Background

The Automated Referral System Has Improved Control Over Complex Taxpayer Questions

Management Information Is Needed to Provide Reasonable Assurance that Eliminating the Program Is Feasible

Recommendation 1:

An Organizational Commitment Is Needed to Provide Reasonable Assurance that Customer Service Goals Are Achieved

Management and Internal Controls Are Needed to Mitigate Risks to Customer Service Goal Accomplishment

The Centralized Quality Review Site Needs to Improve the System Used to Assess the Quality of Responses Provided to Taxpayers

Appendix I – Detailed Objective, Scope, and Methodology

Appendix II – Major Contributors to This Report

Appendix III – Report Distribution List

Appendix IV – Glossary of Terms

Appendix V – The Automated Referral System’s Tax Law Categories and Who Responded to Questions in Each

Background

The Internal Revenue Service’s (IRS) mission is to provide America’s taxpayers top quality service by helping them understand and meet their tax responsibilities and by applying the tax law with integrity and fairness to all. In 1993, the President issued Executive Order 12862, "Setting Customer Service Standards," which set a goal for the Federal government to deliver customer service equal to the best in business.

A taxpayer can contact the IRS’ toll-free telephone system to obtain answers to tax law/procedural questions. A Customer Service Representative (CSR) will converse with a taxpayer to provide an immediate response. If a taxpayer’s question concerns one of several difficult tax law categories, the telephone call is routed to a Customer Service Screener (CSS), and the taxpayer is informed that a specialty group will provide a response within 2 workdays. The CSS inputs the question into the Automated Referral System (R-Mail Program), and the question is immediately available for the specialty group to answer.

During the filing season, the IRS’ Compliance function personnel, located throughout the country, are used to answer these difficult questions. Depending on the taxpayer’s preference, the R-Mail Program response is provided by telephone, by electronic mail (e-mail), or by written correspondence.

We performed the audit at the Headquarters Offices of the Small Business/Self-Employed and Wage and Investment Divisions in Washington, DC, and Atlanta, GA, respectively; the Centralized Quality Review Site in Philadelphia, PA; and Territory Offices in Pittsburgh, PA, and Jacksonville, FL. The audit was performed between October 2000 and April 2001 in accordance with Government Auditing Standards.

The scope of our audit was limited because the database, provided by the IRS, did not include approximately 172,000 of the 892,056 records for Fiscal Year (FY) 2000. Beyond not having the data available for analysis, the limitation did not affect our findings and conclusions. We have requested validated copies of the R-Mail Program database for FY 2000 and 2001 for a more in-depth analysis.

Details of our audit objective, scope, and methodology are presented in Appendix I. Major contributors to this report are listed in Appendix II. In addition, we have included a Glossary of Terms in Appendix IV.

The Automated Referral System Has Improved Control Over Complex Taxpayer Questions

The R-Mail Program is an improvement over the prior paper-reliant system. The prior system required a CSR to transcribe a question from a recorded taxpayer message and fax the message to the IRS’ Compliance function. This system resulted in poor workload controls and management information. In FY 2000, the R-Mail Program was implemented nationwide to replace this paper system.

We found that the R-Mail Program provided expedient service to taxpayers who had complex tax questions that could not be answered immediately by a CSR. During FY 2000, the IRS reported that it processed 892,056 questions with an average response time of 1.6 workdays. This workload consisted of 803,707 telephone responses, 34,553 e-mail responses, and 53,796 written responses.

Because of the complexity of the tax questions, the IRS budgeted professional resources to answer 2 questions an hour at a cost of over $18.9 million, or $34 per call, for the 2001 filing season.

Management Information Is Needed to Provide Reasonable Assurance that Eliminating the Program Is Feasible

For the 2001 filing season, the IRS eliminated four of the R-Mail Program tax law categories. The IRS did not perform any analyses of the effects that answering these complex questions would have on CSRs’ handle time, the adequacy of the specialized training, or the quality of responses pertaining to the four categories.

The IRS’ strategic plan assumptions call for eliminating the Compliance function resources to assist toll-free operations and transferring the responsibility for answering all taxpayer questions to the CSRs. Ultimately, the R-Mail Program will be eliminated and CSRs will answer all taxpayer questions.

