” : .“), ~ IJnikd States General Accounting Office ----. GAO Report to the Chairman, S$%committee on Commerce, Consumer, and Monetary : Affairs, Committee on Government Operations, House of Represerdatives - May 1990 TAX ADMINISTR,ATION Status of IRS’ Test of a Business Information Returns Program 141680 % * -* United States General Accounting Office Washington, D.C. 20548 General Government Division B-238196 May 29,199O The Honorable Doug Barnard, Jr. Chairman, Subcommittee on Commerce, Consumer, and Monetary Affairs Committee on Government Operations House of Representatives Dear Mr. Chairman: This is our second report in response to your May 6, 1987, request to monitor the Internal Revenue Service’s (IRS) plans, time frames, and accomplishments concerning its evaluation of a business document- matching program.’ Like the individual document-matching program, this program could be designed to computer-match businesses’ informa- tion returns to their income tax returns to identify those businesses that may have underreported their income (underreporters) or failed to file tax returns (nonfilers). Third-party sources, like banks, are required to file information returns with IRS to report payments, such as interest and dividends, that they made to individuals. IRSalso requires the filing of information returns on payments made to sole proprietorships and partnerships but does not require them on payments made to most cor- porations. In 1988, IRSreceived about 31 million information returns on payments made to businesses, including corporations, even though these returns were not required. As discussed in our first report, IRS officials recognized that more could Background be done to test the feasibility of a business document-matching program and developed an action plan to do so. The plan contains three initia- tives scheduled for completion by December 31, 1990. IRS plans to (1) develop a cross-reference file to help associate sole proprietors’ informa- tion returns with their tax returns, (2) evaluate the merits of using information returns to identify partnership and corporate nonfilers, and (3) determine the costs and benefits of a document-matching program that could be established if information returns are required for pay- ments made to corporations. This report discusses the progress IRS had made as of September 30, 1989, in completing these action plan initiatives. Page 1 GAO/GGD-9058 Information Returns Program IRS met all of the scheduled completion dates for the first initiative con- Results in Brief cerning sole proprietors. It was 6 months late in completing the second initiative on partnership and corporate nonfilers. Under the third initia- tive, IRSwas 7 months late in completing work on an interim milestone for the computer costs of a corporate document-matching program. Dur- ing the Spring of 1990, IRS expects to issue interim benefit studies, which were due in December 1989. IRS officials said resource constraints and higher priority work contributed to these delays but that this slippage would not affect the December 31, 1990, action plan completion date. Under the first initiative, IRS successfully tested a cross-reference file that will allow IRS to associate sole proprietors’ information returns filed under their employer identification numbers (EIN) with income tax returns filed under their Social Security numbers (SSN). Beginning in 1990, the file will allow IRSto use information returns to identify sole proprietors who may have underreported their income or failed to file income tax returns. Under the second initiative, IRS concluded from its study of 1,610 poten- tial nonfilers that field contact with potential business nonfilers identi- fied with information returns was not cost effective because it obtained only 16 delinquent tax returns per 100 cases. In comparison, IRS obtained 32 returns per 100 cases in its business nonfiler program, known as the stopfiler program. IRS’ study should not have included cases for which it had not completed its investigations of potential nonfilers or for which it had data showing that the business had filed a return. Adjusting the results to exclude these cases yields a higher rate of 22 returns per 100 cases. Accordingly, we concluded that using infor- mation returns to identify potential business nonfilers would comple- ment IRS’ stopfiler program, which does not use information returns. In written comments on our draft report, the Commissioner of Internal Revenue said IRS will continue to explore ways to use information returns to identify business nonfilers and will test the productivity of sending correspondence to potential business nonfilers that IRS identified by using information returns. Under the third initiative, IRS estimated that the computer costs for per- sonnel, space, equipment, and operations would be $166 million over 7 years. These costs are in line with the costs of IRS’ individual document- matching program. IRS is still examining corporate tax returns to obtain the data needed to estimate program benefits. IRS plans to complete its Page 2 GAO/GGD-90-38 Information Returns Program costs and benefits studies on the feasibility of a corporate document- matching program by December 31,199O. Sole proprietors report business income on Schedule C, Profit