This is the accessible text file for GAO report number GAO-02-353 
entitled 'Private Pensions: IRS Can Improve the Quality and Usefulness 
of Compliance Studies' which was released on April 12, 2002. 

This text file was formatted by the U.S. General Accounting Office 
(GAO) to be accessible to users with visual impairments, as part of a 
longer term project to improve GAO products' accessibility. Every 
attempt has been made to maintain the structural and data integrity of 
the original printed product. Accessibility features, such as text 
descriptions of tables, consecutively numbered footnotes placed at the 
end of the file, and the text of agency comment letters, are provided 
but may not exactly duplicate the presentation or format of the 
printed version. The portable document format (PDF) file is an exact 
electronic replica of the printed version. We welcome your feedback. 
Please E-mail your comments regarding the contents or accessibility 
features of this document to Webmaster@gao.gov. 

This is a work of the U.S. government and is not subject to copyright 
protection in the United States. It may be reproduced and distributed 
in its entirety without further permission from GAO. Because this work 
may contain copyrighted images or other material, permission from the 
copyright holder may be necessary if you wish to reproduce this 
material separately. 

United States General Accounting Office: 
GAO: 

Report to Congressional Requesters: 

April 2002: 

Private Pensions: 

IRS Can Improve the Quality and Usefulness of Compliance Studies: 

GAO-02-353: 

Contents: 

Letter: 

Results in Brief: 

Background: 

IRS Audited a Sample of 401(k) Plans to Estimate Noncompliance: 

IRS Study Does Not Generally Provide Accurate Estimates of 401(k) Plan 
Noncompliance: 

IRS Is Planning to Continue Compliance Research: 

Conclusions: 

Recommendations: 

Agency Comments: 

Appendix I: Scope and Methodology: 

Determining the Steps IRS Took to Conduct the 401(k) Study: 

Evaluating the IRS Study on the Prevalence and Types of Noncompliance 
among 401(k) Plans: 

Describing IRS Plans for Compliance Research on Private Pensions: 

Appendix II: Changes in the Laws That Govern Tax-Qualified 401(k) 
Plans: 

Appendix III: Comments from the Internal Revenue Service: 

Appendix IV: GAO Contacts and Staff Acknowledgments: 

GAO Contacts: 

Staff Acknowledgments: 

Abbreviations: 

ASA: American Statistical Association: 

Audit CAP: Audit Correction Agreement Program: 

DC: defined contribution: 

ERISA: Employee Retirement Income Security Act: 

GAO: General Accounting Office: 

IRC: Internal Revenue Code: 

IRS: Internal Revenue Service: 

PWBA: Pension and Welfare Benefits Administration: 

SCP: Self-Correction Program: 

SIMPLE: Savings Incentive Match Plan for Employees: 

VCP: Voluntary Correction Program: 

[End of section] 

United States General Accounting Office: 
Washington, DC 20548: 

April 12, 2002 

The Honorable Max Baucus: 
United States Senate: 

The Honorable Bob Graham: 
United States Senate: 

The Honorable Orrin Hatch: 
United States Senate: 

The Honorable James Jeffords: 
United States Senate: 

To encourage employers to establish and maintain pension and savings 
plans for their employees, the federal government provides 
preferential tax treatment under the Internal Revenue Code (IRC) for 
plans that meet certain requirements. In fiscal year 2002, these tax 
preferences for employer-sponsored pension plans are estimated at 
about $88 billion and represent the largest federal “tax expenditure,” 
exceeding those for either home mortgages or health benefits.[Footnote 
1] 

The Internal Revenue Service (IRS) conducts audits and other 
enforcement and research activities to ensure that tax-qualified 
pension plans satisfy applicable requirements in exchange for 
preferential tax treatment. IRS monitors plan adherence to 
requirements primarily through audits but also through periodic 
compliance studies. Compliance studies are conducted by IRS to better 
design and implement enforcement activities, such as targeted plan 
audits, and provide information that IRS can use to inform pension 
plan sponsors and administrations about key aspects of noncompliance. 
In 2000, IRS released a study on 401(k) plan compliance with certain 
statutory requirements that govern tax-qualified plans.[Footnote 2] 

To conduct the study, IRS selected a sample of 401(k) plans on the 
basis of information that plan sponsors filed with IRS for the 1993 
plan year, and IRS agents answered study questions after they 
completed audits of these plans. The study, including the audits and 
data analysis, was conducted between 1995 and 1998, and IRS published 
the study results in 2000. IRS reported that 44 percent of the plans 
in the study had one or more instances of noncompliance with the 
statutory requirements IRS examined. Members of Congress are concerned 
about whether the stated noncompliance rate accurately describes the 
level of noncompliance among 401(k) plans. 

In response to this concern, you asked us to (1) determine how the IRS 
conducted a study to estimate 401(k) plan noncompliance with the 
requirements of the Internal Revenue Code; (2) assess to what extent 
the study provided accurate estimates of the prevalence and types of 
noncompliance among 401(k) plans; and (3) describe IRS’s plans for 
research on pension plan compliance. 

To address your questions, we reviewed the IRS 401(k) plan compliance 
study published[Footnote 3] in 2000 and work papers related to its 
design and implementation. We also interviewed IRS officials and 
analysts to obtain information on the design, implementation, and data 
analysis methods that IRS used to conduct the study. In addition, we 
compared IRS’s study design, data collection, and analysis with 
published American Statistical Association (ASA) and GAO guidance on 
conducting research studies. Because IRS was not able to provide us 
with documentation, data, or both that support the published results, 
we were not able to assess IRS study results on specific types of 
noncompliance to verify their accuracy or determine whether the 
instances of noncompliance that IRS reported were significant. We 
reviewed IRS work plans in the employee plans area and interviewed IRS 
officials to identify plans for ongoing and future compliance research 
initiatives. We conducted our work between June 2001 and January 2002 
in accordance with generally accepted government auditing standards 
(see appendix I for details on our scope and methodology). 

Results in Brief: 

To conduct the 401(k) plan compliance study, the IRS audited a sample 
of 401(k) plans and collected data to estimate noncompliance with 
certain requirements of the Internal Revenue Code. However, IRS’s 
estimates on 401(k) plan noncompliance were generally inaccurate. IRS 
selected its study sample of 550 plans, each of which indicated a 
401(k) plan feature for the 1993 plan year, from an agency database 
with information on all tax-qualified plans. To produce a 
representative sample, the IRS randomly picked equal numbers of plans 
from small, medium, and large plan size categories.[Footnote 4] To 
obtain information on the characteristics, design features, and 
compliance of these plans, IRS used an already available study check 
sheet, or questionnaire, that included 254 questions to broadly gather 
information on 401(k) plans. Each of the 550 plans in the study was 
assigned to an IRS examiner for audit. IRS examiners who conducted the 
audits were provided with a copy of the questionnaire and were 
instructed to answer the study questions upon completing the audits. 
Once the data were gathered, IRS identified 73 out of the 254 study 
questions that it believed could indicate whether or not a plan was in 
compliance with certain requirements; IRS called these 73 questions “
compliance indicators.” IRS based its published results on an analysis 
of these indicators for 472 of the plans selected.[Footnote 5] IRS 
data analyses produced estimates of the number of plans that failed to 
comply with one or more of their compliance indicators and the 
frequency of these violations. The study did not attempt to provide 
information on the severity of the compliance violations that IRS 
identified, nor did the study determine the number of plan 
participants or the amount of assets associated with these 
noncompliance errors. 

The IRS study does not, in general, provide accurate estimates of the 
overall prevalence and types of noncompliance among 401(k) plans. This 
is primarily because only 27 of the 73 study questions that IRS 
identified as compliance indicators conclusively demonstrated whether 
or not a plan was noncompliant. Consequently, the 44 percent of plans 
reported to have one or more instances of noncompliance is at best an 
upper bound on the extent of noncompliance found in this study. The 
questionnaire was not developed to specifically identify 401(k) plan 
noncompliance, and the answers were not validated as accurately 
demonstrating compliance or noncompliance. Some data that IRS 
collected were inconsistent or inaccurate, because the agents who 
completed the questionnaires did not conduct the audits or complete 
the questionnaires in a uniform manner. Additionally, although IRS 
audited a stratified random sample of 401(k) plans, it combined 
information in such a way that its reported findings could not be 
generalized to the broader population of all 401(k) plans. This is 
because IRS’s data analysis did not use sample weights, which are 
necessary to make tabulations and other estimates when data are 
combined across plan sizes.[Footnote 6] 

IRS is currently planning and conducting research on several types of 
private pension plans to estimate prevalence and types of 
noncompliance. IRS has selected specific types of private pension 
plans—-401(k) plans, sections 403(b) and 457 plans,[Footnote 7] and 
multiemployer plans[Footnote 8]-—on which to conduct compliance 
research and will undertake studies similar in design to the one 
conducted on 401(k) pension plans for each of these plan types. The 
data that IRS collects will be analyzed to determine the prevalence 
and types of noncompliance among the plans studied. After implementing 
initiatives designed to improve the compliance of different plan 
types, IRS will conduct another compliance study and, by comparing 
results to the previous study, will determine the effectiveness of its 
initiatives. IRS officials told us that the agency is adopting lessons 
learned from its 1995 401(k) compliance study to enhance the quality 
and usefulness of future studies. For example, compliance planning 
groups, which involve key stakeholders from across the agency, have 
been assembled to assist in planning and designing specific compliance 
research studies. In addition, IRS told us that the questionnaires for 
upcoming 401(k) plan compliance studies will be pretested to determine 
the usefulness of the study questions in producing data on compliance 
and noncompliance. 

