miss because examination selection methods do not consider possible noncompliance with home equity debt limits. Program: CIPs; The program shows: * Recent years' adjustments to line 10 deductions and other tax changes for a select group of taxpayers with high adjusted gross incomes or high mortgage interest deductions; focus was on acquisition debt; The program does not show: * Noncompliance by taxpayers beyond the narrowly defined groups examined in the projects. Program: All of the above; The program shows: [Empty]; The program does not show: * Information that IRS can use to categorize the reasons that misreporting occurred. For example, none of the programs compile data that isolate misreporting because of violations of the limits on acquisition or home equity debt. Source: GAO analysis of IRS interviews and data. [End of table] Of the four main programs, only NRP projects the amount of mortgage interest deduction misreporting to the population of individual taxpayers. However, because NRP automatically excluded any Schedule A deduction where the Form 1098 amounts matched the Schedule A amounts, it may have underestimated mortgage interest deduction misreporting relating to the home equity and acquisition debt limits. Taxpayers Face a Variety of Problems Trying to Comply with Mortgage Interest Deduction Rules, and IRS Faces Challenges in Detecting Noncompliance: The mortgage interest deduction rules create compliance problems for taxpayers, reflecting the deduction's complexity. The effects of the problems, however, are uneven. Although many taxpayers might encounter few problems, others could face many more. Problems cited by tax practitioners and in our review of articles on deducting home mortgage interest included the following: * Taxpayers need to distinguish between acquisition and home equity debt but did not always do so. * Taxpayers deducted interest on loans exceeding the limitations, including the acquisition, home equity, and two-home limits. * Taxpayers who were subject to the AMT and thus not eligible to deduct home equity interest claimed it nonetheless. * Tax practitioners also missed the limitations and did not comply with the AMT rules. * Depending on the circumstances, some taxpayers and practitioners faced extensive recordkeeping and calculations related to such matters as refinancing, the AMT, business use of the home, other uses of loan proceeds, and the periodic use and repayment of home equity lines of credit. One practitioner told us that completing worksheets was the quick and easy part of work related to the mortgage interest deduction; the most difficult and most time-consuming part was getting taxpayers to locate and provide the proper information on what kind of debt was involved and how loan proceeds were used. Typically, taxpayers provided records piecemeal over time and after many phone calls. * Taxpayers may have been unaware that they might need documentation on matters such as how proceeds of home equity loans are spent to properly determine the amount of the mortgage interest deduction on their tax returns. * Mortgage interest deduction limits based on debt amounts are not directly comparable with the information on Form 1098, which lists interest paid. If taxpayers' debts exceed the limits, taxpayers must calculate how much interest they can deduct. The complexity of the laws that govern the mortgage interest deduction are evident in the guidance IRS has published. Figure 1 is the flowchart in IRS's 16-page instructions to taxpayers--Publication 936: Home Mortgage Interest Deduction--that help taxpayers determine if their mortgage interest is fully deductible.[Footnote 10] It leads taxpayers through as many as seven decision points and still sometimes requires them to consult another part of the publication. Appendix IV provides two examples of the mortgage interest deduction's complexity. Figure 1: IRS Chart Illustrating How to Determine Whether Mortgage Interest Is Fully Deductible as Shown in Publication 936: [Refer to PDF for image: illustration] Instructions: Include balances of all mortgages secured by your main home and second home. Start here: 1) Do you meet the conditions[A] to deduct home mortgage interest? Yes, go to #3. No, go to #2. 2) You cannot deduct the interest payments as home mortgage interest.[D] 3) Were your total mortgage balances $100,700 or less[B] ($50,000 or less if married filing separately) at all times during the year? Yes, go to #4. No, go to #5. 4) Your home mortgage interest is fully deductible. You do not need to read Part II of this publication. 5) Were all of your home mortgages taken out on or before 10-13-87? Yes, go to #4. No, go to #7. 6) Go to Part II of this publication to determine the limits on your deductible home mortgage interest. 7) Were all of your home mortgages taken out after 10-13-87 used to buy, build, or improve the main home secured by that main home mortgage or used to buy, build, or improve the second home secured by that second home mortgage, or both? Yes, go to #9. No, go to #10. 8) Were your grandfathered debt plus home acquisition debt balances $1,000,000 or less[C] ($500,000 or less if married filing separately) at all times during the year? Yes, go to #4. No, go to #6. 9) Were the mortgage balances $1,000,000 or less ($500,000 or less if married filing separately) at all times during the year? Yes, go to #4. No, go to #10. 10) Were your home equity debt balances $100,000 or less ($500,000 or less if married filing separately) at all times during the year? Yes, go to #8. No, go to #6. [A] You must itemize deductions on Schedule A (Form 1040) and be legally liable for the loan. The loan must be a secured debt on a qualified hone. See Part I, Home Mortgage Interest. [B] If all mortgages on your main or second home exceed the home's fair market value, a lower limit may apply. See Home Equity debt limit under Home Equity Debt, Part II. [C] Amounts over the $1,000,000 limit ($500,000 if married filing separately) qualify as home equity debt if they are not more than the total home equity debt limit. See Part II of this publication for more information about grandfathered debt, home acquisition debt, and home equity debt. [D] See Table 2 in Part II of this publication for where to deduct other types of interest payments. Source: Department of the Treasury, Internal Revenue Service, Publication 936: Home Mortgage Interest Deduction: For Use in Preparing 2008 Returns. [End of figure] Addressing Compliance Problems and Complexity Would Likely Require Policy Changes: To alleviate the problems and complexity facing taxpayers complying with the mortgage interest deduction rules, policy makers likely would have to change tax laws. For example, we noted in the past that shifting mortgage interest deduction limits from debt amounts to an interest-limit cap could simplify administration of the deduction. [Footnote 11] Under an interest-limit cap, IRS could more easily compare the amounts reported on Schedule A and Form 1098 to the cap and identify cases for audit or follow-up. Similarly, taxpayers would have only to report the amount of Form 1098 interest that fell under the cap and would not have to figure out which property qualified or how to account for debt limits, depending on how any new legislation was written. However, changing the tax code for qualified residential mortgage interest deductions could have significant tax policy implications. For example, an interest-paid cap could cause taxpayers living in areas with high housing prices to be disadvantaged compared with taxpayers with similar incomes living in areas with low housing prices. A cap could also disadvantage those who borrowed during periods of high interest rates. Assessing such policy changes and whether such consequences would be appropriate are beyond the scope of this report. Because Mortgage Interest Reported on Form 1098 Is Not Directly Comparable to Legal Debt Limits, IRS Must Use Costly Methods to Detect Certain Noncompliance: Because the Form 1098 information report shows the dollar amount of interest a taxpayer paid in a year without regard to the limits on the amount of debt imposed by law, IRS's computer matching program comparing Form 1098 and tax return amounts will not detect certain noncompliance. For example, as already discussed, taxpayers cannot claim a deduction for interest exceeding the $1 million acquisition debt limit or other limitations. However, the annual information report for a taxpayer with a $1.5 million acquisition mortgage would show the interest paid on the entire mortgage instead of the interest on the $1 million under the annual debt limit. Because of this, the information IRS receives from third parties on Form 1098 cannot be used by itself to determine if taxpayers improperly claimed interest on debt in excess of the legal limitations. However, the way to overcome this problem--using examinations to detect noncompliance--is expensive compared to automated matching programs, and its payoff in increased revenue is low. For example, according to IRS examiners, pinpointing which tax returns have a home-equity debt noncompliance issue is very labor intensive. As shown earlier, for non-CIP line 10 examinations by revenue agents and tax compliance officers closed in fiscal year 2008, the median decrease to line 10 was $6,430, according to our analysis of IRS data. At the highest individual tax rate of 35 percent and assuming no offsetting factors, the resulting increase in tax revenue would be about $2,250. Although not an exact comparison, the average tax assessment IRS recommended for field examinations of individual taxpayers was about $19,150 per taxpayer. Given the non-CIP examinations' relatively low payoff, some practitioners we interviewed said the tax code as it relates to the home mortgage interest deduction is unenforceable in a practical sense. Options Exist for Expanding Mortgage Information Reporting and Using Private Sector Data to Enhance Enforcement: Additional information about taxpayers' mortgages could help IRS identify the most productive cases to examine and determine whether taxpayers are claiming the correct amount of mortgage interest deduction. IRS could obtain more helpful information about taxpayers' mortgages by expanding information collected on Form 1098. IRS officials said that in implementing certain additional reporting requirements, the agency would need to meet the terms of the Paperwork Reduction Act, which requires agencies to minimize the paperwork burden they impose on the public and maximize the practical utility of the information they collect.[Footnote 12] IRS officials also said attention would need to be paid to the added costs that expansion would impose on third parties and IRS, such as updating computer systems. Any computer programming changes for IRS and third parties would likely be done only once to accommodate the new information. Form 1098 could be revised to collect the following information: * address of the property secured by the mortgage to which the interest on the form relates; * outstanding mortgage debt balances on the property; * an indicator if the mortgage interest is for a loan that was refinanced during the year; and: * an indicator of whether the mortgage interest relates to an acquisition loan or a home equity loan. Figure 2 shows what a Form 1098 might look like if revised to collect any of this information. We found that some companies that submit Form 1098 statements to IRS already provide some of this information with the Form 1098 statements they send to borrowers. These options could be considered singly or in combination. Further, changes to Form 1098 reporting requirements would need to apply only to future reports to give filers sufficient time to adjust their computer systems to collect and calculate the additional information. Figure 2: Current Form 1098 Information Compared with Options for Additional Mortgage Debt Information: [Refer to PDF for image: illustration] Current From 1098: Recipient's/Lender's name, address, and telephone number: Recipient's federal identification number: Payer's social security number: Payer's Borrower's name: Street address (including apartment number): City, State, and Zip code: Account number (see instructions): 1. Mortgage interest received from payer(s)/borrower(s): 2. Points paid on purchase of principal residence: 3. Refund of overpaid interest: 4. Mortgage insurance premiums: Possible additions for Form 1098: 6. Address of secured property: 7. Beginning and ending principal balances: 8. Refinancing this year? Yes/No: 9. Loan Type: Acquisition; Home equity: Source: GAO analysis of IRS information. [End of figure] Adding the Address of Mortgaged Property to Form 1098: If the address relating to taxpayers' mortgage interest deductions were on the Form 1098 (see figure 2, box 6) and electronically captured in IRS databases, IRS could use an automated process to determine whether the mortgage interest taxpayers claimed corresponded to a qualified residence and was eligible for the deduction. For example, IRS could see if an address reported on Form 1098 matched the address that the taxpayers listed on their Form 1040.[Footnote 13] Tax preparers told us that requiring the property address also would help them prepare returns more accurately because they could use the information to determine if the property securing the debt is the taxpayer's qualified home. Some IRS officials responsible for examination policy whom we interviewed agreed the property address would be useful in selecting returns for examination and in cases where taxpayers have more than one home. We previously recommended that IRS revise Form 1098 to collect the address of the property whose mortgage is reported on Form 1098. [Footnote 14] In response, IRS agreed to consider implementing our recommendation, citing the burden the requirement could place on third parties. Representatives of the mortgage banking industry told us that it would be feasible to report property address information on Form 1098 because mortgage lenders already maintain this information. We also found an example of a lender that provided address information with Form 1098 information that it sent to borrowers. Adding Mortgage Balance Information to the Form 1098: If the beginning and ending mortgage debt balances or average annual debt balances were provided on Form 1098, IRS could electronically identify taxpayers with more than $1 million in mortgage debt (see figure 2, box 7) or identify taxpayers whose mortgage interest deductions appeared out of proportion to their debt amounts. IRS officials said that debt balance information would be particularly helpful in selecting returns for examination if combined with the address information because the additional information would help disentangle mortgage interest deductions of taxpayers with multiple homes. According to a mortgage industry representative, Form 1098 filers could provide debt balance information. We found examples of companies that provided information on mortgage balances on the Form 1098 statements they sent to borrowers. However, industry representatives further stated that companies do not necessarily have accurate balance information at a particular point in time and that average balances would be easier for the industry to report than balances pegged to any specific day, especially for home equity lines of credit. For companies that do not keep average balance information now, they would incur the associated set-up costs of collecting the information in the future. Adding a Check Box to the Form 1098 to Indicate Refinancing: Form 1098 could be redesigned with a check box for filers to show whether a mortgage was refinanced (see figure 2, box 8). This would help IRS identify taxpayers who might be noncompliant with rules specific to refinancing, such as the rule to amortize points paid on the mortgage. A mortgage industry representative said that the burden of a refinancing check box could be reduced if the rule were only to complete the box in the year that the loan was made, eliminating the need for a Form 1098 filer to track over time whether a refinancing had occurred. Also, Form 1098 filers do not uniformly keep any indication in their records as to whether loans are acquisition loans or refinancing. However, because the deductibility of the interest on the cash taken out in a refinancing depends in part on how the money is spent, identifying refinancing would not necessarily eliminate the need for IRS to examine returns. It also would be helpful to IRS if Form 1098 filers could show whether cash was taken out in a refinancing, but such a change might not be able to include the amount of cash. The mortgage industry representative said that any information on the cash taken out would be unreliable and burdensome for Form 1098 filers to get. Form 1098 filers have estimates on cash taken out, not the final figures available to the closing agent. The industry representative also said that Form 1098 filers do not maintain information on whether loan proceeds exceeded the amount of the loan that was refinanced.