Executive Summary

Objective and Scope

Background

Results

Conclusion

Appendix I - Detailed Objective, Scope, and Methodology

Appendix II - Major Contributors to This Report

Appendix III - Report Distribution List

Appendix IV - Management's Response to the Draft Report

Executive Summary

Corporations with assets of $10 million or more voluntarily report billions of dollars in taxes each year. To determine whether corporations comply with tax laws, the Internal Revenue Service (IRS) examines these large corporations under one of two programs.

Approximately 1,700 of the largest and most complex corporations are selected for examination under the Coordinated Examination Program (CEP). In the CEP, the IRS uses a scoring system to select returns for examination. Corporate returns selected for the CEP generally report assets that exceed $250 million and are examined by a team of IRS revenue agents.

This review focused on the approximately 56,000 other large corporations that may be selected for examination under the IRS’ General Examination Program. Unlike the scoring system and team approach used in the CEP, these returns are generally selected for examination using a screening system conducted by experienced revenue agents. The returns are then assigned to other agents for examination.

Approximately one out of every three large corporate returns examined in the general program in the last several years was closed without any additional recommended tax. Consequently, the IRS has been studying ways to improve its selection process. The objective of our audit was to evaluate these IRS efforts to improve its methods for selecting large corporate returns for examination.

Examinations that are closed without any additional recommended taxes can result in the ineffective and inefficient use of both IRS and corporate resources. The IRS believes that its current study efforts could lead to assessing an additional billion dollars or more in taxes each year. While we agree that some improvements in the return selection process may result from these studies, we also identified various problems in how the studies were conducted which indicate that results on this scale may not be realized.

Further, other problems concerning audit results will not be addressed by the results of these studies. These other problems involve conducting examinations with the most recently filed return and better preparing revenue agents to conduct large corporate examinations.

Since 1995, the IRS has been studying ways to improve methods to select large corporate returns for examination. During this period, the IRS completed a Net Operating Loss (NOL) Study that expended approximately 34,000 hours (over $1 million in salaries) examining corporations that reported a NOL deduction. In addition, the IRS is now in the second of four phases of the Discriminate Analysis System (DAS) Study. The DAS Study involves developing and testing a complex mathematical formula to identify more productive returns for examination. The Examination function believes its efforts could lead to assessing a billion dollars or more in taxes. However, various study methodological problems raise questions about whether this much revenue will be realized and whether the examinations associated with the ongoing DAS Study will successfully validate the DAS scoring system.

During Fiscal Years (FY) 1998 and 1999, examinations of large corporate returns were started, on average, 388 days after the returns were available to be examined. These delays in starting examinations, combined with legal and program deadlines, can add pressure to close cases before issues are identified and/or fully developed.

In FYs 1998 and 1999, 47 percent of the large corporate examinations included 2 or more returns. In 60 percent of these multi-year examinations, the corporation had already filed the next return that was due before the examination on the first return filed was started. If the Examination function audited the most recently filed return, there could be more time to either identify and develop issues or examine other corporations. Starting examinations with the most recently filed return could also further reduce the burden on corporate taxpayers that are subjected to multi-year examinations.

The Large and Mid-Size Business Division’s (LMSB) vision is to improve service to the approximately 210,000 corporations and partnerships it serves through a highly skilled staff of experienced revenue agents trained in unique industry tax accounting practices and issues. Industry specific skills will provide corporate customers with better customer service and can allow for more efficient and effective audits.

However, although the IRS provided the 231 revenue agents in our universe with 64,139 hours of training between 1997 and 1999, more training could have been given to these revenue agents on industry issues and accounting practices. Specifically, less than half of these training hours (26,265) dealt with business income tax practices. Moreover, the 26,265 hours that were devoted to business income tax practices included only 1,991 hours of specific industry-related topics.

In the near term, the Commissioner, LMSB, should delay the next phase of the DAS Study until certain problems with the study can be resolved. The LMSB Division should also encourage revenue agents to initiate more examinations on the most recently filed returns instead of ones that have been in the examination stream for extended periods of time.