IRS can take additional steps to ensure that the data it collects and 
the findings it reports are as accurate as possible. In this report, 
we make three recommendations to the Commissioner of the IRS to ensure 
the quality and usefulness of compliance studies, which in turn will 
enhance IRS’s efforts to promote compliance among pension plan 
sponsors. 

IRS provided written comments on a draft of this report and generally 
agreed with our findings, conclusions, and recommendations. IRS notes 
that it has incorporated our recommendations in a current compliance 
study on 401(k) plans. We agree that IRS has taken specific steps to 
improve its current 401(k) plan compliance study and describe these 
steps in our report. In addition to the current 401(k) study, IRS 
should also implement our recommendations throughout its current and 
upcoming compliance study initiatives on 401(k) and other types of 
pension plans. 

Background: 

The IRC permits employers to sponsor defined contribution (DC) 
retirement plans and outlines requirements to which plan sponsors must 
adhere for tax-qualified status. With DC plans, employees have 
individual accounts to which employers, employees, or both make 
periodic contributions. DC plan benefits are based on the 
contributions to, and investment returns on, the individual accounts, 
and the employee bears the investment risk. In some types of DC plans, 
including 401(k), 403(b), and 457 plans and the Savings Incentive 
Match Plan for Employees (SIMPLE), employees may choose to make tax-
deferred contributions instead of receiving the same amount as taxable 
salary.[Footnote 9] 

IRS and the Pension and Welfare Benefits Administration (PWBA) of the 
Department of Labor are primarily responsible for enforcing laws 
related to private DC plans. Under the Employee Retirement Income 
Security Act (ERISA) of 1974, IRS and PWBA jointly enforce standards 
for coverage and participation, for vesting, and for funding that, 
respectively, determine how plan participants become eligible to 
participate in benefit plans, define how participants become eligible 
to earn rights to benefits, and ensure that plans have sufficient 
assets to pay promised benefits. IRS also enforces provisions of the 
IRC that apply to tax-qualified pension plans, including provisions 
under section 401(k) of the Code. PWBA enforces ERISA’s reporting and 
disclosure provisions and fiduciary standards, which concern how 
pension plans should operate in the best interest of participants. 

Since the IRS 401(k) plan compliance study was undertaken in 1995, 
various changes have occurred in certain legal requirements for tax-
qualified status that IRS examined in the study. Certain requirements 
IRS examined in the 401(k) study are no longer applicable to tax-
qualified DC plans or have been materially modified. Also, the IRC has 
since been amended to permit employers to adopt SIMPLE 401(k) plans 
and safe-harbor design methods for 401(k) plans. SIMPLE 401(k) plans 
and safe-harbor designs exempt 401(k) plan sponsors from certain rules 
that apply generally to 401(k) plans.[Footnote 10] However, many of 
the statutory requirements that IRS examined in the 401(k) study have 
not changed materially. We were not able to assess the extent to which 
changes in relevant pension laws and 401(k) plan designs have affected 
the overall prevalence and incidence of noncompliance among the 
population of 401(k) plans (see app. II for more detail on changes in 
relevant pension laws since the study was published). 

IRS groups violations of the IRC and corresponding regulations that 
must be satisfied to achieve tax-qualified status into four categories.
[Footnote 11] 

* Plan Document failure occurs when the language of the plan documents 
does not comply with provisions of the tax Code. 

* Operational failure occurs when the implementation and operation of 
a plan does not comply with provisions of the tax Code. 

* Demographic failure occurs when a plan fails to comply with 
fundamental nondiscrimination requirements faced by all tax-qualified 
plans.[Footnote 12] 

* Employer Eligibility failure occurs when an employer that is not 
allowed to establish a section 401(k) plan, such as a state or local 
government, adopts such a plan. 

IRS issued Revenue Procedure 2001-17 in February 2001 to establish its 
current framework for promoting the compliance of tax-qualified 
pension plans with the applicable requirements of the IRC. This 
framework has evolved since IRS first introduced voluntary compliance 
procedures in the early 1990s. To promote compliance, IRS developed 
the Self-Correction Program (SCP), the Voluntary Correction Program 
(VCP), and the Audit Correction Agreement Program (Audit CAP). SCP is 
used to correct insignificant Operational failures at any time, 
without fee or sanction and without IRS supervision.[Footnote 13] VCP 
allows plan sponsors to voluntarily report and correct all types of 
qualification failures with IRS approval. Upon receiving IRS approval 
of the proposed correction measures, plan sponsors must implement the 
specified corrective measures and pay a compliance fee, one that is, 
on average, much less than the financial sanctions assessed for 
violations identified by IRS audits. The Audit CAP allows plan 
sponsors to correct all types of qualification failures that IRS 
identifies through formal audits. Under Audit CAP, plan sponsors must 
correct all qualification failures and pay a negotiated financial 
sanction commensurate with the nature, extent, and severity of the 
failures. If IRS and the plan sponsor do not reach an agreement with 
respect to the correction of the failure(s), IRS can pursue 
disqualification of the plan for tax purposes.[Footnote 14] 

All IRS audits of tax-qualified employer-sponsored plans are carried 
out under one of two audit programs, the Examination Program or the 
Compliance Research Program. The Examination Program includes a wide 
range of compliance-related activities. These activities include 
auditing based on referrals and computer targeting, training for IRS 
examiners who perform plan audits, and reviewing closed audit cases. 
The Compliance Research Program sponsors studies, such as the 401(k) 
study, to identify and monitor noncompliance among private plans. 
Compliance studies are based on plan audits, which IRS conducts so 
that it can collect study data. IRS is in various stages of planning 
and conducting compliance research on several types of private 
pensions, and intends to use data from these studies to develop more 
effective enforcement and compliance activities. However, plan audits 
conducted under the auspices of the Compliance Research Program 
represent a small proportion of IRS’s total audit activity. In fiscal 
year 2001, IRS plans allocated a total of 1,845 staff days to audits 
for the Compliance Research Program, compared with 33,734 staff days 
allocated to audits for the Examination Program. For fiscal year 2002, 
IRS plans to increase the number of staff days related to compliance 
research activities, but direct examination activities will still 
constitute the majority of IRS’s audit work. 

Audits of employer pension plans are initiated when the IRS selects 
for audit a plan return, or form 5500 filing, from the Return 
Inventory Classification System. A notification letter is sent to the 
plan sponsor with a request for information that the examiner needs to 
complete the audit. IRS examiners complete a process that includes 
interviewing the plan benefits administrator, reviewing plan 
documents, and holding a closing conference to discuss the results of 
the audit with the plan sponsor. If an examiner finds a qualification 
issue, or a failure that can potentially disqualify a plan’s tax-
exempt status, the examiner can resolve the violation through 
correction under IRS’s SCP (Self-Correction Program disposal) or enter 
into a closing agreement with the plan sponsor through the Audit CAP 
(closing agreement disposal). Both of these audit disposal methods 
indicate that the examiner identified a violation that could 
potentially disqualify the plan, but the Audit CAP closing represents 
a more significant disposal than correction under the SCP. 

IRS Audited a Sample of 401(k) Plans to Estimate Noncompliance: 

IRS audited a sample of 401(k) plans to collect data and estimate 
noncompliance with certain requirements of the Internal Revenue Code. 
IRS examiners were provided with a questionnaire to obtain information 
on the compliance of these 401(k) plans after conducting the audits. 
Once the data were gathered, IRS identified 73 study questions that 
could indicate whether or not a plan was in compliance. IRS data 
analyses produced estimates on the number of plans that failed to 
comply in one or more instances, based on the answers to these 73 
compliance indicators. IRS’s original estimates on noncompliance 
decreased after some adjustments were made to its initial analysis. 

