IRS Personnel Administration: Use of Enforcement Statistics in Employee
Evaluations (Letter Report, 11/30/98, GAO/GGD-99-11).
Pursuant to a congressional request, GAO reviewed the Internal Revenue
Service's (IRS) managers' compliance with legislative and policy
prohibitions against using enforcement statistics in employee
evaluations and the requirement that managers certify each quarter
whether violations had occurred, focusing on: (1) the extent to which
IRS' certification process identified violations of law and policy; (2)
IRS employees' perceptions of the use of tax enforcement results in
their annual performance evaluations; (3) supervisors' use of tax
enforcement results in written employee performance evaluations; and (4)
IRS' efforts to revise the certification process.
GAO noted that: (1) for fiscal years 1996 and 1997, district and service
center directors submitted 368 quarterly certifications that reported 11
potential violations; (2) GAO identified several systemic weaknesses
that affected the reliability of the certifications; (3) specifically,
GAO found: (a) some confusion among IRS officials about what constituted
a violation; (b) inadequate guidance about specific actions directors
should take to identify violations; (c) a failure to integrate
performance evaluations and the certification process; and (d) unclear
guidance on sanctions that could be applied against managers for
misusing tax enforcement results or submitting false certifications; (4)
GAO's survey of a statistically representative sample of examination and
collection employees showed a widespread perception that managers
considered enforcement results when preparing annual performance
evaluations; (5) GAO estimated that 75 percent of front-line employees
and 81 percent of group managers perceived that tax enforcement results
affected their most recent performance evaluation; (6) about 70 percent
of front-line employees said they based their perception in part on
information communicated to them verbally in staff meetings or
performance feedback sessions with their managers; (7) and about 36
percent of front-line employees indicated tax enforcement results were
used in their written performance evaluations; (8) 9 percent of
employees received a written evaluation that contained a tax enforcement
result and an estimated 69 percent contained narrative that employees
could have interpreted as inappropriate references to tax enforcement
results but which did not violate IRS guidance; (9) about 41 percent of
the evaluations in GAO's sample mentioned process measures dealing with
the age of the cases in the employee's workload inventory and the number
of cases worked within guidelines established for closing cases; (10)
IRS has undertaken several steps to strengthen the certification
process; (11) although these actions address some of the weaknesses of
the current system, IRS' revised guidance has few examples of the
appropriate and inappropriate use of tax enforcement results in written
evaluations; (12) the revised guidance does not clearly inform managers
about potential sanctions for inappropriate use of tax enforcement
results; and (13) IRS' new independent review process is geared toward
identifying written violations and not violations communicated verbally.
--------------------------- Indexing Terms -----------------------------
REPORTNUM: GGD-99-11
TITLE: IRS Personnel Administration: Use of Enforcement Statistics
in Employee Evaluations
DATE: 11/30/98
SUBJECT: Tax law
Performance measures
Personnel evaluation
Federal employees
Internal controls
Internal audits
Surveys
Sanctions
Noncompliance
Statistical data
******************************************************************
** This file contains an ASCII representation of the text of a **
** GAO report. Delineations within the text indicating chapter **
** titles, headings, and bullets are preserved. Major **
** divisions and subdivisions of the text, such as Chapters, **
** Sections, and Appendixes, are identified by double and **
** single lines. The numbers on the right end of these lines **
** indicate the position of each of the subsections in the **
** document outline. These numbers do NOT correspond with the **
** page numbers of the printed product. **
** **
** No attempt has been made to display graphic images, although **
** figure captions are reproduced. Tables are included, but **
** may not resemble those in the printed version. **
** **
** Please see the PDF (Portable Document Format) file, when **
** available, for a complete electronic file of the printed **
** document's contents. **
** **
** A printed copy of this report may be obtained from the GAO **
** Document Distribution Center. For further details, please **
** send an e-mail message to: **
** **
** **
** **
** with the message 'info' in the body. **
******************************************************************
Cover
================================================================ COVER
Report to the Chairman of the Committee on Ways and Means and the
Chairman of the Subcommittee on Oversight, House of Representatives
November 1998
IRS PERSONNEL ADMINISTRATION - USE
OF ENFORCEMENT STATISTICS IN
EMPLOYEE EVALUATIONS
GAO/GGD-99-11
IRS Personnel Administration
(268823)
Abbreviations
=============================================================== ABBREV
IRS - Internal Revenue Service
NTEU - National Treasury Employees Union
Letter
=============================================================== LETTER
B-279232
November 30, 1998
The Honorable Bill Archer
Chairman, Committee on Ways and Means
House of Representatives
The Honorable Nancy L. Johnson
Chairman, Subcommittee on Oversight
Committee on Ways and Means
House of Representatives
Both the law and Internal Revenue Service (IRS) policy prohibit IRS
managers from using records of tax enforcement results, such as
dollars collected and taxes assessed, to evaluate employee
performance or establish production goals. The underlying reason for
these prohibitions is that employees may try to achieve statistical
benchmarks through inappropriate or abusive enforcement actions, such
as unwarranted seizures or unreasonable tax assessments, instead of
acting strictly on the merits of the case.
This report responds to your request for information on IRS managers'
compliance with legislative and policy prohibitions and the
requirement that managers certify each quarter whether violations had
occurred. Accordingly, this report discusses our evaluation of (1)
the extent to which IRS' certification process identified violations
of law and policy, (2) IRS employees' perceptions of the use of tax
enforcement results in their annual performance evaluations, (3)
supervisors' use of tax enforcement results in written employee
performance evaluations, and (4) IRS' efforts to revise the
certification process.
BACKGROUND
------------------------------------------------------------ Letter :1
Since 1973, IRS Policy Statement P-1-20 has prohibited using records
of tax enforcement results to evaluate enforcement officers or to
impose or suggest production goals or quotas. An enforcement officer
is an employee who exercises judgment with regard to determining tax
liability or the ability to pay. Enforcement officers include
employees directly involved in collection, examination, or criminal
investigation functions as well as appeals officers and reviewers.
Each function has its own set of tax enforcement results. For
example, for the examination function, tax enforcement results
include data such as the number of hours spent per return, amount of
dollars proposed per return, and number of cases closed. For the
collection function, tax enforcement results include the number of
liens, levies, and seizures; amount of dollars collected; and number
of cases closed.
As a general rule, the policy prohibited managers from using any
statistic that measures quantity, time per case, type of disposition,
or dollar value. Instead, employees are to be evaluated on the basis
of the quality of the work, including the use of enforcement tools,
timeliness of action, and application of tax law. However, the
policy statement permitted IRS officials to use statistics on tax
enforcement results when carrying out certain management activities,
such as preparing long-range and financial plans, allocating
resources, and evaluating program effectiveness. In addition,
managers were permitted to use tax enforcement results when
evaluating employee performance during a case review, for example by
noting whether a revenue officer appropriately levied a bank account
given the facts of the specific case. Managers were also permitted
to discuss with employees the number of cases processed, the amount
of time spent on cases, and the kind of results obtained, as long as
the discussion was based on a review of the results that were
appropriate for specific cases processed by the employees and goals
and quotas were not involved.
In 1988, along with other changes, the Taxpayer Bill of Rights\1 also
prohibited the use of records of tax enforcement results to evaluate
employees, but the prohibition only applied to employees directly
involved in collection activities and their immediate supervisors.
The law specified that IRS would be in compliance with the
legislative requirements as long as IRS followed its policy
statement. To help ensure compliance, the law also required each
district director to certify each quarter whether violations had
occurred and the corrective action taken, if needed. Because IRS
service centers also have collection responsibilities, IRS imposed
the same certification requirement on its service center directors.
Generally, the self-certification process requires managers to submit
certifications to the next higher level for review. Quarterly, group
managers are to submit certifications to their branch chiefs, branch
chiefs are to submit certifications to their division chiefs, and
division chiefs are to submit certifications to their directors.
Directors, relying on lower-level managers' certifications, are to
submit their certifications to the Commissioner of Internal Revenue
through the Assistant Commissioner (Collection).
During hearings held by the Senate Committee on Finance in September
1997, witnesses alleged that IRS' focus on enforcement statistics was
encouraging enforcement officers to take unnecessary and illegal
enforcement actions against taxpayers. Prompted by these hearings,
IRS' Internal Audit reviewed the use of enforcement statistics by
collection personnel in 12 districts, as well as at national and
regional offices, and found an atmosphere largely driven by
statistical measures. Internal Audit also found similar conditions
in its July 1998 report on the use of enforcement results by the
examination functions in 12 district offices.
The IRS Restructuring and Reform Act of 1998 (Public Law 105-206,
July 22, 1998) repealed the Taxpayer Bill of Rights prohibitions,
which only covered collection employees, and instead expanded the
prohibitions to cover all employees. The act also expanded the
quarterly certifications to cover each appropriate supervisor.
--------------------
\1 Taxpayer Bill of Rights, Public Law 100-647, Nov. 10, 1988.
RESULTS IN BRIEF
------------------------------------------------------------ Letter :2
For fiscal years 1996 and 1997, district and service center directors
submitted 368 quarterly certifications that reported 11 potential
violations, of which 4 were subsequently determined by IRS to be
actual violations. However, we identified several systemic
weaknesses that affected the reliability of the certifications.