In the long term, we recommend that the LMSB Division develop a consistent approach and provide guidance to its staff when studying ways to improve work processes. The LMSB Division should also ensure revenue agents receive additional training about the unique accounting practices and issues in the industries in which they are working.

Management’s Response: LMSB Division management agreed with our recommendation and delayed the next phase of the DAS Study. They are currently in the process of developing an action plan to validate the DAS formulas that will include the use of outside academic experts. LMSB Division management has directed field personnel to select returns for examination that are currently in the service centers which are the most recently filed returns.

In addition, LMSB Division management believes that under the new organization configuration their Office of Strategy, Research and Planning will be able to establish a consistent approach for evaluating and improving the Division’s workload selection process. The LMSB organization structure concentrates its technical expertise around business segments, which will allow it to improve its ability to offer training in industry-specific accounting practices and issues.

Management’s complete response to the draft report is included as Appendix IV.

Objective and Scope

Our objective was to evaluate the Internal Revenue Service’s (IRS) efforts to improve workload selection methods for large corporate examinations. The audit was performed in accordance with Government Auditing Standards from October 1999 through May 2000. On-site tests were performed in the Illinois, Los Angeles, Manhattan, New England, New Jersey and Southern California Districts. The audit tests were focused in the following areas:

• Assessed the status and initial impact of the IRS’ ongoing changes to improve workload selection methods.
• Reviewed the current approach in six districts for selecting corporate returns for audit.
• Evaluated how well examiners were prepared to audit large corporate returns.

Appendix I outlines the detailed objective, scope, and methodology of this review. Major contributors to this report are listed in Appendix II.

Background

The nation’s largest corporations with assets of $10 million or more voluntarily report billions of dollars in taxes each year. To determine whether they report all the taxes they owe, the IRS examines large corporations under one of two programs.

Approximately 1,700 of the largest and most complex corporations are selected for examination under the Coordinated Examination Program (CEP). In the CEP, the IRS uses a scoring system to select returns for examination. Corporate returns selected for the CEP generally report assets that exceed $250 million and are examined by a team of IRS revenue agents.

This review focused on the approximately 56,000 other large corporations that may be selected for examination under the IRS’ General Examination Program. Unlike the scoring system and team approach used in the CEP, these returns are generally selected for examination using a screening system conducted by experienced revenue agents. The returns are then assigned to other agents for examination.

Approximately one out of every three large corporate returns examined in the general program in the last several years was closed without any additional recommended tax. Consequently, the IRS has been studying ways to improve its selection process.

Results

Examinations that are closed without any additional recommended taxes can result in ineffective and inefficient use of both IRS and corporate resources. The IRS believes that its current study efforts could lead to assessing an additional billion dollars or more in taxes each year. While we agree that some improvements in the return selection process may result from these studies, we also identified various problems in how the studies were conducted which indicate that the intended results may not be realized.

Further, other problems concerning audit results will not be addressed by the results of these studies. These other problems involve conducting examinations with the most recently filed return and better preparing revenue agents to conduct large corporate examinations.

Studies to Improve the Selection Process for Large Corporate Returns May Not Yield Expected Results

Since 1995, the IRS has been studying ways to improve methods to select large corporate returns for examination. During this period, the IRS completed a Net Operating Loss (NOL) Study that expended approximately 34,000 hours (over $1 million in salaries) examining corporations that reported a NOL deduction. In addition, the IRS is now in the second of four phases of the Discriminate Analysis System (DAS) Study. The DAS Study involves developing and testing a complex mathematical formula to identify more productive returns for examination.

The Examination function believes its efforts could lead to assessing a billion dollars or more in taxes. However, various study methodological problems raise questions about whether this much revenue will be realized. Likewise, the examinations associated with the ongoing DAS Study may not successfully validate the DAS scoring system.

The General Accounting Office’s (GAO) Business Process Reengineering Assessment Guide provides project management concepts that can apply to studies. The concepts include:

• Establishing an executive steering committee and project sponsor.
• Forming a qualified, trained, well-led team.
• Establishing a clear team charter that defines goals, resources, constraints, and deliverables.
• Selecting and following a methodology to guide the project.

Our review of the studies identified the following problems.

Estimated milestones and completion dates were not consistently developed and communicated to study participants (DAS Study). An effective performance monitoring process was not in place to help ensure resources were available and/or delivered on time. We obtained the listings of the Phase II DAS test returns containing 545 returns sent to the Illinois, Los Angeles, Manhattan, New England and Southern California districts. Sixty-six percent of the returns (358 of 545) we reviewed in Phase II of DAS were not delivered to the districts in time to ensure examinations could be completed within the 27-month window specified in the IRS guidelines.

The design of sampling plans limits the ability to rely on study results (NOL Study and DAS Study). Fundamental changes to IRS work processes may be made based on relatively small samples of examinations that were neither randomly drawn nor statistically valid. For example, Phase I of DAS called for examining a sample of 1,200 returns during Fiscal Year (FY) 1998 from across the nation. However, a year later, only 164 of the returns were closed. Discussions with IRS officials indicated that most of the 1,200 returns were never examined as part of the study.

Important estimated cost information was not consistently documented (NOL Study and DAS Study). Adequate consideration was not given to the cost and feasibility of implementing recommendations. For example, one recommendation in the NOL Study involves expanding the number of corporate return line items that are transcribed by hand into IRS computers. At an estimated cost of over $6.5 million for programming changes, and considering the ongoing IRS move towards electronic filing, the expense may be cost-prohibitive.

Estimated increases in revenue were misleading (NOL Study and DAS Study). Because the impact on revenue did not consider the outcome of disputed tax adjustments, the actual tax that could get assessed is likely overstated. For example, 21 percent (49 of 234) of returns examined in the NOL Study had disputes that were not resolved at the examination level, but instead, were settled later in the IRS’ Office of Appeals. As a result of Appeals’ settlements, the study results were overstated by $21 million, which represents 68 percent of the total dollars assessed ($31 million).

Impact on compliance was not accurately measured (NOL Study and DAS Study). Although an important outcome of an examination is to change a non-compliant taxpayer into a compliant one, there were no plans to monitor subsequent compliance of the studies’ taxpayers. We analyzed the subsequent year’s filing for the taxpayers included in the NOL Study and found that 12 percent of the taxpayers reported in subsequent years a NOL deduction that was equal to or greater than one that was disallowed during the NOL Study.

Our evaluation of these studies indicates that the fundamental reason these problems occurred is because the Examination function did not have formal guidelines to follow when studying ways to improve its work processes. As a result, important components such as reliable sampling plans, expected costs and benefits, milestones and completion dates, and measures of success were not consistently developed. Reports issued by the GAO in recent years have discussed the importance of developing these components in federal government programs.

We met with LMSB Division officials on May 18, 2000, and learned that the Division has begun managing many of the day-to-day operations that were previously handled by the Examination function – including the DAS Study. The LMSB Division is part of the overall IRS effort to modernize into four operating divisions and has full responsibility for serving the large corporate taxpayer.

During our meeting, LMSB Division officials recognized the problems with the DAS Study and indicated a need to use outside experts to validate the DAS scoring system, as well as a quality review process for the examinations. Officials also pointed out the significant challenges they face in managing multiple priorities related to the Division’s stand up. These priorities include managing stay-in business programs, such as ongoing taxpayer examinations, working on near term improvements in operations, and laying the foundation for future improvements.

Continuing the study under these circumstances would not yield reliable results. Consequently, the costs to continue the study, estimated at $11.3 million in salaries and benefits, would not represent funds well spent by the IRS.

The Commissioner, LMSB, should:

1. In the near term, develop and implement an action plan to ensure that DAS Study returns are timely delivered for examination, and once delivered, more returns are examined. The plan should provide for an outside expert to validate the DAS formulas.
2. Delay conducting additional DAS examinations until the action plan is prepared and approved.
3. In the long term, establish a structured framework for evaluating ways of improving work processes. The framework should provide guidance for designing reliable sampling plans, measuring expected costs and benefits, establishing milestones and completion dates, and measuring results.

Management’s Response: LMSB Division management agreed with all the above recommendations. It is currently in the process of developing an action plan to ensure that DAS scored returns are delivered timely and that the DAS formulas are properly validated using outside experts from academia.

The Division followed our recommendation and delayed the next phase of the DAS Study, and has directed the field to begin selecting returns currently in the service centers. These returns are not DAS scored, but are the most current returns.

In the IRS’ new organizational framework the LMSB Division believes its Office of Strategy, Research and Planning will be able to establish a consistent approach for evaluating and improving the Division’s workload selection process.

Large Corporate Examinations Could Be Started With the Most Recently Filed Return

Once a corporate return is filed, IRS examiners are required by law to complete their examination within a 3-year period. To assist examiners in meeting this responsibility, the IRS’ guidelines recommend that examiners take no more than 27 months to complete corporate examinations. These deadlines, combined with other delays in starting examinations, can add pressure to close cases prematurely, thereby contributing to the number of examinations that are closed without any additional recommended taxes.

To determine how timely examinations are started, we analyzed 12,294 non-CEP corporate examinations with assets of $10 million or greater that were closed in FYs 1998 and 1999. We found that examinations of large corporate returns were started, on average, 388 days after the returns were available to be examined.

Delays can be caused by several factors. First, a significant number of large corporate income tax returns spend, on average, 2 1/2 months in the Statistics of Income (SOI) Program before being examined. While in SOI, return data is input into databases and eventually used by governmental and private entities for analytical purposes. During this time, the returns are unavailable for examination.

Resource constraints can also contribute to delays in starting examinations. Due to the complexity of large corporate issues, the IRS assigns large corporate returns to its most experienced revenue agents. However, they can be involved in ongoing examinations and not readily available to start a new examination. As a result, returns can sit in the unassigned inventory for extended periods of time.

Prior to our review, LMSB Division officials had recognized a need to start large corporate examinations faster and developed a comprehensive plan for testing an electronic filing system for corporate returns. Division officials believe the system will eliminate the need to send returns through the SOI, thereby helping to expedite the start of examinations. They also believe the system could help overcome some of the resource constraints they face by enhancing their ability to address more issues without actually conducting an examination, which could help eliminate more compliant taxpayers before they enter the examination stream.

Until the new electronic filing system is available, an interim solution lies with starting examinations on the most recently filed return. In FYs 1998 and 1999, almost half (47 percent) of the large corporate examinations covered two or more returns. In 60 percent of these multi-year examinations, the corporation had already filed the next return that was due before the examination on the first return filed was started.

Had the compliance problems been addressed on the most recently filed return, there would have been approximately 240,287 more hours ($9.9 million in opportunity costs) with which to identify and develop issues or expand examination coverage to other corporations. Starting examinations on the most recently filed return could also further reduce the burden of multi-year examinations on approximately 1,328 corporate taxpayers.

4. The Commissioner, LMSB, should encourage revenue agents to start more examinations with the last return filed instead of ones that have been in the examination stream for an extended period of time. This could be accomplished through a memorandum to revenue agents and team managers emphasizing the benefits of addressing compliance problems on the most recently filed return.

Management’s Response: LMSB Division management agrees with our recommendation and has asked the field to select the returns currently in the service centers, which are the most recently filed returns.

Revenue Agents Could Be Better Prepared to Conduct Large Corporate Examinations

The LMSB Division envisions having experienced revenue agents trained in unique industry tax accounting practices and issues, and believes they are best equipped to conduct examinations of large corporations. Industry-specific agent skills provide corporate customers with better customer service and can allow for more efficient and effective audits.

We found that experienced revenue agents (GS-13 level) generally conducted large corporate audits. However, the agents in the reviewed universe had received relatively little training on industry-related tax accounting practices and issues. Instead, the agents’ training was focused more on recognizing and developing a broader range of tax issues such as those affecting individuals. Unlike revenue agents in the CEP, the agents in the reviewed universe generally did not continually examine large corporate returns, but were also responsible for examining other types of returns.

The 231 revenue agents from the 6 districts we reviewed received 64,139 hours of training between 1997 and 1999. However, less than half of the training hours (26,265 of 64,139 hours) addressed business income tax practices, and only 1,991 (3 percent) of the total training hours were focused on industry-related issues.

While there is no specific requirement for a certain level of industry-specific training, amounts exceeding 3 percent over a 36-month period are needed to meet the LMSB Division’s goals. The LMSB Division envisions improving service to some 210,000 corporations and partnerships through a highly skilled staff, and industry specialization training would accelerate the Division’s ability to meet this vision.

5. The Commissioner, LMSB, should ensure revenue agents receive additional training about industry-specific accounting practices and issues in which they are working.

Management’s Response: The LMSB Division agrees with our recommendation. Its current Continuing Professional Education efforts are tailored around the business segment concept and contain industry-specific topics. In addition, other courses are in the development stages.

Conclusion

Examinations that are closed without any additional recommended taxes can result in ineffective and inefficient use of both IRS and corporate resources. The IRS believes that its current study efforts could lead to assessing an additional billion dollars or more in taxes each year. While we agree that some improvements in the return selection process may result from these studies, we also identified various problems in how the studies were conducted, which indicate that results on this scale may not be realized.

Further, other problems concerning audit results will not be addressed by the results of these studies. These other problems involve conducting examinations with the most recently filed return, and better preparing revenue agents to conduct large corporate examinations.

Detailed Objective, Scope, and Methodology

Our objective was to evaluate the Internal Revenue Service’s (IRS) efforts to improve workload selection methods for large size corporate examinations. On-site tests were conducted in the IRS’ Headquarters, Illinois, Los Angeles, Manhattan, New England, New Jersey, and Southern California Offices. The specific audit tests included the following steps.

1. Analyzed large corporate examination trends from Fiscal Years 1993 to 1999 data in the Audit Information Management System (AIMS).
2. Reviewed the completed Net Operating Loss (NOL) and the ongoing Discriminate Analysis System (DAS) Studies to assess the adequacy of the planning, implementation and/or reporting phases of the studies.
3. Conducted on-site tests in six district offices and evaluated techniques used to select large corporate returns for examination, any local efforts underway to improve workload selection, the impact of the NOL and DAS Studies, and other training that may not have been reflected on the Automated Corporate Education System (ACES). (Due to time limitations, the New Jersey District tests were not completed or included in our results.)
4. Computed the potential salaries and benefit costs that could be incurred by examining 2,209 large corporate examinations under Phase III and IV of the ongoing DAS Study using the Office of Personnel Management’s 2000 Salary Schedule for a Minneapolis-St. Paul GS-13 (Step 5) revenue agent with a 20 percent benefit package. The Minneapolis-St. Paul Salary Table was used because it equates to the average salaries nationwide.
5. Computed the average elapsed time from when a large corporation entered the examination stream until the examination was started (AIMS Creation Date to Status Code 12 Date of initial returns selected for examination).
6. Computed the average elapsed time from when a business return of a sole proprietor (Activity Codes 535 and 536 with either a Source Code 02 or 65) entered the examination stream until the examination was started (AIMS Creation Date to Status Code 12 date).
7. Analyzed ACES to determine the amount and types of training received between 1997 and 1999 by 231 revenue agents from the 6 district offices who had conducted large corporate examinations.

Major Contributors to This Report

Gordon C. Milbourn III, Associate Inspector General for Audit (Small Business and Corporate Programs)
Philip Shropshire, Director
Frank Dunleavy, Audit Manager
Earl Charles Burney, Senior Auditor
Lawrence R. Smith, Senior Auditor
Jean Kao, Auditor
William Tran, Auditor

Report Distribution List

Deputy Commissioner Operations C:DO
Commissioner, Large and Mid-Size Business Division LM
District Director, Illinois D
District Director, Los Angeles District D
District Director, Manhattan District D
District Director, New England District D
District Director, New Jersey District D
District Director, Southern California District D

Director, Legislative Affairs CL:LA
Director, Office of Program Evaluation and Risk Analysis M:O

National Taxpayer Advocate C:TA
Office of the Chief Counsel CC
Office of Management Controls CFO:A:M
Audit Liaison:

Commissioner, Large and Mid-Size Business Division LM

Management’s Response to the Draft Report

The response has been removed due to its size. To see the complete response, please go to the Adobe PDF version of the report on the TIGTA Public Web Page.