IRS Selected a Sample of 401(k) Plans to Audit: 

In selecting a sample of plans to study, IRS analyzed a database that 
it maintains on the population of tax-qualified plans. This database 
contains records of form 5500 returns that plan sponsors file with the 
IRS and Department of Labor, and IRS identified pension plans that had 
reported a 401(k) feature for the 1993 plan year. IRS identified 
143,535 plans that reported a 401(k) plan feature, but excluded about 
470 plans prior to sample selection, because these plans had no 
participants at the end of 1993 or had recently been audited by IRS. 
This step reduced the population to 142,768 401(k) plans from which 
IRS would select plans to study. 

These remaining 401(k) plans were subdivided evenly by size into three 
groups labeled small, medium, and large plans. To create a sample of 
525 plans, IRS randomly drew equal numbers of plans from these small, 
medium, and large categories. The method that IRS used to create the 
sample of 525 401(k) plans from these categories was basically 
equivalent to drawing a simple random sample in which each plan had an 
equal probability of selection. However, before drawing 175 plans from 
each of these three groups, IRS carved out the 25 largest plans from 
the large-plan category and put these plans into a separate group that 
it called “superlarge plans”; this super large category was selected 
as a 100 percent sample of the largest 401(k) plans. Taken together, 
IRS’s sampling method was intended to produce a representative sample 
from, and reliable results for, the 401(k) plan population. The sample 
of 550 plans was assigned to IRS key district offices, where study 
coordinators were responsible for selecting the plan’s 1994 form 5500 
return and assigning the plan to an IRS examiner for audit. 

Examiners Completed a Study Questionnaire after Auditing Plans: 

IRS examiners were provided a questionnaire to obtain information on 
the compliance of the 401(k) plans and were instructed to complete the 
questionnaire after auditing each plan in the study. IRS examiners’ 
answers to the study questions were based on their plan audits. The 
questionnaire, or check sheet, that IRS used for its study was 
originally developed as part of a broad information-gathering project 
and included 254 questions to obtain information on 401(k) plan 
characteristics, design features, and compliance with certain 
requirements of the IRC. IRS used this available questionnaire to 
collect data relevant to its study objective of measuring 401(k) plan 
compliance. Once the study questionnaires were completed, they were 
sent to IRS Employee Plans headquarters for review and data analysis. 
Prior to its data analysis of 401(k) plan noncompliance, IRS reviewed 
all 550 questionnaires and excluded 78 of them from the analysis 
because the study questionnaire contained insufficient data or because 
the plan erroneously reported a 401(k) plan feature. 

IRS Analysis of 401(k) Plan Noncompliance Based on Compliance 
Indicator Data: 

Once the data on the 472 remaining plans were gathered in Employee 
Plans headquarters, IRS analysts identified 73 out of the 254 
questions on the questionnaire that they believed could indicate 
whether or not a plan was in compliance with certain requirements. 
That is, IRS identified the study questions it expected would provide 
information that a plan was either in compliance or not in compliance 
with certain requirements. These 73 “compliance indicators” became the 
focus of IRS’s analysis in identifying and summarizing the prevalence 
and types of noncompliance among 401(k) plans. The study questions 
that related to compliance issues included a range of items concerning 
certain statutory requirements that apply to all qualified defined 
contribution plans and concerning legal requirements that apply to 
qualified 401(k) plans. For example, the compliance indicators that 
IRS analyzed included items concerning employer contribution 
requirements, coverage rules, nondiscrimination provisions, and limits 
on employee contributions in addition to other important rules and 
requirements that qualified plans must satisfy. 

IRS data analyses identified the number of plans that failed to comply 
with one or more of their compliance indicators. The IRS study 
reported that 44 percent of the 472 plans remaining in the study had 
one or more instances of noncompliance with certain requirements that 
IRS examined; the other 56 percent of the plans were found to have no 
violations. These percentages varied slightly according to plan size 
category. The study reports that 41 percent of small plans, 47 percent 
of medium plans, and 44 percent of large and super-large plans had one 
or more instances of noncompliance. 

IRS also used its noncompliance indicator data to estimate, by 
calculating the number of times specific violations were identified, 
the frequency with which these violations occurred in its study 
sample. IRS analyses counted 251 instances of noncompliance that it 
categorized by requirements to which tax-qualified 401(k) plans should 
adhere. In total, the study publication uses 16 categories, such as 
nondiscrimination, loans, coverage, vesting, and participation, to 
report on various types of noncompliance that IRS found among the 
401(k) plans in the study.[Footnote 15] For each of the compliance 
categories, the study publication reports the total number of 
violations that occurred. According to the study report, the total 
number of violations for each compliance category cannot be correlated 
to the number of plans containing these violations because some plans 
may have contained more than one violation within a category. As a 
result, the study publication does not show how many plans had more 
than one instance of noncompliance in a single category or how many 
plans had more than one type of compliance violation. 

The analysis did not attempt to distinguish instances of noncompliance 
according to the severity of the violation. For the plans that had one 
or more instances of noncompliance, no study questions captured 
information on the insignificance or significance of the violations 
that IRS identified. Nor did the questionnaire include specific items 
on the number of participants (if any) affected and the amount of 
assets (if any) that were represented by the noncompliance errors IRS 
found. The questionnaire did contain items on the total number of plan 
participants and assets, but IRS did not analyze these data in 
relation to its findings on noncompliance. 

With Supplemental Information, IRS Modified Original Findings: 

IRS’s original estimates on 401(k) plan noncompliance decreased after 
IRS made some adjustments to its initial analysis of compliance 
indicator data. Initially, IRS used its noncompliance indicator data 
to produce estimates of 401(k) plan noncompliance. For some plans, 
however, IRS found problems with the data for specific compliance 
indicators. During its analysis, IRS told us that it sometimes 
discovered instances in which data for certain compliance indicators 
were found to be either inaccurate or insufficient to determine 
whether an instance of noncompliance had occurred. However, IRS’s 
discovery of discrepancies in the data was not the result of 
systematically reviewing all the compliance indicator data for each 
plan in the study. Instead, in some of these instances where IRS 
discovered problems with its compliance indicator data, these data 
were compared with information that IRS routinely captures about the 
results of their plan audits. According to IRS, analysts who worked on 
the data analysis met occasionally to review the data recorded on the 
study questionnaires and to determine whether the compliance study 
data were sufficient to identify noncompliance. After comparing the 
compliance indicator data with the other information that IRS collects 
on their audits of these 401(k) plans, the analysts made adjustments 
to the compliance indicator data. However, IRS analysts sometimes 
adjusted the data solely on the basis of their assessments that 
specific compliance indicators were not reliable or sufficient to 
determine whether or not a violation had occurred. Because these 
adjustments were not based on a systematic review of the accuracy and 
sufficiency of the data, we could not determine whether the 
adjustments that IRS made resolved all potential problems with its 
compliance indicator data. 

These adjustments changed noncompliant plans to compliant, and vice 
versa. For example, IRS analysts determined that some 401(k) plans 
with at least one violation of certain nondiscrimination requirements 
were found to be fully compliant once the additional information was 
included in the analysis. Also, some plans that had been included in 
the original estimate of plans with no compliance errors were 
determined to have at least one instance of noncompliance when IRS 
used this extra information to inform its analysis. 

When IRS used these adjustments to supplement the analyses it had 
performed, the total number of compliance violations decreased. At one 
point during its data analysis, IRS estimated that 298 total instances 
of noncompliance had occurred among the plans in the study. However, IRS
’s final estimate of the total number of compliance errors was revised 
downward to 251. As a result of these changes, IRS’s estimate of the 
percentage of 401(k) plans with one or more instances of noncompliance 
decreased from 56 percent to 44 percent. 

IRS Study Does Not Generally Provide Accurate Estimates of 401(k) Plan 
Noncompliance: 

The IRS study did not, in general, provide accurate estimates of the 
overall prevalence and types of noncompliance among 401(k) plans. IRS’
s estimates of noncompliance among 401(k) plans were inaccurate 
primarily because only 27 of the 73 questions that it identified as 
compliance indicators conclusively demonstrated a plan’s 
noncompliance. Also, the reported findings could not be generalized to 
the broader population of all 401(k) plans because the analysis did 
not take into account the sample weights. 