We agree with the IRS’ efforts to reduce the use of Compliance function resources to answer R-Mail Program questions. However, the IRS does not have the essential information it needs to make a decision to transfer complex tax law categories to the toll-free telephone system and eventually eliminate the R-Mail Program. This information is important, because the General Accounting Office’s Standards for Internal Control in the Federal Government state that relevant, reliable, and timely information is needed to achieve objectives.

The IRS did not assess the accuracy rates of the responses for the four categories prior to the transfer to the toll-free telephone system for the 2001 filing season. Further, there is no specific information on the accuracy and productivity rates that the toll-free telephone system assistors attained for these categories. Consequently, the IRS cannot discern whether transferring these categories to the toll-free telephone system was helpful, or harmful.

The rationale for using Compliance function personnel is that they are highly trained and experienced professionals. The R-Mail Program provides ample time to analyze the taxpayer’s question and perform research without the stress of having to respond directly and immediately to the taxpayer on complex issues. Without sufficient time to research answers to complex questions, there is an additional risk that CSRs will provide inaccurate or incomplete responses.

Recommendation

1. In order to assess the feasibility of eliminating the R-Mail Program, the Commissioners of the four divisions need to develop baseline information on the current R-Mail Program tax law categories. This information should include research time, handle time, and wrap time to determine whether these questions are conducive to the toll-free environment.

Management’s Response: Management’s response was due on September 20, 2001. As of September 21, 2001, management had not responded to the draft report.

An Organizational Commitment Is Needed to Provide Reasonable Assurance that Customer Service Goals Are Achieved

The Wage and Investment Division has functional responsibility for maintaining and operating the R-Mail Program’s computer system, and for ensuring that CSRs are trained to process questions pertaining to the R-Mail Program categories. The other three divisions provide Compliance function resources to respond to complex taxpayer questions. The only established goal is to answer a taxpayer’s question within 2 workdays.

None of the managers involved with the R-Mail Program in these divisions accepted full responsibility for overall management. In fact, each operating division was allowed to establish its own workday response goal for the respective tax law categories. We recognize that the IRS organization is based on end-to-end accountability for the taxpayers they serve. However, the same service levels need to be maintained by all divisions.

The reason why managers do not recognize full responsibility is that the R-Mail Program has not been included as an integral part of any division and is considered an interim system until CSRs can answer all questions through the toll-free telephone system. In effect, the R-Mail Program operates as a workload distribution system rather than a fully supported program.

The General Accounting Office’s Standards for Internal Control in the Federal Government require that controls provide reasonable assurance that the objectives of an organization are being achieved. A factor affecting the control environment is the agency’s organizational structure. It provides management’s framework for planning, directing, and controlling operations to achieve agency objectives. A good internal control environment requires that the agency’s organizational structure clearly define key areas of authority and responsibility and establish appropriate lines of reporting.

Without an organizational commitment with management accountability, improved management and internal controls, and dedicated resources, the R-Mail Program cannot be effectively managed. The absence of this organizational commitment has left the Program without an infrastructure that includes goals, objectives, critical success factors, performance measures, and adequate managerial oversight. As a result, the IRS is at greater risk of not achieving its customer service goal, and taxpayers are at risk of not receiving responses to their questions.

Recommendation

2. If the IRS decides to continue with the R-Mail Program, the Commissioners of the four divisions need to fully commit to the Program by structuring it to include goals, objectives, critical success factors, performance measures, and managerial accountability.

Management and Internal Controls Are Needed to Mitigate Risks to Customer Service Goal Accomplishment

The IRS has not established a system of controls to monitor quality, ensure consistency of responses, prevent or deter an employee from closing a case without responding to a taxpayer’s question, and alert management when an excessive number of cases are closed in one day. As a result, the existing control environment does not ensure the Customer Service goal, to serve each taxpayer, is achieved.

The Standards for Internal Control in the Federal Government state that internal control should provide reasonable assurance that the objectives of an organization are being achieved. Also, monitoring should assess the quality of performance and should generally be designed to ensure that this occurs in the course of normal operations. Further, all documentation and records are to be properly managed and maintained.