or Loss Initiative 1: From Business, and/or Schedule F, Farm Income and Expenses. They Associating Sole attach the schedules to their Form 1040, US. Individual Income Tax Proprietors’ Return, which is filed under their SSN. IRSalso receives information returns from third parties, such as banks, to report payments to sole Information Returns proprietors, either under the sole proprietors’ EIN or SSN. As discussed in With Tax Returns our prior report, IRSmet its January 1,1988, milestone for developing a workable cross-reference file so that information returns received under sole proprietors’ EINS could be matched with their tax returns. We tested that file by reviewing 96 sole proprietor returns and found that all the SSNS and EINS associated with these returns were accurately reflected on the file. Examination officials advised us that IRS will use this cross-ref- erence file in 1990 to identify sole proprietors who may have underre- ported their income or failed to file tax returns for tax year 1988. does not use information returns in its current program to identify Initiative 2: Using IRS potential partnership and corporate nonfilers. This program-the Information Returns stopfiler program-only identifies potential nonfilers when these types to Identify of businesses file tax returns in 1 year but not the next year or when businesses receive EINS but do not file required returns. IRS’ stopfiler pro- Partnership and gram does not identify all businesses that never filed a required return, Corporate Nonfilers or may not identify businesses whose filing requirements are not prop- erly reflected on IRS’ Business Master File-the file on which all busi- nesses’ tax return requirements and transactions are maintained. IRS could better identify these potential nonfilers if it used information returns to supplement its current stopfiler program. On June 16, 1989,6 months later than scheduled, IRS’ Collection Division completed its test of using information returns to identify businesses that did not file tax returns. The test was scheduled to be completed on December 31,1988, but was delayed due to changes, some of which we suggested, in the criteria for selecting sample cases, and problems in writing computer programs for the test. Consequently, the nonfiler investigations were not started until December 1988. The test used a sample of 1,6 10 potential business nonfilers in nine district offices: Atlanta, Cincinnati, Chicago, Cleveland, Des Moines, Jacksonville, Los Angeles, San Francisco, and St. Louis. Page 3 GAO/GGD3@38 Information Returm Program To select the sample cases, IRS matched the calendar year 1986 informa- tion returns it had received from third parties to its Business Master File records. Potential nonfiler cases were created when the records showed that the business had not filed a required 1986 tax return. After the sample cases were selected, field staff (i.e. revenue officers) were to con- tact the businesses to determine whether they were required to file tax returns. From the test, IRS Collection officials initially concluded that information returns should not be used to identify potential business nonfilers. IRS gave three reasons to support this conclusion: (1) the Collection Division did not obtain many delinquent returns from the test and, consequently, believed computer-matching information returns with tax returns would not be cost effective; (2) the Business Master File contains inaccurate information, like businesses having multiple EINS or the wrong filing requirements, which creates unproductive business nonfiler cases; and (3) because the law does not require. information return reporting for payments to corporations, document matching cannot be done for all businesses. From our review of IRS’ test, we concluded that even with the impedi- ments, information returns can be used to identify potential business nonfilers. Our review showed that IRS should not have used certain cases in its test. If the cases were eliminated, the test results would compare more favorably to the results IRS gets from its stopfiler program. Also, we believe that inaccuracies in the Business Master File and the fact that information returns are not required for payments to corporations should not prevent IRS from starting to use the information returns it receives to identify potential business nonfilers. IRS’test resulted in getting 262 delinquent tax returns from the 1,610 sample cases. IRS Collection Division officials believed this yield was too marginal, both in the number of returns obtained and tax dollars col- lected, to warrant using information returns to identify business nonfilers. Our review of the 1,610 test cases showed that 394, or 24 percent of the cases, should have been eliminated from IRS’ test results. Specifically, the test results should not have included l 266 cases that were selected for investigation, because the test’s criteria did not screen out all businesses that had filed tax returns on a fiscal year basis; and Page 4 GAO/GGD-90-38 Information Returns Program B238196 . 