Specifically, we found (1) some confusion among IRS officials about
what constituted a violation, (2) inadequate guidance about specific
actions directors should take to identify violations, (3) a failure
to integrate performance evaluations and the certification process,
and (4) unclear guidance on sanctions that could be applied against
managers for misusing tax enforcement results or submitting false
certifications.
While the quarterly certifications showed few violations, our survey
of a statistically representative sample of examination and
collection employees showed a widespread perception that managers
considered enforcement results when preparing annual performance
evaluations. We estimated that 75 percent of front-line employees
(revenue agents, tax auditors, and revenue officers) and 81 percent
of group managers perceived that tax enforcement results affected
their most recent performance evaluation. When asked about various
potential bases for their perception, about 70 percent of front-line
employees, who indicated that tax enforcement results affected their
evaluations, said they based their perception in part on information
communicated to them verbally in staff meetings or performance
feedback sessions with their managers. And about 36 percent of
front-line employees indicated tax enforcement results were used in
their written performance evaluations.
On the basis of our review of the two most recent performance
evaluations for a statistically representative sample of front-line
employees, we estimated that 9 percent of employees received a
written evaluation that, according to our application of IRS' revised
guidelines, contained a tax enforcement result. Although this
incidence rate is low relative to our employee survey results, we
also found that an estimated 69 percent of the evaluations in our
sample contained narrative that employees could have interpreted as
inappropriate references to tax enforcement results but which did not
violate IRS guidance. For example, an estimated 50 percent of the
evaluations in our sample included either general or case-specific
references to enforcement-related activities, such as employees'
overall use of collection tools or auditing techniques, that IRS does
not consider violations. In addition, we estimated that about 41
percent of the evaluations in our sample mentioned process measures
dealing with the age of the cases in the employee's workload
inventory and the number of cases worked within guidelines
established for closing cases, which could be interpreted by
employees as a surrogate for prohibited measures, such as the "number
of cases closed."
As a result of IRS' Internal Audit reviews, IRS has undertaken
several steps to strengthen the certification process, such as (1)
expanding the number of employees covered by certifications, (2)
revising its guidance on the proper use of tax enforcement results,
(3) requiring an annual independent review to verify the accuracy of
the certifications, and (4) requiring the preparer and reviewer to
attest that the tax enforcement results were not improperly used in
the evaluation. Although these actions address some of the
weaknesses of the current system, IRS' revised guidance has few
examples of the appropriate and inappropriate use of tax enforcement
results in written evaluations. In addition, the certification form
includes only a general statement of compliance rather than a
statement specifically detailing what managers are certifying.
Furthermore, the revised guidance does not clearly inform managers
about potential sanctions for inappropriate use of tax enforcement
results. In addition to these potential sources of confusion, IRS'
new independent review process is geared toward identifying written
violations and not violations communicated verbally, which our
employee survey indicates is the most common way tax enforcement
results are communicated.
SCOPE AND METHODOLOGY
------------------------------------------------------------ Letter :3
To determine the extent to which IRS' current certification process
identified violations, we reviewed fiscal year 1996 and 1997
quarterly certifications filed by district and service center
directors--a total of 288 for the district offices and 80 for the
service centers. We also obtained summary information on the
potential violations for the first quarter of fiscal year 1998. To
obtain information on IRS' quarterly certification policy and
procedures, we interviewed officials from IRS' National Office; the
Western Regional Office; the Georgia, Kansas/Missouri, and Northern
California District Offices; and the Atlanta and Fresno Service
Centers. We visited these offices because of their close proximity
to our offices. We also surveyed officials in all districts and
service centers on the quarterly certification process they had in
place as of the last quarter of calendar year 1997. Appendix I
contains a summary of the violations reported in the quarterly
certifications we reviewed.
To determine IRS employees' perceptions of the use of tax enforcement
results, we surveyed a statistically representative sample of 1,104
of 20,974 IRS front-line employees and group managers in the
collection and examination functions. The survey included questions
on whether employees perceived that tax enforcement results
influenced their evaluations, the basis for those perceptions, and
the use of tax enforcement results to establish group goals. We also
analyzed the comments provided voluntarily by IRS employees on the
survey. Appendixes II through VI contain our sampling methodology
and survey results. All results are reported at the 95-percent
confidence level. Sampling errors for all of our estimates are less
than plus or minus 10 percentage points unless otherwise reported.
To determine how tax enforcement results were used in written
employee evaluations, we reviewed the 2 most recent evaluations for a
statistically representative sample of 300 of 18,292 examination and
collection front-line employees. We reviewed the evaluations to
determine whether tax enforcement results were used to rate the
employee in violation of IRS guidance. We discussed potential
violations with IRS officials and considered their opinions on
whether they were violations. We also reviewed the evaluations to
determine whether the narratives used tax enforcement results to
generally characterize employee performance or to discuss specific
case examples. Appendix II contains our sampling methodology and
appendix VII contains the results of our review of employee
evaluations. All results are reported at the 95-percent confidence
level and, unless noted otherwise, the data have sampling errors of
no more than 8 percent.
To evaluate IRS' efforts to revise the certification process, we
discussed IRS proposals with senior-level officials who were involved
in developing the new process. We reviewed the existing and proposed
guidance provided to managers and identified the differences between
the existing and proposed certification processes. We also met with
National Treasury Employees Union (NTEU) officials to obtain their
perspective on the existing and proposed processes. We compared the
proposed changes with the results of our review of the current
process to determine if the limitations we noted were being
addressed.
There were some limitations to our analysis. We did not verify
employees' responses to our survey. However, we discussed these
responses with senior IRS officials. We also did not verify IRS'
personnel database from which we drew our sample.
We did our work between December 1997 and August 1998 in accordance
with generally accepted government auditing standards.
We requested comments on a draft of this report from the Commissioner
of Internal Revenue. The comments we received are in appendix VIII
and are evaluated at the end of this letter.
IRS' CERTIFICATION PROCESS
IDENTIFIED FEW VIOLATIONS BUT
HAD SYSTEMIC WEAKNESSES
------------------------------------------------------------ Letter :4
Our review of quarterly certifications for fiscal years 1996 and 1997
found that directors reported 11 potential violations in 368
quarterly certifications (see app. I). IRS determined that four of
the incidents were actual violations, two were unfounded allegations,
and five were reported because the directors believed the actions
could be misconstrued as violations. However, there are some
indications that the certification process may not have identified
all violations. For example, for the first quarter of fiscal year
1998, directors reported 28 potential enforcement results violations,
more than double the total reported for the prior 8 quarters.
According to IRS officials, the increase in reported potential
violations occurred as a result of the September 1997 Senate hearings
on IRS abuses, which made managers more aware of their
responsibilities for reporting potential violations. IRS had not
completed its investigation of these potential violations at the time
we completed our audit work.
Also, we identified several systemic weaknesses that affected the
reliability of the certification process. First, our analysis of the
potential violations that were reported indicated that there was
confusion among IRS officials about what constituted a violation.
For example, one certification noted that a district director had
asked the National Office official, who was responsible for providing
guidance on the use of tax enforcement results, whether a group could
establish a team goal to increase dollars collected by 10 percent as
part of an organizational pilot project. Although the National
Office official agreed that this was acceptable because neither
awards nor individual evaluations would be based on achieving the
goal, IRS' Chief Counsel subsequently determined that the group goal
was a production quota or goal and, as such, a violation.
Second, IRS guidance did not specify how the certifications were to
be supported. For example, IRS policy did not mandate what action
directors should take to support their certifications, such as
reviewing a sample of evaluations or indicating on the certification
the number and type of documents reviewed. As a result, National
Office officials may not have been in a position to determine whether
directors performed enough oversight to validate their
certifications. We found that officials at all levels with
certification responsibility said that they reviewed some documents
to determine whether managers complied with the tax enforcement
results policy. However, the nature and scope of their review varied
widely. For example, some officials reviewed all performance
evaluations and supporting documents for violations while other
officials reviewed a random sample or reviewed a small, judgmental
sample.
Third, IRS did not integrate the employee performance evaluation and
certification processes. For example, although IRS procedures
require managers to submit an evaluation to a higher level of
management for review and approval before the evaluation is given to
the employee, they do not require the reviewer to assess the
evaluation for potential tax enforcement results violations.
Although the preparers' and reviewers' signatures on the evaluation
form signified that the evaluation complied with procedures for
conducting evaluations, the procedures did not specifically refer to
the tax enforcement results policy or the certification process.
Such a requirement could heighten managerial awareness of policy and
legal prohibitions at the time the evaluation is prepared.
Fourth, IRS guidance did not clearly inform managers that they could
be subject to sanctions or the kinds of sanctions that could be
imposed if they either misused tax enforcement results or signed a
false or misleading certification. Neither the policy guidance nor
the certification included an acknowledgement that managers could be
subject to disciplinary action for the improper use of tax
enforcement records, so managers might not have known the importance
or potential consequences of their actions. Furthermore, the lack of
clear guidance could lead to inconsistent disciplinary actions. The
Guide for Penalty Determinations listed in IRS' Interim Handbook of
Employee Conduct and Ethical Behavior did not specify the types of
disciplinary actions that could be imposed on managers who violate
tax enforcement results policy. According to an IRS official, the
guide is not intended to be an exhaustive listing of all offenses,
and one general category can be used for penalty determinations in
matters not otherwise covered. The penalties for this general
category ranged from a reprimand, to a 14-day suspension, to a
removal. Disciplinary action for the four violations IRS identified
in the fiscal year 1996 and 1997 certifications ranged from
counseling to removing the manager from his/her position.
EMPLOYEES PERCEIVE THAT TAX
ENFORCEMENT RESULTS AFFECT
EVALUATIONS
------------------------------------------------------------ Letter :5
Our employee survey showed a widespread perception on the part of
collection and examination front-line employees that tax enforcement
results are considered by managers when preparing performance
evaluations. We estimated that, overall, about 75 percent of
front-line employees (i.e., revenue agents and tax auditors in the
examination function and revenue officers in the collection function)
and 81 percent of examination and collection group managers believe
that one or more enforcement measures were considered to some extent
by their managers when preparing their most recent performance
evaluation.
For those revenue officers who indicated that tax enforcement results
affected their evaluation, an estimated 58 percent of them cited
"number of cases closed"--a measure of productivity--as a tax
enforcement result influencing their evaluations. Revenue agents and
tax auditors who indicated that tax enforcement results affected
their evaluations cited measures of revenue production more
frequently as factors that influenced their evaluations than did
revenue officers. At least an estimated 61 percent of revenue agents
and tax auditors said that "additional dollars proposed per return"
influenced their evaluations, while an estimated 29 percent of
revenue officers said that "average dollars collected per return"
influenced their evaluations. Appendix III shows the point-estimates
intervals for each type of enforcement result by type of employee and
manager.
An estimated 77 percent of collection group managers who indicated
that tax enforcement results affected their evaluations believed that
the number of cases closed affected their evaluations, compared with
an estimated 39 percent who believed that the average dollars
collected per return affected their evaluations. Examination group
managers who indicated that tax enforcement results affected their
evaluations cited workload measures about as frequently as revenue
production measures. About 74 percent of examination group managers
identified tax enforcement results dealing with the average hours per
return and about 76 percent identified additional dollars proposed
per return as affecting their evaluations.
IRS' quarterly certification process focused on the use of
enforcement results in formal written performance evaluations.
However, our survey results indicated that although many employees
had multiple bases for their perceptions on the use of tax
enforcement statistics, most based their perception on verbal
information received from their managers. We estimated that about 70
percent of front-line employees, who believed that tax enforcement
results affected their evaluations, based their perceptions on verbal
communication with their managers during group meetings and
performance feedback sessions. An estimated 36 percent based their
perception on written information contained in their evaluations.
Likewise, managers who believed that tax enforcement results affected
their evaluations had multiple bases for their perceptions. An
estimated 83 percent based their perception on verbal communication
with their branch and division chiefs, while an estimated 51 percent
based their perception on written communications. Appendix IV
contains detailed information on the basis for front-line employee
and group manager perceptions.
IRS' managers are prohibited from using tax enforcement results to
establish group production quotas or goals because of the possibility
that group goals will be interpreted as individual goals or
expectations. About one-half of front-line employees and group
managers indicated that performance goals were established for their
groups. Where group goals were established, at least two-thirds of
the front-line employees and managers indicated that the goals
included one or more enforcement results. Appendix V provides more
detailed information on which enforcement results were included in
group goals.
In voluntary written comments to our questionnaires, front-line
employees and managers included additional reasons why they believe
enforcement results influenced their performance evaluations. The
comments generally centered on the following scenarios: (1)
supervisors or managers implied that an employee's performance
evaluation was influenced by tax enforcement results; (2) the general
culture or atmosphere of IRS implied that enforcement results
affected performance evaluations; (3) higher-level managers pressured
group managers to increase production or revenue yield; and (4) IRS'
business plan, reports, and other documents emphasized enforcement
results as goals. Front-line employees and managers also made
positive and negative comments about using enforcement and workload
measures to establish group goals or in performance evaluations.
Generally, the number of negative comments made by front-line
employees and managers exceeded the number of positive comments.
Appendix VI contains examples of voluntary comments received from
front-line employees and managers.
EMPLOYEE EVALUATIONS CONTAINED
POLICY VIOLATIONS AND OTHER
REFERENCES TO ENFORCEMENT AND
PRODUCTIVITY-RELATED ACTIVITIES
------------------------------------------------------------ Letter :6
Using IRS guidance for determining whether an evaluation contained a
violation, we estimated that 9 percent of front-line employees
received at least one written performance evaluation containing a
violation in the two most recent annual evaluations. The potential
effect of the violations on future employee performance could vary,
however, depending on how the enforcement results were used in
evaluations. Some violations could influence employees to be more
aggressive in their dealings with taxpayers. For example, the
following two comments taken from evaluations could create an
incentive for employees to more aggressively pursue fraud cases,
liens, and levies.
"You had one possible fraud case during this period. Keep
trying for more fraud cases."
"You do not hesitate to lien or levy and have demonstrated
innovative ideas in your efforts to comply with group goals. .
. Your workload management produces a closure rate that
demonstrates your commitment to group goals."
Both evaluations violations link a positive evaluation to achieving a
higher number of enforcement results, rather than focusing on the
appropriateness of the employees' action.
In other cases, however, the potential impact of the violations on
employee performance was less clear. For example, many of the
violations mentioned closing cases when summarizing employees'
accomplishments, as shown in the following two examples.
"The effectiveness of your planning and time utilization is
demonstrated by the fact that you closed 38 key cases during the
year."
"You have selected the proper technique to detect unreported
income and applied your technique to detect and recognize
indications of fraud. For example, [employee name], during this
fiscal year you have submitted and had accepted six civil fraud
referrals."
For most of the violations we identified, IRS officials said that the
comments would not have been considered violations under the guidance
in effect at the time the evaluations were prepared. However, the
officials agreed that the comments would be considered inappropriate
under IRS' revised guidance, which is discussed in the following
section.
The incidence of violations was relatively low, given that 36 percent
of front-line employees who believed that tax enforcement results
affected their evaluations based their perception on their written
evaluations. Our analysis of employee evaluations found that an
estimated 69 percent of the evaluations contained two types of
narrative that employees could have interpreted as inappropriate
references to tax enforcement results but which did not violate IRS
guidance. The first type of narrative involved the use of general or
case-specific references to enforcement-related activities, such as
the amounts of revenue collected or use of collection tools. These
references, which are not based on statistics such as group goals,
are not violations. We estimated that 50 percent of employees
received evaluations that contained this type of narrative. For
example, one evaluation included the statement "You had one case that
the property was in foreclosure and you were looking at seizing or
redeeming the property but would only get around $23,000. Through
your working with a third-party lending institution, you were able to
secure $30,000. . . ." Another included "[employee name] . . .
have done a very good job in locating assets even in unusual
situations. An example of this would be the research you completed
on a large dollar tax protestor assigned to your inventory." Since
the comments were based on specific cases reviewed, they are not
considered tax enforcement results violations under IRS' guidelines.
However, IRS guidance cautions managers that discussing a tax
deficiency or referencing a dollar amount collected may give the
employee an improper perception that only the size of the deficiency
or the amount collected was the basis for the employee's evaluation.
The second type of narrative that could create misperceptions about
the use of tax enforcement results was the use of the terms "overage"
(the age of the cases in the employee's workload inventory) and
"cycle time" (the number of cases worked within the established
guidelines for closing cases). IRS views these statistics as
measures of process time and responsiveness to taxpayers because they
are indications of timely attention to taxpayer issues rather than
measures of tax enforcement results. We estimate that 41 percent of
employees received evaluations that included either general or
case-specific references to overage cases and cycle time. For
example, one evaluation included the statement "You are making
excellent progress in closing out cases and specifically targeting
overage and high cycle time to minimize these problems." Another
included the statement "You appear to be doing a better job of
meeting the time frames for initial contact and follow-up action on
your cases, but your high overage and potential overage rates still
indicate a need to improve in this area." Although neither cycle time
nor overage cases are defined by IRS as enforcement results, IRS
guidance cautions managers in their use of such terms because of the
potential for either of the terms being interpreted as a prohibited
statistic such as "hours per case." Also, during our discussions with
NTEU officials, they raised concerns about whether front-line
employees and group managers clearly understood the distinction
between the inappropriate use of productivity-related enforcement
measures, such as hours per case, and the appropriate use of overage
and cycle-time data. Appendix VII provides the detailed results for
our review of employee evaluations.
IRS' PROPOSED REVISIONS TO THE
CERTIFICATION PROCESS ADDRESS
SOME WEAKNESSES
------------------------------------------------------------ Letter :7
Because of Internal Audit reports, IRS has undertaken several steps
to strengthen the certification process, including expanding the
number of employees covered, instituting an independent review
process of the certifications, revising guidance, and training
managers.
According to IRS officials, IRS has expanded the certification
process to include all appropriate enforcement supervisors, not just
those working in the collection function. Employees who are to be
covered by the certification include revenue agents, tax auditors,
appeals officers, and criminal investigators. The IRS Restructuring
and Reform Act of 1998 also mandates quarterly certification of all
appropriate enforcement supervisors. IRS is also instituting an
annual independent review to verify the accuracy of the
certifications. The reviews are to be performed annually by
cross-functional management teams and must include a review of
Employee Performance Folders and employee evaluations, which may
include documentation related to the evaluation process--such as
award narratives and case reviews. The review team may also consider
other sources, such as local memos, minutes of meetings, and
grievances submitted by employees.
IRS has also implemented a new employee evaluation process effective
September 1998 that includes a revised form to document a performance
evaluation. The guidance on the new form sent to IRS managers in
August 1998 states that by signing the form the rater and reviewer
certify that enforcement statistics were not used in preparing the
appraisal.
In addition, IRS has revised the handbook that provides guidance on
the appropriate and inappropriate use of tax enforcement results.
The handbook provides guidance on which employees are considered
enforcement officers and examples of common scenarios involving the
proper and improper use of tax enforcement results for each job
classification. The handbook also includes a decision table that
leads managers through a series of four questions to help them
determine whether they are using tax enforcement results
appropriately. This new guidance may help clarify some of the
confusion managers had on what constitutes a tax enforcement result
violation.
Furthermore, IRS trained managers on the new certification process
and handbook. After providing background on the prohibitions, the
training discussed the appropriate use of tax enforcement results and
how the prohibitions affect supervisory activities, such as setting
expectations as well as explaining the self-certification and
independent review processes. To signal its commitment to addressing
issues involving the use of enforcement statistics, IRS involved
senior executives in the training. For example, IRS directed that
district and service center directors would teach the section on
self-certification and the independent review. IRS completed the
training before the new certification process became effective in the
last quarter of 1998.
These initiatives address existing weaknesses that we have identified
by clarifying which employees are covered and the steps directors
should take to independently validate certifications. However,
managers may be confused about how to apply the decision table
contained in the revised guidance to the preparation of written
evaluations. The scenarios included in the revised guidance focus
primarily on the prohibited use of statistical reports to imply group
goals with relatively few examples relating to the appropriate and
inappropriate uses of tax enforcement results in written evaluations.
Providing such examples is particularly important because, as noted
earlier, IRS officials said that most of the violations we identified
would be considered violations under the revised guidance but not
under the guidance in effect before 1998. IRS officials agreed that
expanding the guidance to include additional examples could be
beneficial, especially since managers raised many questions about the
appropriate use of tax enforcement results in written evaluations
during the training on the revised guidance.
Furthermore, IRS has revised its certification form to reflect that
more employees are covered by the certification process, but the form
contains only a vague statement of conformance. As shown in appendix
IX, the certification does not specifically state that first-level
managers have not personally misused records of tax enforcement
results in written performance evaluations they prepared or to set
individual production goals or quotas during their verbal
communications with employees. In the case of second-level managers,
the certifications do not specifically require that managers certify
that tax enforcement results were not used in written performance
evaluations they prepared or reviewed or during their verbal
communications with employees. As a result, the form does not
clearly explain to managers what they are certifying to.
IRS has not informed managers about what specific sanctions can be
imposed for misusing tax enforcement results or submitting misleading
certifications. Neither the revised guidance nor the certification
form refers to potential sanctions.
In addition to these possible sources of confusion, IRS' independent
review process is to focus on reviewing documents and records to
identify written violations. As a result, IRS does not have a
mechanism, such as our employee survey, for monitoring employee
perceptions of how often tax enforcement results are verbally
communicated or whether their written evaluations are influenced by
records of tax enforcement results. Consequently, without an
employee survey, IRS cannot determine whether further clarifications
of the guidance or additional training is needed.
CONCLUSIONS
------------------------------------------------------------ Letter :8
IRS directors reported few violations through the quarterly
certification process in fiscal years 1996 and 1997, and we found an
estimated 9 percent of employees received evaluations that were in
violation of IRS' revised guidance during our review of employee
evaluation files. Nonetheless, our survey of IRS employees indicated
a widespread perception that managers consider tax enforcement
results when preparing performance evaluations. Most of the
employees indicated that the violations occurred during verbal
communications with their supervisors, such as staff meetings or
performance feedback sessions, rather than in their written
performance evaluations. Perceptions based on such verbal
communications could encourage employees to focus on achieving
statistical benchmarks rather than acting solely on the merits of the
case. Furthermore, the use of overage and cycle-time data in
evaluations may further employee perceptions that tax enforcement
results affect their evaluations, because they may be misconstrued as
an enforcement statistic, such as hours per case.
Although IRS is taking steps to strengthen its reporting of
violations, weaknesses remain. IRS' revised guidance has few
examples of the appropriate and inappropriate use of tax enforcement
results in written evaluations. Furthermore, the certification form
does not specifically require managers to certify that tax
enforcement results were not used in written evaluations or used
inappropriately during verbal communications with employees. IRS has
not provided clear guidance on the sanctions for misusing tax
enforcement results or submitting misleading certifications.
Additionally, IRS does not have a mechanism for monitoring how often
violations result from verbal communications or employee perceptions
of whether records of tax enforcement results are influencing their
evaluations.
RECOMMENDATIONS TO THE
COMMISSIONER OF INTERNAL
REVENUE
------------------------------------------------------------ Letter :9
To better ensure managerial accountability for the proper use of tax
enforcement results, we recommend that the Commissioner
-- expand the guidance to include additional examples of the
appropriate and inappropriate use of records of tax enforcement
results in written evaluations,
-- revise the quarterly certification form to specifically state
that tax enforcement results were not used in any written
employee evaluation prepared or reviewed, including appraisals,
awards, or promotion justifications, and that the manager did
not verbally communicate to employees that tax enforcement
results affected their evaluations or were used to set
individual production goals or quotas,
-- revise the penalty guide to specifically list the disciplinary
actions that can be taken for violations, and
-- survey employees periodically to determine whether they perceive
that tax enforcement results were used in written evaluations or
verbally communicated by their supervisors and use the results
to assess whether IRS needs to further clarify the guidance,
provide additional training, or take any other appropriate
action.
Also, to avoid the potential inappropriate use of overage and
cycle-time data, the IRS Commissioner should either designate overage
and cycle-time data as prohibited tax enforcement results or
emphasize in official policies or procedures to front-line employees
and managers how overage and cycle-time data may be used
appropriately.
AGENCY COMMENTS AND OUR
EVALUATION
----------------------------------------------------------- Letter :10
We obtained written comments on a draft of this report from the
Commissioner of Internal Revenue. The Commissioner agreed with all
of our recommendations and described the actions IRS plans to take in
response to our recommendations. Regarding our recommendation that
IRS should either designate overage and cycle-time data as prohibited
tax enforcement results or emphasize in official policies or
procedures to front-line employees and managers how overage and
cycle-time data may be used appropriately, the Commissioner said that
IRS would revise the Internal Revenue Manual to include examples of
the appropriate use of these data. If effectively implemented, the
actions described should help to resolve the problems we identified.
In addition, the Commissioner pointed out that IRS has undertaken a
number of initiatives to address the use of tax enforcement results
in employee performance evaluations. These initiatives included (1)
revising the Internal Revenue Manual provisions concerning the use of
enforcement statistics; (2) providing training to all managers on the
use of enforcement statistics, which included numerous examples of
appropriate and inappropriate language for discussing tax enforcement
results in evaluations; (3) providing an orientation course on the
proper use of statistics for all IRS employees; (4) establishing a
Support Panel to answer questions from managers or employees on the
proper use of enforcement statistics; (5) establishing independent
review teams that are obliged to review documentation in Employee
Performance Files and Employee Evaluations and may, at the director's
discretion, look to other sources of information to ensure potential
violations are uncovered, including group discussions, union comments
or reports, grievances, and minutes of managers' meetings; and (6)
developing a balanced performance measure system to include measures
of customer satisfaction, employee satisfaction, and business
results. The business results measure is to focus on the quality and
quantity of work done. The quantity measures are to consist
exclusively of outcome-neutral production data, such as cases closed
and time per closing.
--------------------------------------------------------- Letter :10.1
We are sending copies of this report to the Ranking Minority Member
of the House Ways and Means Committee; the Ranking Minority Member of
the Subcommittee on Oversight, Committee on Ways and Means; various
other congressional committees; the Secretary of the Treasury; the
Commissioner of Internal Revenue; and other interested parties. We
will also make copies available to others on request.
Please contact me at (202) 512-9110 if you or your staff have any
questions. Major contributors to this report are listed in appendix
X.
James R. White
Director, Tax Policy
and Administration
POTENTIAL VIOLATIONS REPORTED ON
FISCAL YEARS 1996 AND 1997
CERTIFICATIONS
=========================================================== Appendix I
This appendix is a summary list of all potential violations that were
reported on quarterly certifications for fiscal years 1996 and 1997.
No violations were reported in the quarters ending June 30, 1996,
September 30, 1996, and December 31, 1996. The Internal Revenue
Service (IRS) determined that the reported incidents numbered 1, 2,
5, and 11 were actual violations.
1. Quarter Ending 12/31/95. A branch chief discussed enforcement
statistics in a group manager's evaluation. According to a district
office official, the violation, discovered during the next level
review, was a statement to the effect that "Your group led the branch
in the number of seizures." Corrective action included assigning the
individual to another position outside the Collection Division,
assigning an acting branch chief until a permanent replacement could
be brought in, and reviewing guidelines.
2. Quarter Ending 03/31/96. A proposal to redesign the revenue
officer position under an IRS test project recommended that
individual groups negotiate their own production goals. One group
established a goal to increase the amount of dollars collected by 10
percent in the next year. The National Office Coordinator certified
that this goal was not a policy violation because (1) it was
established as a part of a test, and (2) awards or individual
evaluations would not be based on the achievement of the goal.
District officials were still uncertain about whether this goal was
in conflict with policy and requested an opinion from the Office of
Assistant Commissioner (Collection), who in turn requested legal
advice from IRS' Chief Counsel. The Chief Counsel ruled that the
group's goal was a violation because it was a production quota or
goal and corrective action needed to be taken to be in compliance
with policy. The Chief Counsel also cautioned that IRS managers must
have the authority to void goals negotiated under the test project if
the goals violated law or policy.
3. Quarter Ending 03/31/97. A local chapter of the National
Treasury Employees Union (NTEU) raised a concern about a branch chief
who discussed the number of seizures reflected on a monthly activity
report with his group managers. Although this instance was reported
on the quarterly certification, the incident was not considered a
violation because, under the policy, managers can discuss recorded
statistics with employees as long as they do not imply that goals or
quotas are being established. Because the comments were open to
misinterpretation, district management said they reinforced the
policy guidelines.
4. Quarter Ending 03/31/97. An Automated Call Site manager
discussed the low number of calls made during 1 week with his
employees and requested them to quadruple the number of calls. The
manager also commended the group on the number of closures they made
and identified the number of closures he was "going for." Although
this instance was reported, it was not considered a violation because
(1) Automated Call Site calls were not considered an enforcement
statistic and (2) individual goals were not established.
Nonetheless, the manager was counseled and requested to immediately
meet with his team again to emphasize that it was the quality of the
case work that resulted in the closures.
5. Quarter Ending 06/30/97. A group manager prepared a positive
employee evaluation that was based on the amount of dollars
collected. According to a district office official, this violation
was discovered during a branch manager's quarterly policy review.
The group manager was counseled and the evaluation was revised.
6. Quarter Ending 09/30/97. A district reported a pending
allegation in which an employee claimed he was forced to make
seizures. According to a district official, based on subsequent
district and regional offices' investigations, the allegation was
found to be unwarranted.
7. Quarter Ending 09/30/97. A district identified, during a
division and branch chiefs' sample review of employees' performance
binders, a few instances in which performance evaluations and
documents contained enforcement results that could easily be
misconstrued as violations (e.g., "His collection actions resulted in
both the generation of substantial revenue and in the proper
disposition of numerous cases"; "You have processed a number of Offer
In Compromise recommendations"; etc.). Corrective actions included
emphasizing the policy guidelines on the use of tax enforcement
results and requiring more thorough reviews to assess adherence with
policy and the law.
8. Quarter Ending 09/30/97. An acting branch manager distributed a
memorandum to group managers indicating dollars collected were low
and needed to be improved. Although not a violation, the acting
branch manager was advised by the division chief that his memorandum
was inappropriate and could be misconstrued. He was also instructed
to schedule a meeting with the group managers on the real intent of
his message to ensure they understood the focus was to identify
reasons for the current level of performance and find ways to improve
it.
9. Quarter Ending 09/30/97. According to the NTEU, a group manager
shared with his group a report containing revenue officers'
statistical accomplishments, including a recap of the dollars
collected over a 3-year period. As a corrective action, managers
were given instructions that this information should not be shared at
the group level and that branch statistics, comparing groups within a
branch, should not be shared below the branch level. While this was
not classified as a violation at that time, district management
officials said they would consider it a violation under the new draft
certification procedures.
10. Quarter Ending 09/30/97. The NTEU reported an allegation that,
in the presentation of a manager's award, a group manager mentioned
that the revenue officer being honored had made the most seizures in
the group. The manager denied making that statement, and the
incident was not considered a violation. District officials said
that, although not directly related to this incident, several actions
were taken, including holding meetings and briefings to discuss the
policy and the discontinuation of the distribution of statistics.
11. Quarter Ending 09/30/97. The NTEU and an outside party reported
that a group manager placed a routing slip on employees' desks
indicating the amount of dollars that needed to be collected by each
revenue officer each day, week, and month. The group manager was
temporarily reassigned until an investigation and administrative
determination was completed. Instructions were given to branch
chiefs not to provide statistical information below their branch
level. As a general corrective action applicable to these three
incidents, a meeting was held to discuss the importance of adhering
to the policy and law, and guidance was distributed to managers.
GAO SAMPLING METHODOLOGY
========================================================== Appendix II
This appendix discusses the sampling methodology we used to determine
IRS employees' perceptions of the use of tax enforcement results and
how tax enforcement results were used in employee evaluations.
SAMPLE DESIGNS
-------------------------------------------------------- Appendix II:1
To determine IRS employees' perceptions of the use of tax enforcement
results, we conducted statistically representative surveys of five
groups of IRS employees: (1) revenue agents assigned to district
office Examination Divisions and Employee Plans and Exempt
Organizations Divisions, (2) tax auditors assigned to Examination
Divisions, (3) revenue officers assigned to Collection Divisions, (4)
group managers assigned to Examination Divisions and Employee Plans
and Exempt Organizations Divisions, and (5) group managers assigned
to Collection Divisions. To determine how tax enforcement results
were used in employee evaluations, we reviewed performance
evaluations for a separately drawn sample of revenue agents assigned
to district office Examination Divisions and Employee Plans and
Exempt Organizations Divisions, tax auditors assigned to Examination
Divisions, and revenue officers assigned to Collection Divisions.
Both samples were drawn from IRS' personnel database as of January 3,
1998. Because the questionnaires for the employee surveys were not
mailed until March 1998, the results do not include the opinions of
any IRS employees who were in the groups of interest as of January 3,
1998, but were not in these groups at the time the questionnaire was
sent. We assumed that the opinions of these employees, if they would
have been included, would have been the same in the aggregate as
those of employees who were included in the surveys. We sampled a
total of 1,104 IRS employees from the 5 employee groups of interest
from the population of 20,974 employees in the January 3, 1998, IRS
personnel database. We allocated proportionately more of our sample
to Collection Division revenue officers and managers, since they are
directly engaged in collection activities.
For the objective of determining how tax enforcement results were
used in employee evaluations, we reviewed the evaluations of a sample
of 300 IRS employees from the 3 employee groups of interest. By
design, none of the 300 employees selected were the same as the 1,104
employees sampled for the first objective. IRS managers are not
required to write performance narratives for every rating dimension
for every employee each year. To review more narratives, the two
most recent ratings were requested from IRS for each employee in the
sample. The results presented in the report only reflect employees
for whom two performance evaluations were received.
SAMPLE DISPOSITION
-------------------------------------------------------- Appendix II:2
For our survey of employees' perceptions of the use of tax
enforcement results, we received usable responses (those that
indicated the employee was a member of one of the five target groups
at the time the questionnaire was received and responded to at least
some of the questions) from 814 employees, for a response rate of
approximately 74 percent. The disposition of the sampled cases for
our employee survey and review of performance evaluations is shown in
table II.1 and table II.2, respectively.
Table II.1
Disposition of Sample Cases for Our
Survey of the Extent to Which Employees
Perceived Their Ratings Were Based on
Enforcement Results
Initial selection Nonrespondents
------------------------------- --------------------
Number
Number of Eligible Known Unknown
Strata Definition of in selection responden ineligibl eligibili
ID strata universe s ts e ty
------ ----------------- --------- --------- --------- --------- ---------
1 Examination group 1,335 125 100 3 22
managers
2 Collection group 543 175 123 13 39
managers
3 Tax auditors 1,867 125 84 6 35
4 Revenue agents 11,796 320 250 11 59
5 Revenue officers 5,433 359 257 8 94
================================================================================
Total 20,974 1,104 814 41 249
--------------------------------------------------------------------------------
Source: IRS data and GAO sample.
Table II.2
Disposition of Sample Cases for Our
Review of the Extent to Which
Enforcement Results Were Discussed in
Employees' Evaluations
Initial selection Nonrespondents
------------------------------- -------------------------------
Defini Number
tion Number of Eligible Known Unknown
Strata of in selection responden ineligibl eligibili
ID strata universe s ts Eligible e ty
------ ------ --------- --------- --------- --------- --------- ---------
1 Tax 1,867 30 28 1 1
audit
ors
2 Revenu 11,796 135 127 2 6
e
agent
s
3 Revenu 5,433 135 120 6 9
e
offic
ers
================================================================================
Total 19,096 300 275 9 16
--------------------------------------------------------------------------------
Source: IRS data and GAO sample.
CALCULATION OF SAMPLE ESTIMATES
-------------------------------------------------------- Appendix II:3
After weighting the responses to the employee survey to account for
selection probabilities and nonresponse, we were able to estimate the
percentage of IRS employees who perceived that certain tax
enforcement results were or were not used in their evaluations. We
also were able to estimate the percentage of IRS employees for whom
tax enforcement results were used in at least one of their two most
recent employee evaluations.
SAMPLING ERROR
-------------------------------------------------------- Appendix II:4
Because we reviewed a statistical sample of employees, each estimate
developed from the samples had a measurable precision or sampling
error. The sampling error is the maximum amount by which the
estimate obtained from a statistical sample can be expected to differ
from the true population value being estimated. Sampling errors are
stated at a certain confidence level--in this case, 95 percent. This
means that the chances are 19 out of 20 that, if we surveyed all IRS
employees in the groups of interest, the true value obtained for a
question on these surveys would differ from the estimate obtained
from our sample by less than the sampling error for that question.
The sampling errors for our survey of the extent to which employees
perceived that their ratings were based on enforcement results are
less than plus or minus 10 percentage points, unless otherwise
reported. The sampling errors for our sample of the extent to which
enforcement results were discussed on employee evaluations are no
more than plus or minus 8 percentage points.
NONSAMPLING ERROR
-------------------------------------------------------- Appendix II:5
In addition to the reported sampling errors, the practical
difficulties of conducting any survey may introduce other types of
nonsampling errors. For example, differences in how a particular
question is interpreted or the types of individuals that do not
respond can introduce unwanted variability into the survey results.
SURVEY RESULTS ON EMPLOYEE
PERCEPTIONS OF THE EFFECT OF TAX
ENFORCEMENT RESULTS ON EVALUATIONS
========================================================= Appendix III
Our survey asked a representative sample of IRS employees the
question "To what extent, if at all, do you perceive that each of the
following factors was considered by your supervisor or managers in
preparing your most recent performance evaluation?" The "factors"
shown for this question were the tax enforcement results associated
with each employee category. The responses are shown in tables III.1
through III.6.
To compare the perceptions of front-line employees and managers, we
consolidated the responses of front-line employees (revenue agents,
tax auditors, and revenue officers) and managers (Examination
Division and Collection Division group managers) as shown in table
III.1.
Table III.1
Estimated Percent of Front-Line
Employees and Managers Who Perceived
That Tax Enforcement Results Were
Considered by Their Supervisor or
Manager In Preparing Their Most Recent
Performance Evaluation
Percent
----------------------------------
Factor not
considered/
One or more Do not know if
factors factor(s)
Employee group considered considered
---------------------------------- -------------- ------------------
Front-line employees 75 25
Managers 81 19
----------------------------------------------------------------------
Source: GAO analysis of employee survey results.
Table III.2
Estimated Percent of Revenue Agents Who
Perceived Tax Enforcement Results Were
Considered by Their Supervisor or
Manager In Preparing Their Most Recent
Performance Evaluation
Percent
----------------------------------
To no
extent/ To some/ To great/
do not moderate very great
Tax Enforcement Result know\a extent extent
---------------------------------- ---------- ---------- ----------
Average hours spent per return 27 30 43
Additional dollars proposed per 39 24 37
return
Additional dollars proposed per 40 25 36
hour
Number of criminal and civil fraud 57\c 28 15
referrals
No-change rate for examined 53 26 21
returns
Number of work units closed 41 32 27
Number of first contact closures\b 80 12 8
----------------------------------------------------------------------
Note: Percents may not add to 100 percent due to rounding.
\a We assumed the behavior of employees who responded "To no extent"
to be the same as that of employees who responded "Do not know"
because neither category would be encouraged to take inappropriate
case action to receive a positive evaluation.
\b The category "Number of first contact closures" refers to the
number of times the IRS employee was able to close the case based on
information provided during the first contact with the taxpayer.
\c Sampling error is plus or minus 10.15 percent.
Source: GAO analysis of employee survey results.
Table III.3
Estimated Percent of Tax Auditors Who
Perceived Tax Enforcement Results Were
Considered by Their Supervisor or
Manager In Preparing Their Most Recent
Performance Evaluation
Percent
----------------------------------
To no
extent/ To some/ To great/
do not moderate very great
Tax enforcement result know\a extent extent
---------------------------------- ---------- ---------- ----------
Average hours spent per return 22 36\d 42\e
Additional dollars proposed per 38\c 31 31
return
Additional dollars proposed per 37\c 25 38\e
hour
Number of criminal and civil fraud 33 42\d 25
referrals
No-change rate for examined 43\c 38\d 19
returns
Number of work units closed 31 36\d 33
Number of first contact closures\b 33 40\d 26
----------------------------------------------------------------------
Note: Percents may not add to 100 percent because not all
respondents completed each tax enforcement result category.
\a We assumed the behavior of employees who responded "To no extent"
to be the same as that of employees who responded "Do not know"
because neither category would be encouraged to take inappropriate
case action to receive a positive evaluation.
\b The category "Number of first contact closures" refers to the
number of times the IRS employee was able to close the case based on
information provided during the first contact with the taxpayer.
\c Sampling errors range from plus or minus 10.15 percent to plus or
minus 10.40 percent.
\d Sampling errors range from plus or minus 10.07 percent to plus or
minus 10.37 percent.
\e Sampling errors range from plus or minus 10.21 percent to plus or
minus 10.44 percent.
Source: GAO analysis of employee survey results.
Table III.4
Estimated Percent of Revenue Officers
Who Perceived Tax Enforcement Results
Were Considered by Their Supervisor or
Manager in Preparing Their Most Recent
Performance Evaluation
Percent
----------------------------------
To no
extent/ To some/ To great/
do not moderate very great
Tax enforcement result know\a extent extent
---------------------------------- ---------- ---------- ----------
Average dollars collected per 71 18 11
return
Number or returns secured 67 23 10
Number of installment agreements 71 23 6
obtained
Number of installment agreements 80 15 5
defaulted
Number of levies filed 67 15 17
Number of liens filed 66 17 18
Number of seizures 60 16 24
Number of cases closed 41 21 37
Number of cases closed as 64 30 6
"currently not collectable"
Number of fraud referrals 74 18 8
----------------------------------------------------------------------
Note: Percents may not add to 100 percent because not all
respondents completed each tax enforcement result category or due to
rounding.
\a We assumed the behavior of employees who responded "To no extent"
to be the same as that of employees who responded "Do not know"
because neither category would be encouraged to take inappropriate
case action to receive a positive evaluation.
Source: GAO analysis of employee survey results.
Table III.5
Estimated Percent of Examination
Division Group Managers Who Perceived
Tax Enforcement Results Were Considered
by Their Supervisor or Manager In
Preparing Their Most Recent Performance
Evaluation
Percent
----------------------------------
To no
extent/ To some/ To great/
do not moderate very great
Tax enforcement result know\a extent extent
---------------------------------- ---------- ---------- ----------
Average hours spent per return 26 33 41
Additional dollars proposed per 28 32 40
return
Additional dollars proposed per 24 22 54
hour
Number of criminal and civil fraud 37 37 26
referrals
No-change rate for examined 38 43 19
returns
Number of work units closed 42 30 28
Number of first contact closures\b 77 13 10
----------------------------------------------------------------------
\a We assumed the behavior of employees who responded "To no extent"
to be the same as that of employees who responded "Do not know"
because neither category would be encouraged to take inappropriate
case action to receive a positive evaluation.
\b The category "Number of first contact closures" refers to the
number of times the IRS employee was able to close the case based on
information provided during the first contact with the taxpayer.
Source: GAO analysis of employee survey results.
Table III.6
Estimated Percent of Collection Division
Group Managers Who Perceived Tax
Enforcement Results Were Considered by
Their Supervisor or Manager in Preparing
Their Most Recent Performance Evaluation
Percent
----------------------------------
To no
extent/ To some/ To great/
do not moderate very great
Tax enforcement result know\a extent extent
---------------------------------- ---------- ---------- ----------
Average dollars collected per 61 20 19
return
Number or returns secured 61 25 14
Number of installment agreements 61 32 7
obtained
Number of installment agreements 84 11 5
defaulted
Number of levies filed 71 18 11
Number of liens filed 76 17 7
Number of seizures 57 25 18
Number of cases closed 23 35 42
Number of cases closed as 55 34 11
"currently not collectable"
Number of fraud referrals 61 29 9
----------------------------------------------------------------------
Note: Percents may not add to 100 percent because not all
respondents completed each tax enforcement result category.
\a We assumed the behavior of employees who responded "Do not know"
to be the same as that of employees who responded "To no extent."
Source: GAO analysis of employee survey results.
SURVEY RESULTS ON THE BASIS FOR
EMPLOYEE PERCEPTIONS THAT
ENFORCEMENT RESULTS AFFECT
EVALUATIONS
========================================================== Appendix IV
For survey respondents who perceived that their supervisor or manager
considered one or more tax enforcement results in preparing their
most recent performance evaluation, we asked "Why do you believe that
one or more of the factors for which you checked box 2 through 5 [To
some extent, To a moderate extent, To a great extent, To a very great
extent] in [the prior question] influenced your rating?" They were
asked to check all the reasons that applied. The responses are shown
in tables IV.1 through IV.8.
To compare the perceptions of front-line employees and managers, we
consolidated the responses of front-line employees (revenue agents,
tax auditors, and revenue officers) and managers (Examination
Division and Collection Division group managers) as shown in table
IV.1. To compare the importance of verbal and written
communications, we also consolidated the responses indicating verbal
communications (feedback and meetings) and written communications
(performance expectations, performance evaluations, and award
justifications). Two other categories--promotion patterns and a
general "other" category--are reported separately.
Table IV.1
Estimated Percent of Front-Line
Employees and Managers Whose Perception
Was Based on Verbal and/or Written
Communication
Percent
--------------------------
Verbal Written
communicatio communicatio
Employee category n n
------------------------------------------ ------------ ------------
Front-line employees 70 49
Managers 83 51
----------------------------------------------------------------------
Source: GAO analysis of employee survey results.
The responses of front-line employees and managers on the specific
basis for their perception that tax enforcement results affected
their ratings are given in tables IV.2 through IV.4.
Table IV.2
Estimated Percent of Front-Line
Employees and Managers Who Had Various
Bases for Their Perception That Tax
Enforcement Results Influenced Their
Evaluation
Front-
Basis for employees' and managers' perception line
that tax enforcement results influenced employees Managers
evaluation (percent) (percent)
---------------------------------------------- ---------- ----------
One or more factors were mentioned in my 29 27
performance expectations
One or more factors were mentioned in my 36 41
performance evaluation
One or more factors were mentioned in feedback 52 69
I received from my group/branch manager
One or more factors were mentioned by my 56 70
group/branch manager in group meetings
One or more factors were mentioned in an award 13 18
justification
One or more factors were mentioned in a 6 2
promotion justification
Promotion patterns suggest that management 37 19
considers one or more factors
Other -Please describe\a 30 31
----------------------------------------------------------------------
\a The category "other" is discussed in greater detail in appendix
VI.
Source: GAO analysis of employee survey results.
Table IV.3
Estimated Percent of Revenue Agents, Tax
Auditors, and Revenue Officers Who Had
Various Bases for Their Perception That
Tax Enforcement Results Influenced Their
Evaluation
Revenue Tax Revenue
Basis for employees' perception that tax agents auditor officer
enforcement results influenced (percent (percent (percent
evaluation ) ) )
---------------------------------------- -------- -------- --------
One or more factors were mentioned in my 27 30\b 34
performance expectations
One or more factors were mentioned in my 35 29\b 41
performance evaluation
One or more factors were mentioned in 50 48\b 56
feedback I received from my group
manager
One or more factors were mentioned by my 55 74 50
group manager in group meetings
One or more factors were mentioned in an 13 14 11
award justification
One or more factors were mentioned in a 5 1 9
promotion justification
Promotion patterns suggest that 36 23 43
management considers one or more
factors
Other -Please describe\a 29 23 35
----------------------------------------------------------------------
\a The category "other" is discussed in greater detail in appendix
VI.
\b Sampling errors range from plus or minus 10.21 percent to plus or
minus 10.35 percent.
Source: GAO analysis of employee survey results.
Table IV.4
Estimated Percent of Examination and
Collection Division Group Managers Who
Had Various Bases for Their Perception
That Tax Enforcement Results Influenced
Their Evaluation
Examination Collection
Basis for managers' perception that tax managers managers
enforcement results influenced evaluation (percent) (percent)
------------------------------------------ ------------ ------------
One or more factors were mentioned in my 29 23
performance plan (Form 9688)
One or more factors were mentioned in my 44\b 36
performance evaluation
One or more factors were mentioned in 76 58
feedback I received from my branch chief
One or more factors were mentioned by my 71 68
division chief or branch chief in group
meetings
One or more factors were mentioned in an 17 21
award justification
One or more factors were mentioned in a 3 2
promotion justification
Promotion patterns suggest that management 21 16
considers one or more factors
Other -Please describe\a 27 38
----------------------------------------------------------------------
\a The category "other" is discussed in greater detail in appendix
VI.
\b Sampling error is plus or minus 10.64 percent.
Source: GAO analysis of employee survey results.
SURVEY RESULTS ON INCLUSION OF
ENFORCEMENT RESULTS IN GROUP GOALS
=========================================================== Appendix V
Our survey asked a representative sample of IRS employees "To your
knowledge, were performance goals established for your group in
calendar year 1997?"
To compare the perceptions of front-line employees and managers, we
consolidated the responses of front-line employees (revenue agents,
tax auditors, and revenue officers) and managers (Examination
Division and Collection Division group managers) as shown in table
V.1.
Table V.1
Estimated Percent of Front-Line
Employees and Managers Who Had
Performance Goals Established for Their
Group in Calendar Year 1997
Employee category Percent
---------------------------------- ----------------------------------
Front-line employees 44
Managers 49
Revenue agents 44
Tax auditors 46\a
Revenue officers 46
Examination Division group 45
managers
Collection Division group managers 54
----------------------------------------------------------------------
\a Sampling error is plus or minus 10.54 percent.
Source: GAO analysis of employee survey results.
For those employees who had group performance goals, we asked "Did
the performance goals established for your group in calendar year
1997 include any reference to any of the following factors?" The
point estimates and confidence intervals for each employee group are
shown in tables V.2 and V.3.
Table V.2 Estimated Percent of Revenue
Agents, Tax Auditors, and Examination
Group Managers Who Had Performance Goals
Established for Their Group That
Included at Least One Tax Enforcement
Result
Examinatio
Revenue Tax n
agent auditor manager
Tax enforcement result (percent) (percent) (percent)
---------------------------------- ---------- ---------- ----------
Average hours spent per return 77 89 60\d
Additional dollars proposed per 54\b 55\c 53\d
return
Additional dollars proposed per 76 66\c 70\d
hour
Number of criminal and civil fraud 38 59\c 47\d
referrals
No-change rate for examined 50\b 51\c 23\d
returns
Number of units closed 49\b 74\c 28\d
Number of first contact closures\a 11 57\c 7
----------------------------------------------------------------------
\a The category "number of first contact closures" refers to the
number of times the IRS employee was able to close the case based on
information provided during the first contact with the taxpayer.
\b Sampling errors range from plus or minus 10.35 percent to plus or
minus 10.90 percent.
\c Sampling errors range from plus or minus 13.80 percent to plus or
minus 15.90 percent.
\d Sampling errors range from plus or minus 12.21 percent to plus or
minus 14.41 percent.
Source: GAO analysis of employee surveys.
Table V.3
Estimated Percent of Revenue Officers
and Collection Group Managers Who Had
Performance Goals Established for Their
Group That Included at Least One Tax
Enforcement Result
Revenue Collection
officer manager
Tax enforcement result (percent) (percent)
---------------------------------------------- ---------- ----------
Average dollars collected per return 40 43\a
Number of returns secured 33 30\a
Number of installment agreements obtained 11 11
Number of installment agreements defaulted 6 2
Number of levies filed 23 8
Number of liens filed 16 6
Number of seizures 32 15
Number of cases closed 70 70
Number of cases closed as "currently not 24 15
collectable"
Number of fraud referrals 13 8
----------------------------------------------------------------------
\a Sampling errors range from plus or minus 10.01 percent to plus or
minus 10.63 percent.
Source: GAO analysis of employee survey results.
EXAMPLES OF VOLUNTARY EMPLOYEE
COMMENTS INCLUDED ON SURVEYS OF
IRS EMPLOYEES
========================================================== Appendix VI
This appendix gives some examples of voluntary written comments by
front-line employees and group managers to our questionnaires on why
they believe enforcement results influenced their performance
evaluations.\2 The comments generally involved the following
scenarios: (1) supervisors or managers implied that an employee's
performance evaluation was influenced by tax enforcement results; (2)
the general culture or atmosphere of IRS implied that enforcement
results affected performance evaluations; (3) higher level managers
pressured group managers to increase production or revenue yield; and
(4) IRS' business plan, reports, and other documents emphasized
enforcement results as goals. Front-line employees and managers also
wrote about positive and negative effects of using enforcement and
workload measures to establish group goals or in performance
evaluations. Generally, the number of negative comments made by
front-line employees and managers exceeded the number of positive
comments.
--------------------
\2 Of the 814 surveys received, a total of 402 front-line employees
and 151 group managers added voluntary comments.
COMMENTS INDICATING THAT
SUPERVISORS OR MANAGERS IMPLIED
THAT AN EMPLOYEE'S PERFORMANCE
EVALUATION WAS INFLUENCED BY
ENFORCEMENT RESULTS
-------------------------------------------------------- Appendix VI:1
-- "The district office provided monthly statistics to each branch,
rating branches and groups based on `dollars per hour,' `dollars
per return,' `no-change rate,' [and] `collections percent by
exam division.' . . . Even though performance goals were not
established formally based on `dollars per return,' etc., the
statistics were mentioned at each group meeting and pressure was
felt by each agent to assess the most tax at the least amount of
time spent on the case. It was always felt that our evaluations
were based indirectly on the factors mentioned in Item 19
(productivity and enforcement statistics)."
-- "Our emphasis used to be on quality and customer service but
when our district went under Boston, the emphasis changed to
`dollars per hour.' At each group meeting, we would hear about
`stats' and `dollars per hour.' This kind of pressure does tend
to carry over into the work product."
-- "Although numbers [were] never mentioned, [my] group manager
discussed closure of cases, fraud referrals, and defaulted
installment agreements in meetings and in appraisals so it gave
the appearance that [is] what he expected."
-- "We were told by management in group meetings and in town hall
meetings that goals in the form of yield (dollars per hour)
would no longer be set. Yet, at group meetings, our manager
reads our yield from Table 37 [a management information report]
and comments on whether it is good, bad, or fair. Our yield is
obviously still very important to management."
COMMENTS INDICATING THAT THE
GENERAL CULTURE OR ATMOSPHERE
OF IRS IMPLIED THAT ENFORCEMENT
RESULTS AFFECTED PERFORMANCE
EVALUATIONS
-------------------------------------------------------- Appendix VI:2
-- "There is one important goal in any revenue agent's
life--dollars per hour. From the day I started with the Service
until the present, that has been the only constant in the
organization. Quality programs have come and gone, emphasis on
taxpayer service ebbs and flows, lip service to auditing
standards is periodically given. Through it all, any successful
revenue agent knows that low time and high dollars will result
in recognition, promotion, and awards."
-- "It is well known that these factors are considered by upper
management to be important, although not all are mentioned in
evaluations."
-- "Management has never been specific relative to the connection
between promotion and enforcement. However, the perception is
here that the more enforcement minded a revenue officer is, the
greater the chance of promotion."
-- "You are expected (unsaid) to do a seizure every quarter."
COMMENTS INDICATING THAT HIGHER
LEVEL MANAGERS PRESSURED GROUP
MANAGERS TO INCREASE PRODUCTION
OR YIELD
-------------------------------------------------------- Appendix VI:3
-- ". . . [My present manager] was required to post statistics
on most of these items on our library room wall up to just a few
months ago. When Congress started seriously making noises about
statistics, they were taken down. . . . [My manager] would
make reference to these statistics in a general way as pressure
was put on him through his chain of command. The pressure
relative to improving these statistics came primarily by passing
word down from the chief of audit and his assistant chief that
our statistics were not measuring up. . . ."
-- "The manager only works toward faster closures and higher yield
because it has been stressed by the district management."
COMMENTS INDICATING THAT IRS'
BUSINESS PLAN, REPORTS, AND
OTHER DOCUMENTS EMPHASIZED
ENFORCEMENT RESULTS
-------------------------------------------------------- Appendix VI:4
-- "Time and dollars per hour, although less so within the past few
years, has always been a major consideration in `informal'
discussions and meetings. Tables 36 and 37 [management
information reports] were always utilized and disbursed to the
group and discussed with agents, although seldom in writing."
-- "Statistics on dollars collected, number of returns [secured],
number of levies and liens were reported by group on a monthly
basis to the division via the branch chief."
-- "Monthly statistics . . . were distributed regarding levies,
dollars collected, liens, seizures, fraud referrals, currently
not collectible cases, and cases closed."
POSITIVE COMMENTS ON THE USE OF
ENFORCEMENT RESULTS
-------------------------------------------------------- Appendix VI:5
-- "If a revenue agent is doing a good job, these factors
(adjustments, etc.) are present. The revenue agent will have
good adjustments that are technically correct and will also have
no-changes. That is part of the job."
-- "Both time spent on a return and deficiencies are indicators of
effectiveness and efficiency in the audit process. Excessive
time may indicate complexity of issues or inefficiency and
floundering in the audit. The deficiency amount will indicate
effectiveness of audit as compared to similar audits or,
potentially, a lack of issue identification. When used as
indicators of efficiency or effectiveness, these can help direct
employees toward improvement. It's not all bad."
-- "Our job is to collect (hence `collection') dollars and returns.
It is ludicrous to think that a revenue officer who collected no
dollars or returns would get a good performance evaluation.
After all--that is our job!"
-- "I definitely feel pressure to be productive in terms of dollars
per hour, but realistically, that makes sense to me. We should
spend time on issues which will most likely produce revenue and
work them as quickly as possible."
NEGATIVE COMMENTS ON THE USE OF
ENFORCEMENT RESULTS
-------------------------------------------------------- Appendix VI:6
-- "Statistics regarding `dollars per hour' are disseminated at
group meetings. Our group is compared to the branch and the
district. . . . Ten years ago these kinds of statistics were
never discussed at the group level. Due to the constant
referring to these numbers, it is obvious that `dollars per
hour' is the most important concept to management.
Consequently, agents will always strive to get the highest yield
per case. This means that you will disregard adjustments in the
taxpayers favor so as not to reduce the tax yield. Individual
statistics have not been discussed. However, the constant
reminder of the yield causes agents to do anything they can to
get good yield on their cases."
-- "Even though it is made very clear in the . . . region that
there are no individual or group enforcement goals, statistics
are kept and very well known throughout the region.
Specifically, our `mission' is very goal oriented and very
driven by the goals of the Chief of Collections. Very clearly
those groups that have good statistics--specifically the number
of seizures, cases closed, fraud referrals, etc.--receive
special recognition. In some cases, managers receive
performance rewards. The revenue officers who specifically take
enforcement action to resolve cases are rewarded with promotion
even though the enforcement action is not mentioned in the
performance evaluation. Those revenue officers who are most
aggressive using enforcement are those rewarded most quickly
with promotion."
-- "National and regional management have used dollar goals and
quotas for as long as I can remember. It is common knowledge
among employees and first-line managers that `increase voluntary
compliance' equals `dollars per hour.' You get what you
measure."
-- "Within the last couple of years, `dollars collected' was being
conveyed from the division to the group level. This caused
group comparisons and competition. While never conveyed to an
individual within my group that dollars needed to be increased,
it certainly was implied. While this didn't change my case
decisions, I cannot speak for everyone."
RESULTS OF REVIEW OF EMPLOYEE
EVALUATIONS
========================================================= Appendix VII
Table VII.1 provides the results of our review of employee
evaluations in terms of whether the evaluations contained law or
policy violations and case or general discussions of
enforcement-related activity.
Table VII.1
Estimated Percent of Employees Whose
Performance Evaluations Contained
References to Tax Enforcement Results
Percent of employees whose
evaluations included tax
enforcement results
references
----------------------------
Collectio Examinati
Types of tax enforcement results n on
references employees employees Total
---------------------------------------- --------- --------- ======
Law or policy violations 1 12 9
Case-specific discussion of:
1. Overage cases 2 12 6
2. Cycle time 1 2 2
3. Overage cases or cycle time 2 9 7
4. Assessment/Collection measures 33 31 32
A. Amount of revenue collected 14 N/A N/A
B. Use of collection tools 33 N/A N/A
C. Amount of revenue assessed N/A 25 N/A
D. Amount of dollars collected on agreed N/A 12 N/A
cases
Any specific discussion of the above 33 38 37
General discussion of:
1. Overage cases 31 33 32
2. Cycle time 9 29 16
3. Overage cases or cycle time 32 34 33
4. Assessment/collection measures 75 20 36
A. Amount of revenue collected 9 N/A N/A
B. Use of collection tools 74 N/A N/A
C. Amount of revenue assessed N/A 17 N/A
D. Amount of dollars collected on agreed N/A 5 N/A
cases
Any general discussion of the above 78 54 61
Any discussion of overage cases or cycle 38 43 41
time
Any discussion of assessment/collection 75 40 50
measures
Any discussion of any factors above 78 65 69
----------------------------------------------------------------------
Note: N/A means not applicable to this employee category.
Source: GAO analysis of IRS employees' performance evaluations.
(See figure in printed edition.)Appendix VIII
COMMENTS FROM THE INTERNAL REVENUE
SERVICE
========================================================= Appendix VII
(See figure in printed edition.)
(See figure in printed edition.)
(See figure in printed edition.)
(See figure in printed edition.)
(See figure in printed edition.)
EXAMPLES OF CERTIFICATION FORMS
========================================================== Appendix IX
(See figure in printed
edition.)
(See figure in printed
edition.)
MAJOR CONTRIBUTORS TO THIS REPORT
=========================================================== Appendix X
GENERAL GOVERNMENT DIVISION
James A. Bell, Senior Statistician
James M. Fields, Senior Statistician
Stuart M. Kaufman, Senior Social Science Analyst
SAN FRANCISCO OFFICE
Ralph T. Block, Assistant Director
Jonda Van Pelt, Evaluator-in-Charge
Shari Caporale, Senior Evaluator
Suzy Foster, Senior Evaluator
Sam Scrutchins, Senior Evaluator
John N. Zugar, Senior Evaluator
ATLANTA OFFICE
Robert V. Arcenia, Senior Evaluator
Ronald J. Heisterkamp, Evaluator
KANSAS CITY OFFICE
Benjamin Douglas, Senior Evaluator
*** End of document. ***