This is the accessible text file for GAO report number GAO-09-716
entitled 'Pension Benefit Guaranty Corporation: More Strategic Approach
Needed for Processing Complex Plans Prone to Delays and Overpayments'
which was released on September 16, 2009.
This text file was formatted by the U.S. Government Accountability
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.
Report to Congressional Requesters:
United States Government Accountability Office:
GAO:
August 2009:
Pension Benefit Guaranty Corporation:
More Strategic Approach Needed for Processing Complex Plans Prone to
Delays and Overpayments:
GAO-09-716:
GAO Highlights:
Highlights of GAO-09-716, a report to congressional requesters.
Why GAO Did This Study:
As the insurer of over 29,000 private sector defined benefit plans, the
Pension Benefit Guaranty Corporation (PBGC) may be required to assume
responsibility for the plans of a growing number of companies filing
bankruptcy due to the recession. Concerns about PBGC’s benefit
determination process, reductions in benefits due to guarantee limits,
and workers’ retirement security overall led the chairmen and ranking
members of the Senate Health, Education, Labor, and Pensions Committee
and the Senate Finance Committee, among others, to ask GAO to study:
(1) how long it takes PBGC to make benefit determinations;
(2) the extent of overpayments on retirees’ benefits;
(3) how well PBGC communicates with participants; and;
(4) the timeliness and accessibility of the appeals process.
To conduct this study, GAO reviewed PBGC policies and procedures,
analyzed automated data and case files, and interviewed PBGC officials
and certain associations, participants, and their representatives from
among those most affected by the process.
What GAO Found:
GAO’s review of plans terminated with insufficient funds and trusteed
by PBGC during fiscal years 2000 through 2008 revealed that a small
number of complex plans--especially those with large numbers of
participants affected by limits on guaranteed benefit amounts--
accounted for most cases with lengthy delays and overpayments.
Processing times. PBGC completed most participants’ benefit
determinations in less than 3 years, but required more time—up to 9
years—to process determinations for complex plans and plans with
missing data. Nearly three-quarters of the lengthiest processing times
were associated with individuals in just 10 of the 1,089 plans
reviewed. Although PBGC has taken steps to shorten this process, its
initiatives do not address the longest delays.
Overpayments. Although many participants are affected by sizable
benefit reductions, the vast majority are not affected by overpayments.
Moreover, nearly two-thirds of overpayments involved participants in
just 10 plans. In cases with overpayments, PBGC’s policy generally
requires participants’ benefits to be reduced by no more than 10
percent until the amount owed is repaid, but due to participants’ ages,
the full amount often is never recouped.
Figure: A Small Number of Complex Plans Account for Most Delays and
Overpayments:
[Refer to PDF for image: vertical bar graph]
Cases with processing times of 4 or more years (78,553 total cases from
561 plans):
Top 10 plans: 72.8%;
All other plans: 27.2%.
Cases with overpayments (22,623 total cases from 467 plans):
Top 10 plans: 65.3%;
All other plans: 34.7%.
Source: GAO analysis of PBGC data on participants in plans terminated
and trusteed during fiscal years 2000 through 2008.
[End of figure]
Communication. PBGC has made efforts to improve communication, but key
correspondence often did not meet the needs of those in complex plans.
For example, when the process was lengthy, PBGC did not communicate
with some participants for several years. When the benefit calculation
was complicated, PBGC did not provide explanations that could be
understood without further information or assistance.
Appeals. Since restructuring the appeals process in 2003, PBGC has
reduced the average amount of time needed to decide an appeal by almost
a year. However, the agency does not readily provide key information
that would be helpful to participants in deciding whether to pursue an
appeal.
What GAO Recommends:
GAO is recommending that PBGC develop a better strategy for processing
complex plans in order to reduce delays, minimize overpayments, improve
communication with participants, and make the appeals process more
accessible. PBGC generally agreed with the recommendations.
View [hyperlink, http://www.gao.gov/products/GAO-09-716] or key
components. For more information, contact Barbara Bovbjerg at (202) 512-
7215 or bovbjergb@gao.gov.
[End of section]
Contents:
Letter:
Background:
Most Benefit Determinations Are Completed within 3 Years, but Cases
with Complex Plans and Missing Data Take Longer:
Overpayments Have Been Infrequent and Mostly Concentrated in a Few
Complex Plans:
PBGC's Initial Communications Drew Praise, but Later Communications
Were Sometimes Found Lacking:
Restructured Appeals Process Resolves Requests More Efficiently but Key
Information Is Not Readily Provided:
Conclusions:
Recommendations:
Agency Comments and Our Evaluation:
Appendix I: Scope and Methodology for Analysis of Automated and Imaged
Data:
Appendix II: Organizations and Participants Contacted:
Appendix III: The Process for Allocating Assets and Recoveries to
Participant Benefits:
Appendix IV: Limits on Guaranteed Benefits from a Participant's
Perspective:
Appendix V: Profiles of Three Large, Complex Plans:
Appendix VI: Three Categories of Large, Complex Plans:
Appendix VII: Sample Benefit Determination Letter:
Appendix VIII: Comments from the Pension Benefit Guaranty Corporation:
Appendix IX: GAO Contact and Staff Acknowledgments:
Related GAO Products:
Tables:
Table 1: Types of Guaranteed Benefit Limits under ERISA:
Table 2: Characteristics of 10 Large, Complex Plans:
Table 3: Priority Categories for Allocating Participant Benefits:
Table 4: Asset and Recovery Allocation among 10 Large, Complex Plans:
Table 5: Profile of a Bethlehem Steel Terminated Pension Plan:
Table 6: Profile of a Republic Technologies International Terminated
Pension Plan:
Table 7: Profile of a US Airways Terminated Pension Plan:
Table 8: Ten Plans Most Affected by Guaranteed Benefit Limits:
Table 9: Ten Plans Most Affected by Long Processing Times:
Table 10: Ten Plans Most Affected by Overpayments:
Figures:
Figure 1: Growth in Number of Single-Employer Plans under PBGC
Trusteeship (1975-2008):
Figure 2: PBGC's Benefit Determination Process:
Figure 3: Number of Years to Process Benefit Determinations:
Figure 4: Number of Participants and Plans Terminated and Trusteed, by
Fiscal Year (2000-2008):
Figure 5: Average Processing Time, by Fiscal Year of Trusteeship:
Figure 6: Ten Plans with the Greatest Number of Benefit Determinations
Taking 4 Years or Longer:
Figure 7: Proportion of Participants with Estimated Benefits that
Differ from Final Benefits:
Figure 8: Ten Plans with the Greatest Number of Participants Owing PBGC
for Overpayments:
Figure 9: The Effect of Plan Termination on One RTI Retiree's Monthly
Payment:
Figure 10: The Effect of Plan Termination on One Weirton Retiree's
Monthly Payment:
Figure 11: Participants' Ages at the End of Recoupment in Cases with
Overpayments Greater than $10,000:
Figure 12: PBGC's Appeals Process:
Figure 13: Actions Taken on Correspondence Associated with Plans
Trusteed during Fiscal Years 2000-2008:
Figure 14: Average Age of Closed and Pending Appeals, by Fiscal Year:
Figure 15: Outcome of Appeals since Restructuring:
Figure 16: Example Scenario of Section 4044 Asset Allocation to
Participants' Benefits:
Figure 17: Example Scenario of Section 4022(c) Recovery Allocation to
Participants' Benefits:
Figure 18: Understanding the Limits on Guaranteed Benefits:
Abbreviations:
AFL-CIO: American Federation of Labor and Congress of Industrial
Organizations:
ERISA: Employee Retirement Income Security Act of 1974:
OPM: Office of Personnel Management:
PBGC: Pension Benefit Guaranty Corporation:
RTI: Republic Technologies International:
SPARR: Small Plan Average Recovery Ratio:
SSA: Social Security Administration:
SSI: Supplemental Security Income:
USWA: United Steelworkers of America:
[End of section]
United States Government Accountability Office:
Washington, DC 20548:
August 17, 2009:
Congressional Requesters:
The Pension Benefit Guaranty Corporation (PBGC) insures the retirement
income of nearly 44 million workers in over 29,000 private-sector
defined benefit plans. In the event that a company files for bankruptcy
and can no longer fund the benefits promised by its pension plan, the
plan may be terminated and PBGC may be required to step in and assume
responsibility for paying the pension benefits owed by that plan,
subject to certain legal limits. Determining the correct benefit
amounts following a plan's termination is a complex process that can
take years to complete. Concerned about potential reductions to
retirees' benefits, workers' retirement security in the face of
economic volatility, as well as PBGC's actions with respect to the
recent termination of the Republic Technologies International (RTI)
pension plans, you asked GAO to assess PBGC's procedures for
determining benefits following the termination of underfunded plans.
Specifically, this report addresses the following:
* the length of time it takes PBGC to make benefit determinations and
the causes for delays;
* the extent to which overpayments affect retirees' benefits;
* PBGC's communication with participants to keep them informed of
possible impacts on their benefits; and:
* the length of time it takes to obtain an appeals decision and the
accessibility of the appeals process.
To address these topics, we reviewed PBGC policies and procedures,
analyzed automated data, and interviewed PBGC officials knowledgeable
of various stages of the benefit determination process. We reviewed
several reports issued by PBGC's Office of Inspector General in the
late 1990s that provided a description of the benefit determination
process before 2000,[Footnote 1] and focused our study on plans
trusteed since then--that is, during fiscal years 2000 through
2008.[Footnote 2] PBGC provided electronic data on all the participants
in these plans and the status of their benefits as of February 2009. We
also examined plan and participant documents in PBGC's image processing
system, selecting for review those most affected by delays and
overpayments. For details on our methodology, see appendix I. Finally,
we spoke with personnel from employee associations and advocacy groups
who have been involved in some plan terminations, including the Pension
Rights Center, the American Federation of Labor and Congress of
Industrial Organizations (AFL-CIO), and the United Steelworkers, and
with participants and their representatives from six large terminated
plans: Reliance Group Holdings, Reliance Insurance Company, RTI-United
Steelworkers' of America (USWA), RTI-USS/KOBE, United Air Lines (ground
employees), and United Air Lines (pilots). For a complete list of
organizations and participants contacted, see appendix II.
We conducted this performance audit between October 2008 and August
2009, in accordance with generally accepted government auditing
standards. Those standards require that we plan and perform the audit
to obtain sufficient, appropriate evidence to provide a reasonable
basis for our findings and conclusions based on our audit objectives.
We believe the evidence obtained provides a reasonable basis for our
findings and conclusions based on our audit objectives.
Background:
PBGC was created as a government corporation by the Employee Retirement
Income Security Act of 1974 (ERISA)[Footnote 3] to help protect the
retirement income of U.S. workers with private-sector defined benefit
plans by guaranteeing their benefits up to certain legal limits. PBGC
administers two separate insurance programs for these plans: a single-
employer program and a multiemployer program. The single-employer
program covers about 34 million participants in about 28,000 plans. The
multiemployer program covers about 10 million participants in about
1,500 collectively-bargained plans that are maintained by two or more
unrelated employers. If a multiemployer pension plan is unable to pay
guaranteed benefits when due, PBGC will provide financial assistance to
the plan, usually a loan, so that retirees continue receiving their
benefit. However, if the sponsor of a single-employer plan is in
financial distress and does not have sufficient assets to pay promised
benefits, the plan will be terminated and PBGC will likely become the
plan's trustee, assuming responsibility for paying benefits to
participants as they become due, up to the guaranteed benefit limits.
As of the end of fiscal year 2008, PBGC had terminated and trusteed a
total of 3,860 single-employer plans (see figure 1). The single-
employer program is financed through premiums paid by the plan
sponsors, recoveries from the companies formerly responsible for the
plans, and investment income from the assets that PBGC acquires when it
assumes control of a plan. A three-member Board of Directors,
consisting of the Secretaries of the Commerce, Labor, and Treasury, is
charged with providing policy direction and oversight of PBGC's
finances and operations.
Figure 1: Growth in Number of Single-Employer Plans under PBGC
Trusteeship (1975-2008):
[Refer to PDF for image: vertical bar graph]
As of: 1979;
Cumulative number of plans: 586.
As of: 1984;
Cumulative number of plans: 1,207.
As of: 1989;
Cumulative number of plans: 1,744.
As of: 1994;
Cumulative number of plans: 2,437.
As of: 1999;
Cumulative number of plans: 2,877.
As of: 2004;
Cumulative number of plans: 3,545.
As of: 2008;
Cumulative number of plans: 3,860.
Source: GAO analysis of PBGC data, 1975-2008.
[End of figure]
PBGC's Pension Insurance Program Has Been Designated "High Risk":
We designated PBGC's single-employer pension insurance program as "high
risk" in 2003, including it on our list of major programs that need
urgent attention and transformation.[Footnote 4] The program remains a
high-risk concern due to an ongoing threat of losses from the
terminations of underfunded plans. Financially, PBGC's accumulated
deficit totaled $33.5 billion at the end of the second quarter of
fiscal year 2009, a $22.5 billion increase since the end of fiscal year
2008. Additionally, as we concluded in a recent report, PBGC's
governance structure and strategic management need improvement. We
found that PBGC's Board of Directors is limited in its ability to
provide policy direction and oversight, and recommended that the board
be expanded.[Footnote 5] Further, in two additional reports, we
concluded that PBGC lacks a strategic approach to its acquisition and
human capital management needs.[Footnote 6]
The Benefit Determination Process:
Under the single-employer program, if a company's pension plan has
inadequate assets to pay all promised benefits, plan sponsors meeting
certain criteria can voluntarily terminate a plan through a "distress"
termination.[Footnote 7] PBGC may also decide to terminate an
underfunded plan involuntarily to protect plan assets,[Footnote 8] and
PBGC must terminate a plan if assets are insufficient to pay benefits
currently due. In all these situations, PBGC generally becomes the
trustee of the plan and assumes responsibility for paying benefits to
the participants as they become due. Determining participants' benefit
amounts following termination, however, is a complex process (see
figure 2). It begins with gathering extensive data on plans and
individuals' work and personnel histories, and determining who is
eligible for benefits under a plan, which can be complicated if the
company or plan has a history of mergers, elaborate structure, or
missing data. It requires understanding plan provisions that vary from
plan to plan and can be numerous, applying the guarantee limitations to
each individual's benefit, and valuing plan assets and liabilities. If
the participant is already retired, or retires before the process is
complete, PBGC makes payments to the retiree based on an estimate of
the final benefit amount. Once the process is complete, PBGC notifies
each participant of the final benefit amount through a "benefit
determination letter." In cases with a final benefit that is greater
than the estimated amount, retirees are likely due a backpayment for
having been underpaid, which PBGC will repay in a lump sum, with
interest. In cases with a final benefit that is less, the retirees are
likely to have received an overpayment, which they then must repay to
PBGC, with no added interest.
Figure 2: PBGC's Benefit Determination Process:
[Refer to PDF for image: illustration]
Pretermination:
* Monitor underfunded plans;
* Work with plans that face distress terminations.
Initial Trusteeship (Termination):
* Obtain agreement on plan trusteeship;
* Notify participants and request information from retirees;
* Ensure that retirees receive benefit payments and that estimated
payments are reduced to reflect statutory limits.
Audit:
* Gather needed plan documents and participant data;
* Define plan population; build and audit participant database;
* Audit plan assets;
* Determine employer liability.
Benefit Valuation:
* Calculate individual benefits, in accordance with statutory and
regulatory requirements;
* Determine PBGC’s overall benefit liability.
Notification:
* Determine if estimated benefits being paid retirees are correct and
reconcile any differences:
- If benefits have been underpaid, PBGC provides a payment, with
interest;
- If benefits have been overpaid, PBGC takes steps to recoup the
overpaid funds (with no interest);
* Notify participants of the final benefit amount through a “benefit
determination letter;”
* Process participants’ appeals.
Postvaluation Administration (Ongoing):
* Process address changes and death notices;
* As nonretired participants enter retirement, calculate benefit based
on actual retirement and place participants in pay status;
* Respond to participants’ requests regarding their benefits.
Source: GAO analysis of PBGC documents.