[Footnote 15] However, IRS examiners could more easily see if cash were taken out at refinancing and investigate whether rules on refinancing were followed if Form 1098 had address, mortgage debt, and a refinancing check box, because they could see whether increases in the loan balances over time for a particular home took place. Adding an Indicator to the Form 1098 to Distinguish between Home Equity and Acquisition Debt: Tax forms do not require taxpayers or Form 1098 filers to report the types of debt on which mortgage interest deductions are based, even though the mortgage debt limitations for deducting interest are different for acquisition and home equity debt. If Form 1098 filers were required to identify the type of debt associated with the deduction (see figure 2, box 9), IRS might more easily discern whether a taxpayer's deduction exceeded the acquisition or home equity debt limit, especially if combined with reports of the debt amounts. Having Form 1098 filers identify debt types, however, would create challenges. A mortgage industry representative told us that Form 1098 filers do not always have information about debt types, especially if the loans have been sold and re-sold or the original loan has been refinanced. Form 1098 filers do not know whether mortgage refinancing proceeds were used for home improvements, which could qualify as acquisition debt or home equity debt, depending on how the money was used. The representative also said that if a home-equity debt reporting requirement were instituted, the industry would have to introduce costly and burdensome systems. Private Sector Data Might Be Useful to IRS in Detecting Mortgage Interest Noncompliance: IRS already contracts with a private firm to obtain information about taxpayers for routine examination purposes. Other private sector data might also be useful to IRS in detecting mortgage interest noncompliance. For example, we obtained information from SMR Research, one of several companies that analyze loan information that IRS might find useful in determining taxpayer compliance with rules governing the deduction of interest on home equity loans. By comparing homeowners' current mortgage debt for a particular property with earlier debt, SMR Research estimated that up to several million homeowners had loans that might have exceeded the home equity debt limitation. In the aggregate, these homeowners' debts over the debt limitation were several hundred billion dollars.[Footnote 16] SMR Research's data do not show whether taxpayers correctly reported deductions based on home equity loans. IRS would still have to check the returns. Given information underlying the transactions in SMR Research's database and shown in appendix VII, IRS could test the use of this or similar private-sector databases to: * pinpoint taxpayers for examination, * initiate correspondence to taxpayers, or: * conduct outreach to paid preparers. Some IRS CIP examiners told us they had significant findings related to home equity debt, indicating that follow-up might be productive. IRS's Guidance to Taxpayers and Training Materials for Examiners Did Not Indicate the Range of Possible Mortgage Interest Deduction Situations: Taken as a whole, IRS taxpayer guidance--Schedule A and its instructions, Publication 17, Your Federal Income Tax, and Publication 936, Home Mortgage Interest Deduction--generally informed taxpayers that mortgage interest deductions are subject to limits. Even though the guidance was generally sufficient, Schedule A does not explicitly mention the limitations. As shown in appendix II, Schedule A, line 10 asks taxpayers for the "Home mortgage interest and points reported to you on Form 1098." Sometimes, as when an acquisition loan exceeds $1 million, this wording could be problematic because a taxpayer or preparer would need to look beyond the Form 1098 to determine the proper line 10 amount.[Footnote 17] Using the same amount of space on Schedule A as now, line 10 could easily be revised to mention limitations by deleting "reported to you" and using wording such as "Unless limited, home mortgage interest and points on Form 1098."[Footnote 18] Another way to revise Schedule A would be to add something about mortgage interest deduction limitations to the margin near line 10. Other possibilities for changing Schedule A to highlight limitations also exist, but would entail tradeoffs. For example, a check box could be added to the Schedule asking taxpayers if either the $1 million or the $100,000 limitation applies to their loans. Making such a change not only would help taxpayers preparing their own returns but also might prompt paid preparers to more thoroughly interview their clients. However, with the mortgage interest deduction, these types of changes would add some burden to taxpayers whose deductions are not affected by the debt limits. The effectiveness of such change is also unclear. Taxpayers might not read IRS guidance when preparing returns. To counter this possibility but also be mindful of preparation, mailing, and taxpayer service costs involved, IRS could test whether corresponding with some taxpayers would cost-effectively reduce misreporting. Added outreach--through seminars or communications with stakeholders such as paid preparers, tax return software providers, and industry groups--could communicate key rules and common mistakes and be targeted to taxpayers, such as those reporting the deduction above a certain level. Sending correspondence directly to taxpayers also could help inform paid preparers of rules and common mistakes because the taxpayers would likely share it with their preparers, according to representatives of the tax return preparation industry.[Footnote 19] IRS has used outreach programs on mortgage interest deductions in recent years. For instance, IRS's tax tips covered deducting refinancing costs. IRS's guidance on the acquisition debt limit is inconsistent with a prior tax court ruling. Currently, Publication 936 guides the taxpayer to treat any qualified mortgage debt above the $1 million acquisition debt limit as home equity debt subject to the $100,000 limit, regardless of the loan's use. In practice, this means that the taxpayer can deduct the interest on up to $1.1 million in acquisition debt (treating the amount above $1 million as home equity debt even if used to acquire the home). However, a 1997 U.S. Tax Court case ruled that the acquisition debt limit is $1 million and no additional deduction can be taken above the $1 million limit unless the taxpayer truly has home equity debt.[Footnote 20] One tax software company official cited the 1997 case as support for the company's decision to change the guidance in its software to allow interest deductions only up to $1 million in acquisition debt, rather than follow the guidance in Publication 936. In 2008, IRS Chief Counsel began reviewing the inconsistency in the debt limit allowances. However, a completion date for this work is uncertain. Until IRS completes its determination, taxpayers will calculate their deductions using different debt limits depending on whether they, their tax preparers, or their tax software follow the guidance in Publication 936 or the tax court interpretation. Depending on IRS's determination, some taxpayers may not be optimizing their deduction or IRS could be losing revenue from taxpayers over deducting. IRS Examiner Guidance and Training Materials Addressed Mortgage Interest Deduction Limitations but Lacked Instructive Examples of Some Relevant Situations: IRS's examiners' guidance and training materials included information for identifying and calculating home-equity and the acquisition-debt limitations. Overall, examiners we interviewed were satisfied with training and guidance on the mortgage interest deduction. However, IRS's guidance and training materials did not cover problems that may arise in some real-life situations. For example, the guidance does not mention a problem that an examiner said she commonly found: taxpayers owning a third home, waiting to sell one of their first two, and deducting the interest paid on all three. Neither does the guidance address situations that tax professionals described as challenging for IRS to enforce, such as multiple refinancings, mortgages, or home equity lines of credit.[Footnote 21] The absence of such examples could impede examiners who do not deal with mortgage interest deduction issues regularly. An IRS training official told us that updating training materials with relevant examples could be done easily. Also, IRS's examiner training materials reflect the taxpayer guidance with regard to the allowable acquisition debt limit. For example, according to one IRS training exercise, taxpayers could deduct interest on up to $1.1 million in mortgage debt even if they used the $1.1 million just to acquire a home, contradicting the 1997 tax court ruling, discussed above. The difference may lead examiners to inconsistently calculate the tax owed, because some may calculate the acquisition debt limit as $1 million based on the tax court determination and others may allow deductions on up to $1.1 million as instructed by the training. As previously mentioned, IRS Chief Counsel's office is reviewing the inconsistency between the IRS guidance and the tax court ruling. IRS officials said that they have procedures to ensure that once the review is complete, examiners will be told of any clarification. Tax Software Differed in the Ease of Navigating Mortgage Interest Deduction Instructions: The three companies' tax preparation software for individuals that we analyzed differed from each other in how they treated the limitations on the amount of debt for which interest can be deducted.[Footnote 22] These software packages were among the most widely used by individuals.[Footnote 23] One company's initial software screen containing mortgage interest deduction instructions mentioned possible deduction limitations, sending users to Publication 936 for calculations. One of the other companies made changes to its mortgage interest deduction displays in its 2008 version after considering our input on its 2007 version. These changes give more prominence to the deduction limitations and how Publication 936 may be used for calculations. The third company's on-line software had information on the limitations, but users would not find this information unless they looked under "frequently asked questions." This software also led the user to take amounts directly off Form 1098. For 2008, the software added a question to its frequently asked questions area about what taxpayers should do if interest and points exceeded the amount they could deduct. If taxpayers were to click on this question, they would find information about the limitations and the calculation needed. IRS does not know how frequently software packages are associated with misreporting. Such information could help IRS in its compliance efforts. We recently recommended that IRS require software companies to include a software identification number that specifically identifies the software package used to prepare tax returns. IRS plans to implement a software identification system in October 2009.[Footnote 24] Conclusions: Because IRS has little information on taxpayers' mortgage debts, it cannot easily detect when taxpayers are not complying with the deduction limits. Instead, to ensure taxpayer compliance with the statutory requirements, IRS must conduct examinations that are more costly to use than other enforcement methods, such as a matching program. The absence of specific information about the reasons for noncompliance also prevents IRS from making efficient decisions on how to select cases for examination. Simultaneously, the complexity of mortgage interest deduction rules creates problems with some taxpayers trying to comply. These problems are compounded by an inconsistency between IRS guidance and a tax court ruling on acquisition debt limits. Because the limits causing complexity are written in law, statutory changes may be the most direct way to address the challenges of administering the deduction and alleviate the compliance problems. However, consideration of statutory changes was beyond the scope of this report. Nonetheless, within the current statutory framework, opportunities exist for IRS to mitigate some of the problems we identified. Recommendations for Executive Action: We are making seven recommendations. Specifically, we recommend that the Commissioner of Internal Revenue: * revise NRP's case selection system so that a tax return's mortgage interest deduction is not automatically excluded as an examination issue if it matches information reported on Form 1098; * revise Form 1098 to require third parties to provide information on mortgage balances, the address of a home securing a mortgage, and an indicator of whether the mortgage is for a current year refinancing; * investigate whether using information from private sources would be productive in detecting mortgage interest noncompliance, especially for home equity debt; * revise the wording on Schedule A to clearly state that the mortgage interest deduction is subject to limitations; * conduct a test to evaluate whether mortgage interest deduction- related outreach programs to taxpayers and tax return preparers could be a cost-effective way to reduce noncompliance; outreach might include sending correspondence covering key rules and common mistakes or promoting seminars on common types of misreporting; * set a date to complete the Chief Counsel determination on whether the acquisition debt limit is $1 million or $1.1 million when used in combination with the home equity debt limit; and: * revise examiner training materials by adding examples cited as common problems by auditors and paid tax return preparers, such as those involving multiple homes or home-based businesses, and after the Chief Counsel's final determination on the acquisition limit, revise examiner training and the worksheet in guidance to reflect the project's outcome. Agency Comments and Our Evaluation: We received written comments from the Internal Revenue Service on July 23, 2009 (for the full text of the comments, see appendix VIII). IRS agreed with five of our recommendations and agreed to study the other two. It acknowledged that without information about taxpayers' mortgage debts, it cannot easily detect taxpayer noncompliance with the mortgage interest deduction limits. IRS also said that the absence of information about noncompliance prevents it from efficiently deciding how to select cases for review and that the complexity of the rules causes problems for some taxpayers. Regarding our recommendation to revise Form 1098 to include more information on taxpayer mortgages, IRS agreed to study the issue, saying it does not have enough data to support revisions at this time. Because IRS acknowledged in its comments that it does not have information about taxpayers' mortgage debts to easily detect noncompliance, we believe that our recommended revisions to Form 1098 would be cost-effective ways to provide IRS with additional useful information to help it detect noncompliance. Concerning our recommendation to conduct a test to evaluate whether mortgage interest deduction-related outreach programs could be a cost- effective way to reduce noncompliance, IRS said it will study the feasibility of such a test. As agreed with your office, unless you publicly announce its contents earlier, we plan no further distribution of this report until 30 days after its date. At that time, we will send copies to the Secretary of the Treasury, the Commissioner of Internal Revenue, and other interested parties. This report will also be available at no charge on GAO's Web site at [hyperlink, http://www.gao.gov]. For further information regarding this report, please contact me at (202) 512-9110 or at whitej@gao.gov. Contacts for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. Individuals making key contributions to this report may be found in appendix IX. Signed by: James R. White: Director, Tax Issues, Strategic Issues Team: [End of section] Appendix I: Scope and Methodology: To provide information on how IRS detects taxpayers' noncompliance with the home mortgage interest deduction rules and what IRS knows about the extent of noncompliance and to identify problems, if any, that taxpayers face in attempting to comply with the deduction and describe IRS's challenges in detecting mortgage interest deduction noncompliance, we: * reviewed IRS documents and interviewed agency officials to determine how IRS checks taxpayers' compliance with deduction rules and if there were issues affecting its ability to detect noncompliance. This included reviewing information on IRS's compliance initiative projects that focused on the mortgage interest deduction and interviewing IRS examiners from four widely dispersed parts of the country who were involved in the projects. The examiners were selected by IRS officials for their mortgage interest deduction expertise; * obtained information from IRS's 2001 National Research Program (NRP) compliance study of individual taxpayers and data for 2006 through 2008 from its Examination Operational Automation Database (EOAD) and Audit Information Management System (AIMS) to determine examination results about individual taxpayers' compliance with mortgage interest deduction rules; and: * conducted a literature search of relevant tax publications and interviewed non-government experts to enhance our understanding of mortgage interest deduction compliance issues affecting taxpayers and IRS. To assess options to give IRS more information to enforce compliance with mortgage interest deduction rules, we surveyed literature and interviewed industry representatives about third-party information reporting options that would give IRS useful information to improve its ability to enforce mortgage interest deduction rules while minimizing burden on third parties. We discussed the options, including their pros and cons, with industry representatives from organizations such as the Mortgage Bankers Association and the American Institute of Certified Public Accountants and with others such as scholars who have expertise in this area. We also met with IRS officials to obtain their views on the feasibility of the options to improve compliance. Consideration of statutory changes was beyond the scope of the request. To determine if IRS could benefit from using private sector information to detect noncompliance with home equity rules, we obtained statistical information about home equity loans, home equity lines of credit, and cash-out refinancings that have exceeded $100,000 from SMR Research, one of several companies that compile and analyze mortgage data. According to its Web site, SMR