Fewer Than Half of the IRS Compliance Questions Conclusively 
Demonstrated Noncompliance: 

More than half of the study questions that IRS identified to analyze 
401(k) plan compliance were unable to conclusively demonstrate 
noncompliance. We asked IRS analysts involved in the study’s data 
analysis to evaluate the 73 questions that were selected as compliance 
indicators and determine whether these questions could definitively 
demonstrate a compliance violation. In evaluating each of the 
compliance indicators, IRS assessed whether the answers to these 
questions would provide information that was relevant to, or 
suggestive of, noncompliance or in fact demonstrated an instance of 
noncompliance. As a result of this evaluation, the IRS analysts 
identified only 27 questions that could definitively demonstrate 
noncompliance. In contrast, IRS determined that the remaining 46 
questions were not sufficient by themselves to demonstrate 
noncompliance because potential problems rendered these indicators 
less conclusive. Although a positive response was generally sufficient 
to demonstrate compliance, the IRS analysts whom we spoke with told us 
that additional information would be needed to determine whether or 
not negative answers to these questions conclusively indicated 
noncompliance. Consequently, the 44 percent of plans reported to have 
one or more compliance violations is at best an upper-bound estimate 
of the extent of noncompliance found in this study because the 
reported results are not limited to those items with sufficient 
information to identify noncompliance. 

IRS’s compliance indicators were not initially developed to 
specifically identify and substantiate noncompliance among 401(k) 
plans and the answers were not validated as accurately demonstrating 
noncompliance. Instead of formulating study questions that were 
directly relevant and sufficient to demonstrate noncompliance, IRS 
used an already available questionnaire that had been developed as 
part of a broad information-gathering project. This broadly scoped 
research project had been revised to address the narrower objective of 
401(k) plan compliance. Only after administering the check sheet and 
collecting the data did IRS identify the study questions that it 
expected to demonstrate noncompliance. As a result, most of the 254 
questions on the questionnaire were not directly relevant to the study 
objective of estimating noncompliance among 401(k) plans. Also, some 
of the answers expected to demonstrate noncompliance from IRS’s 
analysis of noncompliance indicator data were found to be suggestive, 
rather than demonstrative, of noncompliance. Although IRS, to help 
ensure ease in recording the answers, pretested the software that its 
examiners used to complete the study questionnaires, it did not 
pretest the study questions. Because IRS did not pretest the 
questionnaire for the accuracy and appropriateness of the answers, 
problems with the questionnaire were not identified or remedied before 
the data were collected. For example, answers might have more 
accurately reflected the types of information that were being sought 
if a preliminary evaluation and pretesting of the 73 compliance 
indicators had been used to improve the wording of the questions and 
the instructions provided to the examiners collecting the information. 

We found that the accuracy of IRS estimates was also hampered by the 
lack of adequate training for examiners who filled out the study 
questionnaires after completing the audits. Each field office sent 
representatives to a kickoff conference that provided training for the 
401(k) study. However, the training did not address which study 
questions would be used to distinguish compliance from noncompliance, 
because IRS identified these questions after the data were collected. 
Additionally, IRS told us that uniform audit standards were not 
developed to guide examiners in conducting the audits and in using the 
audit information to answer the study questions. As a result, an IRS 
analyst responsible for the data analysis stated that the 401(k) plan 
audits were not uniform and that some of the data were not collected 
consistently. Further, the representatives trained were not the 
examiners expected to conduct the audits and complete the subsequent 
questionnaires but rather the field office representatives charged 
with managing the local data collection efforts and transmitting data 
to headquarters for analysis. However, the field office 
representatives did not receive information regarding which questions 
would be used to distinguish compliance from noncompliance and thus 
could not relay this information to the auditing examiners. 

Despite the discovery of inaccurate and inconsistent answers, IRS did 
not systematically verify the accuracy of all the data analyzed. 
Instead, the IRS analyst who summarized the study data told us that he 
made some judgmental corrections to obviously incorrect or 
inconsistent answers rather than ordering the relevant closed case 
file or contacting the relevant examiner to obtain valid and accurate 
answers. As a result, some answers to certain study questions were not 
used in IRS’s final estimates of 401(k) plan noncompliance and others 
were used but judgmentally adjusted. In addition, the use of 
additional information to revise estimates of noncompliance was not 
well documented. We could not verify the revisions in IRS estimates, 
because IRS was not able to provide us with a single complete data 
file to check whether its reclassifications of plans as compliant or 
noncompliant were accurate. More complete documentation would have 
helped IRS ensure that it accurately estimated the proportion of plans 
that had one or more compliance errors and the frequency of occurrence 
for specific violations. 

Reported Results Cannot Be Generalized to Broader 401(k) Plan 
Population: 

Not all of the IRS study findings could be generalized to the broader 
population of all 401(k) plans, a fact that makes them less useful. To 
the extent that findings were reported separately for the small, 
medium, large, or super-large groupings, these results are reliable 
estimates for compliance errors of all plans in such groups (other 
data issues notwithstanding). For example, the report estimates that 
53 percent of the 162 medium plans audited had no violations. This 
figure can also be used as an estimate of the percentage of medium-
size plans in the broader population that had no violations (other 
data issues notwithstanding). However, in cases where compliance 
information was aggregated to include results from more than one 
group, such results are not reliable estimates for compliance errors 
of other plans in these groups. IRS sampled all of the super-large 
401(k) plans to ensure their inclusion in the study. Because the super-
large plans were a 100 percent sample and the plans sampled in the 
other plan-size categories each represented about 1,000 plans from the 
total population, combining sample results for these groups without 
weighting them gives the super-large plans more influence in the final 
answer than is warranted by their representation in the total 401(k) 
plan population. Proper weighting of all sample cases is necessary to 
make tabulations and other estimates that can be generalized to the 
broader 401(k) population. In some cases, information from large and 
super-large plans was combined for reporting. In other cases, 
information was combined for all plans studied. For example, the 
report estimates that 56 percent of the 171 large and super-large 
plans studied had no compliance errors and that 56 percent of all 472 
plans studied had no errors. In these cases where IRS has combined 
information for all plans in the study or for two plan-size 
categories, the reported percentages do not represent the percentages 
in the corresponding population of 401(k) plans. If IRS’s analysis had 
accounted for its sampling methodology, it is possible that IRS would 
have produced estimates similar to the reported results of the 401(k) 
study because the reported estimates of the proportion of plans with 
one or more instances of noncompliance were similar across plan-size 
categories. 

Furthermore, the 401(k) study findings cannot be used as estimates of 
noncompliance among the current population of 401(k) plans. To assess 
whether the 401(k) study results reflect the level and types of 
noncompliance among the current population of 401(k) plans, the data 
that support the published results would need further analysis to 
account for changes that have occurred in relevant pension laws since 
the study was undertaken. Although the 401(k) study publication 
describes changes to relevant pension laws that occurred during the 
course of the study, it is not possible to determine how these changes 
have affected noncompliance among 401(k) plans by simply examining the 
study findings. Also, changes have occurred in relevant pension laws 
since the study was published (see app. II for a description of 
changes in relevant pension laws since the study was published). 

IRS Is Planning to Continue Compliance Research: 

IRS is currently planning and conducting research on several types of 
private pension plans to determine the prevalence and types of 
noncompliance. To obtain information on the extent and types of 
noncompliance among these plans, IRS plans to conduct compliance 
studies similar to the one conducted on 401(k) pension plans. After 
implementing initiatives to improve compliance, IRS plans to once 
again collect and analyze similar compliance data to determine the 
effectiveness of its initiatives. In its ongoing research efforts, IRS 
is adopting lessons from its prior compliance study. 

Current Plans Include Compliance Study Initiatives on Several Types of 
Private Pensions: 

IRS is currently planning and conducting compliance research on 
several types of private pension plans. According to IRS officials who 
are involved in IRS enforcement and audit activities, compliance 
research will be used to help plan and implement initiatives that 
address compliance issues among various types of plans. IRS also told 
us that compliance research initiatives could be useful sources of 
information for plan sponsors and administrators, who are encouraged 
by IRS to use voluntary compliance procedures in identifying and 
remedying noncompliance. In addition, IRS uses this information to 
determine issues that are appropriate for published guidance. Ongoing 
compliance research is being conducted according to an overall 
strategy that IRS calls its market segment approach, developed to 
identify compliance issues among various types of tax-qualified 
pensions that employers sponsor. This market segment approach is being 
used by IRS to estimate the level and types of noncompliance among 
specific types of pension plans and to measure the impact of 
initiatives that IRS devises to address noncompliance. IRS has 
selected specific types of private plans for ongoing compliance 
research, including 401(k) plans, sections 403(b) and 457 plans, and 
multiemployer plans. IRS chose these plan types for several reasons, 
such as their prevalence, the significant degree of noncompliance 
known from past audits of these plan types, and/or the need to develop 
experience in conducting audits and compliance research. According to 
IRS, compliance studies for these plan types are in various stages of 
development and implementation. In the future, IRS plans to expand its 
compliance research and initiative development to other types of 
private plans. 