128 cases that were still being investigated by IRS when the test results were completed. By basing its test on the remaining 1,216 cases (1,610 minus 394 cases), IRS would have more accurately assessed the usefulness of information returns because it had completed the investigations on these cases and was able to establish whether the businesses were nonfilers. The 262 delinquent tax returns that were obtained from these 1,216 cases show that information returns can be used for identifying potential business nonfilers. Obtaining 22 delinquent returns for every 100 businesses would supplement IRS’ current business stopfiler program in which IRS obtains 32 delinquent returns for every 100 businesses. Moreover, IRScan improve its results as it eliminates inaccurate Master File data, such as the wrong business filing requirements or multiple EINS, that create unproductive cases. IRS’ actions to correct these inaccu- rate data are based, in part, on recommendations that we made in a May 1988 report on the stopfiler program.2 We found that in about 16 per- cent of the stopfiler cases we reviewed, IRS had not updated the Master File with the results of its stopfiler investigations. We also found that in 16 percent of the stopfiler cases we reviewed, IRS had erroneously issued multiple EINS to the businesses. We recommended that IRS (1) use the results of stopfiler investigations to correct inaccurate filing require- ments on its Business Master File and (2) better research its files to iden- tify businesses with previously issued EINS and better control the issuance of new EINS so that businesses will not be issued more than one. IRS agreed with our recommendations and is taking actions that should reduce the number of unproductive business stopfiler cases that would be generated by matching information returns to the Business Master File. Because information returns are not required on payments made to cor- porations, IRS will be limited in the extent to which it can use computer- matching to identify potential business nonfilers and underreporters. This legal impediment could be overcome with legislation, such as the bill you introduced in May 1988, to require information return filing for interest, dividends, and other income paid to corporations. However, the lack of this requirement should not prevent IRS from starting to use the estimated 31 million information returns filed on businesses for calendar year 1988. ‘TAX ADMINISTRATION: IRS Could Reduce the Number of Unproductive Business Nonfiler Investi- gations (GAO/GGD -88 -77 , May 24,1988). Page 5 GAO/GGD90-38 Information Retums Program ” d - B-238195 We discussed our observations with Collection officials. They agreed that (1) their test results included cases that should not have been used to evaluate the usefulness of information returns, (2) they need to con- tinue to correct the Business Master File, and (3) the information returns currently received could be used to identify businesses that failed to file tax returns. However, these officials said IRS has experi- enced serious budget shortfalls for fiscal years 1989 and 1990, and may not be able to pursue this or any other new program initiative right now. Nevertheless, IRS has agreed to explore ways to use information returns to complement its current program for identifying potential business nonfilers. IRS will test the productivity of sending correspondence to those who appear, on the basis of information returns, to be nonfilers. IRS Examination officials assured us that they will continue to develop Initiative 3: the costs and benefits estimates for a corporate document-matching pro- Determining Costs and gram. They said the estimates would be based on information returns Benefits of a being filed on payments made to corporations, The officials anticipated that they will be able to complete these estimates, in spite of budget Document-Matching shortfalls, because the cost estimates are not going to take much more Promam for money and the benefit estimates are to be based on examinations that will still be done. Corsorations Under this corporate document-matching program initiative, IRS is doing one study to determine the potential costs and three studies to deter- mine the potential benefits of such a program. These studies cover cor- porations with assets of (1) $100 million and over, (2) $10 million to $100 million, and (3) under $10 million. These costs and benefits studies are scheduled to be completed by December 31, 1990. IRS completed its estimates of the various computer costs on August 3, 1989,-7 months later than the scheduled December 3 1,1988, comple- tion date. According to IRS officials, this delay was due to staff being diverted to handle unexpected and urgent problems that arose in other programs. IRS estimated that the computer costs for personnel, space, equipment, and operations would be about $166 million, including $66 million in start-up costs, over 7 years- from fiscal year 1991 through fiscal year Page 6 GAO/GGD9038 Information Returns Program i B-229195 1997.” IRS based these costs on the assumption that this computer- matching program would be implemented in fiscal year 1994, starting with the processing of information returns for tax year 1993. The cost estimates also assumed that legislation would be passed in 1991 to require information returns reporting on interest, dividends, rents, roy- alties, and capital gain income earned by corporations. IRS estimates that if all payers of this income comply with the reporting requirements, it