Research is the nation's largest publisher of industry and market research studies on the home mortgage business, home equity lending, and other consumer loan subjects. We did not study the information of any of the other company with extensive mortgage data because our purpose was only to gather data on the type of information that might be available from private sector sources that potentially could benefit IRS in detecting home equity noncompliance. To assess the reliability of the SMR Research data, we discussed the controls over the database with a responsible company official and reviewed documentation of the controls used to gather the data. We also used information from the American Housing Survey for the United States: 2007, prepared by the U.S. Department of Housing and Urban Development and the U.S. Census Bureau, to confirm the reasonableness of SMR Research information. We found the SMR database reliable for our purposes. To find whether IRS's guidance to taxpayers provided enough information to properly calculate the mortgage interest deduction, we reviewed the guidance to determine whether it mentioned the limitations prominently, i.e., whether the limitations were mentioned directly as opposed to the user being directed to another place. To determine whether IRS's examiners' training and guidance provided enough information on properly calculating the mortgage interest deduction, we compared examiner training and guidance documents with IRS publications and instructions on mortgage interest deductions, as well as the applicable tax laws. To describe how tax return preparation software programs and IRS's Free File program handle the mortgage interest deduction, we checked how the three software packages used by 88 percent of individuals filing electronically treated the debt limitations. We did this by logging on to the software and reviewing the screens pertaining to the home mortgage interest deduction. We also discussed our findings with company officials. We reviewed tax year 2007 versions of each of the software packages and included the versions used for IRS's Free File program. Our results are limited to the versions of the software packages we reviewed and cannot be generalized to others. Free File offers free federal tax return preparation and electronic filing services through these companies and others. In addition, we reviewed documentation showing how software widely used by paid tax return preparers for millions of taxpayers throughout the country handles the mortgage interest deduction. In addition, to respond to your request, we analyzed data from IRS's Statistics of Income (SOI) Division and the Compliance Data Warehouse (CDW) to compile information about year-to-year changes in taxpayers' mortgage interest deductions and Form 1098 filings for tax years 2004 through 2006.[Footnote 25] Our analyses focused on the dollar amounts listed on Schedule A, line 10; whether mortgage interest deductions on Schedule A, line 10 matched Form 1098 amounts; frequency of paid preparer use by those claiming a Schedule A, line 10 deduction; loan account numbers; and amounts of mortgage interest deduction taken on forms other than Schedule A to which we had access, including Schedule C for sole proprietorship businesses, Schedule E for supplemental income and losses, Schedule F for farms, and Form 4835 for farm rental income and expenses.[Footnote 26] For part of our analysis, we used the SOI database to identify a panel of taxpayers from tax year 2004 and then collected tax information about them from the CDW database for tax years 2004 through 2006. These analyses are in appendix V, and margins of error for these tables are provided in appendix VI. We found the IRS databases we used--NRP, EOAD, AIMS, SOI, and CDW-- reliable for the purposes of this report. NRP, AIMS, SOI and CDW are also data sources that we have tested for previous engagements. For EOAD, we reviewed documentation on database procedures, interviewed IRS officials, ran electronic checks, and compared output to other available information for reasonableness purposes. We conducted this performance audit from May 2008 through July 2009 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. [End of section] Appendix II: Schedule A, Itemized Deductions: Figure 3: Internal Revenue Service Schedule A on Which the Home Mortgage Interest Deduction Is Taken: [Refer to PDF for image: illustration] The following are highlighted on Schedule A: Mortgage interest deductions: 10. Home mortgage interest and points reported to you on Form 1098: 11. Home mortgage interest not reported to you on Form 1098. If paid to the person from whom you bought the home, see page A-6 and show the person's name, identifying number, and address: 12. Points not reported to you on Form 1098. See page A-6 for special rules: 13. Qualified mortgage insurance premiums (see page A-6). Source: IRS. [End of figure] [End of section] Appendix III: Examination Results Involving Mortgage Interest Deductions: The following table compares examination results of IRS's compliance initiative projects (CIP) on the mortgage interest deduction with results from its routine examinations (non-CIP). Table 4: Results of IRS's Fiscal Year 2008 Examinations of Schedule A, Line 10 for CIP and non-CIP Examinations: Examination result: Type of change to the line 10 deduction: Decrease to the line 10 deduction; First CIP: Number of examinations: 1,012; Second CIP: Number of examinations: 6,982; Non-CIP: Number of examinations: 11,971. Examination result: Type of change to the line 10 deduction: Increase to the line 10 deduction; First CIP: Number of examinations: 27; Second CIP: Number of examinations: 134; Non-CIP: Number of examinations: 14,279. Examination result: Type of change to the line 10 deduction: No line 10 change; First CIP: Number of examinations: 119; Second CIP: Number of examinations: 890; Non-CIP: Number of examinations: 6,306. Examination result: Type of change to the line 10 deduction: Total; First CIP: Number of examinations: 1,158; Second CIP: Number of examinations: 8,006; Non-CIP: Number of examinations: 32,556. Examination result: Type of change to the tax amount: Increase to the tax amount; First CIP: Number of examinations: 994; Second CIP: Number of examinations: 6,648; Non-CIP: Number of examinations: 26,625. Examination result: Type of change to the tax amount: Decrease to the tax amount; First CIP: Number of examinations: 42; Second CIP: Number of examinations: 233; Non-CIP: Number of examinations: 2,748. Examination result: Type of change to the tax amount: No tax change; First CIP: Number of examinations: 122; Second CIP: Number of examinations: 1,125; Non-CIP: Number of examinations: 3,183. Examination result: Type of change to the tax amount: Total; First CIP: Number of examinations: 1,158; Second CIP: Number of examinations: 8,006; Non-CIP: Number of examinations: 32,556. Examination result: Median change to the line 10 deduction or to the tax amount: Median decrease to the line 10 deduction; First CIP: Number of examinations: $29,359; Second CIP: Number of examinations: $27,940; Non-CIP: Number of examinations: $6,430. Examination result: Median change to the line 10 deduction or to the tax amount: Median increase to the tax amount; First CIP: Dollar amount of the median change: $7,673; Second CIP: Dollar amount of the median change: $8,012; Non-CIP: Dollar amount of the median change: $4,612. Source: GAO analysis of IRS data. Note: This table includes only examinations done by revenue agents or tax compliance officers. [End of table] [End of section] Appendix IV: Examples Illustrating Mortgage Interest Deduction's Complexity: As shown in the examples in this appendix, to determine the amounts of mortgage interest that is deductible on Schedule A, a taxpayer or his or her tax return preparer might need to follow many steps. First, the taxpayer must find all his or her records pertaining to the property's acquisition and determine if the October 13, 1987, cutoff date applies, then locate records of refinancing arrangements affecting acquisition debt, and calculate the appropriate average annual balance. Second, the taxpayer must determine home equity indebtedness, properly tracking, for instance, the use of proceeds from home equity lines of credit, and determining average balances and home equity amounts not deductible for AMT purposes. Third, the taxpayer must know when loan proceeds finance business use of the home, a partly rented home, or personal, business, investment, or municipal bond expenditures because how the loan proceeds are used determines the deductibility of the loan interest on the tax return. The American Institute of Certified Public Accountants gave us details of the following scenario in which married homeowners refinanced their house: * In 1990, a married couple borrowed $150,000 to buy a house. * By January 1, 2008, they had made principal payments of $50,000, leaving an acquisition debt of $100,000. * On January 1, 2008, their house's fair market value was $500,000, and they refinanced their original loan for $400,000. * The new $400,000 mortgage would be allocated as follows: - The first $100,000 would be considered home acquisition debt, and the interest on it would be deductible for the regular tax and for the AMT. - The next $100,000 would be considered home equity debt, equal to the home equity limitation and, therefore, deductible for the regular tax but not for AMT unless it was used to acquire, construct, or improve a main or second home. - The last $200,000 would be subject to "interest tracing" rules, with deductibility depending on how the money was used--for instance, if used for personal or municipal bond purposes, the interest would not be deductible. However, if used for business or investment purposes, the interest may be deductible. The example would have been more complicated if not for the following simplifying assumptions that we made: * The couple owned only one house and never borrowed money to improve it. * They bought it after the October 13, 1987, the grandfather date when rules changed regarding the amount of acquisition debt for which interest is deductible. * This was the first time they refinanced the house. * The refinancing was done on the first day of the year and $1,000 in principal was repaid each month, making monthly mortgage balances easily calculated throughout the year.[Footnote 27] * They had no home equity loans other than a home equity part of the refinancing and had not used a home equity line of credit, the use of whose proceeds they would have had to have tracked. * They never used the house for business or rented any part of it. We used the facts and assumptions of this example to complete the 13 steps in table 5, which is patterned after a worksheet in Publication 936. Table 5 tracks the path taxpayers would follow in figuring their deductible home mortgage interest. Table 5: IRS Worksheet to Figure Home Mortgage Interest: Part I--Qualified loan limit: Number: 1; Computation action: Enter average balance of grandfathered debt; Dollar amount: $0; Comments: The simplifying assumption was that the house was not bought before October 14, 1987. Number: 2; Computation action: Enter average balance for all your home acquisition debt; Dollar amount: $100,000; Comments: The taxpayers had to find records of their original debt and how much principal they had repaid at the time of the refinancing; if there had been second and third mortgages, previous refinancings, additions to the home, or seller-financed mortgages with no financial institution involved, they would have to be considered; under average balance rules for refinancings that include home acquisition and home equity components, principal payments for the refinancing had not yet reduced the $100,000 home acquisition debt. Number: 3; Computation action: Enter $1 million ($500,000 if married filing separately); Dollar amount: $1,000,000; Comments: The taxpayers were filing a joint return. Number: 4; Computation action: Enter the larger of #1 or #3; Dollar amount: $1,000,000; Comments: [Empty]. Number: 5; Computation action: Add #1 and #2; Dollar amount: $100,000; Comments: [Empty]. Number: 6; Computation action: Enter the smaller of #4 or #5; Dollar amount: $100,000; Comments: [Empty]. Number: 7; Computation action: Enter the smaller of $100,000 ($50,000 if married filing separately) or homes' fair market value minus acquisition debt minus grandfathered debt ($500,000 - $100,000 - 0 = $400,000); Dollar amount: $100,000; Comments: The taxpayers were filing a joint return; they would determine the fair market value and retain it for use in any future years' calculations. Number: 8; Computation action: Add #6 + #7 = qualified loan limit; Dollar amount: $200,000; Comments: [Empty]. Part II--Deductible home mortgage interest: Number: 9; Computation action: Enter total of average balances of all mortgages on all qualified homes ($400,000 + $388,000)/2 = $394,000; continue with this table if #8 is less than #9; Dollar amount: $394,000; Comments: Each month, the same amount of the principal--$1,000--was repaid for the only house involved; this calculation would be more complicated if the refinancing had occurred after January 1, or if another mortgage also were incurred, especially since it likely would have begun at a different time of the year. Number: 10; Computation action: Enter interest you paid, including the amount from Form 1098; Dollar amount: $20,000; Comments: The amount from Form 1098 is the amount that some taxpayers might enter directly onto Schedule A. Number: 11; Computation action: Divide #8 by #9, rounding to 3 places; Dollar amount: %0.508; Comments: [Empty]. Number: 12; Computation action: Multiply #10 by #11 and enter result here and on Schedule A; Dollar amount: $10,160; Comments: This is the home mortgage interest deduction for now, but where home equity debt is involved, there could be a home mortgage interest adjustment that would have to be calculated for AMT purposes. Number: 13; Computation action: Subtract #12 from #10; Dollar amount: $9,840; Comments: This is the portion of your home mortgage interest that is not deductible on Schedule A. Other rules might have to be consulted on the deductibility of this interest, depending on whether the applicable loan proceeds were used, for instance, for a trade or business, an investment, a municipal bond, personal purposes, or a combination of these; if the taxpayer used his home for business or rented part of it out, mortgage interest could appear in different parts of the tax return. Source: Department of the Treasury, IRS, Publication 936: Home Mortgage Interest Deduction: For Use in Preparing 2008 Returns; GAO analysis. [End of table] A professional tax preparer also gave us the following example to illustrate the complexity of the mortgage interest deduction. This example may be atypical but reflects a more complicated case than the previous example. The client had more than one home, a mortgage greater than $1 million, a refinancing, and home equity debt. In this case, the practitioner said that it took over 4 hours of tracing transactions to calculate the acquisition and home equity debt for one house, using an Excel spreadsheet to enter amounts from closing statements for each of multiple refinancings. Home equity debt had to be tracked separately because interest on allowable home equity debt is generally not deductible for AMT purposes unless the proceeds are used to buy, build, or improve a home. Acquisition debt after refinancing had to only include the acquisition debt part of the previous loan and closing costs. A deductible interest spreadsheet had to consider beginning and ending loan balances, how loan proceeds were originally used, and total interest paid. [End of section] Appendix V: Descriptive Data on Home Mortgage Interest Deduction Claims and Form 1098 Filings: This appendix presents analyses of Statistics of Income (SOI) Division and Compliance Data Warehouse (CDW) data focusing on Schedule A, line 10 deductions and Form 1098 filings. Tables 6 through 12 use a SOI- derived panel of taxpayers from 2004 through 2006, combined with CDW data on Form 1098 filings, and table 13 uses SOI data exclusively. Table 6 shows that for returns reporting a Schedule A, line 10 deduction for all three years of the study period, about half had mortgage interest from one Form 1098. For example, in 2006, 42.6 percent of taxpayers reporting a mortgage interest deduction had one Form 1098, totaling about $126.5 billion in Schedule A, line 10 deductions. Table 6 also shows that some taxpayers reported a mortgage interest deduction on Schedule A, line 10. This might mean that the taxpayer improperly filled out line 10 on Schedule A or that IRS had not received a corresponding Form 1098 for that particular taxpayer. The number of Form 1098 filings does not reflect the number of homes a taxpayer has because, for example, taxpayers would receive multiple Form 1098 filings in a given year if they sold a home or if their existing mortgage was sold in the secondary market. The margins of error for the tables in this appendix appear in appendix VI. Table 6: Number and Dollar Total of Schedule A, Line 10 Deductions per Number of Form 1098 Filings, 2004 through 2006: 2004: Number of Form 1098 filings: 0; Number of tax returns (millions): 1.5; Percentage: 4.1; Line 10 amounts for tax returns (billions of dollars): $9.3; Percentage: 3.2. Number of Form 1098 filings: 1; Number of tax returns (millions): 16.8; Percentage: 46.1; Line 10 amounts for tax returns (billions of dollars): $114.3; Percentage: 39.3. Number of Form 1098 filings: 2; Number of tax returns (millions): 11.0; Percentage: 30.3; Line 10 amounts for tax returns (billions of dollars): $97.5; Percentage: 33.5. Number of Form 1098 filings: 3; Number of tax returns (millions): 4.2; Percentage: 11.5; Line 10 amounts for tax returns (billions of dollars): $42.2; Percentage: 14.5. Number of Form 1098 filings: 4 or more; Number of tax returns (millions): 2.9; Percentage: 7.9; Line 10 amounts for tax returns (billions of dollars): $27.6; Percentage: 9.5. Number of Form 1098 filings: Total; Number of tax returns (millions): 36.5; Percentage: 100.0; Line 10 amounts for tax returns (billions of dollars): $290.9; Percentage: 100.0. 