IRS officials whom we spoke with said that these compliance studies 
will be similar in overall design to the prior 401(k) study. For the 
various plan types that IRS has identified, IRS will select plans to 
study through sampling or some other mechanism. A study questionnaire 
will be developed to capture information about compliance with certain 
requirements. IRS examiners will audit plans that have been selected 
for the study and will answer study questions on the basis of the 
audits. The data that IRS collects will be analyzed, and the results 
will be used to estimate the extent and types of noncompliance among 
the plans in these studies. 

Study findings will be used by IRS as baseline information about 
noncompliance among the plan types selected for compliance research. 
After implementing initiatives designed to improve the compliance of 
the plan types that were selected for compliance research, IRS will 
conduct a follow-up compliance study to assess the impact of its 
compliance activities and specific initiatives. The follow-up studies 
will be designed to collect data that permit a comparison with 
baseline data from the initial studies of the level and types of 
noncompliance. IRS data analysis and examination of results from both 
the initial and follow-up studies will help IRS determine whether 
overall compliance has improved. 

IRS staff told us that as the results of compliance studies become 
available, IRS will be able to make better assessments of how to use 
compliance study data. For example, IRS has conducted compliance 
research on 403(b) plans that it has used to develop specific outreach 
and education initiatives, including a Web site with information on 
noncompliance and speaking points for IRS examiners who meet with plan 
sponsors and administrators. In addition, IRS plans to use its 
compliance studies to improve the way it conducts audits. For example, 
IRS intends to use the results of compliance studies to develop more 
standardized audit guidelines and targeted audits to better identify 
compliance issues among, and to limit plan audits to those issues 
relevant to, specific types of plans. 

Current IRS Compliance Research Initiatives Incorporate Lessons from 
Prior Study: 

IRS is adopting lessons learned from its prior compliance study to 
enhance the quality and usefulness of ongoing and future compliance 
research initiatives. Through our review of IRS work plans and 
interviews with IRS officials, we identified several aspects of 
current and future IRS compliance studies that are improvements on the 
prior 401(k) study. For example, IRS’s current approach to planning 
compliance research has become more comprehensive. Unlike the 1995 
401(k) study, IRS work plans indicate that “compliance planning groups”
have been assembled for each of the four plan segments on which IRS is 
conducting compliance research. These groups, which include key 
stakeholders from across the agency with expertise in various aspects 
of pension plan compliance, are being used to help IRS formulate 
comprehensive plans for conducting upcoming compliance research. 
According to IRS officials whom we spoke with, IRS will obtain 
guidance and input from its Research and Analysis group to assist with 
the design and implementation of its compliance studies. 

We identified other aspects of compliance studies, in addition to 
better planning, that improve on the prior 401(k) study. In conducting 
upcoming studies, IRS told us that it plans to develop and provide 
enhanced training for examiners who are responsible for auditing the 
plans and recording the study information. For example, IRS plans to 
conduct a training session for IRS examiners who will be assigned to 
conduct 401(k) plan audits for ongoing compliance research. IRS 
officials told us that examiners would receive training on the study 
questionnaires and in how to answer the study questions. In addition, 
part of the training that IRS intends to provide for 401(k) plan 
studies will be based on standardized guidelines that IRS has 
developed for collecting information from 401(k) plan audits. IRS has 
developed standardized audit guidelines for each of the plan types 
that the agency has selected for ongoing compliance research. 
According to IRS, these guidelines will help IRS examiners, including 
examiners involved in compliance studies, collect and record 
information consistently and accurately. 

IRS told us that it intends to incorporate other improvements into its 
upcoming 401(k) plan compliance studies. For example, IRS said that 
examiners who participate in upcoming IRS compliance studies will have 
a role in developing the questionnaires used to collect compliance 
study data, and IRS will pretest compliance study questionnaires to 
help determine their usefulness and the accuracy of the information 
that they are intended to collect. Also, IRS is developing automated 
tools that its examiners will use to record answers to compliance 
study questions. Automated tools that IRS examiners can use to collect 
information during the course of an audit have been developed for the 
401(k) plans but are still in development for other plan segments. 
According to IRS officials, these automated tools will help IRS 
produce work papers to document and verify its compliance study data. 

Conclusions: 

Compliance research studies could play an integral role in IRS’s 
efforts to ensure that tax-qualified pension plans adhere to 
applicable laws and regulations. The findings from such studies can 
provide data on the prevalence and types of noncompliance among 
pension plans, helping IRS shape its enforcement efforts. For example, 
IRS can use compliance study findings to identify key aspects of 
noncompliance among specific types of plans and develop targeted 
audits and other activities to address compliance issues. In recent 
years, IRS enforcement efforts have placed greater emphasis on 
voluntary correction procedures—that is, encouraging plan sponsors to 
correct violations that are discovered. Information on noncompliance 
that is useful and accurate could help improve targeting for audits 
and enhance voluntary compliance initiatives that assist plan sponsors 
in discovering and making such corrections. Compliance research can 
also measure the impact of such efforts to determine whether they are 
effective. 

The more accurate the findings from compliance studies, the better 
able IRS is to ensure that plans are operating in accordance with 
applicable requirements, so that participants receive the coverage and 
benefits to which they are entitled. Compliance study findings can 
help IRS tailor its initiatives to identify, monitor, and address the 
most essential aspects of noncompliance among specific types of 
pension plans and measure whether its activities are effective in 
promoting compliance among plan sponsors. IRS recognizes the need to 
improve the way it conducts compliance studies and is in the process 
of implementing specific steps to improve aspects of planning and 
conducting these studies. Since IRS compliance research is focused on 
other types of plans besides 401(k) plans, it is important that IRS 
consistently implement these steps throughout its ongoing and future 
compliance research initiatives. 

Several shortcomings of the 1995 IRS 401(k) study undercut its 
effectiveness in meeting IRS’s research objective of estimating the 
extent and types of noncompliance among 401(k) plans. These 
shortcomings cut across important components of the 401(k) study, 
including questionnaire design, data collection, and data analysis. 
Whether these and other elements of research are designed and carried-
out in a sound manner help determine the effectiveness of research 
studies in meeting their objectives. For example, the 1995 401(k) 
study questions were not pretested to determine whether they would 
have produced demonstrative data on noncompliance, and examiners who 
completed the study questionnaires were not provided with training on 
answering the questions in an accurate and uniform manner. To ensure 
the accuracy of its findings, IRS will need to build steps into its 
compliance studies that improve the accuracy and usefulness of the 
data that are collected, analyzed, and reported. Additionally, 
documenting a research study can help produce evidence that supports 
the answers to the research questions. Insufficient documentation 
limits the perceived accuracy and the usefulness of a research study. 

Recommendations: 

To ensure the quality and usefulness of ongoing and future compliance 
studies in providing information that enhances IRS’s efforts to 
promote compliance among private pension plans, IRS should take steps 
to improve how it conducts compliance study research. These steps, in 
addition to the agency’s current efforts to improve the quality of 
compliance studies, should be incorporated into all planned compliance 
studies. Accordingly, we are making three recommendations to the IRS 
Commissioner for all future compliance studies. 

We recommend that IRS pretest compliance study questionnaires to 
obtain information on the usefulness and accuracy of the answers in 
achieving IRS’s research objective. 

We also recommend that IRS provide uniform and comprehensive training 
to examiners who participate in compliance studies, so that they will 
know what information is needed to answer the study questions and can 
collect this information consistently and accurately. 

Finally, we recommend that IRS maintain sufficient written or 
electronic documentation to enable it to validate and verify the 
results of compliance studies with evidence; this would allow IRS to 
explain the methods used to analyze study data and arrive at findings. 

Agency Comments: 

We provided a draft of the report to the Commissioner of the IRS and 
the Department of the Treasury. IRS generally agreed with our 
findings, conclusions, and recommendations. In its letter, IRS notes 
that has incorporated our recommendations in a current compliance 
study on 401(k) plans. We agree that IRS has taken specific steps to 
improve its current 401(k) plan compliance study and describe these 
steps in our report. In addition to the current 401(k) study, IRS 
should also implement our recommendations throughout its current and 
upcoming compliance study initiatives on 401(k) and other types of 
pension plans. The IRS also provided us with technical comments, which 
we incorporated as appropriate. IRS’s comments are included in 
Appendix III. 

We are sending copies of this report to the Honorable Paul H. O’Neill, 
Secretary of the Treasury, the Honorable Charles O. Rossotti, 
Commissioner of the IRS, and other interested parties. We will also 
make copies available to others on request. If you or your staff have 
any questions concerning this report, please call me at (202) 512-
7215. Key contributors are listed in appendix III. 

Sincerely Yours, 

Signed by: 

Barbara D. Bovbjerg, Director: 
Education, Workforce, and Income Security Issues: 

[End of section] 

Appendix I: Scope and Methodology: 

Determining the Steps IRS Took to Conduct the 401(k) Study: 

To determine what IRS did to estimate the prevalence and types of 
401(k) plan noncompliance with the requirements of the Internal 
Revenue Code, we reviewed the final 401(k) compliance study report 
that IRS posted to its Web site. In addition, we reviewed the initial 
and interim draft reports that we received from IRS, as well as study-
related work papers, which documented the design, implementation, and 
analysis components of the study. We also interviewed IRS officials in 
the Employee Plans area of IRS’s Tax Exempt and Government Entities 
Division, including officials in the Office of Examinations and the 
Office of Education and Outreach who were responsible for conducting 
and disseminating compliance research on private plans to obtain 
information about how IRS designed and conducted the study. Our work 
focused on identifying and summarizing the major components of the 
401(k) study in relation to key elements of research study methodology 
including the study objective, study design, sample selection, 
questionnaire design, data collection, and data analysis. 

Evaluating the IRS Study on the Prevalence and Types of Noncompliance 
among 401(k) Plans: 

Our evaluation of IRS’s estimates of the prevalence and types of 
noncompliance was limited because IRS was unable to provide us with a 
complete data set or documentation that supports the final study 
results. As a result, we could not assess the usefulness of the study 
in relation to compliance among the broader population of 401(k) plans 
because we did not have data or other documentation that supported IRS 
estimates on specific types of noncompliance. Without this 
information, we could not make the appropriate sample weight 
adjustments to assess IRS estimates of the overall prevalence of 
noncompliance among all plans in the study or within specific plan 
size categories. Furthermore, the lack of a complete data set or 
comprehensive documentation supporting the published results limited 
our ability to reliably assess revisions in IRS estimates of the 
proportion of plans that had one or more compliance errors and the 
frequency with which specific types of errors occurred among the plans 
in the study. Additionally, the inability of IRS to provide closed 
case file information on audited plans limited our ability to assess 
the reliability of the data collected for analysis. 

In light of these limitations, we elected to assess to what extent the 
IRS study provides accurate estimates on 401(k) plan noncompliance by 
evaluating, in relation to published guidance for conducting research, 
how the study was conducted. We evaluated the IRS study using a series 
of brochures on surveys published by the American Statistical 
Association (ASA) and published GAO guidance on methodology and 
program evaluation.[Footnote 16] These published guidelines address 
important elements of research studies, such as sampling and 
questionnaire design. Our evaluation examined and compared the 
sampling methodology, the questionnaire development, the data 
collection process, and the data analysis on which the IRS report is 
based with ASA and GAO guidelines on each of these elements. To 
examine and compare elements of the IRS study with published guidance, 
we collected and reviewed relevant documents such as draft study 
reports, the questionnaire check sheet, and other working papers made 
available by IRS. We also received and examined many electronic data 
files pertaining to the 401(k) study. Additionally, we interviewed IRS 
analysts who were responsible for conducting the data analysis and the 
IRS statistician who assisted with selecting the stratified random 
sample. 

Describing IRS Plans for Compliance Research on Private Pensions: 

To describe IRS’s current efforts in planning and conducting 
compliance research on private pension plans, we: 

* reviewed draft work plans for IRS’s ongoing and future compliance 
research initiatives, including plans for an upcoming 401(k) plan 
compliance study; 

* discussed lessons learned from the prior 401(k) study with IRS 
officials and analysts involved in compliance research initiatives; 

* interviewed IRS officials in the Division of Tax Exempt and 
Government Entities, Employee Plans office to discuss the role of 
compliance research in IRS efforts to promote compliance among plan 
sponsors; 

* reviewed official IRS guidance on agency procedures for identifying 
and remedying compliance violations; and; 

* discussed how compliance research initiatives can inform IRS’ 
voluntary compliance activities with IRS officials. 

We assessed IRS work plans and our discussions with IRS to identify 
and summarize the agency’s overall plans for ongoing and future 
compliance research, including the role of compliance studies. As part 
of our work, we identified lessons learned from the previous 401(k) 
study that IRS has adopted in its plans to design and conduct 
compliance research initiatives. 

[End of section] 

Appendix II: Changes in the Laws That Govern Tax-Qualified 401(k) 
Plans: 

The IRS 401(k) study publication provides information on changes in 
relevant laws that occurred while the study was performed. The summary 
information that IRS includes in its published study report describes 
changes in relevant pension laws since the study was conducted and is 
pertinent up to the time at which the profile was posted on IRS Web 
site. This appendix summarizes and describes key changes in laws that 
apply to tax-qualified 401(k) plans that have occurred since the 
release of the published 401(k) study report, mostly changes arising 
from the Economic Growth and Tax Relief Reconciliation Act (EGTRRA) 
[Footnote 1] as they relate to violation categories identified in the 
study. Our summary of recent changes in applicable laws is grouped by 
the compliance categories that IRS used to present its 401(k) study 
results. 

Neither the IRS 401(k) study report nor this appendix should be 
regarded as a comprehensive explanation of the laws that relate to tax-
qualified pension plans in general and tax-qualified 401(k) plans in 
particular. While this appendix provides context where necessary to 
understand how EGTRRA provisions change certain pension laws, it does 
not provide a history or complete description of the purpose and 
nature of the Internal Revenue Code (IRC) requirements that EGTRRA 
changes. The published 401(k) study report provides more in-depth 
description of the purpose and requirements of the specific IRC 
provisions that IRS examined as part of its 401(k) study. 

A. Distributions Eligible For Rollover Treatment EGTRRA section 636(b) 
mandates that any distribution made upon hardship of an employee will 
not be an eligible rollover distribution. Thus, no assets distributed 
to an employee on account of his or her hardship will be eligible for 
direct rollover to another plan or individual retirement account 
(IRA). Such distributions will therefore be subject to the withholding 
rules applicable to distributions that are not eligible rollover 
distributions. 

Section 401(a)(31) of the Code provides that participants receiving an 
eligible rollover distribution must have the option to have the 
distribution transferred in the form of a direct rollover to another 
eligible retirement plan. If an eligible rollover distribution is not 
transferred by a direct rollover, the distribution is subject to 
withholding at a 20% rate, under section 3405(c)(1). 

Regulations under section 401(k) currently provide that elective (pre-
tax) deferrals under a 401(k) plan can, if the plan provides, be 
distributed (without earnings) in the event of the financial hardship 
of the employee. The regulations provide that a distribution is made 
on account of hardship only if the distribution is made on account of 
an immediate and heavy financial need of the employee and the 
distribution is necessary to satisfy such financial need.[Footnote 18] 
Under pre-EGTRRA law, hardship withdrawals of elective deferral 
amounts under 401(k) plans were not eligible for rollover, while other 
types of hardship distributions (e.g., employer matching contributions 
distributed on account of hardship) were eligible rollover 
distributions. Different withholding rules apply to eligible rollover 
distributions than to distributions that are not eligible rollover 
distributions. 

EGTTRA section 641(c) also provides for an expanded explanation to 
recipients of rollover distributions. This provision requires that the 
rollover notice include a description of the provisions under which 
distributions from the eligible retirement plan receiving the 
distribution may be subject to restrictions and tax consequences which 
are different from those applicable to the plan making the 
distribution. 

Effective for distributions after December 31, 2001, EGTRRA section 
641 allows rollovers among 401(k) plans, 403(b) plans, or governmental 
section 457 plans. EGTRRA section 657 mandates that unless the 
participant elects otherwise, any eligible rollover distribution in 
excess of $1,000 that may be distributed without the participant’s 
consent be automatically rolled over to a designated IRA. This change 
applies to distributions that occur after the Department of Labor 
issues final regulations implementing section 657. 

Section 642(a) of EGTRRA provides that an eligible rollover 
distribution from an IRA may be rolled over to another IRA or an 
eligible retirement plan as long as the amount is transferred no later 
than 60 days after the date the distribution was received. 

Section 642(b)(3) of EGTRRA provides that a distribution from a 
Savings Incentive Match Plan for Employees (SIMPLE) IRA may also be 
rolled over to another SIMPLE IRA. 

Under pre-EGTRRA law, elective (pre-tax) deferrals may not be 
distributed earlier than one of the events described in section 
401(k)(2)(B) or section 401(k)(10). EGTRRA modifies these rules as 
they apply in the case of a corporate transaction, such as an asset or 
stock sale, that results in employees of the seller going to work for 
the buyer. Pre-EGTRRA law permits distribution in the case of certain 
types of transactions but not others. EGTRRA section 646 amends 
section 401(k)(2)(B) by replacing “separation from service” with the 
more lenient standard of “severance from employment.” This generally 
will permit distributions to employees who move from seller to buyer 
in connection with a corporate transaction, unless corresponding 
assets of the seller’s plan move as well. Section 646 of EGTRRA also 
makes conforming changes to section 401(k)(10). The amendments made by 
section 646 apply to distributions made after December 31, 2001. 

B. Nondiscrimination (ADP/ACP): 

EGTRRA section 666 repeals the multiple use test effective for plan 
years beginning after December 31, 2001. 

The multiple use test occurs where a 401(k) plan is subject to both 
the ADP and ACP tests and both tests can only be satisfied using the 
alternative limitations of those tests described under section 
401(k)(3) and section 401(m)(2) (the 2 percentage point limit or the 
200 percent limit). The purpose of the multiple use test is to prevent 
the multiple use of the more generous alternatives for meeting both 
the ACP and the ADP test when certain employees are eligible under 
both a section 401(k) plan and a section 401(m) plan. 

C. Loans: 

EGTRRA section 612 repeals the rule prohibiting loans to sole 
proprietors, partners who own more than 10% of the partnership, and 
shareholders of S corporations who own more than 5% of the S 
corporation effective for years beginning after December 31, 2001. 

D. Contingent Benefits: No change. 

E. Hardship Distributions: 

EGTRRA section 636(a) directs that the regulations under section 
401(k) be revised to permit a participant who receives a hardship 
distribution to resume elective (pre-tax) deferrals 6 months, instead 
of 12 months, after receiving a hardship distribution. This change is 
effective for years beginning after December 31, 2001. 

F. Top-Heavy Requirements: 

EGTRRA section 613 generally simplifies several elements of top-heavy 
testing and their application. First, it simplifies the definition of 
key employee, so that the term includes only individuals who during 
the year in question or the immediately preceding year were officers 
earning over $130,000 (adjusted for cost of living increases), 5% 
owners, or 1% owners earning more than $150,000. Second, it specifies 
that in determining whether or not a plan is top heavy, only 
distributions made within the preceding 1 year, rather than the 
preceding 5 years (except for in-service distributions, for which the 
5-year rule will continue to apply) must be added. Third, it requires 
that matching contributions to a top-heavy plan be counted in 
determining whether nonkey employees have received the required 
minimum benefit. Last, it states that certain plans meeting safe-
harbor requirements applicable to the nondiscrimination rules 
regarding 401(k) and matching contributions will automatically be 
deemed to not be top heavy, and frozen defined benefit plans (with 
respect to which there are no current benefit accruals for current or 
former key employees) will be exempt from certain of the minimum 
accrual requirements. The new rules are effective for years beginning 
after December 31, 2001. 

G. Coverage: 

EGTRRA section 664 directs that the regulations under Code section 
410(b) be revised to allow a 401(k) plan to treat as excludable 
employees the employees of a Code Section 501(c)(3) entity who are 
eligible for a Code section 403(b) arrangement provided that: (1) no 
employee of the 501(c)(3) entity is eligible to participate in a 
401(k) plan; and (2) at least 95 percent of the employees who are not 
employees of the 501(c)(3) entity are eligible to participate in the 
401(k) plan. This change is effective January 1, 1997. 

H. Section 415: 

Under EGTRRA section 611, the $35,000 limit on combined employer and 
employee contributions for defined contribution plans is raised to 
$40,000 (indexed for the cost of living in $1,000 increments). The 25% 
of compensation limit is increased to 100% of compensation. Therefore, 
the new 415(c) limit will be the lesser of (1) 100% of compensation or 
(2) $40,000 (adjusted for cost of living increases). This provision is 
effective for years beginning after December 31, 2001. Catch-up 
contributions are not taken into account in applying the $40,000 limit. 

Section 611(d) of EGTRRA also increases the limit on elective 
contribution under Code section 402(g) from $10,500 in 2001 to $11,000 
in 2002; $12,000 in 2003; $13,000 in 2004; $14,000 in 2005; and 
$15,000 in 2006. The limit is adjusted for increases in the cost of 
living for years after 2006 in $500 increments. 

Section 631 of EGTRRA amends Code section 414 and provides that the 
otherwise applicable dollar limit on elective deferrals under a 401(k) 
plan, 403(b) plan, SEP, or SIMPLE plan, or deferrals under a 
governmental 457 plan will be increased for individuals who have 
attained age 50 before the end of the plan year, and who have 
otherwise already made the maximum permitted deferral under the Code 
or the plan or arrangement. The additional or “catch-up” contribution 
amount under a 401(k) plan, 403(b) plan or 457 plan is $1,000 for 
2002, $2,000 for 2003, $3,000 for 2004, $4,000 for 2005, and $5,000 
for 2006 and thereafter. The limit is adjusted for cost of living 
increases for years after 2006 in $500 increments. These additional 
contributions are for individuals who are age 50 and or older and such 
contributions will not violate the nondiscrimination, top-heavy or 415 
requirements. 

Under Code section 401(a)(17), for years beginning after December 31, 
2001, the amount of compensation that may be taken into account under 
a 401(k) plan is also increased from $150,000 (adjusted for cost of 
living increases to $170,000 in 2001) to $200,000. This limit is 
adjusted for cost of living increases in $5,000 increments. 

I. Nondiscrimination under Section 401(a)(4): No change. 

J. Vesting: 

Under EGTRA section 633, employer matching contributions must vest at 
least as rapidly as under one of two new vesting schedules. These 
schedules provide for faster vesting than the current schedules. The 
first schedule requires 100% vesting after three years of service and 
the second requires 20% vesting after two years of service with an 
additional 20% vesting for each year of service, reaching 100% vesting 
after six years of service. This provision is effective for 
contributions for plan years beginning after December 31, 2001, with a 
delayed effective date for plans maintained pursuant to collective 
bargaining agreements. 

K. Prohibited Transactions: No change. 

L. Plan Asset Rule: 

EGTRRA section 655 modifies the effective date of the rule excluding 
certain elective deferrals (and earnings thereon) from the definition 
of eligible individual account plan by providing that the rule does 
not apply to any elective deferral which is invested in qualifying 
employer securities, qualifying employer real property, or both 
acquired before January 1, 1999. 

M. Partnership Issues: No change. 

N. Participation: No change. 

O. Miscellaneous Limits: 

Under the Taxpayer Relief Act of 1997,[Footnote 19] the former 15% tax 
on excess distributions and the 15% estate tax on excess retirement 
accumulations from qualified retirement plans, tax-sheltered 
annuities, and individual retirement arrangements is repealed. 

P. Miscellaneous Violations: No change. 

[End of section] 

Appendix III: Comments from the Internal Revenue Service 

DEPARTMENT OF THE TREASURY: 
INTERNAL REVENUE SERVICE: 
COMMISSIONER: 
WASHINGTON, D.C. 20224: 
        
April 4, 2002: 

Ms. Barbara D. Bovbjerg: 
Director, Education, Workforce and Income Security Issues: 
United States General Accounting Office: 
Washington, D.C. 20548: 

Dear Ms. Bovbjerg: 

We reviewed your draft report on the quality and usefulness of our 
compliance studies. We find that it fairly and accurately describes 
our efforts to conduct a compliance study of IRC 401(k) plans. While 
the report focuses on some of the deficiencies of our previous study, 
we have learned many lessons from this study that will help us improve 
future compliance studies. 

We have already implemented your recommendations on improving compliance
initiatives in our current compliance study of IRC 401(k) plans. We 
are conducting a pilot training class and examination initiative to 
test automated workpapers that will allow us to more consistently 
conduct examinations of IRC 401(k) plans. The automated workpapers 
will also allow us to electronically store certain examination results 
that will help validate the results of the compliance study. We will 
also test a compliance questionnaire as part of the pilot. We plan to 
conduct a comprehensive IRC 401(k) study in the future, after we 
complete this pilot project. 