Managerial reviews to assess the quality of responses to taxpayers were not performed

Employees assigned to the R-Mail Program worked independently without managerial supervision. During FY 2000, 803,707 (90 percent) of the R-Mail Program responses were provided by telephone. The toll-free telephone system has the capability to monitor incoming calls. However, the R-Mail Program design, which uses outbound calls, does not include monitoring capabilities. This limits managerial oversight.

In addition, a manager at one call site advised that they cannot monitor telephone conversations between R-Mail Program employees and taxpayers because of the local union agreement. At another call site, a manager advised that telephone monitoring was not performed because employees were trusted to complete their R-Mail Program assignments.

Our analysis of the FY 2000 database showed that employees did not always document the response or they entered limited information in the "Response" section of the case file that is displayed on the computer screen. However, managers did not review the responses that were documented in the R-Mail system at either call site.

Without monitoring capabilities, the R-Mail Program has few capabilities to assess quality or to provide employees with evaluative feedback.

Indicators of systemic problems or system abuse went undetected

Managerial reviews or data analyses were not performed on the R-Mail Program database. This condition was caused, in part, by the emphasis on managing the inventory of taxpayer questions primarily for response timeliness. Managers at the call sites advised that they relied primarily upon inventory reports to manage the program.

In our initial analysis of the FY 2000 database, we found indicators of system abuse. Specifically, we identified 97 employees who closed between 50 and 589 cases per day on 411 occasions, involving 34,346 taxpayer questions. The large number of daily closures is in stark contrast to the IRS’ budgeting standard of 16 cases per day.

Referrals of potential system abuse involving these employees are pending. In addition, during the 2001 filing season, we identified indications of misconduct concerning three employees who closed R-Mail Program cases and may not have responded to taxpayers’ questions.

While some of these employees may have had legitimate reasons for closing cases in such high volume, there are no computer system controls to alert management to unusual situations. The R-Mail Coordinators at various locations may analyze the system to identify and close duplicate questions that were inadvertently entered into the system, or for instances when a taxpayer calls more than once with the same question. However, 86 of the 97 employees discussed above were not R-Mail Coordinators.

The IRS was unable to provide a reasonable explanation for how a large number of cases could be properly closed within one day. As a result, there is no reasonable assurance that taxpayers actually received responses to their questions.

One division bypassed basic computer system controls

The IRS Commissioner testified to Congress in April 2000 that one advantage of the R-Mail Program would be improved controls through the elimination of clerical time associated with the logging and filing of paper referrals.

During the 2000 filing season, one division printed copies of each case and assigned the paper copies to employees to respond to taxpayers’ questions. When a sufficient number of cases were completed an employee would close the cases on the R-Mail Program database.

By circumventing basic system controls, this division inefficiently used resources, increased the risk of losing control over paper documents, and risked not answering taxpayers’ questions. Daily and weekly management reports would be inaccurate, and the efficiencies of the computer system were not realized because of the clerical time associated with logging and filing paper referrals. In addition, this division’s R-Mail Coordinator advised that some of the 2000 filing season paper documents were used as training aids, and the documentation was then destroyed.

Recommendation

3. If the IRS decides to continue with the R-Mail Program, the Commissioners of the four divisions need to develop a system of controls that includes:

• Procedures for managerial reviews of all three types of responses.
• Procedures requiring that the information provided on telephone responses be entered in the R-Mail Program database.
• Procedures to allow written correspondence and e-mail responses to be analyzed for trends.
• A computer-generated alert of unusual situations that would allow managers an opportunity to analyze and resolve problems as they occur.

The Centralized Quality Review Site Needs to Improve the System Used to Assess the Quality of Responses Provided to Taxpayers

The Centralized Quality Review Site (CQRS) cannot assess the quality of the actual telephone responses provided to taxpayers because the IRS does not have monitoring capabilities for outbound calls. Over 90 percent of the responses are made through these telephone systems. Instead, the CQRS can only assess the quality of responses to test questions it submits to the R-Mail Program.

The CQRS did not meet the sampling plan requirements that were developed by the IRS’ Statistics of Income (SOI) function. The IRS’ SOI function established the sampling plan parameters for the 2000 filing season with input from the CQRS.