2005: Number of Form 1098 filings: 0; Number of tax returns (millions): 1.4; Percentage: 3.7; Line 10 amounts for tax returns (billions of dollars): $8.9; Percentage: 2.7. Number of Form 1098 filings: 1; Number of tax returns (millions): 16.2; Percentage: 43.4; Line 10 amounts for tax returns (billions of dollars): $114.0; Percentage: 34.9. Number of Form 1098 filings: 2; Number of tax returns (millions): 11.7; Percentage: 31.3; Line 10 amounts for tax returns (billions of dollars): $111.0; Percentage: 34.0. Number of Form 1098 filings: 3; Number of tax returns (millions): 4.6; Percentage: 12.4; Line 10 amounts for tax returns (billions of dollars): $52.2; Percentage: 16.0. Number of Form 1098 filings: 4 or more; Number of tax returns (millions): 3.5; Percentage: 9.3; Line 10 amounts for tax returns (billions of dollars): $40.1; Percentage: 12.3. Number of Form 1098 filings: Total; Number of tax returns (millions): 37.3; Percentage: 100.0; Line 10 amounts for tax returns (billions of dollars): $326.2; Percentage: 100.0. 2006: Number of Form 1098 filings: 0; Number of tax returns (millions): 1.4; Percentage: 3.7; Line 10 amounts for tax returns (billions of dollars): $10.3; Percentage: 2.7. Number of Form 1098 filings: 1; Number of tax returns (millions): 16.4; Percentage: 42.6; Line 10 amounts for tax returns (billions of dollars): $126.5; Percentage: 33.7. Number of Form 1098 filings: 2; Number of tax returns (millions): 12.6; Percentage: 32.6; Line 10 amounts for tax returns (billions of dollars): $135.7; Percentage: 36.1. Number of Form 1098 filings: 3; Number of tax returns (millions): 4.7; Percentage: 12.1; Line 10 amounts for tax returns (billions of dollars): $58.9; Percentage: 15.7. Number of Form 1098 filings: 4 or more; Number of tax returns (millions): 3.4; Percentage: 8.9; Line 10 amounts for tax returns (billions of dollars): $44.2; Percentage: 11.8. Number of Form 1098 filings: Total; Number of tax returns (millions): 38.5; Percentage: 100.0; Line 10 amounts for tax returns (billions of dollars): $375.5; Percentage: 100.0. Source: GAO analysis of IRS data. Note: Percentages may not add to 100 percent because of rounding. [End of table] Table 7 breaks down the pattern of Schedule A, line 10, deduction amounts compared with the number of Form 1098 filings. For example, in 2006, 75 percent of taxpayers with two Form 1098 filings had deduction amounts of $14,325 or less. Table 7: Distribution of Schedule A, Line 10 Deductions in Select Percentiles per Number of Form 1098 Filings, 2004 through 2006: 2004: Number of Form 1098 filings: 0; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 5th: $184; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 10th: $435; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 25th: $2,262; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 50th: $5,248; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 75th: $8,570. Number of Form 1098 filings: 1; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 5th: $1,265; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 10th: $2,158; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 25th: $3,905; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 50th: $6,093; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 75th: $8,839. Number of Form 1098 filings: 2; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 5th: $2,705; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 10th: $3,678; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 25th: $5,568; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 50th: $8,121; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 75th: $11,959. Number of Form 1098 filings: 3; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 5th: $3,494; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 10th: $4,598; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 25th: $6,788; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 50th: $9,986; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 75th: $14,978. Number of Form 1098 filings: 4 or more; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 5th: $4,378; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 10th: $5,883; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 25th: $8,405; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 50th: $12,520; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 75th: $19,014. 2005: Number of Form 1098 filings: 0; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 5th: $243; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 10th: $505; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 25th: $2,271; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 50th: $5,140; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 75th: $8,691. Number of Form 1098 filings: 1; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 5th: $1,342; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 10th: $2,217; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 25th: $3,994; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 50th: $6,211; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 75th: $9,041. Number of Form 1098 filings: 2; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 5th: $2,900; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 10th: $3,917; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 25th: $5,908; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 50th: $8,666; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 75th: $12,648. Number of Form 1098 filings: 3; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 5th: $4,287; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 10th: $5,543; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 25th: $7,743; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 50th: $11,229; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 75th: $16,682. Number of Form 1098 filings: 4 or more; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 5th: $5,704; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 10th: $6,875; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 25th: $9,819; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 50th: $14,630; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 75th: $22,597. 2006: Number of Form 1098 filings: 0; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 5th: $249; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 10th: $490; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 25th: $2,638; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 50th: $5,692; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 75th: $9,555. Number of Form 1098 filings: 1; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 5th: $1,435; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 10th: $2,379; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 25th: $4,229; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 50th: $6,657; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 75th: $9,860. Number of Form 1098 filings: 2; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 5th: $3,170; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 10th: $4,366; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 25th: $6,475; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 50th: $9,524; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 75th: $14,325. Number of Form 1098 filings: 3; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 5th: $4,589; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 10th: $5,825; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 25th: $8,298; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 50th: $12,372; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 75th: $18,957. Number of Form 1098 filings: 4 or more; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 5th: $5,659; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 10th: $7,533; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 25th: $10,766; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 50th: $16,654; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 75th: $26,085. Source: GAO analysis of IRS data. [End of table] Table 8 shows the results of comparing Schedule A, line 10, deduction amounts on tax returns to amounts reported to IRS by third parties on Form 1098. As the table shows, about 65 percent to 67 percent of the comparisons resulted in a match in all 3 years. However, in some situations, our analysis showed that taxpayers' deductions also were greater or less than the amounts on Forms 1098. Over the three-year period, 8.2 percent to 11.1 percent of tax returns' deductions exceeded the Form 1098 amount. This can occur legitimately when, for example, a property is co-owed and one of the owners claiming the deduction does not receive a Form 1098. It can also occur because of taxpayer or Form 1098 filing errors. Deductions can be less than the Form 1098 amounts when taxpayers' debts have exceeded the deduction limits or by error. Table 8: Results of Comparing Total Schedule A, Line 10 Deduction Amounts to Amounts Reported on Form 1098, 2004 through 2006: 2004: Results of comparison: Match; Total number of tax returns (millions): 23.8; Percentage: 65.3; Schedule A, line 10 amount (billions of dollars): $198.5; Difference between Form 1098 amounts and line 10 amounts (millions of dollars): [Empty]. Results of comparison: Deduction exceeds Form 1098 amount; Total number of tax returns (millions): 4.1; Percentage: 11.1; Schedule A, line 10 amount (billions of dollars): $41.6; Difference between Form 1098 amounts and line 10 amounts (millions of dollars): $9,229.0. Results of comparison: Deduction less than Form 1098 amount; Total number of tax returns (millions): 7.1; Percentage: 19.4; Schedule A, line 10 amount (billions of dollars): $41.4; Difference between Form 1098 amounts and line 10 amounts (millions of dollars): $14,962.2. Results of comparison: No Form 1098; Total number of tax returns (millions): 1.5; Percentage: 4.1; Schedule A, line 10 amount (billions of dollars): $9.3; Difference between Form 1098 amounts and line 10 amounts (millions of dollars): $9,280.3. Results of comparison: Total; Total number of tax returns (millions): 36.5; Percentage: 100.0; Schedule A, line 10 amount (billions of dollars): $290.9; Difference between Form 1098 amounts and line 10 amounts (millions of dollars): [Empty]. 2005: Results of comparison: Match; Total number of tax returns (millions): 24.6; Percentage: 65.9; Schedule A, line 10 amount (billions of dollars): $224.7; Difference between Form 1098 amounts and line 10 amounts (millions of dollars): [Empty]. Results of comparison: Deduction exceeds Form 1098 amount; Total number of tax returns (millions): 3.5; Percentage: 9.3; Schedule A, line 10 amount (billions of dollars): $38.0; Difference between Form 1098 amounts and line 10 amounts (millions of dollars): $7,452.0. Results of comparison: Deduction less than Form 1098 amount; Total number of tax returns (millions): 7.9; Percentage: 21.1; Schedule A, line 10 amount (billions of dollars): $54.6; Difference between Form 1098 amounts and line 10 amounts (millions of dollars): $21,240.5. Results of comparison: No Form 1098; Total number of tax returns (millions): 1.4; Percentage: 3.7; Schedule A, line 10 amount (billions of dollars): $8.9; Difference between Form 1098 amounts and line 10 amounts (millions of dollars): $8,903.2. Results of comparison: Total; Total number of tax returns (millions): 37.3; Percentage: 100.0; Schedule A, line 10 amount (billions of dollars): $326.2; Difference between Form 1098 amounts and line 10 amounts (millions of dollars): [Empty]. 2006: Results of comparison: Match; Total number of tax returns (millions): 25.8; Percentage: 67.0; Schedule A, line 10 amount (billions of dollars): $263; Difference between Form 1098 amounts and line 10 amounts (millions of dollars): [Empty]. Results of comparison: Deduction exceeds Form 1098 amount; Total number of tax returns (millions): 3.2; Percentage: 8.2; Schedule A, line 10 amount (billions of dollars): $39.2; Difference between Form 1098 amounts and line 10 amounts (millions of dollars): $7,033.0. Results of comparison: Deduction less than Form 1098 amount; Total number of tax returns (millions): 8.1; Percentage: 21.1; Schedule A, line 10 amount (billions of dollars): $63.0; Difference between Form 1098 amounts and line 10 amounts (millions of dollars): $26,641.5. Results of comparison: No Form 1098; Total number of tax returns (millions): 1.4; Percentage: 3.7; Schedule A, line 10 amount (billions of dollars): $10.3; Difference between Form 1098 amounts and line 10 amounts (millions of dollars): $10,300.0. Results of comparison: Total; Total number of tax returns (millions): 38.5; Percentage: 100.0; Schedule A, line 10 amount (billions of dollars): $375.5; Difference between Form 1098 amounts and line 10 amounts (millions of dollars): [Empty]. Source: GAO analysis of IRS data. Note: Percentages may not add to 100 percent because of rounding. [End of table] Table 9 shows both the distribution of Schedule A, line 10 deduction amounts and the distribution of the differences between Schedule A, line 10 and Form 1098 statements filed for taxpayers who reported a Schedule A, line 10 deduction. For example, in 2006, the median (50th percentile) difference between Form 1098 and Schedule A amounts for taxpayers whose deduction exceeded the Form 1098 amount was $727. Table 9: Percentile Distributions for Schedule A, Line 10 Amounts Compared with Form 1098 Amounts and the Difference between Schedule A, Line 10 Amounts and Form 1098 Amounts, 2004 through 2006: 2004: Schedule A, line 10 amount; Output of comparison: Match; Percentiles in dollars: 5th: $1,725; Percentiles in dollars: 10th: $2,778; Percentiles in dollars: 25th: $4,651; Percentiles in dollars: 50th: $7,129; Percentiles in dollars: 75th: $10,544. Schedule A, line 10 amount; Output of comparison: Deduction exceeds Form 1098 amount; Percentiles in dollars: 5th: $2,506; Percentiles in dollars: 10th: $3,501; Percentiles in dollars: 25th: $5,614; Percentiles in dollars: 50th: $8,740; Percentiles in dollars: 75th: $13,105. Schedule A, line 10 amount; Output of comparison: Deduction less than Form 1098 amount; Percentiles in dollars: 5th: $1,874; Percentiles in dollars: 10th: $2,872; Percentiles in dollars: 25th: $4,686; Percentiles in dollars: 50th: $7,523; Percentiles in dollars: 75th: $11,777. Schedule A, line 10 amount; Output of comparison: No Form 1098; Percentiles in dollars: 5th: $184; Percentiles in dollars: 10th: $435; Percentiles in dollars: 25th: $2,262; Percentiles in dollars: 50th: $5,248; Percentiles in dollars: 75th: $8,570. Difference between Schedule A, line 10 and Form 1098 amounts; Output of comparison: Match; Percentiles in dollars: 5th: [Empty]; Percentiles in dollars: 10th: [Empty]; Percentiles in dollars: 25th: [Empty]; Percentiles in dollars: 50th: [Empty]; Percentiles in dollars: 75th: [Empty]. Difference between Schedule A, line 10 and Form 1098 amounts; Output of comparison: Deduction exceeds Form 1098 amount; Percentiles in dollars: 5th: $30; Percentiles in dollars: 10th: $87; Percentiles in dollars: 25th: $292; Percentiles in dollars: 50th: $669; Percentiles in dollars: 75th: $2,817. Difference between Schedule A, line 10 and Form 1098 amounts; Output of comparison: Deduction less than Form 1098 amount; Percentiles in dollars: 5th: $49; Percentiles in dollars: 10th: $142; Percentiles in dollars: 25th: $559; Percentiles in dollars: 50th: $1,428; Percentiles in dollars: 75th: $3,716. Difference between Schedule A, line 10 and Form 1098 amounts; Output of comparison: No Form 1098; Percentiles in dollars: 5th: $184; Percentiles in dollars: 10th: $435; Percentiles in dollars: 25th: $2,262; Percentiles in dollars: 50th: $5,248; Percentiles in dollars: 75th: $8,570. 2005: Schedule A, line 10 amount; 2004: Output of comparison: Match; Percentiles in dollars: 5th: $1,926; Percentiles in dollars: 10th: $2,981; Percentiles in dollars: 25th: $4,970; Percentiles in dollars: 50th: $7,588; Percentiles in dollars: 75th: $11,427. Schedule A, line 10 amount; 2004: Output of comparison: Deduction exceeds Form 1098 amount; Percentiles in dollars: 5th: $2,333; Percentiles in dollars: 10th: $3,494; Percentiles in dollars: 25th: $5,659; Percentiles in dollars: 50th: $8,834; Percentiles in dollars: 75th: $13,845. Schedule A, line 10 amount; 2004: Output of comparison: Deduction less than Form 1098 amount; Percentiles in dollars: 5th: $1,999; Percentiles in dollars: 10th: $3,129; Percentiles in dollars: 25th: $5,379; Percentiles in dollars: 50th: $8,511; Percentiles in dollars: 75th: $13,400. Schedule A, line 10 amount; 2004: Output of comparison: No Form 1098; Percentiles in dollars: 5th: $243; Percentiles in dollars: 10th: $505; Percentiles in dollars: 25th: $2,271; Percentiles in dollars: 50th: $5,140; Percentiles in dollars: 75th: $8,691. Difference between Schedule A, line 10 and Form 1098 amounts; Output of comparison: Match; Percentiles in dollars: 5th: [Empty]; Percentiles in dollars: 10th: [Empty]; Percentiles in dollars: 25th: [Empty]; Percentiles in dollars: 50th: [Empty]; Percentiles in dollars: 75th: [Empty]. Difference between Schedule A, line 10 and Form 1098 amounts; Output of comparison: Deduction exceeds Form 1098 amount; Percentiles in dollars: 5th: $21; Percentiles in dollars: 10th: $69; Percentiles in dollars: 25th: $263; Percentiles in dollars: 50th: $587; Percentiles in dollars: 75th: $2,335. Difference between Schedule A, line 10 and Form 1098 amounts; Output of comparison: Deduction less than Form 1098 amount; Percentiles in dollars: 5th: $70; Percentiles in dollars: 10th: $215; Percentiles in dollars: 25th: $694; Percentiles in dollars: 50th: $1,761; Percentiles in dollars: 75th: $4,314. Difference between Schedule A, line 10 and Form 1098 amounts; Output of comparison: No Form 1098; Percentiles in dollars: 5th: $243; Percentiles in dollars: 10th: $505; Percentiles in dollars: 25th: $2,271; Percentiles in dollars: 50th: $5,140; Percentiles in dollars: 75th: $8,691. 