[End of figure]
Guarantee Limits, Ceased Accruals, and Overpayments Can Reduce
Benefits:
When single-employer plans are terminated without sufficient assets to
pay all promised benefits, PBGC guarantees participants' benefits only
up to certain limits, specified under statute in ERISA and related
regulations. Participants whose benefits exceed these limits may have
their benefits reduced to the guaranteed amounts, unless the plan has
sufficient assets to pay the nonguaranteed portion of their benefits,
either all or in part.[Footnote 9] These guarantee limits are commonly
referred to as the maximum limit, the phase-in limit, and the accrued-
at-normal limit (see table 1).[Footnote 10] One group often affected by
the application of these limits is made up of those who take early
retirement. The maximum limit is lowered for each year a person retires
before age 65. Also, supplemental benefits--which are typically
provided to early retirees as a bridge to when they become eligible for
Social Security benefits--are eliminated or greatly reduced by the
accrued-at-normal limit. Because many steelworkers and airline pilots
retire before reaching age 65, retirees in these industries are hit
particularly hard by the application of such limits.
Table 1: Types of Guaranteed Benefit Limits under ERISA:
Type of limit: Maximum limit;
Description: The guaranteed benefit cannot exceed the statutory
maximum, adjusted annually, at the time the plan terminates. In 2009,
the maximum is $54,000 per year for a person retiring at age 65 and
with no survivor benefit (that is, a single-life annuity). The maximum
is lower for those retiring under age 65 or with a survivor benefit. 29
U.S.C. § 1322(b)(3); 29 C.F.R. § 4022.24 (2009).
Type of limit: Phase-in limit;
Description: The guaranteed benefit cannot include any benefit increase
implemented through a plan amendment that was made within 1 year of the
date of the plan termination. For benefit improvements that became
effective more than 1 year but less than 5 years prior to the plan's
termination, the guaranteed amount is the larger of 20 percent of the
benefit increase or $20 per month of the increase for each full year
the increase was in effect. 29 U.S.C. § 1322(b)(1) and (7); 29 C.F.R. §
4022.25 (2009).
Type of limit: Accrued-at-normal limit;
Description: The monthly guaranteed benefit cannot be greater than the
monthly benefit provided as a straight-life annuity (that is, a
periodic payment for the life of the retiree, with no additional
payments to survivors) available at the plan's normal retirement age.
The portion of any combined early retirement benefit and supplemental
benefit that exceeds the normal retirement age straight life annuity is
eliminated by this provision. 29 C.F.R. § 4022.21 (2009).
Source: ERISA, PBGC's implementing regulations, and related documents.
[End of table]
PBGC's benefits are set based on the amounts accrued as of the date of
plan termination.[Footnote 11] When a plan terminates, accruals cease.
[Footnote 12] As a result, participants who are not yet retired are
likely to receive lower benefits than what they would have received
under their plans if they had been able to accrue further benefits. For
example, if participants work for the plan sponsor beyond the
termination date, the additional service would not be credited under
that plan. The dollar amount or salary level used to calculate benefits
is also frozen at the level in effect as of the date of plan
termination, which can cause a participant's benefit to be
substantially less than it would have been if the plan had continued.
Participants can also be affected when a plan's termination date occurs
before they become eligible for certain benefits, such as early
retirement or disability benefits.
For retirees and participants who retire prior to completion of the
benefit determination process, estimated benefits are provided that can
sometimes be greater than the final benefit amount, causing an
overpayment. In addition to having benefits reduced due to the
guarantee limits, some retirees' have their monthly benefit reduced
once their benefit amount is finalized because they are required to
repay an overpayment that was incurred while receiving estimated
benefits.
Most participants of terminated plans receive the full amount of the
benefits they have earned under their plans, according to studies
conducted by PBGC. PBGC does not systematically track the number of
participants affected by guaranteed benefit limits or how much these
limits affect benefit amounts; however, PBGC has conducted two studies
on the impact of these limitations in a sample of large plans. The
first study, issued in 1999, found 5.5 percent of participants were
affected by the limits; and the second study, issued in 2008, found
that 15.9 percent were affected. PBGC attributed the increase in the
numbers affected in the second study to the inclusion of several large
plans from the steel and airlines industries. Officials noted that
these plans were more likely to be subject to the limits. Steel plans
often provide supplements and allow retirement with unreduced benefits
after 30 years of service, regardless of age, and airline plans often
allow pilots to retire early and receive generous benefits. Across the
different plans in both studies, participants' reductions in benefits
varied widely, from less than 5 percent for some, to over 50 percent
for others.
Table: PBGC Findings on the Impact of Guaranteed Benefit Limits:
Number of plans included in study:
1999 study: 22;
2008 study: 125.
Number of participants included in study:
1999 study: 90,448;
2008 study: 525,700.
Percentage receiving full plan benefits:
1999 study: 94.5%;
2008 study: 84.1%.
Percentage receiving reduced benefits:
1999 study: 5.5%;
2008 study: 15.9%.
Average amount of reduction (as percentage of participant’s benefit)
for those with a reduction:
1999 study: 16%;
2008 study: 28%.
Source: PBGC.
[End of table]
Most Benefit Determinations Are Completed within 3 Years, but Cases
with Complex Plans and Missing Data Take Longer:
PBGC makes most benefit determinations within 3 years after assuming
trusteeship of a plan. However, complex plans and plans with missing
data have required more time to process--up to 9 years, in some
instances (the full time span we examined). Most of the benefit
determinations that took 4 or more years to process were for
participants[Footnote 13] in just 10 plans. PBGC officials have taken
steps to shorten the benefit determination process, although their
initiatives have focused on ways to expedite processing of
straightforward cases instead of the processing of difficult cases
prone to delays.
PBGC Makes Most Benefit Determinations in Less than 3 Years:
PBGC becomes the trustee of most plans within 10 months of termination
and, once it has assumed trusteeship of a plan, the agency takes
slightly less than 3 years to process most benefit determinations and
notify participants of their final benefit amount. Following a PBGC
Inspector General study, issued in 2000, that found that the majority
of benefit determination letters were sent more than 5 years after PBGC
assumed trusteeship of the plan, PBGC set a corporate goal of issuing
benefit determinations, on average, no more than 3 years after
trusteeship.[Footnote 14] Our review of the benefit determinations for
participants in plans trusteed during fiscal years 2000 through 2008
indicates that PBGC has moved processing times closer to this mark.
Nearly three-quarters of the benefit determinations completed for these
plans were made in 3 years or less (see figure 3). The vast majority of
all completed benefit determinations--95 percent--was processed in less
than 4 years' time. On the other hand, in February 2009, more than
200,000 participants were awaiting benefit determinations that had been
pending for an average of 3 or more years.
Table: Length of Time from Trusteeship to Final Benefit Determination:
Determinations completed (824,718 total):
Maximum: 9.2 years;
Minimum: -0.5 years[A];
Mean: 2.5 years;
Median: 2.6 years.
Determinations pending, as of February 2009 (232,554 total):
Maximum: 9.3 years;
Minimum: 0.3 years;
Mean: 3.3 years;
Median: 3.6 years.
Source: GAO analysis of PBGC data for participants in plans trusteed
during fiscal years 2000 through 2008.
[A] In two atypical cases, PBGC made benefit determinations prior to
trusteeship.
[End of table]
Figure 3: Number of Years to Process Benefit Determinations:
[Refer to PDF for image: stacked vertical bar graph]
Number of years: Less than 1 year;
Percentage of benefit determinations completed: 5.5;
Percentage of benefit determinations pending, as of February 2009: 1.3;
Total: 6.8%.
Number of years: 1-2 years;
Percentage of benefit determinations completed: 17.7;
Percentage of benefit determinations pending, as of February 2009: 3.9;
Total: 21.6%.
Number of years: 2-3 years;
Percentage of benefit determinations completed: 34.1;
Percentage of benefit determinations pending, as of February 2009: 3.1;
Total: 37.2%.
Number of years: 3-4 years;
Percentage of benefit determinations completed: 16.5;
Percentage of benefit determinations pending, as of February 2009:
10.5;
Total: 27.0%.
Number of years: 4-5 years;
Percentage of benefit determinations completed: 3.6;
Percentage of benefit determinations pending, as of February 2009: 0.6;
Total: 4.2%.
Number of years: More than 5 years;
Percentage of benefit determinations completed: 0.6;
Percentage of benefit determinations pending, as of February 2009: 2.6;
Total: 3.2%.
Source: GAO analysis of PBGC data on length of time, from trusteeship
to final benefit determination, for participants of plans terminated
and trusteed during fiscal years 2000 through 2008 (1,057,272 total).
[End of figure]
PBGC practice is to prioritize benefit determinations based on an
individual's retirement status at the time of plan termination. For
example, participants who were retired when their plans terminated
received their benefit determinations in about 2.0 years after PBGC
assumed trusteeship, on average. Participants who had separated from
employment under the plan but had some vested benefits at the time of
the termination received benefit determinations in about 2.8 years, on
average. All other participants received benefit determinations in
about 2.8 years, on average.
Workload Affects Processing Times:
Processing times have varied considerably in any given year, due in
part to the number and size of plans being terminated and trusteed that
year (see figure 4). The number of plans trusteed by PBGC peaked during
2002, 2003, and 2004, although the largest influx of participants
occurred in 2005. The average number of participants per plan is
slightly fewer than 1,000, but some plans have many more. For example,
the Bethlehem Steel plan has nearly 93,000 participants, the LTV Steel
(hourly) plan has about 68,000 participants, and the Kaiser Aluminum
and Chemical Corp. (hourly) plan has just over 10,000 participants.
Figure 4: Number of Participants and Plans Terminated and Trusteed, by
Fiscal Year (2000-2008):
[Refer to PDF for image: vertical bar graph]
Fiscal year: 2000;
Number of Plans: 99;
Number of Participants: 30,091.
Fiscal year: 2001;
Number of Plans: 103;
Number of Participants: 87,864.
Fiscal year: 2002;
Number of Plans: 144;
Number of Participants: 189,183.
Fiscal year: 2003;
Number of Plans: 152;
Number of Participants: 198,739.
Fiscal year: 2004;
Number of Plans: 178;
Number of Participants: 149,248.
Fiscal year: 2005;
Number of Plans: 138;
Number of Participants: 273,586.
Fiscal year: 2006;
Number of Plans: 86;
Number of Participants: 45,759.
Fiscal year: 2007;
Number of Plans: 115;
Number of Participants: 61,913.
Fiscal year: 2008;
Number of Plans: 74;
Number of Participants: 20,889.
Source: GAO analysis of PBGC data on participants in plans terminated
and trusteed during fiscal years 2000 through 2008(1,057,272
participants and 1,089 plans total).
[End of figure]
We found that processing times were longer, on average, for those plans
trusteed in peak years (see figure 5). For example, processing times
generally increased during fiscal years 2002 through 2005.
Figure 5: Average Processing Time, by Fiscal Year of Trusteeship:
[Refer to PDF for image: vertical bar graph]
Fiscal year: 2000;
Average age of case: 2.5 years;
Determinations completed: 29,369;
Determinations pending, as of February 2009: 722.
Fiscal year: 2001;
Average age of case: 2.5 years;
Determinations completed: 86,862;
Determinations pending, as of February 2009: 1,002.
Fiscal year: 2002;
Average age of case: 2.5 years;
Determinations completed: 182,917;
Determinations pending, as of February 2009: 6,266.
Fiscal year: 2003;
Average age of case: 2.7 years;
Determinations completed: 181,255;
Determinations pending, as of February 2009: 17,484.
Fiscal year: 2004;
Average age of case: 2.8 years;
Determinations completed: 141,486;
Determinations pending, as of February 2009: 7,762.
Fiscal year: 2005;
Average age of case: 3.0 years;
Determinations completed: 178,911;
Determinations pending, as of February 2009: 94,675.
Fiscal year: 2006;
Average age of case: 2.7 years;
Determinations completed: 15,652;
Determinations pending, as of February 2009: 30,107.
Fiscal year: 2007;
Average age of case: 1.8 years;
Determinations completed: 7,999;
Determinations pending, as of February 2009: 53,914.
Fiscal year: 2008;
Average age of case: 0.9 years;
Determinations completed: 267;
Determinations pending, as of February 2009: 20,622.
Source: GAO analysis of PBGC data on length of time, from trusteeship
to final benefit determination, for participants of plans terminated
and trusteed during fiscal years 2000 through 2008 (1,057,272 total).
[End of figure]
Plan Complexity and Missing Data Also Adds to Processing Times:
Processing times have also increased with the complexity of plans and
the unavailability of needed data. Obtaining plan documents, gaining
complete participant data, and interpreting plan requirements often
present difficulties. Nevertheless, nearly three-quarters of the
benefit determinations that took 4 or more years to process were for
participants in just 10 of the 1,089 plans terminated and trusteed
during fiscal years 2000 through 2008, as shown in figure 6. These
plans were sponsored by four steel companies, two mining companies, one
other manufacturer, an insurance company, and a construction company.
Figure 6: Ten Plans with the Greatest Number of Benefit Determinations
Taking 4 Years or Longer:
[Refer to PDF for image: horizontal bar graph]
Percentage of all participants with benefit determinations taking 4 or
more years, as of February 2009:
Top 10 plans:
Bethlehem Steel: 32.61%;
LTV Steel Co. (hourly plan): 5.96%;
Horizon NR LLC: 5.55%;
Kaiser Aluminum and Chemical Corp. (hourly plan): 5.35%;
RTI (USWA plan): 4.86%;
J.A. Jones, Inc.: 4.3%;
Kaiser Aluminum and Chemical Corp. (salaried plan): 3.87%;
Cone Mills Corporation: 3.76%;
Weirton Steel: 3.49%;
Reliance Insurance Company: 3.01%;
Total, top 10 plans: 72.8%.
All other plans: 27.24%.
Source: GAO analysis of PBGC data on participants of plans terminated
and trusteed during fiscal years 2000 through 2008.(Cases taking 4 or
more years: 78,553 total participants from 561 plans.)
[End of figure]
Plan Complexity:
We found that a variety of factors had contributed to the complexity of
the 10 plans with these lengthier determinations. One key factor was
the level of difficulty of calculating benefits. For some, a history of
company or plan mergers, or other unusual or complicated benefit
formulas, made determining a participant's benefit more difficult and
added to processing time. For example, the pension plan of Bethlehem
Steel Corporation--which still had some benefit determinations pending
as of February 2009, nearly 6 years after the plan's trusteeship--is a
product of more than 100 company mergers, consolidations, and/or
spinoffs. There are eight major parts to this plan, and three of the
parts have separate hourly and salaried plans. In general, if a plan
has undergone a merger, participants may be covered by different plan
provisions, or participants may transfer between component plans, such
as moving from an hourly to salaried plan. According to PBGC, the
Bethlehem Steel plan required an analysis of more than 30 sets of plan
documents to make benefit determinations for the nearly 93,000
participants.
Unusual or numerous plan provisions have also made benefit
determinations more challenging and, therefore, time consuming. The
Cone Mills Corporation plan consists of three merged plans. In 2001,
the company's plans for long-distance drivers and salaried workers were
merged into its plan for hourly workers. Yet, distinct provisions in
each of the original plans remained in place for their respective
members. It required time for PBGC to understand which participants
belonged to each group and the provisions associated with each
participant.
In other cases, an elaborate plan structure has also made it
challenging for PBGC to determine the availability of plan assets and
to distribute them across different categories of participants'
benefits in the asset allocation process.[Footnote 15] The Kaiser
Aluminum and Chemical Corp. had 26 direct and indirect subsidiaries in
its controlled group and in bankruptcy; 36 subsidiaries not in
bankruptcy; and 13 operating subsidiaries and joint ventures not in the
controlled group or in bankruptcy.[Footnote 16] Kaiser had eight
defined benefit plans, seven of which were trusteed by PBGC, and the
assets for these eight plans were commingled, which added complexity to
PBGC's audit of the plans' net worth.
Benefit guarantee limits contributed to the complexity of several
plans. PBGC must determine, on a participant-by-participant basis, the
level of benefits each is entitled to under ERISA and related
regulations.[Footnote 17] According to PBGC officials, these
calculations can be time consuming when there are a large number of
participants receiving benefit adjustments as a result of these limits.
For example, there were several benefit rate increases in the LTV Steel
(hourly) plan that went into effect within 5 years of the plan's
termination and, therefore, were subject to the phase-in limit. These
included a plant shutdown supplement for certain participants, a
surviving spouse's special payment, and additional continuous service
for participants affected by certain layoffs.[Footnote 18] In total,
there were 35,279 participants whose benefits were affected by the
phase-in limitation under this plan, as well as 4,850 affected by the
accrued-at-normal limit, and 3,644 affected by the maximum limit.