I would like to clarify your statement on page 4 of the report that 
indicates we are currently conducting a compliance study of the "plans 
of large corporations." We are not conducting a "study" of these 
plans. Rather, as part of the IRS redesign efforts, a Phase III 
Modernization Redesign Team proposed establishing an Employee Plans 
Team Audit (EPTA) program. The TEGE Implementation Steering Committee 
approved this program to address the compliance of the large plans 
that affect over 65% of all plan participants and over 65% of plan 
assets. 

Thank you again for the opportunity to respond to your report. If you 
have any questions or concerns, please contact Carol Gold, Director of 
Employee Plans, Tax Exempt and Government Entities Division at (202) 
283-2100. 

Sincerely, 

Signed by: 

Charles O. Rossotti: 

[End of section] 

Appendix IV: GAO Contacts and Staff Acknowledgments: 

GAO Contacts: 

Barbara D. Bovbjerg, (202) 512-7215: 
George A. Scott, (202) 512-5932: 

Staff Acknowledgments: 

In addition to those named above, Jeremy Citro, Gene Kuehneman, Ed 
Nannenhorn, Corinna Nicolaou, and Roger Thomas made key contributions 
to this report. 

[End of section] 

Footnotes: 

1. Joint Committee on Taxation, “Estimates of Federal Tax Expenditures 
for Fiscal Years 2002 2006,” JCS-1-02 (Washington, D.C.: January, 
2002). Pension contributions that fall within statutory limits, as 
well as investment earnings on pension plan assets, are not taxed 
until benefits are paid to plan participants. 

2. A 401(k) plan allows an employee to optionally reduce his or her 
salary, subject to certain limits, in exchange for the opportunity to 
make tax-free contributions to a qualified retirement account, within 
certain limits (26 U.S.C. 401(k)). Investment income earned on section 
401(k) account balances accumulates tax-free until the individual 
withdraws the funds. During the years covered by the IRS 401(k) study, 
only private taxable employers were eligible to establish a 401(k) 
plan. The Small Business Job Protection Act of 1996 permitted tax-
exempt employers to establish 401(k) plans, effective January 1, 1997. 

3. IRS published the study report on its Web site [hyperlink, 
http://www.irs.gov]. 

4. Additionally, the twenty-five largest 401(k) plans were included as 
a 100 percent sample of the super-large category. 

5. Upon reviewing the complete study information for all 550 plans, 
IRS excluded from its analysis 78 plans that either were not 401(k) 
plans or for which IRS had not obtained sufficient data. 

6. A stratified sample divides a population into smaller groups, or 
strata, and draws a unique sample from each stratum. A researcher may 
sample a larger proportion of a particular stratum of interest to 
improve overall precision or to ensure that reliable estimates for 
that particular stratum can be made. When estimates about the entire 
population are to be made, such sample observations should be given 
less weight than observations selected from other strata. Because each 
sample observation from the stratum of interest represents fewer total 
elements of the population than do sample observations from other 
strata, combining sample observations without weighting them gives the 
stratum of interest more influence in the final answer than is 
indicated by its representation in the population. Thus, using weights 
in combining sample data is necessary to make tabulations and other 
estimates that can be generalized to a broader population. 

7. 403(b) and 457 plans are authorized to include salary deferral 
arrangements under the Internal Revenue Code. Similar to 401(k) plans, 
these defined contribution plans permit participants to make tax-
deferred contributions instead of receiving the same amount as taxable 
salary. Whereas 401(k) plans, during the years covered by the IRS 
study, were solely for private, for-profit employers, 403(b) plans are 
available to employers who are tax-exempt under Code section 501(c)(3) 
and public educational organizations and 457 plans are available to 
state or local governments or tax-exempt employers. (26 U.S.C. 403(b); 
26 U.S.C. 457(g)). 

8. Multiemployer plans are provided for under the Internal Revenue 
Code. These plans are arrangements to which more than one employer is 
required to contribute and are maintained pursuant to one or more 
collective bargaining agreements between one or more employee 
organization(s) and more than one employer (26 U.S.C. 414(f)(1)). 

9. Certain taxable or tax-exempt private employer with 100 or fewer 
employees may establish a SIMPLE plan (26 U.S.C. 408(p)). 

10. See the IRS 401(k) study publication for information on 
requirements for SIMPLE 401(k) plans and safe-harbor designs. 

11. IRS also classifies compliance failures into three categories: 
insignificant, significant, and egregious. The IRS outlines seven 
primary criteria in Revenue Procedure 2001-17 for plan sponsors to 
determine whether or not a qualification failure is insignificant or 
significant. These terms generally apply to operational failures only. 

12. Nondiscrimination requirements stipulate that contributions or 
benefits must be apportioned in a nondiscriminatory manner between a 
top group of highly paid employees and owner-employees, and workers 
outside the top group (26 U.S.C. 401(a)(4), 401(a)(26), 410(b)). These 
provisions of the Code applied to 401(k) plans during the IRS study. 
(26 U.S.C. 401(a)(26) no longer applies to DC plans). 

13. Under SCP, pension plan sponsors may also voluntarily self-correct 
significant Operational failures within a limited time frame. 

14. See U.S. General Accounting Office, Pension Plans: IRS Programs 
for Resolving Deviations From Tax-Exemption Requirements, [hyperlink, 
http://www.gao.gov/products/GAO/GGD-00-169] (Washington, D.C.: August 
14, 2000) for additional information on IRS compliance procedures. 

15. See the study publication for the 16 compliance categories that 
IRS used and the total number of violations within each of these 
categories. 

16. U.S. General Accounting Office, Designing Evaluations, [hyperlink, 
http://www.gao.gov/products/GAO/PEMD-10.1.4] (Washington, D.C.: March 
1991); Developing and Using Questionnaires, [hyperlink, 
http://www.gao.gov/products/GAO/PEMD-10.1.7] (Washington, D.C.: 
October 1993); and Using Structured Interviewing Techniques, 
[hyperlink, http://www.gao.gov/products/GAO/PEMD-10.1.5] (Washington, 
D.C.: July 1991). 

17. P.L. 107-16, June 7, 2001. 

18. 26 CFR § 1.401(k) – 1(d). 

19. P.L. 105-34 § 1073. 

[End of section] 

GAO’s Mission: 

The General Accounting Office, the investigative arm of Congress, 
exists to support Congress in meeting its constitutional 
responsibilities and to help improve the performance and accountability 
of the federal government for the American people. GAO examines the use 
of public funds; evaluates federal programs and policies; and provides 
analyses, recommendations, and other assistance to help Congress make 
informed oversight, policy, and funding decisions. GAO’s commitment to 
good government is reflected in its core values of accountability, 
integrity, and reliability. 

Obtaining Copies of GAO Reports and Testimony: 

The fastest and easiest way to obtain copies of GAO documents at no 
cost is through the Internet. GAO’s Web site [hyperlink, 
http://www.gao.gov] contains abstracts and fulltext files of current 
reports and testimony and an expanding archive of older products. The 
Web site features a search engine to help you locate documents using 
key words and phrases. You can print these documents in their entirety, 
including charts and other graphics. 

Each day, GAO issues a list of newly released reports, testimony, and 
correspondence. GAO posts this list, known as “Today’s Reports,” on its 
Web site daily. The list contains links to the full-text document 
files. To have GAO e-mail this list to you every afternoon, go to 
[hyperlink, http://www.gao.gov] and select “Subscribe to daily E-mail 
alert for newly released products” under the GAO Reports heading. 

Order by Mail or Phone: 

The first copy of each printed report is free. Additional copies are $2 
each. A check or money order should be made out to the Superintendent 
of Documents. GAO also accepts VISA and Mastercard. Orders for 100 or 
more copies mailed to a single address are discounted 25 percent. 
Orders should be sent to: 

U.S. General Accounting Office: 
441 G Street NW, Room LM: 
Washington, D.C. 20548: 

To order by Phone: 
Voice: (202) 512-6000: 
TDD: (202) 512-2537: 
Fax: (202) 512-6061: 

To Report Fraud, Waste, and Abuse in Federal Programs Contact:
Web site: [hyperlink, http://www.gao.gov/fraudnet/fraudnet.htm]: 
E-mail: fraudnet@gao.gov: 
Automated answering system: (800) 424-5454 or (202) 512-7470: 

Public Affairs: 

Jeff Nelligan, managing director, NelliganJ@gao.gov: 
(202) 512-4800: 
U.S. General Accounting Office: 
441 G Street NW, Room 7149:
Washington, D.C. 20548: 

[End of document]