2006: Schedule A, line 10 amount; Output of comparison: Match; Percentiles in dollars: 5th: $2,091; Percentiles in dollars: 10th: $3,200; Percentiles in dollars: 25th: $5,324; Percentiles in dollars: 50th: $8,245; Percentiles in dollars: 75th: $12,728. Schedule A, line 10 amount; Output of comparison: Deduction exceeds Form 1098 amount; Percentiles in dollars: 5th: $2,754; Percentiles in dollars: 10th: $3,891; Percentiles in dollars: 25th: $6,193; Percentiles in dollars: 50th: $9,918; Percentiles in dollars: 75th: $15,722. Schedule A, line 10 amount; Output of comparison: Deduction less than Form 1098 amount; Percentiles in dollars: 5th: $2,161; Percentiles in dollars: 10th: $3,448; Percentiles in dollars: 25th: $5,716; Percentiles in dollars: 50th: $9,173; Percentiles in dollars: 75th: $14,937. Schedule A, line 10 amount; Output of comparison: No Form 1098; Percentiles in dollars: 5th: $249; Percentiles in dollars: 10th: $490; Percentiles in dollars: 25th: $2,638; Percentiles in dollars: 50th: $5,692; Percentiles in dollars: 75th: $9,555. Difference between Schedule A, line 10 and Form 1098 amounts; Output of comparison: Match; Percentiles in dollars: 5th: [Empty]; Percentiles in dollars: 10th: [Empty]; Percentiles in dollars: 25th: [Empty]; Percentiles in dollars: 50th: [Empty]; Percentiles in dollars: 75th: [Empty]. Difference between Schedule A, line 10 and Form 1098 amounts; Output of comparison: Deduction exceeds Form 1098 amount; Percentiles in dollars: 5th: $17; Percentiles in dollars: 10th: $76; Percentiles in dollars: 25th: $287; Percentiles in dollars: 50th: $727; Percentiles in dollars: 75th: $2,708. Difference between Schedule A, line 10 and Form 1098 amounts; Output of comparison: Deduction less than Form 1098 amount; Percentiles in dollars: 5th: $67; Percentiles in dollars: 10th: $235; Percentiles in dollars: 25th: $717; Percentiles in dollars: 50th: $1,939; Percentiles in dollars: 75th: $5,180. Difference between Schedule A, line 10 and Form 1098 amounts; Output of comparison: No Form 1098; Percentiles in dollars: 5th: $249; Percentiles in dollars: 10th: $490; Percentiles in dollars: 25th: $2,638; Percentiles in dollars: 50th: $5,692; Percentiles in dollars: 75th: $9,555. Source: GAO analysis of IRS data. Note: No values are shown for matches between Form 1098 and Schedule A, line 10 amounts when the value was $3 or less, to account for rounding errors. [End of table] Our analysis also examined the change in mortgage interest deduction amounts from 2004 through 2006, for those taxpayers who reported a deduction in all 3 years. Table 10 shows the amount of change from 2004 to 2005 and then from 2005 to 2006. For example, the median (50th percentile) changes for tax returns that had increases from both 2004 to 2005 and 2005 to 2006 were $1,496 and $1,691, respectively. Table 10: Year-to-Year Changes in Amount Deducted on Schedule A, Line 10, 2004 through 2006: Increase all years; Number of returns: 8,148,964; Percentile of change of line 10 deduction amounts in dollars: Year: 2004-2005; Percentile of change of line 10 deduction amounts in dollars: 5th: $103; Percentile of change of line 10 deduction amounts in dollars: 10th: $216; Percentile of change of line 10 deduction amounts in dollars: 25th: $584; Percentile of change of line 10 deduction amounts in dollars: 50th: $1,496; Percentile of change of line 10 deduction amounts in dollars: 75th: $3,550. Increase all years; Number of returns: 8,148,964; Percentile of change of line 10 deduction amounts in dollars: Year: 2005-2006; Percentile of change of line 10 deduction amounts in dollars: 5th: $134; Percentile of change of line 10 deduction amounts in dollars: 10th: $272; Percentile of change of line 10 deduction amounts in dollars: 25th: $694; Percentile of change of line 10 deduction amounts in dollars: 50th: $1,691; Percentile of change of line 10 deduction amounts in dollars: 75th: $4,157. Increase, then lower; Number of returns: 5,586,917; Percentile of change of line 10 deduction amounts in dollars: Year: 2004-2005; Percentile of change of line 10 deduction amounts in dollars: 5th: $90; Percentile of change of line 10 deduction amounts in dollars: 10th: $170; Percentile of change of line 10 deduction amounts in dollars: 25th: $480; Percentile of change of line 10 deduction amounts in dollars: 50th: $1,378; Percentile of change of line 10 deduction amounts in dollars: 75th: $3,946. Increase, then lower; Number of returns: 5,586,917; Percentile of change of line 10 deduction amounts in dollars: Year: 2005-2006; Percentile of change of line 10 deduction amounts in dollars: 5th: $65; Percentile of change of line 10 deduction amounts in dollars: 10th: $110; Percentile of change of line 10 deduction amounts in dollars: 25th: $255; Percentile of change of line 10 deduction amounts in dollars: 50th: $794; Percentile of change of line 10 deduction amounts in dollars: 75th: $2,291. Lower, then increase; Number of returns: 5,294,566; Percentile of change of line 10 deduction amounts in dollars: Year: 2004-2005; Percentile of change of line 10 deduction amounts in dollars: 5th: $65; Percentile of change of line 10 deduction amounts in dollars: 10th: $103; Percentile of change of line 10 deduction amounts in dollars: 25th: $272; Percentile of change of line 10 deduction amounts in dollars: 50th: $779; Percentile of change of line 10 deduction amounts in dollars: 75th: $1,904. Lower, then increase; Number of returns: 5,294,566; Percentile of change of line 10 deduction amounts in dollars: Year: 2005-2006; Percentile of change of line 10 deduction amounts in dollars: 5th: $99; Percentile of change of line 10 deduction amounts in dollars: 10th: $193; Percentile of change of line 10 deduction amounts in dollars: 25th: $509; Percentile of change of line 10 deduction amounts in dollars: 50th: $1,399; Percentile of change of line 10 deduction amounts in dollars: 75th: $3,743. Lower, all years; Number of returns: 8,530,243; Percentile of change of line 10 deduction amounts in dollars: Year: 2004-2005; Percentile of change of line 10 deduction amounts in dollars: 5th: $61; Percentile of change of line 10 deduction amounts in dollars: 10th: $85; Percentile of change of line 10 deduction amounts in dollars: 25th: $151; Percentile of change of line 10 deduction amounts in dollars: 50th: $334; Percentile of change of line 10 deduction amounts in dollars: 75th: $761. Lower, all years; Number of returns: 8,530,243; Percentile of change of line 10 deduction amounts in dollars: Year: 2005-2006; Percentile of change of line 10 deduction amounts in dollars: 5th: $63; Percentile of change of line 10 deduction amounts in dollars: 10th: $89; Percentile of change of line 10 deduction amounts in dollars: 25th: $152; Percentile of change of line 10 deduction amounts in dollars: 50th: $315; Percentile of change of line 10 deduction amounts in dollars: 75th: $672. Source: GAO analysis of IRS data. [End of table] Table 11 shows the comparison of the year-to-year change in the number of Forms 1098 taxpayers received with the dollar amount of change on Schedule A, line 10, for the corresponding years. Only those taxpayers with tax returns filed in all 3 years of our study are included in this analysis. For example, the table shows that 581,950 tax returns had a declining number of Form 1098 filings from both 2004 to 2005 and 2005 to 2006 ("Down all years"). For those taxpayers, the dollar amount of their deductions declined by $2,471 from 2004 to 2005 for those at the 50th percentile and decreased by $1,834 from 2005 to 2006. An increase in deduction dollar amounts can occur when Form 1098 filings decreased, for example, as in the year after a taxpayer sells one of his two homes, and purchases a new home that has a larger mortgage interest deduction amount than the previous two homes combined. Table 11: Year-to-Year Changes in the Number of Form 1098 Filings, with Distribution of Change Amounts on Schedule A, Line 10, 2004 through 2006: Down all years; Number of returns: 581,950; Percentile distribution of change in dollar amounts: Year: 2004-2005; Percentile distribution of change in dollar amounts: 5th: [Empty]; Percentile distribution of change in dollar amounts: 10th: [Empty]; Percentile distribution of change in dollar amounts: 25th: [Empty]; Percentile distribution of change in dollar amounts: 50th: [Empty]; Percentile distribution of change in dollar amounts: 75th: [Empty]. Down all years; Number of returns: 581,950; Percentile distribution of change in dollar amounts: Year: 2005-2006; Percentile distribution of change in dollar amounts: 5th: $112; Percentile distribution of change in dollar amounts: 10th: $210; Percentile distribution of change in dollar amounts: 25th: $685; Percentile distribution of change in dollar amounts: 50th: $1,834; Percentile distribution of change in dollar amounts: 75th: $4,752. Down, then up; Number of returns: 1,374,241; Percentile distribution of change in dollar amounts: Year: 2004-2005; Percentile distribution of change in dollar amounts: 5th: $136; Percentile distribution of change in dollar amounts: 10th: $270; Percentile distribution of change in dollar amounts: 25th: $627; Percentile distribution of change in dollar amounts: 50th: $1,608; Percentile distribution of change in dollar amounts: 75th: $4,011. Down, then up; Number of returns: 1,374,241; Percentile distribution of change in dollar amounts: Year: 2005-2006; Percentile distribution of change in dollar amounts: 5th: $122; Percentile distribution of change in dollar amounts: 10th: $244; Percentile distribution of change in dollar amounts: 25th: $743; Percentile distribution of change in dollar amounts: 50th: $2,000; Percentile distribution of change in dollar amounts: 75th: $5,040. Down, then the same; Number of returns: 3,272,246; Percentile distribution of change in dollar amounts: Year: 2004-2005; Percentile distribution of change in dollar amounts: 5th: $107; Percentile distribution of change in dollar amounts: 10th: $194; Percentile distribution of change in dollar amounts: 25th: $520; Percentile distribution of change in dollar amounts: 50th: $1,307; Percentile distribution of change in dollar amounts: 75th: $3,088. Down, then the same; Number of returns: 3,272,246; Percentile distribution of change in dollar amounts: Year: 2005-2006; Percentile distribution of change in dollar amounts: 5th: $74; Percentile distribution of change in dollar amounts: 10th: $108; Percentile distribution of change in dollar amounts: 25th: $215; Percentile distribution of change in dollar amounts: 50th: $581; Percentile distribution of change in dollar amounts: 75th: $1,727. More, then fewer; Number of returns: 3,499,626; Percentile distribution of change in dollar amounts: Year: 2004-2005; Percentile distribution of change in dollar amounts: 5th: $113; Percentile distribution of change in dollar amounts: 10th: $241; Percentile distribution of change in dollar amounts: 25th: $683; Percentile distribution of change in dollar amounts: 50th: $1,822; Percentile distribution of change in dollar amounts: 75th: $4,594. More, then fewer; Number of returns: 3,499,626; Percentile distribution of change in dollar amounts: Year: 2005-2006; Percentile distribution of change in dollar amounts: 5th: $121; Percentile distribution of change in dollar amounts: 10th: $233; Percentile distribution of change in dollar amounts: 25th: $646; Percentile distribution of change in dollar amounts: 50th: $1,731; Percentile distribution of change in dollar amounts: 75th: $4,358. Up all years; Number of returns: 742,782; Percentile distribution of change in dollar amounts: Year: 2004-2005; Percentile distribution of change in dollar amounts: 5th: $94; Percentile distribution of change in dollar amounts: 10th: $224; Percentile distribution of change in dollar amounts: 25th: $567; Percentile distribution of change in dollar amounts: 50th: $1,735; Percentile distribution of change in dollar amounts: 75th: $4,585. Up all years; Number of returns: 742,782; Percentile distribution of change in dollar amounts: Year: 2005-2006; Percentile distribution of change in dollar amounts: 5th: $201; Percentile distribution of change in dollar amounts: 10th: $414; Percentile distribution of change in dollar amounts: 25th: $1,305; Percentile distribution of change in dollar amounts: 50th: $3,508; Percentile distribution of change in dollar amounts: 75th: $8,065. Up, then the same; Number of returns: 1,791,402; Percentile distribution of change in dollar amounts: Year: 2004-2005; Percentile distribution of change in dollar amounts: 5th: $92; Percentile distribution of change in dollar amounts: 10th: $173; Percentile distribution of change in dollar amounts: 25th: v514; Percentile distribution of change in dollar amounts: 50th: $1,328; Percentile distribution of change in dollar amounts: 75th: $3,154. Up, then the same; Number of returns: 1,791,402; Percentile distribution of change in dollar amounts: Year: 2005-2006; Percentile distribution of change in dollar amounts: 5th: $95; Percentile distribution of change in dollar amounts: 10th: $187; Percentile distribution of change in dollar amounts: 25th: $477; Percentile distribution of change in dollar amounts: 50th: $1,386; Percentile distribution of change in dollar amounts: 75th: $3,702. Same, then fewer; Number of returns: 1,362,878; Percentile distribution of change in dollar amounts: Year: 2004-2005; Percentile distribution of change in dollar amounts: 5th: $82; Percentile distribution of change in dollar amounts: 10th: $161; Percentile distribution of change in dollar amounts: 25th: $501; Percentile distribution of change in dollar amounts: 50th: $1,370; Percentile distribution of change in dollar amounts: 75th: $3,589. Same, then fewer; Number of returns: 1,362,878; Percentile distribution of change in dollar amounts: Year: 2005-2006; Percentile distribution of change in dollar amounts: 5th: $92; Percentile distribution of change in dollar amounts: 10th: $166; Percentile distribution of change in dollar amounts: 25th: $487; Percentile distribution of change in dollar amounts: 50th: $1,268; Percentile distribution of change in dollar amounts: 75th: $3,293. Same, then up; Number of returns: 2,959,900; Percentile distribution of change in dollar amounts: Year: 2004-2005; Percentile distribution of change in dollar amounts: 5th: $65; Percentile distribution of change in dollar amounts: 10th: $93; Percentile distribution of change in dollar amounts: 25th: $211; Percentile distribution of change in dollar amounts: 50th: $559; Percentile distribution of change in dollar amounts: 75th: $1,591. Same, then up; Number of returns: 2,959,900; Percentile distribution of change in dollar amounts: Year: 2005-2006; Percentile distribution of change in dollar amounts: 5th: $89; Percentile distribution of change in dollar amounts: 10th: $172; Percentile distribution of change in dollar amounts: 25th: $535; Percentile distribution of change in dollar amounts: 50th: $1,530; Percentile distribution of change in dollar amounts: 75th: $3,823. Same, then the same; Number of returns: 12,121,270; Percentile distribution of change in dollar amounts: Year: 2004-2005; Percentile distribution of change in dollar amounts: 5th: $60; Percentile distribution of change in dollar amounts: 10th: $89; Percentile distribution of change in dollar amounts: 25th: $176; Percentile distribution of change in dollar amounts: 50th: $429; Percentile distribution of change in dollar amounts: 75th: $1,104. Same, then the same; Number of returns: 12,121,270; Percentile distribution of change in dollar amounts: Year: 2005-2006; Percentile distribution of change in dollar amounts: 5th: $65; Percentile distribution of change in dollar amounts: 10th: $96; Percentile distribution of change in dollar amounts: 25th: $185; Percentile distribution of change in dollar amounts: 50th: $441; Percentile distribution of change in dollar amounts: 75th: $1,130. Source: GAO analysis of IRS data. [End of table] Table 12 shows an analysis of Form 1098 by mortgage loan account numbers. We performed this analysis to show more precisely how frequently taxpayers deducted mortgage interest on the same loan from year to year, assuming the same loan number in different years referred to the same loan. For example, from 2004 and 2005, about 25 million taxpayers had one matching account number on their Forms 1098. The third chart shows taxpayers who had matching accounts throughout the 3- year period. For example, 574,173 taxpayers had 3 matching account numbers on their Forms 1098 in 2004, 2005, and 2006. Taxpayers with zero matching account numbers likely means that they deducted interest on different loans in different years. Table 12: Numbers of Matching Loan Account Numbers on Forms 1098, 2004 through 2006: Account number matches, 2004-2005: 0; Number of returns: 12,937,920; Percentage: 28.0. Account number matches, 2004-2005: 1; Number of returns: 24,779,618; Percentage: 53.6. Account number matches, 2004-2005: 2; Number of returns: 7,081,903; Percentage: 15.3. Account number matches, 2004-2005: 3; Number of returns: 1,005,476; Percentage: 2.2. Account number matches, 2004-2005: 4 or more; Number of returns: 460,467; Percentage: 1.0. Account number matches, 2005-2006: 0; Number of returns: 11,002,815; Percentage: 23.8. Account number matches, 2005-2006: 1; Number of returns: 24,773,661; Percentage: 53.5. Account number matches, 2005-2006: 2; Number of returns: 8,625,659; Percentage: 18.6. Account number matches, 2005-2006: 3; Number of returns: 1,228,589; Percentage: 2.7. Account number matches, 2005-2006: 4 or more; Number