[Footnote 19]
Qualified domestic relations orders have also contributed to the
complexity of making a benefit determination. When participants have
domestic relations orders related to child support, alimony payments,
and marital property rights, some portion of, or all of, a
participant's pension benefits may be assigned to a spouse, former
spouse, child, or other dependent. In these cases, PBGC must determine
whether the order is a qualified domestic relations order, a process
which can entail a detailed review of legal documents. Although nearly
two-thirds of the plans we examined did not have any participants with
qualified domestic relations orders, several of the ten plans
associated with the lengthiest processing times had numerous
participants with such orders. For example, the Bethlehem Steel plan
included 904 participants with qualified domestic relations orders, and
the LTV Steel (hourly) plan included 609.
Missing Data:
The condition of plan and participant data is also a key factor
affecting processing times. When a plan terminates, PBGC tries to
obtain all plan documents, such as the original plan, plan amendments,
and, if applicable, negotiated agreements with unions, as well as
personnel and payroll data. To do so with the termination of a large,
complex plan, PBGC auditors have usually visited sponsor locations to
collect data and contacted the plan's actuarial staff, administrators,
or others responsible for managing the plan's assets. When the plan's
administration is decentralized, this process involves collecting
records from different locations in the course of many site visits. For
example, over a 2-month period, a PBGC audit team visited Bethlehem
Steel facilities in Sparrows Point, MD; Bethlehem, PA; Coatesville, PA;
Steelton, PA; Lackawanna, NY; and Burns Harbor, IN to collect records.
Data were not always available in electronic form. The Bethlehem Steel
Lackawanna facility, for example, did not use an electronic
recordkeeping system, so PBGC collected more than 20,000 hard-copy
employee record cards from the site.
According to PBGC officials, plan sponsors have frequently diverted
resources away from actuarial and information technology services
during rough financial periods, causing records maintenance to
deteriorate before PBGC is able to take over the plan. In such
situations, data become difficult to locate, key personnel with
knowledge of the data leave the organization, and data systems may be
inaccessible. Additionally, the data PBGC is able to collect has often
been incomplete. As a result, PBGC actuaries sometimes have to make
assumptions about which plan provisions apply to whom when estimating
the plan's assets and liabilities, and calculating individual
participants' benefits. When processing the Weirton Steel plan, for
example, PBGC was required to calculate benefits for some participants
whose average monthly earnings were missing. A PBGC official told us
that they sometimes use collective bargaining agreements and board
resolutions, even if their legality cannot be verified, if those
documents provide the best information available.
To avoid situations where data are missing or in poor condition, PBGC
officials told us they generally try to obtain data prior to taking
over a plan. In most situations, they will quickly try to assess the
location and condition of plan records, and take steps to preserve the
records in the event that PBGC takes over the plan. However, officials
acknowledged that negotiations between PBGC and plan sponsors prior to
trusteeship have sometimes deterred them from using their access
authority[Footnote 20] to secure records until after actually becoming
the trustee. For example, the RTI case involved a lengthy legal
deliberation over the plan's termination date, and while this
litigation was ongoing prior to trusteeship, PBGC's case processing
division did not pursue documents from RTI prior to trusteeship, on the
advice of the agency's and company's counsel at the time. PBGC
officials noted that when aspects of termination are being contested,
it is not uncommon for company officials to be unwilling to share
information until after PBGC's trusteeship is official. In the RTI
case, by the time the court case was resolved and PBGC became the
trustee, a new owner had assumed control of the personnel files,
documentation needed to determine benefit entitlement had been purged,
and only one person remained with working knowledge of the RTI pension
plan.[Footnote 21]
New Initiatives to Shorten the Benefit Determination Process Do Not
Address Longest Delays:
PBGC officials have taken steps to shorten the benefit determination
process, although these initiatives do not specifically address complex
cases. Rather, PBGC officials said that their initiatives are intended
to process straightforward cases more quickly so that staff can
concentrate on those that are difficult. Specifically, PBGC adopted a
simplified data validation process to speed the processing of plans
with fewer than 200 participants. They decided that the validation
process used for large plans, which involves a full electronic data
audit and a review of all data elements by an auditor, was unnecessary
for smaller plans, which have fewer participants and less data, making
any errors highly visible. PBGC has also prioritized benefit
determinations for retirees who have been receiving benefits for some
time. Such determinations are more straightforward because these
retirees are less likely to have their benefits reduced by the
guarantee limits. These efforts help PBGC to avoid unnecessary delays
in straightforward cases. PBGC does not, however, target its changes on
complex plans with benefit determinations most prone to lengthy delays.
Nor does PBGC set benchmarks for complex cases or goals for decreasing
the processing time for these cases. Officials acknowledged that the
current tracking of timeliness focuses on average processing times
only.
Overpayments Have Been Infrequent and Mostly Concentrated in a Few
Complex Plans:
Overpayments have been infrequent and the impact on benefit amounts has
been generally minor. As with the cases that required lengthy
processing times, most of the cases in which overpayment occurred have
been concentrated in a small number of plans. These tended to be large
plans with large numbers of retirees, as well as plans whose total
asset values were difficult to determine or anticipate. Meanwhile,
PBGC's requirement for repayment of overpayments is highly amortized,
thereby limiting the amount of money that PBGC will recoup. By
comparison, some other federal agencies have more aggressive repayment
policies, but more liberal waiver policies for cases of hardship.
Overpayments Are Infrequent and Generally Have Limited Impact on
Benefits:
Overpayments generally occur when a plan retiree receives estimated
benefits while PBGC is in the process of making benefit determinations
and the final benefit amount is less than the estimated benefit amount.
Our review of plans terminated and trusteed during fiscal years 2000
through 2008 found that this happened only in a small percentage of
cases (see figure 7). Of the 1.1 million participants in plans
terminated and trusteed during fiscal years 2000 through 2008, more
than half were not yet retirees and, therefore, did not receive
estimated benefits before the benefit determination process was
complete. For most who were retirees, the estimated benefit amount
received did not change when finalized. Of those whose benefit amount
did change when finalized, about half received a benefit that was
greater and half received a benefit that was less (about 3 percent of
total participants in these plans, overall).
Figure 7: Proportion of Participants with Estimated Benefits that
Differ from Final Benefits:
[Refer to PDF for image: pie-chart]
No change: 28.4%;
Final benefit greater: 3.1%;
Final benefit less: 3.4%;
No estimated benefit: 54.9%;
Final benefit pending: 10.2%.
Source: GAO analysis of PBGC data on participants of plans terminated
and trusteed during fiscal years 2000 through 2008(1,057,272 cases
total).
[End of figure]
According to PBGC data on recoupments, 22,623 participants in plans
terminated and trusteed during fiscal years 2000 through 2008 owed PBGC
for overpayments. These amounts varied widely--from less than $1 to
more than $150,000--but our analysis of PBGC data suggests that most
owed less than $3,000.[Footnote 22] Since in most cases PBGC recoups no
more than 10 percent of a participant's final benefit each
month,[Footnote 23] the impact on the participant's benefit was
limited. Per individual, the median benefit reduction due to recoupment
was about $16 a month, or about 3 percent of the monthly payment
amount, on average. Per case, the median amount that had been repaid,
as of February 2009, was $365.
Table: Recoupment Cases:
Number of participants with overpayments: 22,623.
Estimated total amount of overpayments: $100,000,000;
Maximum: over $150,000;
Minimum: under $1;
Mean: about $4,400;
Median: about $2,500.
Estimated total amount recouped, as of February 2009: $13,000,000;
Monthly benefit reductions due to recoupments:
Maximum: $930;
Minimum: under $1;
Mean: about $24;
Median: about $16.
Monthly benefit reductions as percentage of payment:
Maximum: 50%;
Minimum: less than 1%;
Mean: about 3%;
Median: about 2%.
Source: GAO analysis of PBGC data on participants in plans terminated
and trusteed during fiscal years 2000 through 2008.
[End of table]
A Few Plans Accounted for Most Cases with Overpayments:
Of the 1,089 plans terminated and trusteed during fiscal years 2000
through 2008, just 10 accounted for more than 65 percent all cases of
overpayment (see figure 8).[Footnote 24] Nine of these 10 plans were
sponsored by steel companies trusteed by PBGC from 2001 to 2003. When
PBGC assumes responsibility for a plan, retirees generally continue to
receive an estimated benefit that is the same as what they had been
receiving, unless PBGC determines they are subject to any of the
guarantee limits, and that their estimated payments need to be reduced
to reflect these limits. In such cases, overpayments can occur for two
basic reasons: (1) there is a period of time when the retiree's
estimated benefit has not yet been reduced to reflect applicable
limits; and (2) the retiree's estimated benefit is adjusted to reflect
applicable limits, but the estimate is still greater than the benefit
amount that is ultimately determined to be correct once the benefit
determination process is complete.
Figure 8: Ten Plans with the Greatest Number of Participants Owing PBGC
for Overpayments:
[Refer to PDF for image: horizontal bar graph]
Percentage of all participants unidentified as owing PBGC for
overpayment, as of February 2009:
Top 10 plans:
LTV Steel Co. (hourly plan): 19.63%;
Bethlehem Steel: 10.99%;
Weirton Steel: 8.83%;
RTI (USWA plan): 6.67%;
National Steel (hourly plan): 4.75%;
LTV Steel Co. (salary plan): 4.04%;
RTI (USS/KOBE plan): 3.23%;
Outboard Marine (employees plan): 2.71%;
LTV Steel Mining Co.: 2.42%;
Northwestern Steel and Wire (plan A): 1.99%;
Total: 65.3%.
All other plans: 34.7%.
Source: GAO analysis of PBGC data on participants of plans terminated
and trusteed during fiscal years 2000 through 2008.(Cases with
overpayments: 22,623 total participants from 467 plans.)
[End of figure]
As summarized in table 2, of the 10 plans with the greatest number of
overpayments, 9 also had large numbers of participants, including many
who were subject to the guarantee limits and who were retired and
receiving estimated benefits.[Footnote 25] In addition, all these plans
had assets or recoveries allocated to pay some, but not all, of
retirees' nonguaranteed benefits, which are generally some of the first
nonguaranteed benefits to be paid from the allocation process--before,
for example, future retirees' nonguaranteed benefits.[Footnote 26]
According to PBGC officials, uncertainty about how much a plan's assets
or recoveries will be able to contribute toward a retiree's benefit
that the agency does not guarantee, under law, can make it difficult to
calculate an accurate benefit amount until the benefit determination
process is complete.
Table 2: Characteristics of 10 Large, Complex Plans:
Plan sponsor (name)[B]: LTV Steel Co. (hourly plan);
Total number of participants: 68,124;
Number subject to guarantee limits[A]: Accrued-at-normal limit: 4,850;
Number subject to guarantee limits[A]: Maximum limit: 3,644;
Number subject to guarantee limits[A]: Phase-in limit: 35,279;
Number receiving estimated benefits: 46,007;
Sufficient assets or recoveries to pay some nonguaranteed benefits[C]:
Yes--retirees' nonguaranteed benefits partially paid;
Total number of participants with overpayments: 4,442.
Plan sponsor (name)[B]: Weirton Steel;
Total number of participants: 9,757;
Number subject to guarantee limits[A]: Accrued-at-normal limit: 1,342;
Number subject to guarantee limits[A]: Maximum limit: 2,482;
Number subject to guarantee limits[A]: Phase-in limit: 672 [D];
Number receiving estimated benefits: 6,915;
Sufficient assets or recoveries to pay some nonguaranteed benefits[C]:
Yes--retirees' nonguaranteed benefits partially paid;
Total number of participants with overpayments: 1,997.
Plan sponsor (name)[B]: RTI (USWA plan);
Total number of participants: 4,289;
Number subject to guarantee limits[A]: Accrued-at-normal limit: 941;
Number subject to guarantee limits[A]: Maximum limit: 11;
Number subject to guarantee limits[A]: Phase-in limit: 1,693;
Number receiving estimated benefits: 2,257;
Sufficient assets or recoveries to pay some nonguaranteed benefits[C]:
Yes--retirees' nonguaranteed benefits partially paid;
Total number of participants with overpayments: 1,508.
Plan sponsor (name)[B]: National Steel (hourly plan);
Total number of participants: 10,433;
Number subject to guarantee limits[A]: Accrued-at-normal limit: 40;
Number subject to guarantee limits[A]: Maximum limit: 1,113;
Number subject to guarantee limits[A]: Phase-in limit: [E];
Number receiving estimated benefits: 6,272;
Sufficient assets or recoveries to pay some nonguaranteed benefits[C]:
Yes--retirees' nonguaranteed benefits partially paid;
Total number of participants with overpayments: 1,075.
Plan sponsor (name)[B]: LTV Steel Company Inc. (salary plan);
Total number of participants: 13,450;
Number subject to guarantee limits[A]: Accrued-at-normal limit: 45;
Number subject to guarantee limits[A]: Maximum limit: 7;
Number subject to guarantee limits[A]: Phase-in limit: 1,401;
Number receiving estimated benefits: 9,606;
Sufficient assets or recoveries to pay some nonguaranteed benefits[C]:
Yes--retirees' nonguaranteed benefits partially paid;
Total number of participants with overpayments: 913.
Plan sponsor (name)[B]: RTI (USS/KOBE plan);
Total number of participants: 2,299;
Number subject to guarantee limits[A]: Accrued-at-normal limit: 874;
Number subject to guarantee limits[A]: Maximum limit: 51;
Number subject to guarantee limits[A]: Phase-in limit: 7;
Number receiving estimated benefits: 1,356;
Sufficient assets or recoveries to pay some nonguaranteed benefits[C]:
Yes--retirees' nonguaranteed benefits partially paid;
Total number of participants with overpayments: 730.
Plan sponsor (name)[B]: Outboard Marine (employees plan);
Total number of participants: 9,744;
Number subject to guarantee limits[A]: Accrued-at-normal limit: 780;
Number subject to guarantee limits[A]: Maximum limit: 159;
Number subject to guarantee limits[A]: Phase-in limit: [E];
Number receiving estimated benefits: 4,797;
Sufficient assets or recoveries to pay some nonguaranteed benefits[C]:
Yes--retirees' nonguaranteed benefits partially paid;
Total number of participants with overpayments: 614.
Plan sponsor (name)[B]: LTV Steel Mining Co.;
Total number of participants: 3,416;
Number subject to guarantee limits[A]: Accrued-at-normal limit: 643;
Number subject to guarantee limits[A]: Maximum limit: 381;
Number subject to guarantee limits[A]: Phase-in limit: 1,099;
Number receiving estimated benefits: 2,383;
Sufficient assets or recoveries to pay some nonguaranteed benefits[C]:
Yes--retirees' nonguaranteed benefits partially paid;
Total number of participants with overpayments: 548.
Plan sponsor (name)[B]: Northwestern Steel & Wire (plan A);
Total number of participants: 3,576;
Number subject to guarantee limits[A]: Accrued-at-normal limit: 1,023;
Number subject to guarantee limits[A]: Maximum limit: 533;
Number subject to guarantee limits[A]: Phase-in limit: 0;
Number receiving estimated benefits: 2,812;
Sufficient assets or recoveries to pay some nonguaranteed benefits[C]:
Yes--retirees' nonguaranteed benefits partially paid;
Total number of participants with overpayments: 450.
Plan sponsor (name)[B]: US Airways Inc. (pilots' plan);
Total number of participants: 7,050;
Number subject to guarantee limits[A]: Accrued-at-normal limit: 51;
Number subject to guarantee limits[A]: Maximum limit: 5,171;
Number subject to guarantee limits[A]: Phase-in limit: 175;
Number receiving estimated benefits: 1,501;
Sufficient assets or recoveries to pay some nonguaranteed benefits[C]:
Yes--retirees' nonguaranteed benefits partially paid;
Total number of participants with overpayments: 111.
Source: GAO analysis of PBGC data.
[A] Participants may be subject to more than one type of limitation, so
the three columns cannot be added together to determine a total.
[B] Bethlehem Steel was not included in this table because valuation of
the plan was not complete and data on numbers subject to the guarantee
limits were not available as of February 2009.
[C] For further details about the allocation process, see appendix III.
[D] This represents the number of participants subject to the phase-in
limitation who were in pay status as of the date of plan termination.
No data were readily available on the number of participants subject to
this limitation who retired (or will retire) after the date of plan
termination.