The CQRS could not adhere to the sampling plan requirements because of other conflicting priorities and inadequate resources to meet all quality review responsibilities. The CQRS was only able to process 468 of the 875 test questions and did not request that the SOI Statistical Support Section analyze and interpret the sampling results from the 2000 filing season. As a result, the IRS is unable to make a reasonable determination of the R-Mail Program’s accuracy level.

This situation arose because the CQRS had to request e-mail responses to the majority of its test questions, due to the fact that it only had sufficient resources and time to install one additional non-IRS telephone line. If the number of telephone lines were significantly increased, quality measurement would be more representative of the primary method of providing responses to taxpayer questions processed through the R-Mail Program.

During the 2000 filing season, the CQRS submitted 468 test questions to assess the quality of the R-Mail Program’s responses. It did not receive responses to 33 (7 percent) of 468 test questions. The IRS did not review the 33 cases to determine why the responses were not received by the CQRS, even though the R-Mail Program database showed that responses were provided.

The IRS Statistical Sampling Handbook provides guidance on developing a sampling and data collection plan. The handbook states that when developing a sampling plan, a decision on how to handle non-responses needs to be made.

The IRS cannot provide reasonable assurance that taxpayers are provided responses to their questions. The fact that the CQRS did not receive responses to seven percent of its test questions could indicate systemic problems or system abuse. The CQRS speculated that the problem was caused by employees not being familiar with the R-Mail Program or a training problem concerning e-mail transmissions.

However, potential system abuse was not considered an alternative reason for not receiving responses to the test questions. In reality, the IRS does not have any information concerning the actual cause(s) of this problem. The CQRS "no response" cases were not elevated to responsible R-Mail Program management officials for analysis and resolution. Consequently, the problems identified during the 2000 filing season could have continued to be a problem in the 2001 filing season.

Recommendations

If the IRS decides to continue with the R-Mail Program, the Commissioners of the four divisions should:

1. Develop long-term strategic plans that provide an effective and efficient means for managers and the CQRS to monitor telephone conversations.
2. Identify and implement a quality measurement process for the R-Mail Program that provides sufficient information on the quality levels achieved.

The CQRS should:

3. Develop procedures to timely report accuracy levels, systemic weaknesses, and other problems or trends to the appropriate levels of management.

Detailed Objective, Scope, and Methodology

The overall objective of the audit was to determine if the Internal Revenue Service (IRS) established an effective system to provide quality responses to taxpayer questions received through the Automated Referral System (R-Mail Program). The following steps were performed to accomplish the audit objective.

1. To determine the effectiveness of the IRS’ efforts to ensure the accuracy and timeliness of information provided via R-Mail Program responses to taxpayer telephone questions, we:

1. Reviewed the strategic plans to reduce and eventually eliminate the R-Mail Program in order to provide immediate responses to taxpayer telephone questions.

1. Evaluated the adequacy of the overall planning assumptions and the feasibility of effectively implementing these changes.
2. Reviewed program goals, performance measures, critical success factors, and the follow-up system for monitoring progress.
3. Reviewed the minutes of the Risk Management Group meetings for Fiscal Year (FY) 2000 that ceased operations on September 30, 2000. Reviewed the minutes of the FY 2001 Joint Operations Committee that commenced operations October 1, 2000.
4. Reviewed the National Taxpayer Advocate’s FY 2000 Annual Report to Congress to determine additional taxpayer problems that directly effect the R-Mail Program.

1. Evaluated the adequacy of the decision support system developed by the Director, Customer Account Services, Wage and Investment Division, for processing taxpayer questions through the R-Mail Program.

1. Evaluated the process for providing taxpayers with accurate and timely tax information. We evaluated the methods used to control, assign, and complete cases in accordance with established standards.
2. Evaluated the process for assuring the timeliness of responses. Currently, the goal is to provide a response to a taxpayer question within two business days.
3. Evaluated quality control review procedures and the process used to ensure that a sufficient number of responses are reviewed to provide a statistically valid measurement of the quality of tax information provided to taxpayers.