of returns: 634,661; Percentage: 1.4. Number of matching account numbers in all years: 0; Number of returns: 21,249,738; Percentage: 45.9. Number of matching account numbers in all years: 1; Number of returns: 19,929,360; Percentage: 43.1. Number of matching account numbers in all years: 2; Number of returns: 4,284,053; Percentage: 9.3. Number of matching account numbers in all years: 3; Number of returns: 574,173; Percentage: 1.2. Number of matching account numbers in all years: 4 or more; Number of returns: 228,059; Percentage: 0.5. Source: GAO analysis of IRS data. Note: Tables' percentages may not add to 100 percent because of rounding. [End of table] Table 13 shows the distribution of taxpayers' mortgage interest deductions on Schedule A and on other schedules. Non-Schedule A interest deductions may relate to forms such as Schedule C for small business, Schedule E for rental property, and Schedule F for farming. Table 13: Number of Individual Tax Returns with Schedule A Mortgage Interest Deductions and Non-Schedule A Mortgage Interest Deductions, 2004 through 2006: Mortgage interest deduction types: Only Schedule A; 2004: Number of tax returns: 33,531,542; 2004: Percentage: 26.2; 2005: Number of tax returns: 34,239,756; 2005: Percentage: 26.3; 2006: Number of tax returns: 35,341,431; 2006: Percentage: 26.4. Mortgage interest deduction types: Only non-Schedule A; 2004: Number of tax returns: 2,313,995; 2004: Percentage: 1.8; 2005: Number of tax returns: 2,297,572; 2005: Percentage: 1.8; 2006: Number of tax returns: 2,248,038; 2006: Percentage: 1.7. Mortgage interest deduction types: Schedule A and non-Schedule A; 2004: Number of tax returns: 3,360,344; 2004: Percentage: 2.6; 2005: Number of tax returns: 3,477,998; 2005: Percentage: 2.7; 2006: Number of tax returns: 3,574,322; 2006: Percentage: 2.7. Mortgage interest deduction types: No mortgage interest; 2004: Number of tax returns: 88,871,934; 2004: Percentage: 69.4; 2005: Number of tax returns: 90,021,638; 2005: Percentage: 69.2; 2006: Number of tax returns: 92,720,476; 2006: Percentage: 69.3. Mortgage interest deduction types: Total; 2004: Number of tax returns: 128,077,815; 2004: Percentage: 100.0; 2005: Number of tax returns: 130,036,964; 2005: Percentage: 100.0; 2006: Number of tax returns: 133,884,266; 2006: Percentage: 100.0. Source: GAO analysis of IRS data. Note: Percentages may not add to 100 percent because of rounding. [End of table] [End of section] Appendix VI: Margins of Error Tables for Estimates in Appendix V: The following tables provide the margins of error for the corresponding statistics reported in the previous appendix. For example, table 14 corresponds with table 6 in appendix V. Table 14: Margins of Error for Number and Dollar Total of Schedule A, Line 10 Deductions per Number of Form 1098 Filings, 2004 through 2006: 2004: Number of Form 1098 filings: 0; Number of tax returns (millions): [A]; Percentage: 0.85; Line 10 amounts for tax returns (billions of dollars): $0.78. Number of Form 1098 filings: 1; Number of tax returns (millions): [A]; Percentage: 0.63; Line 10 amounts for tax returns (billions of dollars): $2.02. Number of Form 1098 filings: 2; Number of tax returns (millions): [A]; Percentage: 0.58; Line 10 amounts for tax returns (billions of dollars): $2.38. Number of Form 1098 filings: 3; Number of tax returns (millions): [A]; Percentage: 0.39; Line 10 amounts for tax returns (billions of dollars): $1.75. Number of Form 1098 filings: 4 or more; Number of tax returns (millions): [A]; Percentage: 0.30; Line 10 amounts for tax returns (billions of dollars): $1.52. Number of Form 1098 filings: Total; Number of tax returns (millions): [A]; Percentage: [A]; Line 10 amounts for tax returns (billions of dollars): $2.66. 2005: Number of Form 1098 filings: 0; Number of tax returns (millions): [A]; Percentage: 0.68; Line 10 amounts for tax returns (billions of dollars): $0.61. Number of Form 1098 filings: 1; Number of tax returns (millions): [A]; Percentage: 0.45; Line 10 amounts for tax returns (billions of dollars): $1.90. Number of Form 1098 filings: 2; Number of tax returns (millions): [A]; Percentage: 0.42; Line 10 amounts for tax returns (billions of dollars): $1.81. Number of Form 1098 filings: 3; Number of tax returns (millions): [A]; Percentage: 0.29; Line 10 amounts for tax returns (billions of dollars): $1.52. Number of Form 1098 filings: 4 or more; Number of tax returns (millions): [A]; Percentage: 0.24; Line 10 amounts for tax returns (billions of dollars): $1.53. Number of Form 1098 filings: Total; Number of tax returns (millions): [A]; Percentage: [A]; Line 10 amounts for tax returns (billions of dollars): $2.47. 2006: Number of Form 1098 filings: 0; Number of tax returns (millions): [A]; Percentage: 0.63%; Line 10 amounts for tax returns (billions of dollars): $0.72. Number of Form 1098 filings: 1; Number of tax returns (millions): [A]; Percentage: 0.44; Line 10 amounts for tax returns (billions of dollars): $1.71. Number of Form 1098 filings: 2; Number of tax returns (millions): [A]; Percentage: 0.42; Line 10 amounts for tax returns (billions of dollars): $2.15. Number of Form 1098 filings: 3; Number of tax returns (millions): [A]; Percentage: 0.29; Line 10 amounts for tax returns (billions of dollars): $1.74. Number of Form 1098 filings: 4 or more; Number of tax returns (millions): [A]; Percentage: 0.23; Line 10 amounts for tax returns (billions of dollars): $1.74. Number of Form 1098 filings: Total; Number of tax returns (millions): [A]; Percentage: [A]; Line 10 amounts for tax returns (billions of dollars): $2.60. Source: GAO analysis of IRS data. [A] The values of these margins are less than 0.01. [End of table] Table 15: Margins of Error for Distribution of Schedule A, Line 10 Deductions in Select Percentiles per Number of Form 1098 Filings, 2004 through 2006: 2004: Number of Form 1098 filings: 0; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 5th: $75; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 10th: $83; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 25th: $437; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 50th: $399; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 75th: $517. Number of Form 1098 filings: 1; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 5th: $80; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 10th: $84; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 25th: $89; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 50th: $88; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 75th: $145. Number of Form 1098 filings: 2; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 5th: $153; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 10th: $108; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 25th: $122; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 50th: $150; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 75th: $222. Number of Form 1098 filings: 3; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 5th: $325; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 10th: $278; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 25th: $228; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 50th: $329; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 75th: $619. Number of Form 1098 filings: 4 or more; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 5th: $501; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 10th: $433; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 25th: $331; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 50th: $683; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 75th: $959. 2005: Number of Form 1098 filings: 0; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 5th: $56; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 10th: $59; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 25th: $313; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 50th: $242; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 75th: $400. Number of Form 1098 filings: 1; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 5th: $64; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 10th: $67; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 25th: $57; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 50th: $66; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 75th: $93. Number of Form 1098 filings: 2; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 5th: $107; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 10th: $97; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 25th: $102; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 50th: $115; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 75th: $164. Number of Form 1098 filings: 3; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 5th: $178; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 10th: $174; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 25th: $172; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 50th: $217; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 75th: $352. Number of Form 1098 filings: 4 or more; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 5th: $260; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 10th: $297; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 25th: $321; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 50th: $511; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 75th: $629. 2006: Number of Form 1098 filings: 0; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 5th: $72; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 10th: $87; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 25th: $280; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 50th: $355; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 75th: $645. Number of Form 1098 filings: 1; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 5th: $74; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 10th: $65; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 25th: $62; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 50th: $82; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 75th: $123. Number of Form 1098 filings: 2; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 5th: $115; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 10th: $87; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 25th: $100; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 50th: $102; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 75th: $173. 2004: Number of Form 1098 filings: 3; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 5th: $223; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 10th: $226; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 25th: $274; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 50th: $299; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 75th: $490. 2004: Number of Form 1098 filings: 4 or more; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 5th: $365; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 10th: $264; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 25th: $328; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 50th: $519; Schedule A, line 10 deduction distribution by certain percentiles in dollars: 75th: $685. Source: GAO analysis of IRS data. [End of table] Table 16: Margins of Error for Results of Comparing Total Schedule A, Line 10 Deduction Amounts to Amounts Reported on Form 1098, 2004 through 2006: 2004: Result of comparison: Match; Total number of tax returns (millions): [A]; Percentage: 0.58; Schedule A, line 10 amount (billions of dollars): $2.49; Difference between Form 1098 and line 10 amounts (millions of dollars): [Empty]. Result of comparison: Deduction exceeds Form 1098 amount; Total number of tax returns (millions): [A]; Percentage: 0.40; Schedule A, line 10 amount (billions of dollars): $1.95; Difference between Form 1098 and line 10 amounts (millions of dollars): $1,034.82. Result of comparison: Deduction less than Form 1098 amount; Total number of tax returns (millions): [A]; Percentage: 0.45; Schedule A, line 10 amount (billions of dollars): $1.53; Difference between Form 1098 and line 10 amounts (millions of dollars): $759.44. Result of comparison: No Form 1098; Total number of tax returns (millions): [A]; Percentage: 0.85; Schedule A, line 10 amount (billions of dollars): $0.78; Difference between Form 1098 and line 10 amounts (millions of dollars): $778.34. Result of comparison: Total; Total number of tax returns (millions): [A]; Percentage: [A]; Schedule A, line 10 amount (billions of dollars): $2.66; Difference between Form 1098 and line 10 amounts (millions of dollars): [Empty]. 2005: Result of comparison: Match; Total number of tax returns (millions): [A]; Percentage: 0.42; Schedule A, line 10 amount (billions of dollars): $2.09; Difference between Form 1098 and line 10 amounts (millions of dollars): [Empty]. Result of comparison: Deduction exceeds Form 1098 amount; Total number of tax returns (millions): [A]; Percentage: 0.27; Schedule A, line 10 amount (billions of dollars): $1.74; Difference between Form 1098 and line 10 amounts (millions of dollars): $1,213.18. Result of comparison: Deduction less than Form 1098 amount; Total number of tax returns (millions): [A]; Percentage: 0.34; Schedule A, line 10 amount (billions of dollars): $1.43; Difference between Form 1098 and line 10 amounts (millions of dollars): $833.85. Result of comparison: No Form 1098; Total number of tax returns (millions): [A]; Percentage: 0.68; Schedule A, line 10 amount (billions of dollars): $0.61; Difference between Form 1098 and line 10 amounts (millions of dollars): $614.00. Result of comparison: Total; 2004: Total number of tax returns (millions): [A]; Percentage: [A]; Schedule A, line 10 amount (billions of dollars): $2.47; Difference between Form 1098 and line 10 amounts (millions of dollars): [Empty]. 2006: Result of comparison: Match; Total number of tax returns (millions): [A]; Percentage: 0.41; Schedule A, line 10 amount (billions of dollars): $2.46; Difference between Form 1098 and line 10 amounts (millions of dollars): [Empty]. Result of comparison: Deduction exceeds Form 1098 amount; Total number of tax returns (millions): [A]; Percentage: 0.25; Schedule A, line 10 amount (billions of dollars): $1.45; Difference between Form 1098 and line 10 amounts (millions of dollars): $394.38. Result of comparison: Deduction less than Form 1098 amount; Total number of tax returns (millions): [A]; Percentage: 0.33; Schedule A, line 10 amount (billions of dollars): $1.64; Difference between Form 1098 and line 10 amounts (millions of dollars): $1,026.84. Result of comparison: No Form 1098; Total number of tax returns (millions): [A]; Percentage: 0.63; Schedule A, line 10 amount (billions of dollars): $0.72; Difference between Form 1098 and line 10 amounts (millions of dollars): $717.90. Result of comparison: Total; Total number of tax returns (millions): [A]; Percentage: [A]; Schedule A, line 10 amount (billions of dollars): $2.60; Difference between Form 1098 and line 10 amounts (millions of dollars): [Empty]. Source: GAO analysis of IRS data. [A] The values of these margins are less than 0.01. [End of table] Table 17: Margins of Error for Percentile Distributions for Schedule A, Line 10 Amounts Compared with Form 1098 Amounts and the Difference between Schedule A, Line 10 Amounts and Form 1098 Amounts, 2004 through 2006: 2004: Schedule A, line 10 amount: Output of comparison: Match; Percentiles in dollars: 5th: [Empty]; Percentiles in dollars: 10th: [Empty]; Percentiles in dollars: 25th: [Empty]; Percentiles in dollars: 50th: [Empty]; Percentiles in dollars: 75th: [Empty]. Schedule A, line 10 amount: Output of comparison: Deduction exceeds Form 1098 amount; Percentiles in dollars: 5th: $290; Percentiles in dollars: 10th: $227; Percentiles in dollars: 25th: $192; Percentiles in dollars: 50th: $281; Percentiles in dollars: 75th: $613. Schedule A, line 10 amount: Output of comparison: Deduction less than Form 1098 amount; Percentiles in dollars: 5th: $228; Percentiles in dollars: 10th: $208; Percentiles in dollars: 25th: $200; Percentiles in dollars: 50th: $239; Percentiles in dollars: 75th: $400. Schedule A, line 10 amount: Output of comparison: Match: No Form 1098; Percentiles in dollars: 5th: $75; Percentiles in dollars: 10th: $83; Percentiles in dollars: 25th: $437; Percentiles in dollars: 50th: $399; Percentiles in dollars: 75th: $517. Difference between line 10 and Form 1098 amounts; Output of comparison: Match: Percentiles in dollars: 5th: [Empty]; Percentiles in dollars: 10th: [Empty]; Percentiles in dollars: 25th: [Empty]; Percentiles in dollars: 50th: [Empty]; Percentiles in dollars: 75th: [Empty]. Difference between line 10 and Form 1098 amounts; Output of comparison: Deduction exceeds Form 1098 amount; Percentiles in dollars: 5th: $10; Percentiles in dollars: 10th: $17; Percentiles in dollars: 25th: $24; Percentiles in dollars: 50th: $94; Percentiles in dollars: 75th: $301. Difference between line 10 and Form 1098 amounts; Output of comparison: Deduction less than Form 1098 amount; Percentiles in dollars: 5th: $15; Percentiles in dollars: 10th: $32; Percentiles in dollars: 25th: $57; Percentiles in dollars: 50th: $83; Percentiles in dollars: 75th: $233. Difference between line 10 and Form 1098 amounts; Output of comparison: No Form 1098; Percentiles in dollars: 5th: $75; Percentiles in dollars: 10th: $83; Percentiles in dollars: 25th: $437; Percentiles in dollars: 50th: $399; Percentiles in dollars: 75th: $517. 