[E] Various changes in benefits were described as subject to the phase
in limit in these plans (9 for National Steel and 17 for Outboard
Marine), but no listings were provided to indicate the number of
participants affected.
[End of table]
Finally, a lengthy benefit determination process can exacerbate the
impact of inaccurate estimates. The total overpayment can become
substantial over a long period of time, even if the difference between
the estimated and final monthly benefit amount is small. Also, when
plans are terminated involuntarily, there can sometimes be lengthy
delays before PBGC reduces estimated benefits to reflect guarantee
limits. Among the 10 plans with the most overpayments, all were
involuntary terminations, and we found that the length of time between
plan termination and when estimated benefits were adjusted to reflect
guarantee limits varied widely. In some cases, estimated benefits were
adjusted within 9 months of termination, while in other cases, more
than 6 years elapsed before estimated benefits were adjusted--and in
general, the longer the delays, the larger the overpayments. In
contrast, when plans are terminated at the sponsor's request as
distress terminations, the sponsors are required to impose these limits
themselves so that participants' benefits are reduced as of the date of
termination.[Footnote 27]
The following examples illustrate how the above circumstances can
combine to create large numbers of cases with overpayments among some
plans. We chose these two case examples from among the cases sampled in
the 10 plans with the most overpayments to illustrate the two types of
situations that can result in overpayments outlined previously: (1)
delayed adjustment of the retiree's estimated benefit to reflect
applicable limits; and (2) timely, but inaccurate adjustment of the
retiree's estimate to reflect applicable limits. We also chose these
two case examples specifically because they had similar benefit amounts
prior to termination.
In the RTI (USWA) plan, four large groups of participants were affected
by the guarantee limits: (1) those with six different types of
temporary supplements who were subject to the accrued-at-normal limit;
[Footnote 28] (2) former Bar Technologies employees whose benefits were
subject to a $20 or 20 percent phase-in limit;[Footnote 29] (3) those
who retired or will retire with 30 years of service and were subject to
a $60 or 60 percent phase-in; and (4) those who retired under the early
retirement program whose benefits were subject to a $60 or 60 percent
phase-in. To explore the impact of guarantee limits on the retirees who
incurred overpayments, we randomly selected 5 participants from among
the 1,693 subject to the phase-in limits, and found that all were
retirees who had their benefits reduced between 19 percent and 63
percent from what they had been receiving prior to termination. In
three cases, estimated benefits were adjusted to reflect these limits
2.3 years after termination, but in two cases, estimated benefits were
not adjusted prior to issuance of the benefit determination letter,
which took place more than 6 years after termination. Due to inaccurate
estimated benefits that were paid over several years, all 5 had
incurred overpayments, ranging from $2,000 to about $57,000, and as a
result, their benefits were reduced further to recoup the amounts owed.
The effect on the monthly payment for one RTI retiree, whom PBGC
overpaid by a total of $23,986, is illustrated in figure 9. Ultimately,
this retiree's payment was reduced by almost two-thirds, mostly due to
guarantee limits.
Figure 9: The Effect of Plan Termination on One RTI Retiree's Monthly
Payment:
[Refer to PDF for image: stacked vertical bar graph]
6/14/02: Plan termination;
9/30/03: Date of trusteeship.
Final benefit determination amount: $562.
2002 estimated monthly benefit:
Payment amount: $1,504;
Portion of payment amount that is an overpayment: $942.
2003 estimated monthly benefits:
Payment amount: $1,504;
Portion of payment amount that is an overpayment: $942.
10/1/04 estimated benefit reduced due to limits:
Payment amount: $402[A].
7/1/05 estimated benefit increased to correct an error:
Payment amount: $506.
2006 estimated monthly benefit:
Payment amount: $506.
2007 estimated monthly benefit:
Payment amount: $506.
8/1/08 final monthly benefit:
Payment amount: $506;
Payment reduction to recoup overpayment: $56.
Source: GAO analysis of PBGC data.
[A] During this time, participant was underpaid. Amount of underpayment
was factored in when the total net overpayment amount was determined.
[End of figure]
In the Weirton plan, we found that large numbers of participants were
subject to the accrued-at-normal limits due to various plan
supplements; and were subject to the phase-in limits due to seven
different types of benefit changes made within 5 years before plan
termination. In addition, many participants were subject to the maximum
limits, in part due to the aggregate limit imposed when participants
are involved in more than one terminated plan[Footnote 30] (many
participants had worked previously for National Steel or other PBGC-
trusteed plans). We reviewed five randomly-selected cases from among
the 1,342 participants who were subject to the accrued-at-normal limit
and found that all were retirees whose estimated benefit amounts were
inaccurate for at least part of the period involving the benefit
determination process. One case resulted in an underpayment, with a
backpayment of $11,384 to be repaid to the retiree, plus interest. The
other four cases resulted in overpayments, ranging from $3,200 to just
over $6,000, with reductions in benefit payments to recoup the amounts
overpaid. In contrast with the five sampled RTI participants, these
retirees had their benefits adjusted more quickly to reflect the
guarantee limits so that, in general, the overpayments incurred were
not as large. All 4 had their estimated benefits adjusted in less than
9 months. The effect on one Weirton retiree's monthly payment is
illustrated in figure 10. As was the case in the previous example, this
retiree's payment was ultimately reduced by nearly one-half, mostly due
to guarantee limits.
Figure 10: The Effect of Plan Termination on One Weirton Retiree's
Monthly Payment:
[Refer to PDF for image: stacked vertical bar graph]
10/21/03: Plan termination;
11/12/03: Date of trusteeship.
Final benefit determination amount: $821.
2003 estimated monthly benefit:
Payment amount: $1,521;
Portion of payment amount that is an overpayment: $700.
7/1/04 estimated benefit reduced due to limits:
Payment amount: $821.
2005 estimated monthly benefit:
Payment amount: $821.
2006 estimated monthly benefit:
Payment amount: $821.
12/17/07 final monthly benefit:
Payment amount: $821.
3/13/08 benefit adjusted due to overpayment:
Payment amount: $783;
Payment reduction to recoup overpayment: $38.
Source: GAO analysis of PBGC data.
[End of figure]
Large Overpayments Are Not Fully Recouped:
Our analysis of PBGC data indicates that the overpayments owed by
participants in plans terminated and trusteed during fiscal years 2000
through 2008 totaled almost $100 million.[Footnote 31] Of this total,
about $14 million had been recouped, as of February 2009. However,
PBGC's policy of restricting recoupments to no more than 10 percent of
the recipient's monthly benefit results in a long amortization period
for collection that can well exceed normal life expectancies. Since
PBGC does not pursue further collection from a participant's estate
once a retiree (and any beneficiary) dies, a substantial portion of
these overpayments will not be repaid. Specifically, for many of these
individuals, it was projected that these debts would not be fully paid
until the year 2099, PBGC's arbitrary cutoff. Nearly 60 percent of
those with future recoupments would not finish repaying these debts
until the year 2020 and beyond. We analyzed the ages of retirees and/or
beneficiaries at their projected end date of recoupment for all cases
involving overpayments greater than $10,000. Although these cases
accounted for fewer than 10 percent of those with overpayments, the
amounts they owed accounted for more than 40 percent of total
recoupments. We found that about 60 percent of these individuals would
be age 80 or older, and over 30 percent would be age 100 or older, when
their debts to PBGC would be fully repaid (see figure 11). The life
expectancy for those age 65 in 2009 is estimated to be 82 to 87 years.
[Footnote 32]
Figure 11: Participants' Ages at the End of Recoupment in Cases with
Overpayments Greater than $10,000:
[Refer to PDF for image: pie-chart]
Under age 80: 39.7%;
Age 80 to 99: 29.2%;
Age 100 and over: 31.1%.
Source: GAO analysis of PBGC data on participants in plans terminated
and trusteed during fiscal years 2000 through 2008 with recoupments
greater than $10,000 (2,035 cases total).
Note: In 173 of these cases, the end date was the arbitrary cutoff of
2099, so ages at the actual end of recoupment would be greater.
[End of figure]
Some Federal Agencies Reduce Payments More Aggressively, but Also Grant
More Waivers for Overpayments:
Once overpayments have been made, finding the right balance between
agency fiscal responsibility and fairness to participants can be
difficult to achieve. Compared with PBGC's policy on overpayments,
[Footnote 33] federal agencies such as the Social Security
Administration (SSA) and the Office of Personnel Management (OPM)
generally allow larger reductions to benefits when recouping
overpayments, but their policies also give much greater prominence to
waivers. PBGC policy stipulates that in cases with an ongoing payment,
recoupment of an overpayment may not be waived unless the monthly
reduction would be less than $5.[Footnote 34] Waivers for hardship are
to be considered only in cases for which there is no ongoing payment to
the participant. According to the agency's general counsel and
subsequent comments from agency officials, since the outset of 2009,
PBGC has been receiving hardship waiver requests in recovery cases at
more than twice the rate received the prior year.
In contrast, both SSA and OPM policies on overpayments allow hardship
consideration for cases with ongoing payments. For overpayment of
Social Security benefits, SSA will withhold the full amount of the
benefit each month until the overpayment is fully recouped. However, in
its fact sheet on overpayments with respect to Social Security benefits
and Supplemental Security Income (SSI) benefits, available on its Web
site, SSA devotes over half the document to detailing the steps
participants should take if they wish to either appeal or request a
waiver. For SSI benefits, SSA will withhold 10 percent of the maximum
federal benefit rate each month,[Footnote 35] but the beneficiary can
request a lesser withholding amount, subject to SSA approval. Further,
if the beneficiary disagrees with the overpayment, he or she can appeal
or request that collection be waived.[Footnote 36]
Similarly, OPM's policy guidance on overpayments of retirement benefits
devotes over half the document to the subject of waivers. Under law,
OPM is directed not to recover overpayments when the beneficiary bears
no responsibility for the overpayment and requiring repayment would be
"against equity and good conscience."[Footnote 37] In deciding whether
to grant a waiver, errors or delays by OPM may be considered, along
with financial hardship or any other basis for equity that OPM deems
appropriate. Just the last 7 pages of this 34-page policy guide are
devoted to policies on collections. These policies call for
overpayments of federal employee retirement benefits to be collected in
one lump sum, whenever feasible. If one lump-sum payment is not
feasible and recoupment is by installment, the payments are to be
sufficient in size and frequency to recoup the debt in no more than 3
years. The standard rate of collection is 10 percent of the net monthly
annuity or $50 per month, whichever is higher; but if a 10 percent
reduction will not result in full recoupment within 3 years, the
reduction rate can be increased up to 50 percent.[Footnote 38]
PBGC's Initial Communications Drew Praise, but Later Communications
Were Sometimes Found Lacking:
PBGC's initial communications with participants shortly following
termination--especially its on-site information sessions--generally
drew praise from the pension advocacy groups and union representatives
we interviewed. These groups' concerns with PBGC's communication
efforts most often focused on the long gaps between contacts when the
benefit determination process was lengthy and the complicated
calculations that accompanied letters notifying participants of
significant benefit reductions.
PBGC's Initial Notification and On-Site Information Sessions Generally
Drew Praise:
PBGC's first communication with participants is generally a letter
informing them that their pension plan has been terminated and that
PBGC has become the plan trustee.[Footnote 39] Shortly thereafter, this
letter is generally followed by a more detailed letter with a packet of
materials, including a DVD with an introduction to PBGC and frequently-
asked questions about how the benefit determination process works. PBGC
officials refer to this as a "welcome" package. Additionally, for large
plans likely to have many participants affected by the guarantee
limits, PBGC will hold on-site information sessions shortly after plan
termination. PBGC also operates a customer service center with a toll-
free number that participants can call if they have questions, provides
a Web site for workers and retirees with detailed information about
plans and benefits, and sends participants a newsletter with
information about PBGC once or twice per year.[Footnote 40]
Nearly all pension advocacy groups and union representatives we spoke
with[Footnote 41] praised PBGC's efforts to hold information sessions
with the larger plans. One union representative commended PBGC staff
for going out into the field to talk with participants and answer
questions even though participants are going to be angry. Other union
representatives commented that they have been impressed by PBGC's staff
for staying at these sessions until they have answered every
participant's questions. While these sessions are generally viewed as
helpful, some pension rights advocates noted that the information
presented is difficult for participants to understand, and may not have
the same meaning when talked about in generalities as when they later
receive notices concerning their specific benefits. Also, since not
everyone may attend these events, these advocates believe it is
important for all the information presented at the sessions to be
provided through written communication as well.
PBGC's customer service center and Web site received mixed reactions
from the pension rights advocates and union representatives we
interviewed. A few noted that some of their members reported receiving
good service from the toll-free number while others found the service
frustrating or useless. One union representative said that the center's
staff use PBGC terminology, which may be different from the plan and
benefit language that is familiar to their members. However, other
groups we spoke with were generally more positive regarding their own
direct communications with PBGC staff, describing PBGC staff as
forthcoming and responsive to their inquiries. Similarly, the groups we
interviewed generally found the information on PBGC's Web site useful,
but they expressed doubt that this would be the case for most of their
members. They noted that many people whose plans are taken over by PBGC
are not accustomed to using a computer or do not have access to the
internet, and that some do not feel comfortable relying on information
they find on a Web site.
Long Gaps in Communication until Completion of the Process Raised Some
Concerns:
Following the initial contacts, PBGC generally does not communicate
with participants again until the benefit determination process is
complete, which in some cases can stretch into years.[Footnote 42]
Among the participants' files we examined when the benefit
determination process took 4 or more years, we found that there often
was no contact from PBGC for most of this time. For example, we
examined the files of five randomly selected Bethlehem Steel
participants whose benefit determinations were still pending as of
February 2009,[Footnote 43] and found that--aside from one instance of
an acknowledgment of a form submitted by one participant--PBGC had not
communicated with these participants for more than 5 years. The last
PBGC-initiated communications were dated late 2003 or early 2004.
Some of the pension advocacy groups and union representatives we spoke
with said that these long periods without communication are problematic
for participants for several reasons. For example, retirees whose
benefits are subject to the guarantee limits but who continue to
receive their higher plan-level benefits for long periods of time may
come to expect that these higher amounts are permanent, and then they
are surprised when--years later--their benefits are suddenly reduced.
Even for participants who are not yet receiving benefits, the lack of
communication about the likely amount of their final benefits makes it
difficult to plan for retirement. Some groups noted that PBGC does not
always provide realistic time frames for completing the benefit
determination process, and does not periodically update participants on
the status of benefit processing. Two groups suggested it would be
helpful if PBGC provided updates at least every 6 months.
Complicated Benefit Calculations Are Not Adequately Explained:
When participants are notified of a payment amount--whether estimated
or final--PBGC's letters generally provide only limited explanations
for why the amount may be different from the amount provided under
their plan. In complex plans, when benefit calculations are
complicated, the letters do not adequately explain why benefits are
being reduced, and although benefit statements are generally attached,
the logic and math involved can be difficult even for pension experts.
The standard language used in these letters to explain a different
estimated amount states: "We have adjusted the amount of your benefit
because there are legal limits on how much we can pay." The standard
language used to explain a different final benefit amount states: "Your
final monthly benefit of [amount] is the amount that the PBGC is
legally allowed to pay you. It was calculated by determining the
benefit you are entitled to in your plan and then applying the limits
spelled out in federal pension law." These letters generally provide no
specific information about which limits apply or why. However, enclosed
with each benefit letter is a detailed attachment that shows the line-
by-line calculations leading to the benefit amount, referred to as a
"benefit statement." In the participant files we reviewed, these
benefit statements ranged in length from 2 to 8 pages, and were very
difficult to understand. In some cases, there were as many as 20 to 30
different line items that required making comparisons between the items
to understand the logic of the calculations. (See sample letter
provided in appendix VII.)
Some pension advocates and union representatives we spoke with said
that they found the explanations in these letters to be too vague and
generic, and that the letters did not provide enough information
specific to the individual's circumstances to be helpful. This was
especially true in cases where participants were shocked or confused by
a large benefit reduction. Moreover, some said they did not think most
participants would be able to understand the accompanying benefit
statements without additional information and assistance--especially
for complex cases, according to one advocate. At the same time, they
were generally sympathetic to the difficulty of communicating such
complicated information. As one advocate acknowledged, for the letters
to be accurate, they have to be complicated; this may just be "the
nature of the beast." Nevertheless, they said that PBGC could take some
steps to improve the letters. For example, for those likely to incur
overpayments, they suggested providing an example of how the recoupment
process works. For those with complex benefit statements, they
suggested that PBGC provide more text to help explain each step of the
calculations, and include referrals to pension rights groups for
obtaining additional information and assistance.
In addition, we found a number of errors in the correspondence with
participants, although we reviewed only a small sample of letters for
participants in certain complex plans. For example, we found a number
of cases with corrected benefit determination letters and other
correspondence that had been sent to rectify various errors, such as
the failure to account for overpayments, or inaccurate end dates for
recoupment. We also identified some errors in the payment amounts or
other information in the letters that we brought to PBGC's attention to
be corrected.[Footnote 44]
PBGC has developed more than 500 letter formats--in both English and
Spanish--to address the myriad of situations that may arise in the
benefit determination process. Nevertheless, PBGC officials
acknowledged that their standard letter formats may not always meet the
needs of participants, especially those in complex plans, and they
recently undertook a project to review and update their letters to try
to better meet participant needs.
According to PBGC officials, in September 2008, they began rolling out
about 50 different versions of key letters to fit different
circumstances. They also noted that the amount of detail and length of
the benefit statements has varied over time--sometimes longer,
sometimes shorter. Most recently, they have tended toward longer. They
commented, however, that they are not sure it makes a difference either
way, because for the most part, participants react to the benefit
amount, not to the steps PBGC has used to arrive at the amount.
Finally, they also noted that while the benefit amounts in the letters
are verified by actuaries, the letters are prepared manually by Field
Benefit Administration staff, using the standard formats, and until
recently, these letters were not reviewed. Beginning in early 2009,
however, plan analysts have started to review the letters before
mailing.
Restructured Appeals Process Resolves Requests More Efficiently but Key
Information Is Not Readily Provided:
Since streamlining its appeals process in 2003, PBGC has responded more
quickly to correspondence sent to its Appeals Division (see figure 12).
It has reduced the average amount of time to decide an appeal by almost
a year and has cut the average amount of time needed to resolve all
appeals-related inquiries in half. At the same time, most appeals
docketed since 2003 have not resulted in appellants receiving higher
benefit amounts. A lack of understanding on the part of participants
about how their benefits are calculated may contribute to unnecessary
appeals.
Figure 12: PBGC's Appeals Process:
[Refer to PDF for image: illustration]
Incoming correspondence:
Within 45 days of receiving a final benefit determination, participants
may appeal to PBGC’s Appeals Board or request additional time to file
an appeal.[A] PBGC acknowledges the receipt of incoming correspondence
with a letter and decides if it should be docketed as an appeal based
on whether the correspondence raises a question about how the plan was
interpreted, how the law was interpreted, or the practices of the
plan’s sponsor.
Docketed as Appeal:
The appeal is assigned to a member of PBGC’s board of appeals and an
analyst, who review information on the case. The board member
determines whether there is a precedent-setting legal issue involved.
Board Decision (if legal issue involved):
Three of the five board members discuss and vote on any precedent-
setting cases.
Board Member Decision (if legal issue not involved):
Board members make independent decisions on cases that are not
precedent-setting.
Appeals Closed:
Once a decision is made, affected parties are notified. If the decision
results in a change in the benefit determination, the Benefits
Administration and Payment Department is notified.
Change in Benefit Determination:
As a result of an appeal decision or because of new information
received, PBGC’s Benefits Administration and Payment Department may
calculate a new benefit amount for any affected parties.
Incoming correspondence Not Docketed as Appeal:
PBGC officials categorize correspondence that is not docketed as an
appeal.
Treated as extension request:
PBGC grants the appellant an additional 45 days to submit an appeal.
Referred internally:
Correspondence may be referred to the Benefits Administration and
Payment Department or to the Appeals Division.
Other:
Some correspondence may not require a response.
Source: GAO analysis of PBGC documents.
[A] If an appeal is filed, PBGC's benefit determination will not take
effect until the Appeals Board issues a decision; in the meantime, PBGC
continues to pay retirees benefits at the estimated level.
[End of figure]
Triage Approach Has Streamlined Appeals Process:
PBGC's appeals process was restructured in 2003 to create a triage
system that makes more efficient use of agency resources and resolves
cases more quickly. Previously, PBGC treated nearly every
correspondence sent to its Appeals Division as an appeal. The agency
now evaluates correspondence to determine if it raises a question about
how the plan was interpreted, how the law was interpreted, or the
practices of the plan's sponsor and dockets correspondence as an appeal
if it meets these criteria based on regulations.[Footnote 45] In
analyzing appeals correspondence associated with plans trusteed by PBGC
from fiscal year 2000 to fiscal year 2008, we found that since 2003,
PBGC docketed as an appeal less than one-third of the correspondence
received by the Appeals Division (see figure 13). Correspondence
concerning corrections to personal data, such as a participant's date
of hire or length of service, is now directed to PBGC's Benefits
Administration and Payment Department (Benefits Department) so that a
corrected benefit determination can be issued more expeditiously.
Additionally, in instances where a potential appellant requests a more
detailed explanation of his or her benefit determination, the Benefits
Department can quickly provide a detailed explanation based on its
familiarity with the benefit calculation and relevant participant data.
Further, under this triage approach, the Appeals Board staff, rather
than the Appeals Board, responds to appeals received before a benefit
determination has been issued or to claims that PBGC's recovery of
overpayments create a financial hardship and should be waived.
Figure 13: Actions Taken on Correspondence Associated with Plans
Trusteed during Fiscal Years 2000-2008:
[Refer to PDF for image: two pie-charts]
Correspondence received prior to Fiscal Year 2003:
Docketed as appeal: 90.2%;
Treated as an extension request: 9.8%.
Correspondence received Fiscal Years 2003-2009:
Referred to the Appeals Division[A]: 12.4%;
Other[B]: 16.4%;
Treated as an extension request: 16.4%;
Referred to the Benefits Administration and Payment Division[C]: 25.7%;
Docketed as appeal: 29.0%.
Source: GAO analysis of PBGC data on participants in plans terminated
and trusteed during fiscal years 2000 through 2008.
Note: Data reflect multiple correspondences associated with individual
cases.
[A] Referrals to the Appeals Division include appeals that are filed
too early or requests to have PBGC's recovery of overpayments waived.
[B] Other includes correspondence that requires no action or is sent in
response to a PBGC request, as well as Freedom of Information Act
requests.
[C] Referrals to the Benefits Administration and Payment Division
include corrections to personal information and requests for benefit
explanations.
[End of figure]
Since streamlining the appeals process, PBGC has reduced its response
time for appeals and other appeals-related inquiries without increasing
the size of its appeals staff. According to agency data, PBGC reduced
its average time for closing docketed appeals from 2.3 years to 1.4
years since implementing this triage approach. In fact, since fiscal
year 2005, PBGC has averaged a response time of less than 10 months
(see figure 14). PBGC has also reduced the average age of pending
appeals from about 2 years to less than 9 months, since implementing
its triage approach. We also found, on examining the 14,545 appeals-
related correspondences associated with plans trusteed from fiscal year
2000 to fiscal year 2008, that PBGC responded to all correspondence in
an average of less than 4 months after 2002 (fiscal years 2003 through
2009), as compared to an average of about 8 months prior to 2003
(fiscal years 2000 through fiscal year 2002). However, there were also
852 cases of correspondence which had been pending for an average of
nearly 7 months, as of April 2009.
Figure 14: Average Age of Closed and Pending Appeals, by Fiscal Year:
[Refer to PDF for image: vertical bar graph]
Fiscal year: 2000;
Average age, closed appeals: 1.9 years;
Average age, pending appeals: 1.9 years.
Fiscal year: 2001;
Average age, closed appeals: 2.3 years;
Average age, pending appeals: 2.1 years.
Fiscal year: 2002;
Average age, closed appeals: 2.8 years;
Average age, pending appeals: 1.9 years.
Fiscal year: 2003;
Average age, closed appeals: 2 years;
Average age, pending appeals: 1.7 years.
Fiscal year: 2004;
Average age, closed appeals: 1.4 years;
Average age, pending appeals: 0.8 years.
Fiscal year: 2005;
Average age, closed appeals: 0.7 years;
Average age, pending appeals: 0.5 years.
Fiscal year: 2006;
Average age, closed appeals: 0.8 years;
Average age, pending appeals: 0.7 years.
Fiscal year: 2007;
Average age, closed appeals: 0.9 years;
Average age, pending appeals: 0.5 years.
Fiscal year: 2008;
Average age, closed appeals: 0.8 years;
Average age, pending appeals: 0.5 years.
Source: GAO analysis of PBGC data on participants in plans terminated
and trusteed during fiscal years 2000 through 2008.
Note: Data provide a snapshot of the average age of closed and pending
appeals at the end of each fiscal year.
[End of figure]
The procedural requirements of the appeals process do not appear to
present barriers to appellants. Appellants are to provide a specific
reason for their appeals and submit them within 45 days of their
benefit determinations.[Footnote 46] Of the 3,637 closed appeals we
examined, only 37 were closed because the appellant did not conform to
a procedural requirement.[Footnote 47] Additionally, PBGC readily
grants extensions. Within the correspondence we examined, PBGC granted
2,371 extension requests during fiscal years 2000 through 2008.
Most Appeals Have Not Resulted in Higher Benefit Amounts for
Appellants:
More than 80 percent of appeals resulted in appellants receiving no
increase in their benefit amounts. Of the 4,337 correspondences that
were docketed as appeals since the beginning of fiscal year 2003, 3,637
had been decided as of April 2009. In most of these cases, the appeal
decision resulted in no change to the participant's benefit
determination amount (see figure 15). However, appellants received a
higher benefit amount in 18 percent of the cases.[Footnote 48] For
example, in one of the successful appeals, a Bethlehem Steel
participant submitted copies of his medical records with his appeal,
convincing the Appeals Board that he was eligible to receive a
"permanent incapacity" benefit. In another case, a participant in the
US Airways Inc. (pilots) plan had US Airways Inc. furnish documentation
to PBGC that his date of hire had been adjusted as the result of a
lawsuit, and with this new date of hire, PBGC considered the
participant vested.
Figure 15: Outcome of Appeals since Restructuring:
[Refer to PDF for image: pie-chart]
No effect on benefit determination: 81.0%;
More favorable outcome than benefit determination: 18.5%;
Less favorable outcome than benefit determination: 0.5%.
Source: GAO analysis of agency data on participants in plans terminated
and trusteed during fiscal year 2003 to fiscal year 2008 with a closed
appeal. There were 700 appeals that had not yet been closed, as of
April 2009.
[End of figure]
In cases with no change in the participant's benefit determination
amount, the amount of overpayment can grow significantly during an
appeal. While cases are appealed, PBGC typically places a hold on any
change in benefit until the appeal is resolved. Thus, in cases where
the benefit determination amount is less than the estimated amount, the
participant may continue to receive the higher estimated amount during
an appeal. If the lower amount is ultimately upheld, we found that
these continued higher payments could add significantly to the amount
of the participant's overpayment--more than $10,000 in some cases.
PBGC Does Not Readily Provide Key Information that Could Help Avoid
Unnecessary Appeals:
Although some appellants have successfully used the appeals process to
increase their benefits, PBGC is not readily providing key information
that would be helpful to participants in deciding whether or not to
pursue an appeal. For example, the information PBGC provides on how it
arrives at its benefit calculations can be difficult for potential
appellants to understand. Plan provisions and guarantee limitations are
often complicated, and it may be difficult for the average individual
to interpret PBGC's benefit calculations, especially for complex plans.
Based on Appeals Board findings, it appears that participants sometimes
file appeals because they do not understand how the guarantee
limitations affect their benefits. For example, the Appeals Board
denied one Weirton Steel participant's appeal by explaining that the
participant's estimated benefit included a temporary supplement that,
ultimately, was not payable due to the accrued-at-normal limitation. In
another case, the Appeals Board concluded that an Outboard Marine
participant simply did not understand PBGC's benefit statement and
explained the accrued-at-normal, maximum, and phase-in limitations,
while denying the participant's appeal.
Even pension counselors and union representatives, who are
knowledgeable about pensions and have experience filing appeals with
PBGC, had difficulty understanding the materials provided to
participants about their benefits . Several of the pension counselors
and union representatives we interviewed told us that they have
established contacts at PBGC who help them understand benefit
determinations in appeals cases, and they, in turn, help convey this
information to the participants they serve. Some have even held three-
way calls with PBGC's customer service center and participants, so that
they can help participants understand the information provided by PBGC.
Additionally, representatives from the pension counseling centers we
spoke with have actuarial support they consult for help interpreting
complicated benefit calculations.[Footnote 49] In some cases, by
assisting participants in understanding their benefit calculations
better, pension counselors told us they can also help participants
avoid unnecessary appeals.
Some of those we interviewed also told us that a complete understanding
of a participant's benefit determination--which is important for an
effective appeal--cannot be obtained from a benefit determination
letter alone. Several of these pension counselors and union
representatives commented that they routinely file Freedom of
Information Act[Footnote 50] requests, on a participant's behalf, to
obtain more information about a participant's case from PBGC when
preparing an appeal because there is not sufficient information in the
benefit determination letter. Although PBGC provides a guide on how to
use these requests on its Web site, PBGC's communications materials
about the appeals process do not provide a description of how
individuals can gain access to PBGC's full benefit calculation records
through a Freedom of Information Act request.
Conclusions:
The current economic downturn has already brought a new influx of
pension plan terminations to PBGC, and more are expected to follow.
While our findings reveal a reasonably good record of processing
beneficiary cases and assuming responsibility for the payment of
benefits since 2000, the loss of jobs at this time, as well as the
impending retirement of the baby boom generation, leave little room for
anything short of high performance. This means acting as quickly and as
efficiently as possible to value and allocate plan assets; to expedite
the calculation of estimated benefits to reflect guarantee limits, as
well as final benefit amounts; and to keep plan participants well-
informed throughout the benefit determination process. Workers and
retirees in terminated plans who stand to lose as much as one-half or
more of their long-anticipated retirement income will likely have to
make painful financial adjustments, and due consideration in helping to
ease that pain is warranted.
The calculation of benefits according to complicated provisions that
vary by plan is a challenging task. It becomes more so with the delays
that can occur in valuing the assets of large and complex plans and
determining how those assets are to be allocated among different
groups. However, the likelihood of lengthy processing for some plans is
not unpredictable, and while PBGC has taken steps to expedite the
processing of small and simpler plans, its approach to large and
complex plans appears less than strategic. The hope of freeing up staff
to handle complex plans by processing others more quickly will probably
not be sufficient by itself for tackling difficult plans in the near
future. Absent a calculated effort to anticipate and plan for such
terminations, the heretofore modest number of beneficiaries caught in a
protracted process could, indeed, grow in the next few years.
While overpayments to those already in retirement have been infrequent,
delays clearly exacerbate them. Moreover, the failure to communicate
more often and clearly with participants awaiting a final determination
can be disconcerting--especially when they receive the news that their
final determination is "surprisingly" less than they anticipated, or
when retirees learn that the estimated interim benefit they had been
receiving was too high and that they owe money. PBGC's long recoupment
period--which can be even further elongated by an appeal--may be a
consolation to such retirees; however, the agency itself stands to lose
considerable sums under this policy. This is another peril for an
agency that may well be dealing with an increasing number of plan
failures. Clearer and more frequent communication with plan
participants, including quicker and responsible adjustments to
estimated benefits, more information about how their benefits are
calculated, and where to find help if they wish to appeal, would better
manage expectations, help people plan for their future, avoid
unnecessary appeals, and earn good will in a trying time for all.
Recommendations:
To improve PBGC's benefit determination process, a more strategic
approach is needed to prepare for and manage the calculation of benefit
amounts and communications with participants in cases involving large,
complex plans. Specifically, we recommend that:
* PBGC should set goals for timeliness and monitor the progress made in
finalizing benefit determinations for large, complex plans separately
from other plans.
* To reduce the number and size of overpayments in large, complex
plans, PBGC should prioritize the calculation of estimated benefits for
retirees subject to the guarantee limits and adjust estimates, as
needed, throughout the benefit determination process. To reduce
increased overpayments due to appeals, PBGC should prioritize the
processing of appeals for those already receiving benefits and should
consider implementing the final benefit determination for retirees
during the appeals process.
* PBGC should develop improved procedures for adapting and reviewing
letters to participants in large, complex plans, such as by (1)
providing more specific information in letters to participants who
receive benefit reductions describing which limits were applied and
why; (2) ensuring all letters to participants involving benefit
reductions are reviewed for accuracy and coherence before being sent;
and (3) establishing processes to more frequently communicate with
participants who are experiencing delays in receiving final benefits
determinations.