1. Reviewed the process for providing quality review results to the R-Mail Program sites and the usefulness of the quality review information that was provided to the various IRS divisions.
2. Reviewed the process for providing adequate electronic reference material to the Compliance function employees.

1. Assessed the Compliance function plans for increasing productivity and the timeliness of responses.

1. Reviewed the methods for assigning questions to employees.
2. Reviewed the methods for assigning Compliance function personnel to respond to R-Mail Program questions throughout the filing season (Territory Area Offices in Pittsburgh, PA and Jacksonville, FL).

1. Analyzed the R-Mail Management Information Reports that were used to manage the inventory of questions.

Major Contributors to This Report

Gordon C. Milbourn III, Assistant Inspector General for Audit (Small Business and Corporate Programs)

Philip Shropshire, Director

Edmond G. Watt, Audit Manager

Richard J. Turner, Senior Auditor

Benjamin Hawkins, Auditor

Britt Molitoris, Auditor

John Ochal, Auditor

Report Distribution List

Commissioner N:C

Deputy Commissioner N:DC

Commissioner, Large and Mid-Size Business Division LM

Commissioner, Small Business/Self-Employed Division S

Commissioner, Tax Exempt and Government Entities Division T

Commissioner, Wage and Investment Division W

Director, Compliance, Small Business/Self-Employed Division S:C

Director, Compliance, Wage and Investment Division W:C

Director, Customer Account Services, Small Business/Self-Employed Division S:CAS

Director, Customer Account Services, Tax Exempt and Government Entities Division T:CAS

Director, Customer Account Services, Wage and Investment Division W:CAS

Director, Joint Operations Center W:CAS:JOC

Chief Counsel CC

National Taxpayer Advocate TA

Director, Legislative Affairs CL:LA

Director, Office of Program Evaluation and Risk Analysis N:ADC:R:O

Office of Management Controls N:CFO:F:M

Audit Liaisons:

Commissioner, Small Business/Self-Employed Division S

Commissioner, Tax Exempt and Government Entities Division T

Commissioner, Wage and Investment Division W

Glossary of Terms

Abuse – The General Accounting Office’s Government Auditing Standards defines abuse as the conduct of a government program falling far short of societal expectations for prudent behavior.

Centralized Quality Review Site (CQRS) – The Internal Revenue Service’s (IRS) corporate monitoring system for its main toll-free telephone lines. The role of the CQRS is to provide a single product review system that provides qualitative and quantitative data for use in determining whether the IRS is fulfilling its three goals of increasing voluntary compliance, reducing taxpayer burden, and improving quality-driven productivity and customer satisfaction.

Compliance – A term used to describe the IRS’ Collection and Examination functions whose mission is to ensure taxpayer compliance through collecting delinquent taxes and examining tax returns to determine correct tax liabilities.

Customer Service Representative (CSR) – An IRS employee who may perform work on the telephones, as well as paper inventory resulting from telephone contacts.

E-Mail – Electronic text mail used for the transmission of messages sent from a computer terminal or computer system.

Filing Season – The period from January 1 through April 15 of each calendar year during which most taxpayers typically file their individual income tax returns.

Fiscal Year (FY) – A 12-month period that ends on the last day of any month except December. The IRS’ fiscal year runs from October 1 of a calendar year through September 30 of the following calendar year.

Handle Time – The amount or percentage of time that CSRs spend talking on an inbound call and performing after-call work.

R-Mail Coordinator – Serves as a liaison between the IRS divisions. The coordinators ensure sufficient resources are available to answer the inventory of questions.

Toll-Free Telephone System – The IRS has a variety of toll-free telephone lines to assist taxpayers in meeting their Federal tax obligations (tax law inquiries – 800-829-1040, questions regarding notices / bills received from IRS – 800-829-8815, and inquiries on tax refund status – 800-808-4262).

Wrap Time – A work status during which a CSR is completing work related to the prior call, is preparing for the next call or requires time to relieve a stressful call situation.

The Automated Referral System’s Tax Law Categories and Who Responded to Questions in Each

Compliance = Questions answered by Compliance function employees.

CSRs = Customer Service Representatives answered these questions through the toll-free telephone system.