2005: Schedule A, line 10 amount; Output of comparison: Match; Percentiles in dollars: 5th: [Empty]; Percentiles in dollars: 10th: [Empty]; Percentiles in dollars: 25th: [Empty]; Percentiles in dollars: 50th: [Empty]; Percentiles in dollars: 75th: [Empty]. Schedule A, line 10 amount; Output of comparison: Deduction exceeds Form 1098 amount; Percentiles in dollars: 5th: $228; Percentiles in dollars: 10th: $253; Percentiles in dollars: 25th: $184; Percentiles in dollars: 50th: $232; Percentiles in dollars: 75th: $442. Schedule A, line 10 amount; 2004: Output of comparison: Deduction less than Form 1098 amount; Percentiles in dollars: 5th: $195; Percentiles in dollars: 10th: $142; Percentiles in dollars: 25th: $160; Percentiles in dollars: 50th: $213; Percentiles in dollars: 75th: $339. Schedule A, line 10 amount; 2004: Output of comparison: No Form 1098; Percentiles in dollars: 5th: $56; Percentiles in dollars: 10th: $59; Percentiles in dollars: 25th: $313; Percentiles in dollars: 50th: $242; Percentiles in dollars: 75th: $400. Difference between line 10 and Form 1098 amounts; Output of comparison: Match; Percentiles in dollars: 5th: [Empty]; Percentiles in dollars: 10th: [Empty]; Percentiles in dollars: 25th: [Empty]; Percentiles in dollars: 50th: [Empty]; Percentiles in dollars: 75th: [Empty]. Difference between line 10 and Form 1098 amounts; Output of comparison: Deduction exceeds Form 1098 amount; Percentiles in dollars: 5th: $5; Percentiles in dollars: 10th: $13; Percentiles in dollars: 25th: $18; Percentiles in dollars: 50th: $33; Percentiles in dollars: 75th: $187. Difference between line 10 and Form 1098 amounts; Output of comparison: Deduction less than Form 1098 amount; Percentiles in dollars: 5th: $19; Percentiles in dollars: 10th: $26; Percentiles in dollars: 25th: $32; Percentiles in dollars: 50th: $81; Percentiles in dollars: 75th: $192. Difference between line 10 and Form 1098 amounts; Output of comparison: No Form 1098; Percentiles in dollars: 5th: $56; Percentiles in dollars: 10th: $59; Percentiles in dollars: 25th: $313; Percentiles in dollars: 50th: $242; Percentiles in dollars: 75th: $400. 2006: Schedule A, line 10 amount; Output of comparison: Match; Percentiles in dollars: 5th: [Empty]; Percentiles in dollars: 10th: [Empty]; Percentiles in dollars: 25th: [Empty]; Percentiles in dollars: 50th: [Empty]; Percentiles in dollars: 75th: [Empty]. Schedule A, line 10 amount; Output of comparison: Deduction exceeds Form 1098 amount; Percentiles in dollars: 5th: $217; Percentiles in dollars: 10th: $203; Percentiles in dollars: 25th: $199; Percentiles in dollars: 50th: $264; Percentiles in dollars: 75th: $583. Schedule A, line 10 amount; Output of comparison: Deduction less than Form 1098 amount; Percentiles in dollars: 5th: $167; Percentiles in dollars: 10th: $164; Percentiles in dollars: 25th: $164; Percentiles in dollars: 50th: $213; Percentiles in dollars: 75th: $390. Schedule A, line 10 amount; Output of comparison: No Form 1098; Percentiles in dollars: 5th: $72; Percentiles in dollars: 10th: $87; Percentiles in dollars: 25th: $280; Percentiles in dollars: 50th: $355; Percentiles in dollars: 75th: $645. Difference between line 10 and Form 1098 amounts: Output of comparison: Match; Percentiles in dollars: 5th: [Empty]; Percentiles in dollars: 10th: [Empty]; Percentiles in dollars: 25th: [Empty]; Percentiles in dollars: 50th: [Empty]; Percentiles in dollars: 75th: [Empty]. Difference between line 10 and Form 1098 amounts: Output of comparison: Deduction exceeds Form 1098 amount; Percentiles in dollars: 5th: $6; Percentiles in dollars: 10th: $14; Percentiles in dollars: 25th: $19; Percentiles in dollars: 50th: $83; Percentiles in dollars: 75th: $232. Difference between line 10 and Form 1098 amounts: Output of comparison: Deduction less than Form 1098 amount; Percentiles in dollars: 5th: $20; Percentiles in dollars: 10th: $31; Percentiles in dollars: 25th: $39; Percentiles in dollars: 50th: $89; Percentiles in dollars: 75th: $220. Difference between line 10 and Form 1098 amounts: Output of comparison: No Form 1098; Percentiles in dollars: 5th: $72; Percentiles in dollars: 10th: $87; Percentiles in dollars: 25th: $280; Percentiles in dollars: 50th: $355; Percentiles in dollars: 75th: $645. Source: GAO analysis of IRS data. [End of table] Table 18: Margins of Error for Year-to-Year Changes in Amount Deducted on Schedule A, Line 10, 2004 through 2006: Increase all years; Years: 2004-2005; Number of sample tax returns: 1,232.72; Percentile of change of line 10 deduction amounts in dollars: 5th: $16; Percentile of change of line 10 deduction amounts in dollars: 10th: $20; Percentile of change of line 10 deduction amounts in dollars: 25th: $30; Percentile of change of line 10 deduction amounts in dollars: 50th: $64; Percentile of change of line 10 deduction amounts in dollars: 75th: $195. Increase all years; Years: 2005-2006; Number of sample tax returns: 1,232.72; Percentile of change of line 10 deduction amounts in dollars: 5th: $18; Percentile of change of line 10 deduction amounts in dollars: 10th: $24; Percentile of change of line 10 deduction amounts in dollars: 25th: $41; Percentile of change of line 10 deduction amounts in dollars: 50th: $75; Percentile of change of line 10 deduction amounts in dollars: 75th: $200. Increase, then lower; Years: 2004-2005; Number of sample tax returns: 1,170.13; Percentile of change of line 10 deduction amounts in dollars: 5th: $17; Percentile of change of line 10 deduction amounts in dollars: 10th: $18; Percentile of change of line 10 deduction amounts in dollars: 25th: $34; Percentile of change of line 10 deduction amounts in dollars: 50th: $102; Percentile of change of line 10 deduction amounts in dollars: 75th: $224. Increase, then lower; Years: 2005-2006; Number of sample tax returns: 1,170.13; Percentile of change of line 10 deduction amounts in dollars: 5th: $7; Percentile of change of line 10 deduction amounts in dollars: 10th: $10; Percentile of change of line 10 deduction amounts in dollars: 25th: $20; Percentile of change of line 10 deduction amounts in dollars: 50th: $58; Percentile of change of line 10 deduction amounts in dollars: 75th: $176. Lower, then increase; Years: 2004-2005; Number of sample tax returns: 1,215.27; Percentile of change of line 10 deduction amounts in dollars: 5th: $7; Percentile of change of line 10 deduction amounts in dollars: 10th: $8; Percentile of change of line 10 deduction amounts in dollars: 25th: $25; Percentile of change of line 10 deduction amounts in dollars: 50th: $48; Percentile of change of line 10 deduction amounts in dollars: 75th: $101. Lower, then increase; Years: 2005-2006; Number of sample tax returns: 1,215.27; Percentile of change of line 10 deduction amounts in dollars: 5th: $18; Percentile of change of line 10 deduction amounts in dollars: 10th: $18; Percentile of change of line 10 deduction amounts in dollars: 25th: $36; Percentile of change of line 10 deduction amounts in dollars: 50th: $101; Percentile of change of line 10 deduction amounts in dollars: 75th: $235. Lower, all years; Years: 2004-2005; Number of sample tax returns: 1,219.63; Percentile of change of line 10 deduction amounts in dollars: 5th: $4; Percentile of change of line 10 deduction amounts in dollars: 10th: $3; Percentile of change of line 10 deduction amounts in dollars: 25th: $7; Percentile of change of line 10 deduction amounts in dollars: 50th: $14; Percentile of change of line 10 deduction amounts in dollars: 75th: $42. Lower, all years; Years: 2005-2006; Number of sample tax returns: 1,219.63; Percentile of change of line 10 deduction amounts in dollars: 5th: $4; Percentile of change of line 10 deduction amounts in dollars: 10th: $4; Percentile of change of line 10 deduction amounts in dollars: 25th: $6; Percentile of change of line 10 deduction amounts in dollars: 50th: $14; Percentile of change of line 10 deduction amounts in dollars: 75th: $27. Source: GAO analysis of IRS data. [End of table] Table 19: Margins of Error for Year-to-Year Changes in the Number of Form 1098 Filings, with Distribution of Change Amounts on Schedule A, Line 10, 2004 through 2006: Down all years; Years: 2004-2005; Number of tax returns: 629; Percentile distribution of change in dollar amounts: 5th: $69; Percentile distribution of change in dollar amounts: 10th: $119; Percentile distribution of change in dollar amounts: 25th: $179; Percentile distribution of change in dollar amounts: 50th: $424; Percentile distribution of change in dollar amounts: 75th: $1,039. Down all years; Years: 2005-2006; Number of tax returns: 629; Percentile distribution of change in dollar amounts: 5th: $48; Percentile distribution of change in dollar amounts: 10th: $81; Percentile distribution of change in dollar amounts: 25th: $159; Percentile distribution of change in dollar amounts: 50th: $532; Percentile distribution of change in dollar amounts: 75th: $892. Down, then up; Years: 2004-2005; Number of tax returns: 996; Percentile distribution of change in dollar amounts: 5th: $56; Percentile distribution of change in dollar amounts: 10th: $44; Percentile distribution of change in dollar amounts: 25th: $99; Percentile distribution of change in dollar amounts: 50th: $164; Percentile distribution of change in dollar amounts: 75th: $387. Down, then up; Years: 2005-2006; Number of tax returns: 996; Percentile distribution of change in dollar amounts: 5th: $51; Percentile distribution of change in dollar amounts: 10th: $63; Percentile distribution of change in dollar amounts: 25th: $141; Percentile distribution of change in dollar amounts: 50th: $179; Percentile distribution of change in dollar amounts: 75th: $597. Down, then the same; Years: 2004-2005; Number of tax returns: 1,027; Percentile distribution of change in dollar amounts: 5th: $17; Percentile distribution of change in dollar amounts: 10th: $31; Percentile distribution of change in dollar amounts: 25th: $54; Percentile distribution of change in dollar amounts: 50th: $94; Percentile distribution of change in dollar amounts: 75th: $306. Down, then the same; Years: 2005-2006; Number of tax returns: 1,027; Percentile distribution of change in dollar amounts: 5th: $10; Percentile distribution of change in dollar amounts: 10th: $10; Percentile distribution of change in dollar amounts: 25th: $21; Percentile distribution of change in dollar amounts: 50th: $58; Percentile distribution of change in dollar amounts: 75th: $164. More, then fewer; Years: 2004-2005; Number of tax returns: 1,196; Percentile distribution of change in dollar amounts: 5th: $28; Percentile distribution of change in dollar amounts: 10th: $39; Percentile distribution of change in dollar amounts: 25th: $61; Percentile distribution of change in dollar amounts: 50th: $132; Percentile distribution of change in dollar amounts: 75th: $337. More, then fewer; Years: 2005-2006; Number of tax returns: 1,196; Percentile distribution of change in dollar amounts: 5th: $24; Percentile distribution of change in dollar amounts: 10th: $47; Percentile distribution of change in dollar amounts: 25th: $65; Percentile distribution of change in dollar amounts: 50th: $152; Percentile distribution of change in dollar amounts: 75th: $298. Up all years; Years: 2004-2005; Number of tax returns: 930; Percentile distribution of change in dollar amounts: 5th: $60; Percentile distribution of change in dollar amounts: 10th: $72; Percentile distribution of change in dollar amounts: 25th: $178; Percentile distribution of change in dollar amounts: 50th: v257; Percentile distribution of change in dollar amounts: 75th: $752. Up all years; Years: 2005-2006; Number of tax returns: 930; Percentile distribution of change in dollar amounts: 5th: $140; Percentile distribution of change in dollar amounts: 10th: $113; Percentile distribution of change in dollar amounts: 25th: $292; Percentile distribution of change in dollar amounts: 50th: $662; Percentile distribution of change in dollar amounts: 75th: $906. Up, then the same; Years: 2004-2005; Number of tax returns: 1,186; Percentile distribution of change in dollar amounts: 5th: $23; Percentile distribution of change in dollar amounts: 10th: $31; Percentile distribution of change in dollar amounts: 25th: $77; Percentile distribution of change in dollar amounts: 50th: $143; Percentile distribution of change in dollar amounts: 75th: $380. Up, then the same; Years: 2005-2006; Number of tax returns: 1,186; Percentile distribution of change in dollar amounts: 5th: $32; Percentile distribution of change in dollar amounts: 10th: $33; Percentile distribution of change in dollar amounts: 25th: $59; Percentile distribution of change in dollar amounts: 50th: $206; Percentile distribution of change in dollar amounts: 75th: $528. Same, then fewer; Years: 2004-2005; Number of tax returns: 958; Percentile distribution of change in dollar amounts: 5th: $23; Percentile distribution of change in dollar amounts: 10th: $41; Percentile distribution of change in dollar amounts: 25th: $58; Percentile distribution of change in dollar amounts: 50th: $168; Percentile distribution of change in dollar amounts: 75th: $564. Same, then fewer; Years: 2005-2006; Number of tax returns: 958; Percentile distribution of change in dollar amounts: 5th: $27; Percentile distribution of change in dollar amounts: 10th: $47; Percentile distribution of change in dollar amounts: 25th: $79; Percentile distribution of change in dollar amounts: 50th: $169; Percentile distribution of change in dollar amounts: 75th: $379. Same, then up; Years: 2004-2005; Number of tax returns: 1,370; Percentile distribution of change in dollar amounts: 5th: $9; Percentile distribution of change in dollar amounts: 10th: $11; Percentile distribution of change in dollar amounts: 25th: $21; Percentile distribution of change in dollar amounts: 50th: $68; Percentile distribution of change in dollar amounts: 75th: $172. Same, then up; Years: 2005-2006; Number of tax returns: 1,370; Percentile distribution of change in dollar amounts: 5th: $15; Percentile distribution of change in dollar amounts: 10th: $31; Percentile distribution of change in dollar amounts: 25th: $62; Percentile distribution of change in dollar amounts: 50th: $129; Percentile distribution of change in dollar amounts: 75th: $331. Same all years; Years: 2004-2005; Number of tax returns: 1,361; Percentile distribution of change in dollar amounts: 5th: $4; Percentile distribution of change in dollar amounts: 10th: $5; Percentile distribution of change in dollar amounts: 25th: $8; Percentile distribution of change in dollar amounts: 50th: $18; Percentile distribution of change in dollar amounts: 75th: $52. Same all years; Years: 2005-2006; Number of tax returns: 1,361; Percentile distribution of change in dollar amounts: 5th: $4; Percentile distribution of change in dollar amounts: 10th: $4; Percentile distribution of change in dollar amounts: 25th: $7; Percentile distribution of change in dollar amounts: 50th: $19; Percentile distribution of change in dollar amounts: 75th: $54. Source: GAO analysis of IRS data. [End of table] Table 20: Margins of Error for Number of Matching Loan Account Numbers on Forms 1098, 2004 through 2006: Account number matches, 2004 to 2005: 0; Number of tax returns: 1,677.24; Percentage: 0.51. Account number matches, 2004 to 2005: 1; Number of tax returns: 1,624.59; Percentage: 0.56. Account number matches, 2004 to 2005: 2; Number of tax returns: 1,124.92; Percentage: 0.39. Account number matches, 2004 to 2005: 3; Number of tax returns: 490.38; Percentage: 0.40. Account number matches, 2004 to 2005: 4 or more; Number of tax returns: 225.27; Percentage: 0.28. Account number matches, 2004 to 2005: Total; Number of tax returns: 1,387.08; Percentage: [A]. Account number matches, 2004 to 2005: 0; Number of tax returns: 1,506.23; Percentage: 0.48. Account number matches, 2004 to 2005: 1; Number of tax returns: 1,664.22; Percentage: 0.56. Account number matches, 2004 to 2005: 2; Number of tax returns: 1,278.87; Percentage: 0.43. Account number matches, 2004 to 2005: 3; Number of tax returns: 562.73; Percentage: 0.42. Account number matches, 2004 to 2005: 4 or more; Number of tax returns: 283.62; Percentage: 0.31. Account number matches, 2004 to 2005: Total; Number of tax returns: 1,387.08; Percentage: [A]. Account number matches, 2004 to 2005: 0; Number of tax returns: 1,633.34; Percentage: 0.56. Account number matches, 2004 to 2005: 1; Number of tax returns: 1,474.68; Percentage: 0.56. Account number matches, 2004 to 2005: 2; Number of tax returns: 964.55; Percentage: 0.31. Account number matches, 2004 to 2005: 3; Number of tax returns: 463.57; Percentage: 0.41. Account number matches, 2004 to 2005: 4 or more; Number of tax returns: 197.94; Percentage: 0.30. Account number matches, 2004 to 2005: Total; Number of tax returns: 1,387.08; Percentage: [A]. Source: GAO analysis of IRS data. [A] The values of these margins are less than 0.01. [End of table] Table 21: Margins of Error for Numbers of Individual Income Tax Returns with Schedule A Mortgage Interest Deductions and Non-Schedule A Mortgage Interest Deductions, 2004 through 2006: Only Schedule A; 2004: Number of tax returns: 1,439.72; 2004: Percentage: 0.26%; 2005: Number of tax returns: 1,053.89; 2005: Percentage: 0.19; 2006: Number of tax returns: 989.20; 2006: Percentage: 0.18. Only non-Schedule A; 2004: Number of tax returns: 663.07; 2004: Percentage: 0.27; 2005: Number of tax returns: 548.72; 2005: Percentage: 0.25; 2006: Number of tax returns: 494.11; 2006: Percentage: 0.23. Schedule A and non-Schedule A; 2004: Number of tax returns: 572.83; 2004: Percentage: 0.25; 2005: Number of tax returns: 459.87; 2005: Percentage: 0.22; 2006: Number of tax returns: 441.82; 2006: Percentage: 0.21. No mortgage interest; 2004: Number of tax returns: 2,797.03; 2004: Percentage: 0.26; 2005: Number of tax returns: 1,806.69; 2005: Percentage: 0.19; 2006: Number of tax returns: 1,712.74; 2006: Percentage: 0.18. Total; 2004: Number of tax returns: 1,988.22; 2004: Percentage: [A]; 2005: Number of tax returns: 1,382.43; 2005: Percentage: [A]; 2006: Number of tax returns: 1,306.17; 2006: Percentage: [A]. Source: GAO analysis of IRS data. [A] The values of these margins are less than 0.01: [End of table] [End of section] Appendix VII: SMR Research's Data on Home Equity Debt and How IRS Might Use Similar Data for Enforcement: Table 22 describes the three types of home equity loans--lump sum home equity loans, home equity lines of credit, and refinancings--that SMR Research included in its calculations of home equity loan debt that exceeded $100,000, the home equity debt limitation.