will receive an additional 216 million business information returns. IRS currently receives about 1 billion information returns for individuals and sole proprietors. IRS estimated that for fiscal year 1995-the first full year of program operations-these computer costs will be about $22 million. These costs include all relevant computer-matching cost elements, like computer equipment and personnel costs, that IRS incurs in the individual docu- ment-matching program. About $9.5 million of the $22 million estimate is for program enhancements, like automating the underreporter pro- gram process, that are not included in IRS’ individual computer-matching program costs. Factoring out these enhancements shows that the com- puter costs are in line with the costs of the individual program. IRS offi- cials estimate that computer costs for the individual document-matching program in tax year 1984-the most current year for which cost figures are available-were about $14.5 million. IRSis scheduled to complete its final cost estimates for the entire corpo- rate document-matching program by December 31, 1990. IRS officials said they expect to meet this deadline. They also said they would be meeting with representatives from the securities and banking industries to obtain additional data and views on the impacts on these industries if information returns were required on payments made to corporations. IRSExamination officials said they would consider getting the views of representatives of corporations and the tax preparer community on any difficulties in complying with tax return changes that would be needed for a document-matching program. For example, taxpayers filing on a fiscal year basis may have to complete a schedule that reconciles their calendar year information return income to the same income on their fiscal year tax returns. “According to IRS officials, these costs do not account for IRS’ Tax System Modemlzation effort, which is scheduled for the 1990s. This effort’s potential impacts on the computer costs are not known. Page 7 GAO/GGD-99-38 Information Returns Program B-238196 To measure benefits, IRS is doing two studies of corporations with assets of $10 million and over-one on those with assets of $10 million to $100 million and the other on those with assets of $100 million and over. These studies are using returns from IRS’ ongoing corporate examina- tions. IRS is treating these as regular examinations, except that the reve- nue agents have to determine whether the taxpayers reported the income shown on information returns that third-party sources had vol- untarily filed. The amount of unreported income the revenue agents identify from reviewing the information returns in these examinations will be used to estimate the benefits of a corporate document-matching program. According to the action plan, IRS was to submit a final report on the results of its examination of corporations with $10 million to $100 mil- lion in assets and an interim report on corporations with assets of $100 million and over by December 31, 1989. IRS officials said these reports will not be finished until the Spring of 1990 because of problems with the computer databases used in these studies. IRS officials said the other benefit study, which required an interim report by December 31,1989, and involves tax returns from about 13,000 corporations with assets of less than $10 million, is also expected to be finished during the Spring of 1990. This study is based on IRS’ peri- odic Taxpayer Compliance Measurement Program-a program through which IRS estimates taxpayers’ compliance with the tax laws by thor- oughly examining a statistical sample of tax returns. To monitor IRS’progress on the benefit studies, we reviewed 100 cases to determine how well the revenue agents documented their examinations to show whether the corporations reported the income shown on their information returns. We found that 12 of the 50 cases (24 percent) with assets of $10 million and over, and 32 of the 50 cases (64 percent) with assets of less than $10 million had no information returns associated with the cases being examined. Since these benefit studies are to iden- tify noncompliance by using information returns, we suggested to IRS Examination officials that they analyze cases that have no information returns separately from cases that have information returns. The offi- cials agreed to do a separate benefit analysis on cases with information returns. They also said they are gathering data on the amount of unre- ported income that revenue agents found in cases for which the agents had no information returns. As a result, these officials said they will be able to analyze these cases to determine how much unreported income Page 8 GAO/GGD-90-38 Information Returns Program r ” B238195 could have been detected if they had information returns on this unre- ported income. To monitor IRS’ progress in developing a business document-matching Objective, Scope,and program, we Methodology . interviewed IRS officials responsible for the plan’s three initiatives; . reviewed a judgmentally selected sample of 96 of the 369 sole proprie- tors’ income tax returns that IRSused in testing the accuracy of its file for cross-referencing sole proprietors’ EINS and SSNS, to determine if IRS’ evaluation of the