* PBGC should provide information or resources to help participants in
large, complex plans better understand their benefit calculations and
also to avoid any unnecessary appeals. Specifically, PBGC's benefit
determination letters should provide information, such as how
participants can obtain additional information by using the Freedom of
Information Act or other resources.
Agency Comments and Our Evaluation:
We obtained written comments on a draft of this report from PBGC's
acting director, which are reproduced in appendix VIII. PBGC also
provided technical comments, which are incorporated into the report
where appropriate. In addition, we provided copies of the draft report
to the Departments of Commerce, Labor, and Treasury.
In response to our draft report, PBGC generally concurred with our
recommendations and outlined actions the agency has under way or plans
to take in order to address each topic of concern. With respect to the
first recommendation, PBGC agreed and noted that the agency has started
to implement steps for tracking and monitoring tasks associated with
processing large, complex plans. While we are pleased to learn of these
steps being initiated, we would like to emphasize the importance of
setting goals for processing large, complex plans and reporting
progress toward meeting those goals separately from other plans. With
respect to the second recommendation, PBGC agreed and commented that it
generally already identifies and prioritizes cases where adjustments to
estimated benefits are likely, but will continue to look for ways to
improve its processes. Moreover, despite possible legal concerns with
implementing final benefit determinations prior to completion of the
appeals process, the agency is willing to explore options for making
earlier benefit adjustments, when appropriate. With respect to the
third recommendation, PBGC agreed and noted that the agency is revising
the guidelines for its benefit statements to better communicate the
complexities of PBGC benefits and to better manage expectations of plan
participants. The comments state that the agency will evaluate and make
necessary modifications to its letter review process, as well as
examine ways to more frequently and clearly communicate with
participants experiencing delays in receiving final benefit
determinations. Finally, with respect to the fourth recommendation,
PBGC agreed to amend its appeals brochure to include information about
accessing records through Freedom of Information Act requests.
As agreed with your staff, unless you publicly announce its contents
earlier, we plan no further distribution of this report until 30 days
after its issue date. At that time, we will send copies of this report
to the Acting Director of PBGC, the Secretary of Labor, the Secretary
of the Treasury, and other interested parties. In addition, the report
will be available at no charge on the GAO Web site at [hyperlink,
http://www.gao.gov]. If you or your staff have any questions concerning
this report, please contact me at (202) 512-7215 or bovbjergb@gao.gov.
Contact points for our Offices of Congressional Relations and Public
Affairs may be found on the last page of this report. GAO staff who
made key contributions to this report are listed in appendix IX.
Signed by:
Barbara D. Bovbjerg:
Director, Education, Workforce, and Income Security Issues:
List of Requesters:
The Honorable Max Baucus:
Chairman:
The Honorable Charles E. Grassley:
Ranking Member:
Committee on Finance:
United States Senate:
The Honorable Edward M. Kennedy:
Chairman:
The Honorable Michael B. Enzi:
Ranking Member:
Committee on Health, Education, Labor, and Pensions:
United States Senate:
The Honorable Barbara A. Mikulski:
Chairman:
The Honorable Richard Burr:
Ranking Member:
Subcommittee on Retirement and Aging:
Committee on Health, Education, Labor, and Pensions:
United States Senate:
The Honorable Evan Bayh:
The Honorable Sherrod Brown:
The Honorable Robert P. Casey, Jr.
The Honorable Richard J. Durbin:
The Honorable Arlen Specter:
United States Senate:
[End of section]
Appendix I: Scope and Methodology for Analysis of Automated and Imaged
Data:
To assess the timeliness and results of the Pension Benefit Guaranty
Corporation's (PBGC) benefit determination process, we obtained
automated data from PBGC on all plans terminated and trusteed during
fiscal years 2000 through 2008, as well as data for all individuals
associated with those plans. Three different data sets were provided:
(1) a plan level data set, (2) an individual level data set with
benefit data, and (3) an individual level data set with appeals data.
The plan level data set, including 1,089 plans total, was comprised of
three component groups:
* Group A - plans for which the valuation of assets and liabilities had
been completed, as of February 2009 (909 plans).
* Group B - plans for which the valuation of assets and liabilities had
not been completed, as of February 2009 (83 plans). When actual values
were not yet available, estimated values for assets and/or liabilities
were provided. The participant count for these plans was based on
audited data.
* Group C - plans for which the valuation of assets and liabilities was
not completed, as of February 2009 (97 plans). Estimated values for
assets and/or liabilities were only available for some plans. The
participant count for these plans was based on preliminary data.
We analyzed the plan level data to determine the length of time it
takes PBGC to complete the valuation of a plan's assets and
liabilities, on average. We also analyzed the plan level data to
identify various plan characteristics, such as the fiscal year when
trusteed by PBGC and the extent to which participants' benefits are
affected by legal guarantee limits. PBGC does not systematically track
the number of participants affected by one or more of the three types
of guaranteed benefit limits specified under the Employee Retirement
Income Security Act of 1974 (ERISA) and related regulations--which
include maximum, phase-in, and accrued-at-normal limits--or how much
these limits affect participants' benefit amounts. However, PBGC does
systematically track each plan's total benefit liabilities and the
amount PBGC owes, taking into account the guarantee limits. The
difference between these two amounts (referred to as the amount of
"unfunded nonguaranteed benefits") provides an indicator of the
magnitude of the impact of guarantee limits on participants within each
plan. If the amounts are the same, it means that no participants had
benefits reduced due to these limits. If total liabilities are greater,
it means that at least one participant had benefits reduced due to
these limits.
The individual level data set, with benefit data as of February 2009,
included 1,487,679 individuals associated with 1,057,272 primary
participants (the person who had earned the pension). The most common
reasons for multiple individuals per case were situations where a
portion of the pension was to be shared between the primary participant
and another individual with a qualified domestic relations order
(referred to in the data set as an "alternate payee"), or situations
where a primary participant had died and the pension was being paid to
a beneficiary. In our analyses, we aggregated the data so that the
characterization of each case reflected the data for the primary
participant, as well as all other individuals associated with that
primary participant, as appropriate for the data element being
analyzed.
We analyzed the individual level data on benefits, by case, to
determine the length of time it takes PBGC to make benefit
determinations and the extent to which overpayments affect retirees'
benefits. To assess the time required for processing, we began by
identifying all those participants whose benefit determinations had
been completed. We then examined the length of time between the date
the participant's plan was trusteed and the date PBGC first issued a
final benefit determination letter to the participant. (Subsequent
benefit determination letters are sometimes issued when corrections are
needed or when a participant successfully appeals.) For participants
whose benefit determinations were still pending, we calculated the
length of time between the plan's trusteeship and February 18, 2009,
when these data were provided, to determine how long the determinations
had been awaiting completion. We also analyzed the length of time to
process benefit determinations by participants' retirement status at
the time the plan terminated.
To determine the proportion of participants possibly affected by
overpayments, we first identified all those who had received estimated
benefits and then compared the earliest available estimated benefit
amount with the final benefit amount, by case, tabulating whether the
difference was positive (indicating a likely overpayment) or negative
(indicating a likely underpayment). Because estimated benefit amounts
may be adjusted over time, and because the records on estimated
benefits had sometimes been overwritten or deleted, we were not able to
use these data to determine with certainty whether or not an
overpayment or underpayment had been incurred, or the amounts involved.
Instead, to assess the amount of overpayments incurred and the effect
of repaying these debts on participants' benefits, we analyzed the data
on recoupments. First, we identified all those who were listed as
having amounts recouped to date, by case. We then used the available
data on projected benefit reductions, which included the amount of
monthly reduction and the start date and end date for that reduction
amount (sometimes involving up to four different reduction amounts) to
calculate the amounts yet to be recouped. We determined the total
amount of the overpayments, by case, by combining the data PBGC
provided on amounts recouped to date with our calculation of amounts
yet to be recouped.
Based on a review of selected records in PBGC's image processing system
for cases with the largest overpayments, it appears that these data are
reliable for identifying whether a case has an overpayment, but not as
reliable for determining the total amount of overpayments. We were able
to verify that the participant with the largest overpayment, according
to our analysis of these data, was correct: an LTV participant with an
overpayment of about $152,000. Also, we found that the amounts
calculated using these data were very close (within 2 percent) of the
overpayment amounts in the records for 15 of the 24 cases reviewed--
differences small enough to be explained by rounding. However, in the
remaining 9 cases, the amounts calculated varied significantly from
those in the records--some greater, some less. We investigated the 3
most egregious differences and found that all 3 were due to data entry
errors in the PBGC data set. In 2 cases, PBGC officials told us that
the end date for recoupment had been entered as 12/1/2099 by default,
which was not correct. They said that they would implement a system fix
to prevent inappropriate use of this default in the future. In the
third case, we found that the monthly payment amount had been
inadvertently entered as the monthly reduction amount. None of these
errors had resulted in inaccurate payments to participants, since all
involved future recoupment amounts. However, it appears that the
reliability of these data for calculating total overpayment amounts is
limited.
We also analyzed the individual level data, by case, to identify
various case characteristics, aggregating the data together for all
individuals associated with the same case. These characteristics
included the final benefit amount (with and without any benefit
reduction due to recoupment), and the projected age of the youngest
individual at the end of recoupment for cases with overpayments greater
than $10,000.
We then combined the plan level data and individual level data, by
case, to determine the number of individuals and cases associated with
each plan, and identify those plans with the most cases that took 4 or
more years to provide a final benefit determination, and the most cases
with overpayments. We also used these data to generate lists of cases
for more detailed reviews of documents in PBGC's image processing
system and examine more closely the cases that took the longest to
provide a benefit determination and that had the largest overpayments
and benefit reductions.
In addition to the automated data, PBGC maintains records that are
individually scanned into an image processing system. The types of
documents we reviewed in PBGC's image processing system included both
plan documents and participant records. On the plan level, we reviewed
documents for the plans most affected by guarantee limits, by delays in
processing, and by overpayments (see appendix VI). For the 10 plans
ranking highest in each of these categories, we typically reviewed the
"actuarial case memo," which summarizes all the steps taken to obtain
records and determine the value of assets and liabilities for each plan
terminated and trusteed by PBGC. We then selected five of these plans
for more detailed review of participant records in order to illustrate
key trends identified in our analysis of the automated data. These five
plans were: Bethlehem Steel, LTV Steel, RTI-United Steelworkers of
America (USWA), US Airways, and Weirton Steel. For Bethlehem Steel, we
randomly selected five participants from among those participants whose
benefit determinations were still pending. For each of the other four
plans, we randomly selected five participants from the lists of
participants provided in the plans' actuarial memos. Then, for each of
these participants, we typically reviewed all letters sent to the
participant, all benefit calculation documents, and the internal
correspondence among PBGC staff about the case. We reviewed the letters
to participants only to determine if they accurately conveyed
information documented elsewhere in the files. We did not attempt to
verify PBGC calculations of benefit amounts.
Finally, to assess the length of time it takes PBGC to provide a
decision when a participant appeals, we examined PBGC data on the
average time to close docketed appeals and the average age of pending
appeals, by fiscal year, 2000 through 2008. We also analyzed the 14,545
appeals-related correspondences associated with plans terminated and
trusteed during fiscal years 2000 through 2008 so that we could make
comparisons between PBGC's average response time, both before and after
its restructuring of the appeals process. Data reflect multiple
correspondences associated with individual cases. For correspondences
that were pending, we calculated the amount of time between when PBGC
received these correspondences and April 13, 2009, when we received
these data. To describe PBGC's triaging system, which was implemented
in fiscal year 2003, we analyzed PBGC's action taken code for each
correspondence and aggregated these results into two groups: those
correspondences received during fiscal years 2000 through 2002, and
those received during fiscal years 2003 through 2008. Of the 4,337
correspondences that were docketed as appeals, 3,495 had been decided
as of April 2009. We then tabulated the data on the outcomes of closed
appeals and the reasons why these appeals were closed, which were coded
to indicate whether a change to the benefit amount occurred.
[End of section]
Appendix II: Organizations and Participants Contacted:
Organizations:
American Federation of Labor and Congress of Industrial Organizations
(AFL-CIO) [hyperlink, http://www.aflcio.org] A voluntary federation of
56 national and international labor unions, representing 11 million
members in a variety of industries.
Air Line Pilots Association [hyperlink, http://www.alpa.org] The
largest airline pilot union in the world, representing nearly 54,000
pilots at 36 U.S. and Canadian airlines.
Association of Flight Attendants-Communications Workers of America
[hyperlink, http://www.afanet.org] The world's largest flight attendant
labor union, organized by flight attendants for flight attendants,
representing over 55,000 flight attendants at 20 airlines.
New England Pension Assistance Project [hyperlink,
http://www.pensionaction.org/nepap.htm] One of the six regional
projects funded by the Administration on Aging to provide free pension
counseling services. Initially, the project served only Massachusetts
residents, but in 1998, it expanded to help residents of the six-state
New England region: Connecticut, Maine, Massachusetts, New Hampshire,
Rhode Island, and Vermont.
Ohio Pension Rights Center [hyperlink,
http://www.proseniors.org/oh_pension.html] Part of one of the six
regional projects funded by the Administration on Aging to provide free
pension counseling services. The Ohio Pension Rights Center shares a
grant with the Michigan Pension Rights Project and provides all types
of pension assistance to people in Michigan, Ohio, Pennsylvania,
Kentucky, and Tennessee.
Pension Rights Center [hyperlink, http://www.pensionrights.org]
Provides legal consultation and training to the six regional projects
funded by the Administration on Aging to provide pension counseling
services for individuals who need help in understanding and enforcing
their pension and retirement savings plan rights.
United Steelworkers[Footnote 51] [hyperlink, http://www.usw.org/] The
largest industrial labor union in North America, representing 1.2
million current and retired workers in industries that include primary
and fabricated metals, mining, chemicals, paper, glass, rubber,
transportation, utilities, container industries, pharmaceuticals, call
centers, and health care.
Participants:
Members of the Reliance Group Holdings Inc. plan and the Reliance
Insurance Company plan, which were among the plans most affected by
long processing times.
Members of the Republic Technologies International USWA and USS/KOBE
plans, which were among the plans most affected by the guarantee
limits, long processing times, and/or overpayments.
Members of United Air Lines ground employees plan and pilots' plan.
[End of section]
Appendix III: The Process for Allocating Assets and Recoveries to
Participant Benefits:
Upon the termination of a single-employer plan, plan assets are
identified, valued, and then allocated to participant benefits, in
accordance with the provisions in ERISA, section 4044.[Footnote 52] In
addition to plan assets, any monies from company assets that PBGC
recovers for unfunded benefit liabilities are allocated to participant
benefits, in accordance with the provisions in ERISA, section 4022(c).
[Footnote 53]
Section 4044--Allocation of Plan Assets:
The amount of plan assets available to pay for participant benefits
includes all plan assets remaining after the subtraction of all prior
or current liabilities paid or payable from the plan. This amount
includes the value of the collectible portion of any due and unpaid
employer contributions.[Footnote 54] Liabilities include expenses, fees
and other administrative costs, and benefit payments due before the
allocation date. For plans terminated and trusteed by PBGC, assets are
valued and the allocation determined based on liabilities as of the
termination date.[Footnote 55]
Plan assets available to pay for benefits under the plan are allocated
to participant benefits according to six priority categories, as
described in Table 3. Assets are allocated to each priority category in
succession, beginning with priority category 1. If the plan has
sufficient assets to pay for all benefits in a priority category, the
remaining assets are allocated to the next lower priority category.
This process is repeated until all benefits in priority categories 1
through 6 have been provided or until all available plan assets have
been allocated. Most private sector defined benefit plans do not
require or allow participant contributions. Thus, in most trusteed
plans, asset allocation begins with the benefits in priority category
3, that is, the benefits of those retired or eligible to retire 3 years
before the plan terminated. However, it should be noted that assets are
allocated based on type of benefit, not retirement status, and that
many participants have benefits in more than one category.
Table 3: Priority Categories for Allocating Participant Benefits:
Priority category 1:
Accrued benefits derived from voluntary employee contributions.
(According to PBGC, such benefits are "extremely rare" among private
sector defined benefit plans.)[A]
Priority category 2:
Accrued benefits derived from voluntary employee contributions.
(According to PBGC, such benefits are "extremely rare" among private
sector defined benefit plans.)
Priority category 3:
Annuity benefits that have been in pay status for at least 3 years
before the plan’s termination date, or could have been in pay status
for at least 3 years before the plan’s termination date had the
participant chosen to retire at his or her earliest possible retirement
date; however, benefits subject to the phase-in limitation (that is,
benefit increases made within the last 5 years) are excluded. These
benefits can be either guaranteed or nonguaranteed.
Priority category 4:
Other guaranteed benefits, and certain nonguaranteed benefits.[B]
Priority category 5: Other vested nonguaranteed benefits that a
participant is entitled to under the plan; however, benefits that
result solely due to the termination of the plan--which are deemed
"forfeitable"--are excluded.
Priority category 6:
All other benefits under the plan. This category includes nonvested
benefits and "grow-in" benefits, which are benefits that are provided
in some situations where the company continues to operate after the
plan is terminated.
Source: GAO analysis of PBGC documents.
[A] However, one PBGC official noted that in the General Motors
salaried plan, employees have the option to make contributions, and
that if PBGC were to take over the plan, this would add a great deal of
complexity to the benefit determination process.
[B] Specifically, the nonguaranteed benefits included in priority
category 4 are those that are nonguaranteed because they are subject to
the aggregate benefits limitation for participants in more than one
plan that has been terminated with insufficient funds, or because they
are subject to special provisions applicable to substantial owners
(that is, those owning more than 10 percent of the company).
[End of table]
Except for priority category 5, which includes benefits subject to the
phase-in limit, if the plan assets available for allocation to any
priority category are insufficient to pay for all benefits in that
priority category, those assets are distributed among the participants
according to the ratio that the value of each participant's benefit or
benefits in that priority category bears to the total value of all
benefits in that priority category. If the plan assets available for
allocation to priority category 5 are insufficient to pay for all
benefits in that category, the assets are allocated by date of plan
amendment, oldest to newest, until all plan assets available for
allocation have been exhausted. Within each priority category, once the
amount of assets to be allocated to each participant has been
determined, assets are allocated first to the participant's "basic-
type" benefits (which include benefits that are guaranteed by PBGC, or
that would be guaranteed but for the maximum and phase-in limits), and
then to the participant's "nonbasic-type" benefits (which include all
other benefits).
Plan assets are distributed according to the process described above
until all have been allocated. Thus, to the extent plan assets are
available for allocation under this scheme, some participants may have
some or all their nonguaranteed benefits paid. For example, in the
scenario illustrated in figure 16, sufficient plan assets are available
to cover all priority category 3 guaranteed and nonguaranteed benefits,
as well as a portion of priority category 4 guaranteed benefits. PBGC
would then pay the remaining guaranteed benefits in priority category
4, but all remaining benefits (that is, priority categories 5 and 6
benefits, which are all nonguaranteed benefits), would not be paid, and
participants would have their benefits reduced accordingly, unless
there are recoveries of company assets that can be allocated to
benefits, as discussed below.
Figure 16: Example Scenario of Section 4044 Asset Allocation to
Participants' Benefits:
[Refer to PDF for image: vertical bar graph]
Priority Category: PC-1;
No data.
Priority Category: PC-2;
No data.
Priority Category: PC-3;
Guaranteed benefits: 60;
Nonguaranteed benefits: 20;
Portion of benefits paid for by plan assets: 80.
Priority Category: PC-4;
Guaranteed benefits: 70;
Portion of benefits paid for by plan assets: 41.
Priority Category: PC-5;
Nonguaranteed benefits: 22.
Priority Category: PC-6;
Nonguaranteed benefits: 12.
Source: Adapted from PBGC materials.
[End of figure]
Section 4022(c)--Allocation of Recovered Company Assets:
Section 4022(c), added to ERISA in 1987,[Footnote 56] requires PBGC to
share with participants a portion of its recoveries resulting from an
employer liability claim against the plan sponsor and other liable
parties, usually in bankruptcy. As a result, a portion of participants'
losses of unfunded nonguaranteed benefits can be paid. Where a plan's
unfunded nonguaranteed benefits exceed $20 million, the total amount
paid under §4022(c) depends on PBGC's actual recoveries in that case.
In all other cases, the amount paid is determined by an average of
PBGC's recoveries over a 5-year period.[Footnote 57]
PBGC allocates the participants' portion of the §4022(c) amount, as
described above, to participants' unfunded nonguaranteed benefits using
the same priority categories and procedures outlined above for the
§4044 asset allocation process. The allocation begins with the highest
priority category in which there are unfunded nonguaranteed benefits,
and then to each lower priority category, in succession. If the plan
§4022(c) amount to be allocated in a particular priority category is
not sufficient to pay all the unfunded nonguaranteed benefits in that
category, the amount will be allocated within the category as described
above for the §4044 allocation process. As noted by one employee group
we spoke with, it is more advantageous for participants for assets to
be considered recoveries allocated under §4022(c) than plan assets
allocated under §4044, because recoveries are shared between PBGC and
participants. For example, to continue with the scenario introduced
above, if company assets are recovered, some would be allocated to pay
a portion of the guaranteed benefits in priority category 4 that PBGC
would pay to participants regardless, and some would be allocated to
pay a portion of priority category 5 nonguaranteed benefits that would
not have been paid otherwise (see figure 17).
Figure 17: Example Scenario of Section 4022(c) Recovery Allocation to
Participants' Benefits:
[Refer to PDF for image: vertical bar graph]
Priority Category: PC-1;
No data.
Priority Category: PC-2;
No data.
Priority Category: PC-3;
Guaranteed benefits: 60;
Nonguaranteed benefits: 20;
Portion of benefits paid for by plan assets: 80.
Priority Category: PC-4;
Guaranteed benefits: 70;
Portion of benefits paid for by plan assets: 41;
Portion of benefits paid for by recovered company assets: 6.
Priority Category: PC-5;
Nonguaranteed benefits: 22;
Portion of benefits paid for by recovered company assets: 5.
Priority Category: PC-6;
Nonguaranteed benefits: 12.
Source: Adapted from PBGC materials.
[End of figure]
To help illustrate this process, we gathered data from plan records
about § 4044 and § 4022(c) asset allocation for 10 large, complex
plans. The results are summarized in table 4.
Table 4: Asset and Recovery Allocation among 10 Large, Complex Plans:
Plan sponsor (name): LTV Steel Company Inc. (hourly plan);
Results of §4044 asset allocation: 66.08% of PC3;
Results of §4022(c) recovery allocation[A]: 100% of PC3 and 28.76% of
one subgroup of PC5 (phase-in limit losses).
Plan sponsor (name): LTV Steel Company Inc. (salary plan);
Results of §4044 asset allocation: 65.38% of PC3;
Results of §4022(c) recovery allocation[A]: 100% of PC3 (maximum limit
losses) and 0.25% of PC5 (accrued-at-normal and phase-in limit losses).
Plan sponsor (name): LTV Steel Mining Company, Inc.;
Results of §4044 asset allocation: 100% of PC3;
Results of §4022(c) recovery allocation[A]: 3.20% of PC5.
Plan sponsor (name): National Steel Corporation (hourly plan);
Results of §4044 asset allocation: 76.69% of PC3;
Results of §4022(c) recovery allocation[A]: 0.
Plan sponsor (name): Northwestern Steel & Wire (plan A);
Results of §4044 asset allocation: 100% of PC3;
Results of §4022(c) recovery allocation[A]: 0.
Plan sponsor (name): Outboard Marine (employees' plan);
Results of §4044 asset allocation: 100% of PC3 and 62.51% of PC4;
Results of §4022(c) recovery allocation[A]: 9.21% of one subgroup of
PC5 (phase-in limit losses).
Plan sponsor (name): Republic Technologies International (USS/KOBE
plan);
Results of §4044 asset allocation: 85.63% of PC3;
Results of §4022(c) recovery allocation[A]: 0.
Plan sponsor (name): Republic Technologies International (USWA plan);
Results of §4044 asset allocation: 100% of PC3;
Results of §4022(c) recovery allocation[A]: 0.
Plan sponsor (name): US Airways Inc. (pilots' plan);
Results of §4044 asset allocation: 100% of PC3;
Results of §4022(c) recovery allocation[A]: 0.19% of PC5.
Plan sponsor (name): Weirton Steel;
Results of §4044 asset allocation: 70.83% of PC3;
Results of §4022(c) recovery allocation[A]: 0.
Source: Based on PBGC data on selected large, complex plans from among
those listed in appendix VI with completed financial valuations, as of
February 2009.
[A] The amount of assets allocated as recoveries under § 4022(c) is
determined differently, depending on the amount of the plan's unfunded
nonguaranteed benefits. If that amount exceeds $20 million, the total
amount paid under § 4022(c) depends on PBGC's actual recoveries. For
example, this was the case for both Northwestern Steel and RTI-USWA,
and there were no recoveries to allocate. Alternatively, if the amount
of unfunded nonguaranteed benefits is less than $20 million, the total
amount paid under § 4022(c) depends on an average of PBGC's recoveries
over a 5-year period (known as the SPARR calculation). For example, the
SPARR calculation was used for both LTV Steel Mining and US Airways,
and a small amount of assets were allocated to priority category 5
benefits in each of these plans.
[End of table]
[End of section]
Appendix IV: Limits on Guaranteed Benefits from a Participant's
Perspective:
The statutory and regulatory limits on guaranteed benefits can be
difficult to understand for many participants. The following schematic
distills the application of these limits into a series of questions,
one for each type of limit: phase-in, accrued-at-normal, and maximum.
Figure 18: Understanding the Limits on Guaranteed Benefits:
[Refer to PDF for image: illustration]
Is the full amount of your benefit guaranteed?
1) Was your benefit increased in the last 5 years?
No: Go to #2;
Yes:
Your benefit is not likely to be fully guaranteed due to the “phase-in”
limit. The portion of a benefit increase that is guaranteed is reduced
for each year it was not in effect during the last 5 years.
2) Did you receive any supplemental benefits?
No: Go to #3;
Yes:
Your benefit is not likely to be fully guaranteed due to the “accrued-
at-normal” limit. Supplemental benefits that exceed the retirement
benefit provided at normal retirement age are not guaranteed.
3) Is your benefit amount greater than the maximum set by law for your
age at retirement and type of benefit?
No: Go to #4;
Yes:
Your benefit is not likely to be fully guaranteed due to the “maximum
guarantee” limit. The amount of guaranteed benefits is limited by an
amount set by law, and is lower for those retiring before age 65 or
with survivor benefit.
4) Your benefit is likely to be fully guaranteed.
Source: GAO analysis of ERISA, PBGC’s implementing regulations, and
related documents.
[End of figure]
[End of section]
Appendix V: Profiles of Three Large, Complex Plans:
We selected three terminated pension plans to profile as examples of
large, complex plans: Bethlehem Steel, RTI (USWA), and US Airways
(pilots) (see appendix VI). All three were among the 10 plans most
affected by the guarantee limits. In addition, both Bethlehem Steel and
RTI (but not US Airways) were among the 10 plans most affected by
processing delays and by overpayments.[Footnote 58]
Table 5: Profile of a Bethlehem Steel Terminated Pension Plan:
Plan sponsor: Bethlehem Steel Corporation.
Plan name: Pension Plan of Bethlehem Steel Corporation:
Date of plan termination: 12/18/2002.
Date of trusteeship: 4/30/2003.
Date plan valuation complete: Valuation not complete as of February
2009.
Number of individuals associated with the plan:151,991.
Number of cases (primary participants):92,924.
Total amount of unfunded nonguaranteed benefits: $537,500,000.
Context for termination:
In 2000, the principal activities of Bethlehem Steel were the
production, manufacture, and sale of a wide variety of steel mill
products, including hot rolled, cold rolled, and coated sheets; tin
mill products; carbon and alloy plates; specialty booms; carbon and
alloy bars; and large diameter pipe. Bethlehem Steel also had iron ore,
lake shipping, railroad and trucking operations, but steel products
accounted for 92 percent of its revenue in 2000. On October 15, 2001,
Bethlehem Steel filed for bankruptcy under Chapter 11 of the U.S.
Bankruptcy Code. The company entered into an agreement, which was
ratified on April 22, 2003, that provided for the sale of substantially
all of its assets to a third party. The company is no longer in
operation. Over the years, the Bethlehem Steel plan was the product of
a culmination of over 100 mergers, consolidations, and or spinoffs. At
the date of the plan's termination, PBGC and the Milliman Actuaries
identified 8 major component parts: Bethlehem Steel Corporation Steel,
Bethlehem Steel Corporation Shipbuilding, Bethlehem Steel Corporation
Railroad, 1976 Great Lakes Hourly, Bethlehem 1957 Hourly, Washington
Steel Hourly 1998, Bethlehem Lukens Plate Hourly - 1998, and Lukens
Salaried - 1998. Three of the components, the Bethlehem Steel
Corporation Steel, the Bethlehem Steel Corporation Shipbuilding and the
Bethlehem Steel Corporation Railroad, have separate hourly and salaried
plans.
Groups affected by guarantee limits: Data not available as of February
2009.
PBGC processing time (from trusteeship to issuance of benefit
determination letters):
* Number of cases taking 4 or more years (including cases pending as of
February 2009): 25,619;
Maximum: 5.8 years;
Minimum: 7.6 months;
Mean: 2.7 years;
Median: 2.3 years.
Overpayments (total amount, per case)[A]:
* Estimated total amount of overpayments for plan: $11.1 million;
Maximum: (N/A)[B];
Minimum: less than $1;
Mean: about $4,400;
Median: about $3,500.
Recoupments (monthly benefit reduction due to recoupment, per
individual):
* Number of cases with recoupments: 2,487;
Maximum: $660;
Minimum: less than $1;
Mean: $28;
Median: $23.
Appeals:
Cases with appeals: 186;
Appeals cases pending as of April 2009: 33;
Appeals closed: 153;
Closed appeals resulting in a higher benefit amount: 48;
Closed appeals resulting in a lower benefit amount: 1;
Closed appeals resulting in no change in benefits: 104.
Source: GAO analysis of PBGC data and documents.
[A] Data reliability issues prevented us from conducting a more
definitive analysis of total overpayment amounts. For a more detailed
discussion of these data limitations, see appendix I.
[B] Amount of maximum overpayment for this case, based on calculations
using data provided in PBGC's individual level data set: $58,903.
Amount of overpayment for this case, verified in case records: $99,032.
Discrepancy was due to a legal dispute between the participant and his
ex-wife (who had a qualified domestic relations order) and appeals that
were not yet resolved and entered into the system at the time we
obtained our data set.
[End of table]
Table 6: Profile of a Republic Technologies International Terminated
Pension Plan:
Plan sponsor: Republic Technologies International (RTI).
Plan name: Republic Technologies International LLC - USWA Defined
Benefit Plan.
Date of plan termination: 6/14/2002;
Date of trusteeship: 9/30/2003;
Date plan valuation complete: 12/6/2007;
Number of individuals associated with the plan: 6,929;
Number of cases (primary participants): 4,289;
Total amount of unfunded nonguaranteed benefits: $77,901,131.
Context for termination:
After Republic Technologies International, LLC (RTI), once a leading
domestic producer of special bar quality steel products, declared
bankruptcy and was unable to pay pension benefits, PBGC initiated an
involuntary termination of 4 plans administered by RTI. While in
bankruptcy proceedings, RTI agreed to sell a substantial portion of its
assets to a new company that intended to hire many of RTI's employees,
but that did not want responsibility for the pension plans. Two of the
plans covered participants represented by the United Steelworkers (USW)
and included provisions for shutdown benefits, which allow participants
who meet certain age and service requirements to receive an immediate
unreduced early retirement benefit. USW reached an agreement with RTI
to consider the sale of RTI's assets to the new company a "shutdown"
under the plans, thereby triggering the provisions for shutdown
benefits. Concerned about the impact of the proposed sale on the plans
and its own financial condition, PBGC sought to terminate the plan
prior to the asset sale. This earlier termination date would preclude
PBGC from having to pay the shutdown benefits, because participants
cannot earn additional benefits after a plan terminates. USW raised the
issue in court, seeking a termination date after the asset sale.
Although a lower court found in favor of USW, setting a termination
date after the asset sale, a circuit court overturned this decision and
established the termination date prior to the asset sale. As a result,
PBGC was not obligated to pay the shutdown benefits. However, during
the time between the lower and higher courts' decisions, PBGC began
determining participants' eligibility for shutdown benefits and paying
groups of individuals shutdown benefits. When the termination date
changed, PBGC stopped paying these individuals shutdown benefits and
sought repayment of the shutdown benefits already dispensed.