[Footnote 28] Table 22: Types of Transactions Included in SMR Research's Home Equity Debt Calculations: Type of transaction: Lump sum home equity loans of greater than $100,000; Comment about the type of transaction: These loans exclude second mortgages taken out at the time of purchasing a house. Type of transaction: Home equity lines of credit greater than the amount on which the average amount borrowed would yield $100,000; Comment about the type of transaction: These lines of credit exclude second mortgages taken out at the time of purchasing a house. According to SMR Research, the average draw-down rate for home equity lines of credit was 50.16 percent at June 30, 2008. Based on that rate, the line of credit that would result in a draw-down amount more than the $100,000 limit for deductible interest would be $199,362. Type of transaction: Refinancings in which cash of more than $100,000 was taken out beyond the homeowner's original loan amount; Comment about the type of transaction: Because the $100,000 limitation for deducting interest on home equity debt refers in a refinancing to $100,000 more than the outstanding mortgage principal balance and because the original loan amount often exceeds the outstanding balance, SMR Research captured only some of the cash amounts taken out at refinancing that exceeded the limitation. SMR considered all of a homeowner's outstanding refinancings and compared the sum of them to the loan amount that the homeowner originally borrowed. Source: GAO analysis of SMR Research information. [End of table] Although SMR Research data do not include Social Security numbers useful in data matching, its data include (1) homeowner names and addresses; (2) loan dollar amounts useful for determining if loan limitations for deductible interest were exceeded; and (3) other information useful in audit selection or other compliance activities. An example of this other SMR Research information that might be useful for compliance purposes is information showing that more than a third of the cash-out refinancing dollar amounts and the home equity loan dollar amounts not known to be associated with lines of credit were in just 14 counties. Pinpointing homeowners with the largest home equity loans or who live in one of a few counties with taxpayers having large proportions of the nation's home equity loan dollar value might ease IRS's burden in matching large amounts of information to its own databases. However, we recognize that matching databases using names and addresses is much more difficult than using Social Security numbers and that its viability would have to be tested. For instance, IRS could run a name and address match for a particular geographic location and follow up with only those taxpayers whose information matched without extensive effort. If the test were to show a low return on investment or did not deserve a higher priority than competing efforts, it could be abandoned. IRS follow-up could be examinations, possibly targeted at geographic areas or taxpayers with the largest potential overdeductions; correspondence to taxpayers as described in an earlier section; or outreach to the tax return preparation community. Outreach to preparers could inform them that the home equity area is being scrutinized by IRS and, if taxpayers' returns had not properly considered the $100,000 or fair market value limitations, amended returns might be warranted. Non- examination efforts might be preferable to resource-intensive examinations because the amounts of possible extra tax per noncompliant taxpayer might be relatively small. For example, if a taxpayer with an 8-percent interest rate on a $200,000 home equity loan deducted interest on the entire loan, the overdeduction would be about $8,000, or about $2,000 in taxes for a taxpayer in the 25-percent bracket. [End of section] Appendix VIII: Comments from the Internal Revenue Service: Department Of The Treasury: Internal Revenue Service: Deputy Commissioner: Washington, D.C. 20224: July 22, 2009: Mr. Jim White: Director, Tax Issues: U.S. Government Accountability Office: 441 G Street, NW: Washington, DC 20548: Dear Mr. White: We have reviewed your draft report titled: Home Mortgage Interest Deduction: Despite Challenges Presented by Complex Tax Rules, IRS Could Enhance Enforcement and Guidance (Job Code 450675). We appreciate your comprehensive review of our systems and your analysis of our compliance efforts. As the report acknowledges, because we have little information on taxpayers' mortgage debts, taxpayer noncompliance with deduction limits cannot easily be detected. Indeed, the absence of specific information about the reasons for noncompliance also prevents us from making efficient decisions on how to select cases for examination. We also acknowledge that the complexity of mortgage- interest deduction rules creates problems for some taxpayers trying to comply. We agree there are opportunities for us to mitigate some of the problems identified in your report. To this end, we will revise the NRP case selection criteria, study the feasibility of revising Form 1098, and modify the Schedule A caution to reflect the limits in the mortgage interest deduction. We will also enhance our outreach to external stakeholders on this issue and update training materials to ensure common problems cited are addressed. We will review your suggestion regarding the use of commercial information. It should be noted that vendors cannot provide social security numbers which would be critical to the effective use of the data. This could constrain the benefits and affect any cost/benefit study. Responses to your specific recommendations are enclosed. We appreciate the continued and valuable support from you and your staff on this issue. If you have any questions, or if you would like to discuss this response in more detail, please contact Christopher Wagner, Commissioner, Small Business/Self-Employed Division at (202) 622-0600. Sincerely, Signed by: Linda E. Stiff: Enclosure: [End of letter] Enclosure: GAO Recommendations and IRS Responses to GAO Draft Report: Home Mortgage Interest Deduction: Despite Challenges Presented by Complex Tax Rules, IRS Could Enhance Enforcement and Guidance (Job Code 450675): Recommendation: Revise NRP's case selection system so that a tax return's mortgage interest deduction is not automatically excluded as an examination issue if it matches information from Form 1098 reports. Comments: We agree to revise the NRP case selection criteria to include consideration of the mortgage interest deduction limitations. Recommendation: Revise Form 1098 to require third parties to provide information on mortgage balances, the address of a home securing a mortgage, and an indicator of whether the mortgage is for a current year refinancing. Comments: We agree to study this issue since we do not have sufficient data to support such a revision at this time. Our Research staff will provide data regarding the volume and magnitude of noncompliance in this area. If the data supports a revision to Form 1098, we will implement the recommended change. Recommendation: Investigate whether using information from private sources would be productive in detecting mortgage interest noncompliance, especially for home equity debt identified. Comments: The IRS will investigate using commercial information to "detect" home mortgage/equity debt. However, the inability of vendors to provide social security numbers is critical to the effective use of the data. The lack of social security numbers could constrain the benefits and affect any cost/benefit study. Additionally, it should be noted that the information GAO recommends that IRS obtain from private sources is available through the information (internal/external) systems we currently utilize. Recommendation: Revise the wording on Schedule A to clearly state that the mortgage interest deduction is subject to limitations. Comments: We agree with this recommendation. We will replace the caution on the left side of Schedule A about personal interest with a caution about the limits on the mortgage interest deduction. Recommendation: Conduct a test to evaluate whether mortgage interest deduction-related outreach programs to taxpayers and tax return preparers could be a cost effective way to reduce noncompliance; outreach might include sending correspondence covering key rules and common mistakes or promoting seminars on common types of misreporting. Comments: The IRS already provides information regarding mortgage interest deduction through Publication 936 - Home Mortgage Interest Deduction, Tax Topics, the IRS website, stakeholder interaction, etc. We will enhance our outreach to external stakeholders, including financial industry partners, tax professional organizations, small business associations, etc., concerning this issue. SB/SE Communications, Liaison and Disclosure will also pursue the feasibility of a test, with the assistance of SB/SE Research, to evaluate whether outreach in this area would reduce noncompliance. Recommendation: Set a date to complete the Chief Counsel determination on whether the acquisition debt limit is $1 million or $1.1 million when used in combination with the home equity debt limit. Comments: The Associate Chief Counsel (Income Tax and Accounting) has set September 30, 2009, as the target date to issue guidance on this issue. Recommendation: Revise examiner training materials by adding examples cited as common problems by auditors and paid tax return preparers, such as those involving multiple homes or homebased businesses, and after the Chief Counsel's final determination on the acquisition limit, revise examiner training and the worksheet in guidance to reflect the project's outcome. Comments: Small Business/Self Employed Examination training materials will be updated to include additional examples; however; we can not cover all possible scenarios. We will ensure the two common problems cited will be addressed. Once Chief Counsel makes a final determination, the training material will be updated accordingly. Wage & Investment Discretionary Examination will also review and assess their examiner training materials to determine if any revisions are warranted. [End of section] Appendix IX: GAO Contact and Staff Acknowledgments: GAO Contact: James R. White, (202) 512-9110, whitej@gao.gov: Acknowledgments: In addition to the contact named above, Charlie Daniel, Assistant Director; Amy Bowser; Sara Daleski; Eric Gorman; Lawrence Korb; Karen O'Conor; Anne Stevens; and John Zombro made key contributions to this report. [End of section] Footnotes: [1] The most expensive income tax expenditures are reduced rates of tax on dividends and long-term capital gains (about $148 billion of forgone revenue) and the exclusion of employer contributions for health care, health insurance premiums, and long-term care insurance (about $127 billion). [2] We use the term "routine" to cover examinations that include the mortgage interest deduction but are not part of an IRS compliance initiative project that focused on it. [3] Pub. L. No. 99-514 (Oct. 22, 1986). [4] Omnibus Budget Reconciliation Act of 1987, Pub. L. No. 100-203 (Dec. 22, 1987). [5] According to information published in a 2009 Federal Reserve Bulletin, homeowners used about 40 percent of the equity borrowed from their homes in 2007 for home improvements. See Brian K. Bucks, Arthur B. Kennickell, Traci L. Mach, and Kevin B. Moore, "Changes in U.S. Family Finances from 2004 to 2007: Evidence from the Survey of Consumer Finances," Federal Reserve Bulletin, vol. 95, Feb. 12, 2009. [6] These conditions include whether the points were paid in connection with the purchase or improvement of the principal residence securing the mortgage, the payment of points was an established business practice in the area where the loan was made, and the amount paid for the points did not exceed the amount generally charged for points in that area. [7] For instance, the taxpayer might have paid home mortgage interest to the person selling the home, and that person did not have to supply a Form 1098. [8] Because of resource constraints and expected return on the work, IRS does not pursue all of the discrepancies it identifies. [9] About 64 percent of tax returns with a Schedule A, line 10 deduction were done by a paid preparer in 2006. The margin of error for this estimate was less than 1 percent. [10] Department of the Treasury, Internal Revenue Service, Publication 936: Home Mortgage Interest Deduction: For Use in Preparing 2008 Returns. [11] GAO, Tax Policy: Many Factors Contributed to the Growth in Home Equity Financing in the 1980s, [hyperlink, http://www.gao.gov/products/GAO/GGD-93-63] (Washington, D.C.: Mar. 25, 1993). [12] Paperwork burden is defined as the time spent reading and understanding a request for information, as well as the time spent developing, compiling, recording, reviewing, and providing that information. [13] This procedure would not be useful for determining whether taxpayers correctly deducted mortgage interest on a second home. [14] GAO, Tax Gap: Actions That Could Improve Rental Real Estate Reporting Compliance, [hyperlink, http://www.gao.gov/products/GAO-08-956] (Washington, D.C.: Aug. 28, 2008). [15] The mortgage banking industry representative also said that Form 1098 filers would need guidance to know how to report loan modifications in which loan terms are changed to avoid foreclosure and which are not considered refinancing by the industry but could be by IRS. [16] See appendix VII for the types of transactions that SMR Research included in its estimate. [17] It is unclear how many taxpayers are affected by the $1 million debt limit. [18] Although (1) a Schedule A note near line 10 says to "See page A-5" of the instructions, (2) Form 1098 has an asterisk saying that the Form's interest amount might not be fully deductible, and (3) most taxpayers might not be affected by the limitations, changing the wording on Schedule A could enhance compliance. [19] IRS could tell taxpayers that (1) a lack of knowledge about mortgage interest deduction limitations has come to its attention, (2) they should learn if the limitations apply to them and if they have been compliant, (3) they should amend returns if warranted, and (4) IRS will spot-check some taxpayers to ensure compliance. To be credible, IRS would actually need to check some returns. If this approach were considered too onerous on the public, communication with taxpayers and practitioners could be strictly prospective, providing guidance on what to consider when filing the next return. [20] Pau v. Commissioner, T.C. Memo 1997-43. [21] Taxpayers with home-based businesses also have complex returns when mortgage interest payment deductions appear partly on Schedule A and partly on a business return like Schedule C. [22] In 2007, over 39 million (or 28 percent) of the approximately 138 million individual income tax returns filed were prepared by individuals using tax software. [23] We limited our data analysis to the top three software companies- -TaxACT, TaxCut, and TurboTax--because they account for 88 percent of all returns filed electronically by individuals and accepted by IRS. We also reviewed several tax software packages for paid preparers. However, unlike the software for the general public, these programs are not designed to lead users through different calculations. Instead the preparer software is designed for those with professional expertise. [24] GAO, Tax Administration: Many Taxpayers Rely on Tax Software and IRS Needs to Assess Associated Risks, [hyperlink, http://www.gao.gov/products/GAO-09-297] (Washington, D.C.: Feb. 25, 2009). [25] The SOI analysis discussed here was not part of the SOI statistics presented in table 2, which includes data obtained from published SOI materials. [26] Form 8829, Expenses for Business Use of Your Home, also can be used to report a mortgage interest deduction, but mortgage interest reported on this form is not included in the SOI database and consequently was not part of our analysis. [27] Publication 936 also has an example of consistent principal payments per month, but we recognize that principal payments often differ from month to month. [28] Another source of information showing there are potentially large numbers of home equity loans above the $100,000 limit for deducting interest is the American Housing Survey for the United States: 2007, prepared by the U.S. Department of Housing and Urban Development and the U.S. Census Bureau. The Survey shows 197,000 owner-occupied housing units were refinanced in which owners received cash of $100,000 or more at settlement. These homeowners had situations similar to the third type of transactions in table 14--refinancings with more than $100,000 taken out. The Survey data compared the most recent loan refinancing to the previous loan amount, not to the original loan amount, if other refinancings had occurred. SMR Research's data account for the possibility that homeowners refinanced more than once, each time taking cash out and adding to their total home equity debt. [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 Web site [hyperlink, http://www.gao.gov]. Each weekday, GAO posts newly released reports, testimony, and correspondence on its Web site. To have GAO e-mail you a list of newly posted products every afternoon, 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 Web site, [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. To Report Fraud, Waste, and Abuse in Federal Programs: Contact: Web site: [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: Ralph Dawn, Managing Director, dawnr@gao.gov: (202) 512-4400: U.S. Government Accountability Office: 441 G Street NW, Room 7125: Washington, D.C. 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, D.C. 20548: