TREASURY INSPECTOR GENERAL FOR TAX ADMINISTRATION

Using Soft Notices to Address Reporting Discrepancies Has Merit, but Cost and Benefit Questions Remain

September 9, 2011

Reference Number: 2011-30-091

This report has cleared the Treasury Inspector General for Tax Administration disclosure review process and information determined to be restricted from public release has been redacted from this document.

Phone Number | 202-622-6500

Email Address | TIGTACommunications@tigta.treas.gov

Web Site | http://www.tigta.gov

HIGHLIGHTS

USING SOFT NOTICES TO ADDRESS REPORTING DISCREPANCIES HAS MERIT, BUT COST AND BENEFIT QUESTIONS REMAIN

Highlights

Final Report issued on September 9, 2011

Highlights of Reference Number: 2011-30-091 to the Internal Revenue Service Commissioners for the Small Business/Self-Employed Division and the Wage and Investment Division.

IMPACT ON TAXPAYERS

Guided by its strategic plan, the Internal Revenue Service’s (IRS) vision is to make it easier for taxpayers to fulfill their civic responsibility to pay taxes by providing them with world-class service, taking proactive steps to better understand issues from the taxpayer’s perspective, and reducing taxpayer burden. Consequently, it will be important for the IRS to understand and minimize the time and resources taxpayers spend dealing with soft notices to increase the likelihood of achieving its vision for improving service to taxpayers.

WHY TIGTA DID THE AUDIT

The overall objective of this review was to determine whether the Soft Notice Initiative (the Initiative) in the Automated Underreporter Program was effectively implemented and managed to provide IRS officials with reliable information for deciding whether the use of soft notices should be expanded, modified, or terminated. Briefly stated, the notices are designed to serve as an educational tool, encourage self-correction, and improve voluntary compliance. As such, IRS officials consider soft notices an alternative approach for addressing compliance issues outside the IRS’s traditional processes in the Automated Underreporter Program.

If determined to be successful, the Initiative could result in permanently using soft notices in the IRS Automated Underreporter Program to address a large number of taxpayers each year who would not ordinarily be contacted by the IRS due to resource constraints. Given this possibility, reliable cost-benefit information may help alleviate concerns that could be raised about why the IRS risks burdening taxpayers and incurring the costs of sending soft notices that result in little or no tax revenue or impact on taxpayer behavior.

WHAT TIGTA FOUND

Overall, the IRS did an adequate job planning, implementing, and managing the Initiative through its first pilot test. In addition, the 16 measures the IRS used to evaluate the Initiative were accurately computed as reported to its executives in the Fiscal Year 2010 Soft Notice Report.

However, TIGTA’s evaluation showed that the taxpayer burden measures are not presenting as complete a picture as they could because they do not include any information on the time or money spent by the nearly 14,000 taxpayers who did not respond to the soft notice. Further, the IRS has yet to commit to taking specific actions or establishing a specific date defining how or when it will ensure all costs are quantified and used in determining the net benefit from using soft notices in the Automated Underreporter Program.

WHAT TIGTA RECOMMENDED

TIGTA recommended that the Director, Campus Compliance Services, Small Business/
Self-Employed Division, and the Director, Compliance, Wage and Investment Division, implement plans to (1) obtain a more complete picture of the time and costs taxpayers are spending on soft notices and (2) determine the net benefit of using soft notices in the Automated Underreporter Program as an alternative approach for addressing compliance issues.

In their response to the report, IRS officials agreed with the recommendations and plans to continue soliciting feedback from taxpayer responses and evaluating the net benefits derived from soft notices. Although IRS officials agreed with the recommendations, their corrective actions do not address the conditions that gave rise to the recommendations.

September 9, 2011

MEMORANDUM FOR COMMISSIONER, SMALL BUSINESS/SELF-EMPLOYED DIVISION

COMMISSIONER, WAGE AND INVESTMENT DIVISION

FROM: Michael R. Phillips /s/ Michael R. Phillips

Deputy Inspector General for Audit

SUBJECT: Final Audit Report – Using Soft Notices to Address Reporting Discrepancies Has Merit, but Cost and Benefit Questions Remain (Audit # 201030043)

This report presents the results of our review to determine whether the Soft Notice Initiative in the Automated Underreporter Program was effectively implemented and managed to provide Internal Revenue Service officials with reliable information for deciding whether the Soft Notice Initiative should be expanded, modified, or terminated. This audit was conducted as part of our Fiscal Year 2011 Annual Audit Plan and addresses the major management challenge of Tax Compliance Initiatives.

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

Copies of this report are also being sent to the Internal Revenue Service managers affected by the report recommendations. Please contact me at (202) 622-6510 if you have questions or Margaret E. Begg, Assistant Inspector General for Audit (Compliance and Enforcement Operations), at (202) 622‑8510.

Table of Contents

Background

Results of Review

Overall, the Tax Year 2007 Soft Notice Pilot Test Was Adequately Planned, Implemented, and Managed

The Time and Costs Associated With Soft Notices Need to Be Measured More Precisely

Recommendations 1 and 2:

Appendices

Appendix I – Detailed Objective, Scope, and Methodology

Appendix II – Major Contributors to This Report

Appendix III – Report Distribution List

Appendix IV – Automated Underreporter Soft Notice Initiative Objectives and Measures

Appendix V – Example of a Computer Paragraph 2057 Notice

Appendix VI – Management’s Response to the Draft Report

Abbreviations

AUR	Automated Underreporter
CP DCI	Computer Paragraph Data Collection Instrument
FTE	Full-Time Equivalent
IRS SB/SE	Internal Revenue Service Small Business/Self-Employed
TIGTA	Treasury Inspector General for Tax Administration
W&I	Wage and Investment

Background

One purpose of the Automated Underreporter Program (hereafter referred to as the AUR or the AUR Program) is to resolve income discrepancies between the information taxpayers report to the Internal Revenue Service (IRS) on tax returns and related information employers and financial institutions provide the IRS on information returns. Once discrepancy cases are identified, the IRS decides how many cases it believes it has sufficient resources to investigate out of the total number identified.

When selected, AUR cases are distributed to seven IRS campuses[1] where tax examiners manually review each case. After analyzing the tax returns, tax examiners are sometimes able to immediately resolve the discrepancies, in which case no further actions are taken. For the remaining AUR cases, tax examiners may request additional information from taxpayers by sending a Computer Paragraph (CP) 2000 notice.[2]

If the taxpayer provides supporting documentation and the tax examiner determines the information resolves the discrepancy, the case is closed with no changes to the taxpayer’s account. However, if the examiner determines the information does not resolve the discrepancy and the taxpayer agrees with the examiner’s determination, the IRS will assess additional tax based on the CP 2000 notice and close the case as agreed. When a taxpayer does not agree or does not respond to the CP 2000 notice within the required time period, a Statutory Notice of Deficiency[3] will be issued to assess additional tax.

Because IRS resources cannot investigate all areas of noncompliance, including AUR discrepancies, the IRS is increasingly using alternative approaches to resolve compliance issues outside its traditional processes. In the AUR Program, where millions of discrepancy cases are not investigated each year, the IRS is involved in a multiyear initiative (the Initiative) to determine if soft notices can address AUR underreporting discrepancies that are not investigated due to resource constraints.

The soft notices are called a CP 2057[4] and do not require that the taxpayer pay more tax, provide documentation, or even respond to the IRS. Although the notice requests the taxpayer file an amended tax return if appropriate, it is not required. Instead, the notices are designed to serve as an educational tool, encourage self-correction, and improve voluntary compliance. By comparison, the CP 2000 is primarily focused on compliance and indicates that the IRS is proposing a tax change and that the taxpayer may owe additional tax. If determined to be successful, the Initiative could result in permanently using soft notices as an alternative approach for addressing compliance issues among a large number of taxpayers each year who would not ordinarily be contacted by the IRS due to resource constraints.

This review was the second phase of audit work conducted by the Treasury Inspector General for Tax Administration (TIGTA) to determine whether the Initiative was effectively implemented and managed to provide IRS officials with reliable information for deciding whether the Initiative should be expanded, modified, or terminated. Audit work in the first phase[5] was focused on IRS planning activities for the Initiative while the second phase centered on how well plans were implemented and managed and if reliable results were produced.

The review was performed at the IRS Wage and Investment (W&I) Division Headquarters in Atlanta, Georgia, and IRS Small Business/Self-Employed (SB/SE) Division Headquarters in New Carrollton, Maryland, during the period October 2010 through April 2011. Except for not auditing the IRS Individual Master File[6] that was used to validate the accuracy and reliability of the data, we conducted this performance audit 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 objective. We believe that the evidence obtained provides a reasonable basis for our findings and conclusions based on our audit objective. Detailed information on our audit objective, scope, and methodology is presented in Appendix I. Major contributors to the report are listed in Appendix II.

Results of Review

Overall, the IRS did an adequate job planning, implementing, and managing the Initiative through its first pilot test. However, the outcomes from the pilot test raised questions about the costs incurred and benefits realized that the IRS needs to answer before deciding to use soft notices on a much larger scale or permanent basis as an alternative approach for addressing compliance issues in the AUR Program.

Overall, the Tax Year 2007 Soft Notice Pilot Test Was Adequately Planned, Implemented, and Managed

One of the best practices for developing and implementing new business processes, or for improving existing ones, is to establish an overall approach that contains detailed steps for carrying out the various phases of an initiative. For example, the Government Accountability Office developed and used a 20-step approach to evaluate earlier improvement initiatives in the IRS SB/SE Division. The approach is based on its Business Process Reengineering Assessment Guide[7] and discussions with managers in private industry as well as in other Federal agencies. According to the Government Accountability Office, the 20 steps included in its approach help ensure potential obstacles are considered in planning, problems are pinpointed and addressed through pilot testing, and results are evaluated accurately.

In the first phase of our work, we used the 20-step approach as criteria to assess how closely the IRS team responsible for conducting the Initiative considered each of the recommended steps and reported that, except for 2 areas, the team did a good job of planning the Initiative. Importantly, the team developed an evaluation plan that contained 16 measures aligned within 6 objectives to assess the results from its first pilot test. Developing and documenting how to evaluate results in the planning stage was an important step because it ensured the data necessary for decisionmaking purposes would be available and collected once testing began.

Documentation the team provided to us during our review shows it conducted two pilot tests. According to the Government Accountability Office’s 20-step approach, pilot testing is critical because it provides opportunities to evaluate the soundness of the methods being used, make necessary adjustments to correct potential problem areas, and obtain the support needed to make program decisions such as whether to use soft notices in the AUR Program on a permanent basis. The team’s first pilot was called the Tax Year[8] 2007 Soft Notice Pilot Test and involved sending 13,330 notices to taxpayers served by the W&I Division and 15,331 notices to taxpayers served by the SB/SE Division.

To manage the resulting correspondence and inquiries from the 28,661 taxpayers who received notices, as well as to collect and analyze data called for in the pilot’s evaluation plan, the team developed and populated a standardized data collection instrument (DCI).[9] They also supplemented taxpayer response data with information from the AUR Program and the Individual Master File. The AUR staff reviewed these data and concluded soft notices should be used to expand compliance coverage and potentially improve recurring taxpayer behavior.

We validated the accuracy of the 16 measures (populated in the DCI and subsequently reported to IRS executives in the team’s Fiscal Year 2010 Soft Notice Report)[10] by following the team’s methodology for measuring each outcome and then comparing each outcome to source information. Figure 1 shows only minor differences (2 percent or less) for all but one of the measures we validated to source information. Appendix IV contains a detailed description for each of the 16 measures.

Figure 1: Tax Year 2007 Soft Notice Pilot Test Results

*Measures*	*SB/SE Division*	*W&I Division*	IRS Totals	TIGTA Analysis	Percentage Difference
*OBJECTIVE 1 – CHANGE TAXPAYER BEHAVIOR*
1. Subsequent Year Recurring Discrepant Behavior	5,273	3,293	8,566	8,578	<1.0
2. Subsequent Year Recurring Similar Behavior	1,603	1,139	2,742	2,742	0.0
*OBJECTIVE 2 – CORRECT CURRENT YEAR TAXPAYER BEHAVIOR*
3. 1040X[11] – Fully Agreed With Soft Notice Issues	818	965	1,783	1,609	9.8[12]
4. 1040X – Fully Addressed Soft Notice Issues	1,762	1,368	3,130	3,172	1.3
5. 1040X – Partially Addressed Soft Notice Issues	949	356	1,305	1,326	1.6
*OBJECTIVE 3 – LIMIT THE IMPACT ON IRS REVENUE*
6. Case Revenue Generated (average per case)	$182	$137	$161	$159	1.2
7. Yield Versus Calculated Tax Change	($22,223)	($10,966)	($16,987)	($16,988)	<1.0
8. Dollars Per Full-Time Equivalent (FTE)[13]	$1,484,420	$791,543	$1,103,172	$1,077,265	2.3
*OBJECTIVE 4 – LIMIT THE IMPACT ON IRS RESOURCES*
9. Telephone Calls	2,332	3,301	5,633	5,630	<1.0
10. Screening Cost Savings (in hours)	3,129	1,713	4,842	4,842	0.0
11. Letter Responses	1,896	787	2,683	2,682	0.0
12. 1040X Responses	4,666	4,471	9,137	9,201	<1.0
13. Undeliverables (notices returned)	75	130	205	205	0.0
*OBJECTIVE 5 – INCREASED TAXPAYER COVERAGE*
14. Productive Unit of Effort for Sending Soft Notices, Expanding Coverage[14]	1.61	2.04	3.65	3.70	1.4
*OBJECTIVE 6 – MINIMIZE TAXPAYER BURDEN*
15. No Change Rate	15%	11%	13%	13%	0.0
16. Preparers Required	2,595	987	3,582	3,580	<1.0

Source: Our analysis of the IRS’s draft Fiscal Year 2010 Soft Notice Report, dated July 30, 2010.

To add perspective and meaning to the outcomes from the first pilot, the team created two controls groups that contained similar discrepancy characteristics as the soft notice cases. Taxpayers in the first control group were treated as regular AUR Program cases and were considered for issuance of a CP 2000. Those in the second group, the Do Nothing Group, went untreated, meaning the IRS made no contact with the taxpayers concerning the income discrepancies. The specific outcomes achieved are categorized below by objective.

1. Change Taxpayer Behavior. Regardless of whether a soft notice, a CP 2000, or no notice was issued, approximately 30 percent of taxpayers subsequently filed an amended tax return to address a potential discrepancy that was identified through third-party information reporting.

2. Correct Current Year Taxpayer Behavior. A regular AUR Program case was far more likely to result in a taxpayer filing an amended tax return in the current year than a soft notice case. The 28,661 soft notices issued resulted in 9,201 taxpayers (32 percent) voluntarily changing their tax return to address the discrepancies identified in the soft notices. In contrast, the CP 2000 notices issued to 27,135 taxpayers in the AUR Control Group resulted in adjustments to the accounts of 17,830 taxpayers (66 percent).

3. Limit the Impact on IRS Revenue. Soft notice cases generated significantly less revenue than regular AUR Control Group cases. On average, each AUR Control Group case generated $729 more in additional taxes than a soft notice case. Each IRS employee assigned to work AUR Program cases could be expected to generate 106 percent more revenue annually than an employee working soft notice cases.

4. Limit the Impact on IRS Resources. Unlike the manual review of CP 2000 notices, soft notices were not reviewed before they were issued. As a result, the team estimated the issuance of the 28,661 soft notices saved 4,842 labor hours, or about 2.33 FTEs. The purpose of the manual review of CP 2000 notices is to eliminate erroneous discrepancies. A discrepancy, for example, can be identified by IRS computers, but a manual review may identify something the computer could not, such as an income item mistakenly reported on the incorrect line of a tax return.

5. Increased Taxpayer Coverage. Soft notices provide the IRS with a less expensive way to expand its compliance coverage than CP 2000 notices. In terms of labor, the manual review and administration of 27,135 CP 2000 notices required 11 FTEs, each of whom worked about 1.2 cases per hour. In contrast, the 28,661 soft notices required about 4.23 FTEs, each of whom worked about 3.3 cases per hour.

6. Minimize Taxpayer Burden. A high no-change rate means a significant amount of resources are being devoted to unproductive notices and compliant taxpayers are being unnecessarily burdened by the notices. Neither the TIGTA nor IRS officials know what the actual no-change rate should be for soft notices because the normal screening procedures were not applied. However, the IRS determined that no tax change was justified for 13 percent of the taxpayers who responded to the soft notices, while 2.26 percent of the tax liabilities of AUR Program cases were not changed.

In considering the significance of the outcomes from the Tax Year 2007 Soft Notice Pilot Test and their applicability to deciding the future use of soft notices in the AUR Program, it is important to recognize that this test was the first of two soft notice pilot tests. Based on the outcomes of the Tax Year 2007 Soft Notice Pilot Test, the team made adjustments to improve how to best use soft notices as an alternative approach for addressing compliance issues in the AUR Program. Consequently, the outcomes from the later pilot may better achieve the stated objectives for the Initiative and, if so, provide additional support data for making an informed decision about whether soft notices should be incorporated as an alternative compliance treatment on a permanent basis.

It is also important to recognize that the methodology used to measure the burden the soft notices imposed on taxpayers was based solely on the correspondence received from taxpayers. Because nearly 50 percent (13,991 of 28,661) of taxpayers issued soft notices did not respond, the taxpayer burden measures are not providing as complete of a picture as they could because there is no information on the amount of time or money these taxpayers may have spent dealing with the notices.

The Time and Costs Associated With Soft Notices Need to Be Measured More Precisely

Improving service to taxpayers is one of the highest priorities for the IRS. Guided by its strategic plan, the IRS’s vision is to make it easier for taxpayers to fulfill their civic responsibility to pay taxes by providing them with world-class service, taking proactive steps to better understand issues from a taxpayer’s perspective, and reducing taxpayer burden. Consequently, it will be important for the IRS to understand and minimize the time and costs that taxpayers spend dealing with soft notices to increase the likelihood of achieving its vision for improving service to taxpayers.

One of the initial steps that needs to be taken in understanding and minimizing taxpayer burden is assessing the amount of time and money a particular activity imposes. However, as IRS researchers have reported, measuring burden is inherently challenging because it can mean different things to different types of taxpayers (e.g., individuals versus businesses) depending upon whether a taxpayer is seeking advice about a particular tax issue, preparing a return, or dealing with an IRS audit or notice.

To estimate the costs and time taxpayers spent dealing with soft notices, the team used two measures. The first measure involved counting the number of taxpayer accounts marked as having used a paid preparer in the DCI. This measure reasonably assumes these taxpayers incurred a fee for the service of having their returns amended by a tax preparer. The second measure involved counting the number of telephone calls, letters, or amended tax returns recorded in the DCI from taxpayers who indicated no adjustment was warranted to their tax return. This measure reasonably assumes that the time spent by the taxpayers reviewing records and responding to IRS was unnecessary.

While the two measures have merit for determining the burden imposed on taxpayers who responded by filing an amended tax return or contacting the IRS by letter or telephone, they do not provide a complete picture because the burden on the 13,991 taxpayers who chose not to respond to the soft notice remains unknown and could be substantial. To illustrate this point, in Fiscal Year 2009, the AUR Program closed approximately 4.5 million cases. The screening process resolved 18.4 percent of the computer-identified discrepancy cases without contacting taxpayers. As a result, the IRS avoided unnecessarily burdening approximately 828,000 taxpayers with a CP 2000 notice that would have likely resulted in no change being made to the taxpayer’s account. Because the soft notice cases were not manually reviewed before they were issued, but were selected from the same population as the AUR Program cases, it is reasonable to assume that 18.4 percent (2,574) of the 13,991 taxpayers who did not respond may have had to deal with a notice that should not have been issued. It is also reasonable to assume that some percentage of the nonrespondents sought advice from a tax preparer before deciding not to respond. Our analysis shows that 9,182 (66 percent) of the 13,991 taxpayers who did not respond to the soft notice used a paid tax preparer to complete their Tax Year 2007 income tax return.

The net benefit of using soft notices as an alternative approach for addressing compliance issues in the AUR Program needs to be determined

Our August 2010 report included our evaluation of the team’s planning efforts for the Initiative and identified two recommendations that required corrective action by the IRS. The first recommendation involved the IRS defining what it would consider a success within each of the 16 measureable items and/or how much weight the various measures will have in determining the overall success of the Initiative. The second recommendation entailed the IRS ensuring that all costs it has or will incur are quantified and used in determining the net benefit of implementing soft notices as an alternative approach for addressing compliance issues in the AUR Program.

The IRS committed to taking corrective actions in response to the first recommendation by January 2012. However, the IRS did not commit to taking specific actions or establishing a specific date defining how or when it will ensure all costs are quantified and used in determining the net benefit from using soft notices as an alternative approach for addressing compliance issues in the AUR Program. As we reported in August 2010, there are costs that have yet to be fully considered, such as the expenses associated with a contractor who was involved in the design, development, documentation, execution, and assessment activities of the Initiative.

By ensuring all benefits and costs are captured and analyzed, the team can establish an important control for assuring IRS senior executives that key issues were considered before recommending to use soft notices on a much larger scale or permanent basis as an alternative approach for addressing compliance issues in the AUR Program. Moreover, reliable cost-benefit information is a critical component of a sound assessment methodology and may help alleviate concerns that could be raised about why the IRS risks burdening taxpayers and incurring the costs of sending soft notices that result in little or no tax revenue or impact on taxpayer behavior.

Recommendations

The Director, Campus Compliance Services, SB/SE Division, and the Director, Compliance, W&I Division, should coordinate actions to develop and implement a plan:

Recommendation 1: To obtain a more complete picture of the time and costs taxpayers are spending on soft notices.

Management’s Response: IRS officials agreed with this recommendation and will continue to solicit feedback on the CP 2057 notices from taxpayer responses, the Office of Taxpayer Communication, and IRS Advisory Council to improve their processes.

Office of Audit Comment: Although IRS officials agreed with the recommendation, their corrective actions do not address the condition that gave rise to the recommendation. Specifically, the corrective actions do not include developing and implementing a plan that will measure the time and costs the notices are imposing on the taxpayers who are not responding to the notices. As indicated in the report, these could be substantial and taking such steps would provide the IRS with an opportunity to reinforce its commitment to better understand issues from the taxpayer’s perspective and reduce taxpayer burden.

Recommendation 2: For determining the net benefit of using soft notices in the AUR Program as an alternative approach for addressing compliance issues.

Management’s Response: IRS officials agreed with this recommendation and stated that they have determined that soft notices are beneficial as a self correction method and have educational benefit. They will also continue to evaluate the net benefits derived from the use of soft notices.

Office of Audit Comment: Although IRS officials agreed with the recommendation, their corrective actions do not address the condition that gave rise to the recommendation. Specifically, the corrective actions do not include developing and implementing a plan that will ensure all costs are quantified and used in determining the net benefit from using soft notices as an alternative approach for addressing compliance issues in the AUR Program. As indicated in the report, reliable cost-benefit information is a critical component of a sound assessment methodology and may help alleviate concerns that could be raised about why the IRS risks burdening taxpayers and incurring the costs of sending soft notices that result in little or no tax revenue or impact on taxpayer behavior.

Appendix I

Detailed Objective, Scope, and Methodology

The overall objective of this review was to determine whether the AUR Program Soft Notice Initiative (the Initiative) was effectively implemented and managed to provide IRS officials with reliable information for deciding whether the Initiative should be expanded, modified, or terminated. As part of the review, we relied on databases provided to us by the IRS and did not audit the source systems. Unless otherwise noted, our limited tests of the reliability of data obtained from the IRS did not identify any errors. We tested the reliability of the data by scanning the data received for blank, incomplete, illogical, or improper data. We also performed reviews of amended tax returns to determine the accuracy of information recorded in the DCI.

To accomplish our overall objective, we:

I. Identified the methodology used to measure the results of the Initiative.

A. Secured a copy of the draft Fiscal Year 2010 Soft Notice Report, which provides final results and conclusions from the Tax Year[15] 2007 Soft Notice Pilot Test as well as preliminary findings from Tax Year 2008.

B. Interviewed W&I Division’s AUR Program management and obtained the methodologies for each of the 16 measures used to evaluate the Tax Year 2007 Soft Notice Pilot Test.

C. Secured an extract of the DCIs for Tax Years 2007 and 2008 that were used to support the results in the draft Fiscal Year 2010 Soft Notice Report.

D. Used the information gathered in Steps I. A., B., and C. to identify any changes to case selection made for subsequent years’ tests.

II. Determined if the IRS has effectively measured whether soft notice treatment has changed taxpayer behavior.

A. Verified that the preliminary measures supplied to us during the previous Phase I review[16] for the following objectives were still valid.

· Change Taxpayer Behavior.

· Correct Current Year Taxpayer Behavior.

B. Analyzed the accuracy of the taxpayer behavior measures and compared our figures to the results in the draft Fiscal Year 2010 Soft Notice Report.[17]

1. For the Subsequent Year Recurring Discrepant Behavior measure, obtained the AUR repeater file[18] (17,054,422 records) for Tax Year 2008 from AUR Program management and matched the records to Tax Year 2007 samples for the Soft Notice Group (28,661 taxpayers), the AUR Control Group (30,819[19] taxpayers), and the Do Nothing Group (30,760 taxpayers) and identified the number and percentage of those taxpayers who repeated underreporting behavior in Tax Year 2008. [(Soft Notice Group – 8,578 (30 percent); AUR Control Group – 8,892 (29 percent); Do Nothing Group – 8,502 (28 percent)].

Note: The pilot team created two control groups that contained similar discrepancy characteristics as the group that received the soft notices. Taxpayers in the AUR Control Group were treated as regular AUR Program cases and were considered for issuance of a CP 2000. Those in the Do Nothing Group went untreated, meaning the IRS made no contact with the taxpayers concerning the income discrepancies.

2. For the Subsequent Year Recurring Similar Behavior measure, matched the AUR repeater file for Tax Year 2008 to Tax Year 2007 samples for the Soft Notice Group, the AUR Control Group, and the Do Nothing Group and identified the number of those taxpayers who repeated underreporting behavior in the same AUR category in Tax Year 2008. [Soft Notice Group – 2,742 (10 percent); AUR Control Group – 2,866 (9 percent); Do Nothing Group – 3,121 (10 percent)].

3. For the 1040X[20] – Fully Agreed With Soft Notice Issues measure, queried the DCI data to identify records marked as fully agreed. We also matched 17,704 taxpayers who received a soft notice but had no record in the DCI against the Individual Master File[21] to identify 1,609 individuals who had filed an amended tax return fully agreeing with the income discrepancies. In addition, we compared those results to the 27,135 notices sent to taxpayers in the AUR Control Group, which resulted in 11,224 assessments that fully agreed with the income discrepancies.

4. For the 1040X – Fully Addressed Soft Notice Issues measure, queried the DCI data and identified 3,172 records marked as fully addressed.

5. For the 1040X – Partially Addressed Soft Notice Issues measure, queried the DCI data and identified 1,326 records marked as partially addressed.

III. Determined if the IRS effectively measured the impact of the Soft Notice Initiative on revenues and costs.

A. Verified that the preliminary measures supplied to us during the previous Phase I review for the following objectives were still valid.

· Limit the Impact on IRS Revenue.

· Limit the Impact on IRS Resources.

· Increased Taxpayer Coverage.

B. Analyzed the accuracy of the measures used to determine the impact of soft notices on IRS revenues and compared our figures to the results in the draft Fiscal Year 2010 Soft Notice Report.

1. For the Case Revenue Generated measure, we:

a. Queried the DCI data[22] to total the net balance due and refund amounts for the Soft Notice Group for the W&I Division ($854,398) and the SB/SE Division ($1,815,942).

b. Identified the taxpayers receiving soft notices who did not have a record in the DCI (8,530 for the W&I Division and 9,174 for the SB/SE Division) and matched them to the Individual Master File to identify and total all tax adjustment transactions occurring at least 4 weeks after the issuance of the notices ($914,176 for the W&I Division and $972,317 for the SB/SE Division).

c. Combined the DCI and Individual Master File totals to calculate the total case revenue generated of $4,556,833 and computed the average per case by dividing it by the 28,661 notices issued.

d. Obtained case revenue results for the AUR Control Group by matching the 27,135 records with CP 2000 notices sent to the Individual Master File to calculate total revenue of $24,084,251 for 17,620 accounts with adjustments.

2. For the Yield Versus Calculated Tax Change measure, used the case revenue data gathered in Step III.B.1 to identify the total tax change for the entire Soft Notice Group and compared it to the expected tax change identified on the AUR case selection database provided by the AUR Program. For the expected tax change, we computed the total expected tax change for the 28,661 cases ($491,452,135) using the DCI and AUR case selection database data provided by IRS. We then computed the difference between the expected tax change and actual revenue generated ($4,556,833), which is $486,895,302, and computed the average Yield Versus Calculated Tax Change by dividing by 28,661 ($16,988). We validated the accuracy of the case selection database by contacting an AUR Program analyst for confirmation.

3. For the Dollars Per FTE[23] measure, obtained statistical data for time applied by employees to Soft Notice Group cases from Work Planning and Control[24] reports. We then combined those figures with IRS-provided FTE figures for processing amended tax returns and computed the total FTE staff years for working soft notices (4.23). We then divided the total case revenue dollars ($4,556,833) by the 4.23 FTEs to compute the $1,077,265 per FTE.

C. Analyzed the accuracy of the measures used to determine the impact of soft notices on IRS resources and compared our figures to the results in the draft Fiscal Year 2010 Soft Notice Report.

1. For the Screening Cost Savings measure, obtained the AUR Program’s standard historical screening rates for the W&I Division (7.78 cases per hour) and the SB/SE Division (4.9 cases per hour). We then divided the total number of soft notices mailed (13,330 for the W&I Division and 15,331 for the SB/SE Division) by the historical screening rate for each division and combined them to compute the 4,842 screening hours saved.

2. For the Telephone Calls measure, queried the DCI data for the Soft Notice Group and counted the 5,630 records with telephone calls.

3. For the Letter Responses measure, queried the DCI data for the Soft Notice Group and counted the 2,682 records where letters had been received.

4. For the 1040X Responses measure, queried the DCI data for the Soft Notice Group and counted the 4,616 records where amended tax returns had been received. We also matched the 24,045 Soft Notice Group records where no amended tax return was recorded in the DCI to Individual Master File data to identify an additional 4,585 amended tax returns for the Soft Notice Group.

5. For the Undeliverables measure, queried the DCI data for the Soft Notice Group and counted the 205 records where the soft notice was returned undeliverable.

D. Analyzed the accuracy of the measure used to determine the impact that soft notices have on increasing taxpayer coverage and compared our figures to the results in the draft Fiscal Year 2010 Soft Notice Report.

1. For the Productive Unit of Effort for Sending Soft Notices, Expanding Coverage measure, obtained Work Planning and Control data and the AUR Program’s process for computing the measure and recomputed total FTE staff days needed by the AUR Program to work soft notice responses and telephone calls (2.04 FTEs for the W&I Division and 1.66 FTEs for the SB/SE Division). This measure does not include the FTEs for the processing of amended tax returns.

2. Computed Dollars Per FTE for the AUR Control Group by dividing the $24,084,251 in revenue by 10.854 FTEs we computed that was needed for processing, which resulted in $2,218,929 per FTE. We then divided this figure by the $1,077,265 per Soft Notice FTE to determine that the AUR Control Group generated 106 percent more dollars per FTE than the Soft Notice Group.

IV. Determined if the IRS effectively measured the impact of the Initiative on taxpayer burden.

A. Verified that the preliminary measures supplied to us during the previous Phase I review for the “Minimize Taxpayer Burden” objective were still valid. We identified that a Screenout Rate measure originally planned is no longer being used as a measure because there was no screening done on the soft notices.

B. Analyzed the accuracy of the measures used to determine the impact that soft notices had on taxpayer burden and compared our figures to the results in the draft Fiscal Year 2010 Soft Notice Report .

1. For the No-Change Rate measure, queried the DCI data for the Soft Notice Group (28,661 taxpayers) and counted 3,730 (13 percent) taxpayers that justifiably had no tax change made to their account.

2. For the Preparers Required measure, queried the DCI data for the Soft Notice Group and counted the 3,580 taxpayers identified where a preparer was used.

3. We matched the 13,991 soft notice taxpayers who did not respond to the soft notice against tax return data on the TIGTA Data Center Warehouse[25] and identified that 9,182 (66 percent) of these taxpayers used a paid preparer to file their original 2007 tax return.

V. Determined if the measures were sufficient to provide an accurate evaluation of the pilot’s results to IRS executives. We discussed the results of our analysis with AUR Program management and obtained agreement to the facts.

Internal controls methodology

Internal controls relate to management’s plans, methods, and procedures used to meet their mission, goals, and objectives. Internal controls include the processes and procedures for planning, organizing, directing, and controlling program operations. They include the systems for measuring, reporting, and monitoring program performance. We determined the following internal controls were relevant to our audit objective: the IRS’s procedures for measuring revenues and costs for the Initiative to determine its cost effectiveness. We evaluated these controls by interviewing management and analyzing IRS data by comparing our analysis of IRS databases to the results of the draft Fiscal Year 2010 Soft Notice Report measures.

Appendix II

Major Contributors to This Report

Margaret E. Begg, Assistant Inspector General for Audit (Compliance and Enforcement Operations)

Frank Dunleavy, Director

Alan Lund, Audit Manager

Matthew Schimmel, Lead Auditor

Carole Connolly, Senior Auditor

Sharon Summers, Senior Auditor

William Tran, Senior Auditor

Appendix III

Report Distribution List

Commissioner C

Office of the Commissioner – Attn: Chief of Staff C

Deputy Commissioner for Services and Enforcement SE

Deputy Commissioner, Small Business/Self-Employed Division SE:S

Deputy Commissioner, Wage and Investment Division SE:W

Director, Campus Compliance Services, Small Business/Self-Employed Division SE:S:CCS

Director, Communications, Liaison, and Disclosure, Small Business/Self-Employed Division SE:S:CLD

Director, Compliance, Wage and Investment Division SE:W:CP

Director, Reporting Compliance, Wage and Investment Division SE:W:CP:RC

Director, Strategy and Finance, Wage and Investment Division SE:W:S

Chief, Performance Improvement, Wage and Investment Division SE:W:S:PRA:PEI

Chief Counsel CC

National Taxpayer Advocate TA

Director, Office of Legislative Affairs CL:LA

Director, Office of Program Evaluation and Risk analysis RAS:O

Office of Internal Control OS:CFO:CPIC:IC

Audit Liaisons:

Commissioner, Small Business/Self-Employed Division SE:S

Commissioner, Wage and Investment Division SE:W

Appendix IV

Automated Underreporter Soft Notice Initiative Objectives and Measures

AUR Objective	Measure
Change Taxpayer Behavior	1. Subsequent Year Recurring Discrepant Behavior – The number of taxpayers who failed to correctly report third-party income in the following year.
Change Taxpayer Behavior	2. Subsequent Year Recurring Similar Behavior – The number of taxpayers who failed to correctly report the same type of third-party income in the following year.
Correct Current Year Taxpayer Behavior	3. 1040X[26] – Fully Agreed With Soft Notice Issues – The taxpayer filed an amended tax return and the tax change was the same as estimated by the IRS in the DCI.
	4. 1040X – Fully Addressed Soft Notice Issues – The taxpayer filed an amended tax return and addressed the underreported income. However, the tax owed does not match the IRS’s estimates.
	5. 1040X – Partially Addressed Soft Notice Issues – The taxpayer only addressed some of the underreported income in the soft notice.
Limit the Impact on IRS Revenue	6. Case Revenue Generated – The revenue generated per soft notice mailed.
	7. Yield Versus Calculated Tax Change – The amount of tax change expected by the IRS compared to what was on the amended tax return.
	8. Dollars Per FTE[27] – Computation of revenue generated per 1 staff year.

AUR Objective	Measure
Limit Impact on IRS Resources	9. Telephone Calls – Count of calls includes multiple telephone calls from the same taxpayer.
	10. Screening Cost Savings – Number of staff hours saved by not having to manually screen all the cases prior to issuing soft notices.
	11. Letter Responses – Count of letters received from taxpayers.
	12. 1040X Responses – Count of Tax Year[28] 2007 amended tax returns received from taxpayers in the DCI and Individual Master File.
	13. Undeliverables – Count of notices returned by the United States Postal Service that could not be delivered.
Increased Taxpayer Coverage	14. Productive Unit of Effort for Sending Soft Notices, Expanding Coverage – Resources expended (in staff years) to handling telephone calls, letter responses, and amended tax return responses for the soft notice sample.
Minimize Taxpayer Burden	15. No Change Rate – The percentage of cases where the taxpayer’s original tax return was correct and no changes needed to be made to the taxpayer’s account.
Minimize Taxpayer Burden	16. Preparers Required – The percentage of taxpayers who used a paid preparer to file an amended tax return.

Source: Our definitions of measures in the IRS’s draft Fiscal Year 2010 Soft Notice Report, dated, July 30, 2010.

Appendix V

Example of a Computer Paragraph 2057 Notice^{^[29]}

The picture was removed due to its size. To see the picture, please go to the Adobe PDF version of the report on the TIGTA Public Web Page.

Appendix VI

Management’s Response to the Draft Report

DEPARTMENT OF THE TREASURY

INTERNAL REVENUE SERVICE

ATLANTA. GA 30308

COMMISSIONER

WAGE AND INVESTMENT DIVISION

August 9, 2011

MEMORANDUM FOR MICHAEL R. PHILLIPS

DEPUTY INSPECTOR GENERAL FOR AUDIT

FROM: for Richard Byrd, Jr. /s/Peggy Bogadi

Commissioner, Wage and Investment Division

SUBJECT: Draft Audit Report – Using Soft Notices to Address Reporting Discrepancies Has Merit, but Cost and Benefit Questions Remain (Audit # 201030043)

We have reviewed the subject draft report and appreciate the recognition of the performance of the IRS in planning, implementing, and managing the Soft Notice Initiative through its first pilot test. Soft Notices^{^[30]} provide the IRS with a compliance alternative that is a less expensive way to expand its coverage than the Computer Paragraph (CP) 2000 ^{^[31]}notices issued by the Automated Underreporter (AUR) program. The use of soft notices as an alternative compliance treatment is beneficial as a self correction method and as a means by which to educate taxpayers.

We are constantly mindful of taxpayer burden issues and recognize the need to further determine the taxpayer burden and the net benefit of implementing soft notices to increase voluntary compliance. With regard to your first recommendation, it is important to note that determining taxpayer burden for those who fail to respond to a taxpayer notice is difficult, as it is generally unclear the level of effort expended by a taxpayer in making a decision not to respond. Taxpayer burden can mean different things to different types of taxpayers (e.g., individuals versus businesses) depending upon whether a taxpayer is seeking advice about a particular tax issue, preparing a return, or dealing with an IRS audit or notice. We continue to solicit and use feedback from the practitioner community through a variety of avenues, such as the Internal Revenue Service Advisory Council, along with feedback obtained from the IRS Communications and Liaison function. This feedback allows us to improve our communications with taxpayers and their representatives, and to assess and minimize the burden imposed by the Soft Notice Program.

The results of the program indicate a tax revenue benefit. As shown in Figure 1 of your report, for Tax Year 2007, over 6,000 taxpayers corrected their current-year tax returns, resulting in more than $1 million in revenue that would not have been collected otherwise. As we discussed with the Audit Team, the Soft Notice Program supplements our compliance activities by providing coverage to a segment of the taxpayer population that we cannot reach through traditional means. Limited resources prevent the examination of all questionable returns, thus requiring decisions to be made regarding the balancing of appropriate coverage with the work perceived to have the greatest potential for change.

Income tax returns most likely to have unreported income are initially identified by computer matching of tax return data to third-party information. The returns identified are then filtered through an inventory selection tool which analyzes hundreds of additional attributes and attribute combinations. Both processes are updated regularly in an effort to reduce unnecessary taxpayer contacts. The CP 2057, underreporter soft notice contacts, are identified using the same methods. While we acknowledge there may be some taxpayer burden in the receipt of any notice, we are confident that the comprehensive selection criteria minimize this burden.

The Soft Notice Programs for Tax Years 2007 and 2008 involved approximately 30,000 soft notice contacts for each year. Tax Year 2009 was an interim year where a similar volume of notices were issued and the AUR function did not monitor contact results. The long-term plan for the Soft Notice Program includes expanding the volume of soft notices issued, beginning with Tax Year 2010. This expansion will, of course, be dependent on adequate funding. The soft notice recipients will continue to be identified using the AUR program criteria and the notice will be generated by the AUR program.

Attached are our comments to your recommendations. If you have any questions, please contact me, or a member of your staff may contact Jim Clifford, Director, Reporting Compliance, Wage and Investment Division at (404) 338-8983.

Attachment

The Director, Campus Compliance Services, SB/SE Division and the Director, Compliance, W&I Division, should coordinate actions to develop and implement a plan:

RECOMMENDATION 1: To obtain a more complete picture of the time and costs taxpayers are spending on soft notices.

CORRECTIVE ACTION:

We agree with this recommendation. We acknowledge that taxpayers may seek professional advice and/or spend time reconciling their tax returns to address the issues identified by soft notices. As stated in your report, IRS researchers have confirmed measuring burden is inherently challenging. We will continue to solicit feedback on the Computer Paragraph (CP) 2057 Notices from taxpayer responses, feedback from the Office of Taxpayer Communication and the Internal Revenue Service Advisory Council. Those comments and suggestions have been utilized in the revisions made to the CP 2057 notice for the coming year. We will continue to use this feedback as a way to improve our processes.

Implementation Date:

Implemented and Ongoing

Responsible Official:

Director, Reporting Compliance, Wage and Investment Division

Director, Campus Reporting Compliance, Small Business/Self-Employed Division

Corrective Action Monitoring Plan

The IRS will monitor this corrective action as part of our internal management control system.

RECOMMENDATION 2: For determining the net benefit of using soft notices in the AUR Program as an alternative approach for addressing compliance issues.

CORRECTIVE ACTION:

We agree with this recommendation. We have determined that the soft notice is a beneficial as a self correction method and has educational benefits. We have developed a plan, dependent upon funding allocations, beginning with Tax Year 2010, to expand the use of soft notices. We will continue to monitor results and evaluate the net benefits derived from the use of soft notices.

Implementation Date:

January 15, 2013

Responsible Official:

Director, Reporting Compliance, Wage and Investment Division

Corrective Action Monitoring Plan

The IRS will monitor this corrective action as part of our internal management control system.

[1] The data processing arm of the IRS. The campuses process paper and electronic submissions, correct errors, and forward data to the Computing Centers for analysis and posting to taxpayer accounts.

[2] The CP 2000 notice is an IRS letter sent to a taxpayer to resolve discrepancies between income, credits, and/or deductions claimed on a tax return and those reported by a third party, as well as to propose an additional tax assessment.

[3] An IRS letter sent to taxpayers notifying them of an increase in the amount of taxes they owe.

[4] See Appendix V for an example of the CP 2057 notice.

[5] Plans for Evaluating the Use of Soft Notices in Addressing Underreporting Can Be Enhanced (Reference Number 2010-30-089, dated August 27, 2010).

[6] The IRS database that maintains transactions or records of individual tax accounts.

[8] A 12-month accounting period for keeping records on income and expenses used as the basis for calculating the annual taxes due. For most individual taxpayers, the tax year is synonymous with the calendar year.

[9] The DCI is a web-based collection instrument which was used to collect data about responses from taxpayers who received soft notices (Soft Notice Group) or CP 2000 notices (AUR Control Group) during the pilot.

[10] The report’s results were generated from the sample of taxpayers who were sent a soft notice due to underreported income on their Tax Year 2007 tax return. This is a draft report, with the final report scheduled to be issued in August 2011.

[11] Amended U.S. Individual Income Tax Return (Form 1040X).

[12] The percentage difference in the 1040X – Fully Agreed With Soft Notice measure can be attributed to the different methodologies used by the TIGTA and the IRS. The TIGTA’s analysis included only the soft notices that were mailed to taxpayers, whereas the IRS used the total of soft notices selected, which contained more records.

[13] To compute the TIGTA total for the Dollars Per FTE measure, we divided the total case revenue by 4.23, which is the total number of FTEs including those for Submission Processing staff to process amended returns. See Appendix I for details. An FTE is a measure of labor hours in which 1 FTE is equal to 8 hours multiplied by the number of compensable days in a particular fiscal year. For Fiscal Year 2009, 1 FTE was equal to 2,088 staff hours. For Fiscal Year 2010, 1 FTE is equal to 2,088 staff hours.

[14] This analysis does not include the time and resources incurred by Submission Processing staff to process amended tax returns.

[15] A 12-month accounting period for keeping records on income and expenses used as the basis for calculating the annual taxes due. For most individual taxpayers, the tax year is synonymous with the calendar year.

[16] Plans for Evaluating the Use of Soft Notices in Addressing Underreporting Can Be Enhanced (Reference Number 2010-30-089, dated August 27, 2010).

[17] See Appendix IV for a detailed description of the 16 measures.

[18] The AUR repeater file identifies recent trends in underreporting behavior for all taxpayers with income discrepancies.

[19] The 30,819 taxpayers selected for the AUR Control Group included 27,135 who were sent a CP2000 notice and 3,684 who the IRS closed without contact.

[20] Amended U.S. Individual Tax Return (Form 1040X).

[21] The IRS database that maintains transactions or records of individual tax accounts.

[22] The 29,528 records in the DCI include activity related to responses from taxpayers for both the Soft Notice Group and the AUR Control Group.

[23] A measure of labor hours in which 1 FTE is equal to 8 hours multiplied by the number of compensable days in a particular fiscal year. For Fiscal Year 2009, 1 FTE was equal to 2,088 staff hours. For Fiscal Year 2010, 1 FTE is equal to 2,088 staff hours.

[24] A management system designed to relate workload to staffing levels.

[25] A centralized storage facility used by the TIGTA to maintain critical historical data extracted from IRS databases needed for analysis during audits and investigations.

[26] Amended U.S. Individual Tax Return (Form 1040X).

[27] A measure of labor hours in which 1 FTE is equal to 8 hours multiplied by the number of compensable days in a particular fiscal year. For Fiscal Year 2009, 1 FTE was equal to 2,088 staff hours. For Fiscal Year 2010, 1 FTE was equal to 2,088 staff hours.

[28] A 12-month accounting period for keeping records on income and expenses used as the basis for calculating the annual taxes due. For most individual taxpayers, the tax year is synonymous with the calendar year.

[29] All dates, monetary, and taxpayer identifying information contained in this example are hypothetical.

[30] The soft notices, formerly known as CP 2057 Notice, do not require that the taxpayer pay more tax, provide documentation, or even respond to the IRS. Although the notice requites the taxpayer to file an amended tax return, if appropriate, it is not required. Instead, the notices are designed to serve as an educational tool, encourage self-correction, and improve voluntary compliance.

[31] The CP 2000 is an IRS letter sent to a taxpayer to resolve discrepancies between income, credits, and/or deductions claimed on a tax return and those reported by a third party, as well as to propose and additional tax assessment.