test results was accurate; . reviewed 1,610 business nonfiler investigations that IRSstudied in decid- ing on the merits of using information returns for identifying business nonfilers, to determine whether the results supported IRS’ conclusions; . reviewed IRS estimates of the potential computer costs for a corporate document-matching program and compared these estimates to the costs of IRS’individual document-matching program; and . analyzed 100 closed corporate examination cases that IRS is using to esti- mate the benefits of a corporate document-matching program, to deter- mine whether the examinations followed up on all information return income. We did our work from April through September 1989 at the IRS National Office in Washington, DC., the Fresno Service Center, and the San Fran- cisco District Office using generally accepted government auditing standards. The Commissioner of Internal Revenue commented on a draft of our Agency Comments and report in an April 6, 1990, letter, in which he stated his general agree- Our Evaluation ment with its content. He said IRS will continue to explore how informa- tion returns can be used to identify business nonfilers. He said that because of concerns about the costs and benefits of assigning field staff to contact potential business nonfilers, IRS first will test whether attempting to contact nonfilers by correspondence would be productive. Further, he proposed that we change our title to clarify that IRS is still testing the corporate document-matching program. He said any deci- sions on establishing the program will not be made until IRS completes its cost-benefit analysis of document matching for corporate information returns. The report has been updated to reflect these comments. Page 9 GAO/GGD90-38 Information Returns Program I II iK ‘-, / / *- B-238196 We support IRS’proposed test of corresponding with potential business nonfilers, We look forward to working with agency staff responsible for designing the test and for exploring ways that information returns could supplement IRS’current program to identify certain business nonfilers. As agreed, unless you publicly announce its contents earlier, we plan no further distribution of this report until 30 days from the issuance date. At that time, we will send copies to IRS,interested congressional commit- tees, and other interested parties upon request. We will continue to monitor IRS’ progress in assessing the feasibility of a business document-matching program and will keep the Subcommittee apprised of the status of IRS’ action plan initiatives. Major contributors to this report are listed in appendix II. If you have any questions, please contact me on 272-7904. Sincerely yours, Paul L. Posner Associate Director, Tax Policy and Administration Issues Page 10 GAO/GGDW38 Information Retums Program Page 11 GAO/GGD9@3S Information Returns Program r L %> - Contents Letter Appendix I Comments From the Internal Revenue Service Appendix II 16 Major Contributors to General Government Division, Washington, D.C. 15 San Francisco Regional Office 15 This Report Abbreviations EIN employer identification number IRS Internal Revenue Service SSN Social Security number Page 12 GAO/GGD90-38 Information Returns Program Y Page 13 GAO/GGD-90-38 Information Returns Program f Appendix I ,1/. t CommentsFrom the Internd RevenueService ’ DEPARTMENT OF THE TREASURY INTERNAL REVENUE SERVICE WASHINGTON, D.C. 20224 Mr. Richard L. Fogel Assistant Comptroller General United States General Accounting Office Washington, D.C. 20548 Dear Mr. Fogel : We have reviewed your recent draft report entitled, “Tax Administration: Status of IRS’ Efforts to Establish a Business Information Returns Program” and are in general agreement with the report’ 8 content. Regarding the feasibility of a non-filer program using voluntarily filed information returns pertaining to businesses, we will continue to explore how these returns can be used. Current information indicates that assignment of these cases for field contact may not be cost beneficial. However, we plan to redesign the test in order to attempt to contact non-filers by correspondence. We would also propose that the title of the report be changed to reflect that we are testing rather than establishing a Business Information Returns Program. Until we have completed the cost/benefit analysis of a corporate document matching program, we cannot determine whether it is feasible to establish such a program. Thank you for the opportunity to provide our comments. Best regards. Sincerely, Fred T. Goldberg, Page 14 GAO/GGD-99-38 Information Retums Program Appendix II lV!fajorContributors to This Report General overnment Gerald Stankosky, Assistant Director, Tax Policy and Administration Issues Division. Washington, Mark J. Gillen, Assignment Manager Tom Short, Evaluator Ralph T. Block, Evaluator-in-Charge San Francisco Kathleen E. Seymour, Evaluator Regional Off ice Y (268402) Page 15 GAO/GGDSO-SS Iniormath iMum Prognm h = C. c e. S = 5 e = ‘=a. . . . -. ‘I / - -- First,-(Iass Mail It’ostagt~ XL E’ws Paid GAO Permit, No. GlOO
Tax Administration: Status of IRS' Test of a Business Information Returns Program
Published by the Government Accountability Office on 1990-05-29.
Below is a raw (and likely hideous) rendition of the original report. (PDF)