Groups affected by guarantee limits: Accrued-at-normal limit;
Number: 941;
Description: Groups with various types of supplemental benefits.
Groups affected by guarantee limits: Maximum limit;
Number: 11;
Description: (no description provided).
Groups affected by guarantee limits: Phase-in limit;
Number: 1,693;
Description: Groups affected by the following three changes:
1. The RTI settlement agreement effective 8/2/98: $60/60% phase-in;
2. The resolution relating to various retirement plan changes
(effective dates varied): $60/60% phase-in;
3. The restatement effective 9/8/98: $60/60% phase-in.
PBGC processing time (from trusteeship to issuance of benefit
determination letters):
* Number of cases taking 4 or more years (including cases pending as of
February 2009): 3,819;
Maximum: 5.3 years;
Minimum: 3.5 years;
Mean: 4.5 years;
Median: 4.6 years.
Overpayments (total amount, per case):[A]
* Estimated total amount of overpayments for plan: $13.4 million;
Maximum: about $56,700[B];
Minimum: less than $70;
Mean: about $8,900;
Median: about $5,600.
Recoupments (monthly benefit reduction due to recoupment, per
individual):
* Number of cases with recoupments: 1,508;
Maximum: $117;
Minimum: less than $1;
Mean: $28;
Median: $28.
Appeals:[C]
Cases with appeals: 375;
Appeals cases pending as of April 2009: 372;
Appeals closed: 3;
Closed appeals resulting in a higher benefit amount: 1;
Closed appeals resulting in a lower benefit amount: 0;
Closed appeals resulting in no change in benefits: 2.
Source: GAO analysis of PBGC data and documents.
[A] Data reliability issues prevented us from conducting a more
definitive analysis of total overpayment amounts. For a more detailed
discussion of these data limitations, see appendix I.
[B] Amount of maximum overpayment for this case, based on calculations
using data provided in PBGC's individual level data set: $56,745.
Amount of overpayment for this case, verified in case records: $56,772.
[C] According to PBGC officials, as of June 2009, there were 376 RTI
(USWA) appeals. Of these, 160 appeals were pending and 216 appeals had
been closed. Of the 216 appeals closed, one resulted in a higher
benefit amount and 215 resulted in no change in benefits.
[End of table]
Table 7: Profile of a US Airways Terminated Pension Plan:
Plan sponsor: US Airways Inc.
Plan name: Retirement Income Plan for Pilots of US Airways Inc.
Date of plan termination: 3/31/2003;
Date of trusteeship: 3/31/2003;
Date plan valuation complete: 8/3/2006;
Number of individuals associated with the plan: 8,990;
Number of cases (primary participants): 7,050;
Total amount of unfunded nonguaranteed benefits: $1,692,381,669.
Context for termination:
US Airways began as All American Aviation in May 1939 as the first
airmail service for many small western Pennsylvania and Ohio Valley
communities. It evolved through various mergers and name changes until
becoming US Airways in 1997, under the parent holding company of US
Airways Group, Inc. In 2002, it was the seventh largest US air carrier,
transporting passengers, property and mail on a network focused
primarily in the Northeast, with some international operations. Over a
9-month period in 2002, US Airways and US Airways Express served almost
47 million passengers. In 2000, a general economic downturn led US
Airways to propose a merger with United Air Lines, however, the merger
proposal was blocked. Subsequently, US Airways' weak economic position
was exacerbated by the events of September 11, 2001, and in March 2002,
a new management team was hired to formulate a restructuring plan. To
implement the plan, significant labor, aircraft lease, and vendor cost
concessions were negotiated. However, in August 2002, US Airways Group,
Inc., filed voluntary petitions in bankruptcy court seeking
reorganization relief under Chapter 11, and in January 2003, US Airways
notified PBGC of its intent to terminate its Pilots' Retirement Income
Plan (and seven other plans covering various employees of its eight
wholly-owned subsidiaries), citing "a serious funding shortfall.";
After US Airways gained the consent of the pilots' union (the Air Line
Pilots Association, International) to terminate the plan, the company
entered into an agreement with the PBGC setting March 31, 2003, as the
plan's termination date. Prior to the termination date, US Airways, as
the pretermination plan administrator, was responsible for revising
benefit payments under the plan from then-current levels to what it
estimated would be the amount of benefits that would be covered by plan
assets or guaranteed by the PBGC following termination. US Airways
completed these determinations and informed plan participants of their
estimated post-termination benefits by letters dated March 28, 2003.
Groups affected by guarantee limits: Accrued-at-normal limit;
Number: 51;
Description: Former Shuttle Plan participants who had a "non-level
benefit stream."
Groups affected by guarantee limits: Maximum limit;
Number: 5,171;
Description: Five groups were affected by the maximum limit:
1. Non-Shuttle deferreds. Participants who were not in pay status and
were not members of the Shuttle Plan prior to its 2/1/2000 merger into
US Airways (4,409 affected);
2. Shuttle deferreds. Participants who were not in pay status and were
members of the plan prior to the merger (59 affected);
3. Enhanced Retirement Incentive Program (ERIP) retirees. Participants
and beneficiaries who retired under this 1998 program (339 affected);
4. Non-ERIP retirees. Pre-termination participants and beneficiaries
other than the ERIP retirees (327 affected);
5. Total and Permanent Disability participants (37 affected).
Groups affected by guarantee limits: Phase-in limit;
Number: 175;
Description: Groups affected by the following four changes:
1. The merger of the Shuttle Plan into the US Airways Pilots' Plan
effective 2/1/2000: $60/60% phase-in (76 affected);
2. The cost-of-living adjustment increase for former Piedmont
participants effective at the beginning of each calendar year: $/%
varied (97 affected);
3. Increases in benefit limits under Internal Revenue Code section 415
and compensation limits under section 401(a)(17) effective more than 5
years before termination, but subsequent automatic increases subject to
the limit (1 affected);
4. Benefit increases provided by ERIP effective 5/1/1998: $80/80% phase
in (1 affected).
PBGC processing time (from trusteeship to issuance of benefit
determination letters):
* Number of cases taking 4 or more years (including cases pending as of
February 2009): 866;
Maximum: 5.8 years;
Minimum: 3.4 months;
Mean: 3.6 years;
Median: 3.5 years.
Overpayments (total amount, per case):[A]
* Estimated total amount of overpayments for plan: $1.0 million;
Maximum: about; $95,000[B];
Minimum: less than $20;
Mean: about $8,900;
Median: about $4,300.
Recoupments (monthly benefit reduction due to recoupment, per
individual):
* Number of cases with recoupments: 111;
Maximum: $930;
Minimum: less than $1;
Mean: $93;
Median: $48.
Appeals:
Cases with appeals: 871;
Appeals cases pending as of April 2009: 15;
Appeals closed: 856;
Closed appeals resulting in a higher benefit amount: 104;
Closed appeals resulting in a lower benefit amount: 1;
Closed appeals resulting in no change in benefits: 751.
Source: GAO analysis of PBGC data and documents.
[A] Data reliability issues prevented us from conducting a more
definitive analysis of total overpayment amounts. For a more detailed
discussion of these data limitations, see appendix I.
[B] Amount of maximum overpayment for this case, based on calculations
using data provided in PBGC's individual level data set: $95,090.
Amount of overpayment for this case, verified in case records: $95,218.
[End of table]
[End of section]
Appendix VI: Three Categories of Large, Complex Plans:
Table 8: Ten Plans Most Affected by Guaranteed Benefit Limits:
Plan sponsor (name): Delta Air Lines Inc.;
Total number of participants: 13,435;
Unfunded nonguaranteed benefits: $2,958,936,274;
Percentage of total unfunded nonguaranteed benefits
($8,522,175,078)[A]: 34.7%;
Cumulative percentage: 34.7%[B].
Plan sponsor (name): US Airways Inc. (pilots' plan);
Total number of participants: 7,050;
Unfunded nonguaranteed benefits: $1,692,381,669;
Percentage of total unfunded nonguaranteed benefits
($8,522,175,078)[A]: 19.9%;
Cumulative percentage: 54.6%.
Plan sponsor (name): United Air Lines Inc. (management, administrative,
and public contact plan);
Total number of participants: 46,645;
Unfunded nonguaranteed benefits: $744,800,000;
Percentage of total unfunded nonguaranteed benefits
($8,522,175,078)[A]: 8.7%;
Cumulative percentage: 63.3%[B].
Plan sponsor (name): LTV Steel Company Inc. (hourly plan);
Total number of participants: 68,124;
Unfunded nonguaranteed benefits: $672,467,408;
Percentage of total unfunded nonguaranteed benefits
($8,522,175,078)[A]: 7.9%;
Cumulative percentage: 71.2%.
Plan sponsor (name): Bethlehem Steel Corp.;
Total number of participants: 92,924;
Unfunded nonguaranteed benefits: $537,500,000;
Percentage of total unfunded nonguaranteed benefits
($8,522,175,078)[A]: 6.3%;
Cumulative percentage: 77.5%[B].
Plan sponsor (name): United Air Lines (flight attendants plan);
Total number of participants: 28,416;
Unfunded nonguaranteed benefits: $273,600,000;
Percentage of total unfunded nonguaranteed benefits
($8,522,175,078)[A]: 3.2%;
Cumulative percentage: 80.7%[B].
Plan sponsor (name): Weirton Steel;
Total number of participants: 9,757;
Unfunded nonguaranteed benefits: $205,022,166;
Percentage of total unfunded nonguaranteed benefits
($8,522,175,078)[A]: 2.4%;
Cumulative percentage: 83.1%.
Plan sponsor (name): National Steel Corporation (hourly plan);
Total number of participants: 10,433;
Unfunded nonguaranteed benefits: $149,076,504;
Percentage of total unfunded nonguaranteed benefits
($8,522,175,078)[A]: 1.7%;
Cumulative percentage: 84.9%.
Plan sponsor (name): RTI (USWA);
Total number of participants: 4,289;
Unfunded nonguaranteed benefits: $77,190,131;
Percentage of total unfunded nonguaranteed benefits
($8,522,175,078)[A]: 0.9%;
Cumulative percentage: 85.8%.
Plan sponsor (name): National Steel Corporation (retirement plan);
Total number of participants: 5,783;
Unfunded nonguaranteed benefits: $73,293,054;
Percentage of total unfunded nonguaranteed benefits
($8,522,175,078)[A]: 0.9%;
Cumulative percentage: 86.7%.
Source: GAO analysis of PBGC data and documents as of February 2009 for
participants in plans terminated and trusteed during fiscal years 2000
through 2008.
[A] Total calculated based on the 668 plans with data indicating a
balance of unfunded nonguaranteed benefits (for 362 plans, data
indicated there were no unfunded nonguaranteed benefits; and for 59
plans, the data were insufficient to do the analysis).
[B] Data on these plans are based on estimates as of February 2009, as
the financial valuations of the plans were not yet complete.
[End of table]
Table 9: Ten Plans Most Affected by Long Processing Times:
(Continued From Previous Page)
Plan sponsor (name): Bethlehem Steel Corp.;
Total number of participants: 92,924;
Number of participants with benefit determinations that took more than
4 years: 25,619;
Percentage of total number of participants with benefit determinations
that took more than 4 years (78,553): 32.6%;
Cumulative percentage: 32.6%.
Plan sponsor (name): LTV Steel Company Inc. (hourly plan);
Total number of participants: 68,124;
Number of participants with benefit determinations that took more than
4 years: 4,678;
Percentage of total number of participants with benefit determinations
that took more than 4 years (78,553): 6.0%;
Cumulative percentage: 38.6%.
Plan sponsor (name): Horizon NR, LLC;
Total number of participants: 4,722;
Number of participants with benefit determinations that took more than
4 years: 4,356;
Percentage of total number of participants with benefit determinations
that took more than 4 years (78,553): 5.5%;
Cumulative percentage: 44.1%.
Plan sponsor (name): Kaiser Aluminum and Chemical Corp. (hourly plan);
Total number of participants: 10,300;
Number of participants with benefit determinations that took more than
4 years: 4,201;
Percentage of total number of participants with benefit determinations
that took more than 4 years (78,553): 5.3%;
Cumulative percentage: 49.5%.
Plan sponsor (name): RTI (USWA plan);
Total number of participants: 4,289;
Number of participants with benefit determinations that took more than
4 years: 3,819;
Percentage of total number of participants with benefit determinations
that took more than 4 years (78,553): 4.9%;
Cumulative percentage: 54.3%.
Plan sponsor (name): J.A. Jones, Inc.;
Total number of participants: 5,514;
Number of participants with benefit determinations that took more than
4 years: 3,380;
Percentage of total number of participants with benefit determinations
that took more than 4 years (78,553): 4.3%;
Cumulative percentage: 58.6%.
Plan sponsor (name): Kaiser Aluminum and Chemical Corp. (salaried
plan);
Total number of participants: 5,299;
Number of participants with benefit determinations that took more than
4 years: 3,042;
Percentage of total number of participants with benefit determinations
that took more than 4 years (78,553): 3.9%;
Cumulative percentage: 62.5%.
Plan sponsor (name): Cone Mills Corporation;
Total number of participants: 6,365;
Number of participants with benefit determinations that took more than
4 years: 2,956;
Percentage of total number of participants with benefit determinations
that took more than 4 years (78,553): 3.8%;
Cumulative percentage: 66.3%.
Plan sponsor (name): Weirton Steel;
Total number of participants: 9,757;
Number of participants with benefit determinations that took more than
4 years: 2,741;
Percentage of total number of participants with benefit determinations
that took more than 4 years (78,553): 3.5%;
Cumulative percentage: 69.8%.
Plan sponsor (name): Reliance Insurance Company;
Total number of participants: 7,280;
Number of participants with benefit determinations that took more than
4 years: 2,366;
Percentage of total number of participants with benefit determinations
that took more than 4 years (78,553): 3.0%;
Cumulative percentage: 72.8%.
Source: GAO analysis of PBGC data and documents as of February 2009 for
participants in plans terminated and trusteed during fiscal years 2000
through 2008.
[End of table]
Table 10: Ten Plans Most Affected by Overpayments:
Plan sponsor (name): LTV Steel Company Inc. (hourly plan);
Total number of participants: 68,124;
Number of participants with recoupments: 4,442;
Percentage of total number of participants with recoupments (22,623):
19.6%;
Cumulative percentage: 19.6%.
Plan sponsor (name): Bethlehem Steel Corp.;
Total number of participants: 92,924;
Number of participants with recoupments: 2,487;
Percentage of total number of participants with recoupments (22,623):
11.0%;
Cumulative percentage: 30.6%.
Plan sponsor (name): Weirton Steel;
Total number of participants: 9,757;
Number of participants with recoupments: 1,997;
Percentage of total number of participants with recoupments (22,623):
8.8%;
Cumulative percentage: 39.5%.
Plan sponsor (name): RTI (USWA plan);
Total number of participants: 4,289;
Number of participants with recoupments: 1,508;
Percentage of total number of participants with recoupments (22,623):
6.7%;
Cumulative percentage: 46.1%.
Plan sponsor (name): National Steel Corporation (hourly plan);
Total number of participants: 10,433;
Number of participants with recoupments: 1,075;
Percentage of total number of participants with recoupments (22,623):
4.8%;
Cumulative percentage: 50.9%.
Plan sponsor (name): LTV Steel Company Inc. (salary plan);
Total number of participants: 13,450;
Number of participants with recoupments: 913;
Percentage of total number of participants with recoupments (22,623):
4.0%;
Cumulative percentage: 54.9%.
Plan sponsor (name): RTI (USS/KOBE plan);
Total number of participants: 2,299;
Number of participants with recoupments: 730;
Percentage of total number of participants with recoupments (22,623):
3.2%;
Cumulative percentage: 58.1%.
Plan sponsor (name): Outboard Marine;
Total number of participants: 9,744;
Number of participants with recoupments: 614;
Percentage of total number of participants with recoupments (22,623):
2.7%;
Cumulative percentage: 60.8%.
Plan sponsor (name): LTV Steel Mining Company, Inc.;
Total number of participants: 3,416;
Number of participants with recoupments: 548;
Percentage of total number of participants with recoupments (22,623):
2.4%;
Cumulative percentage: 63.3%.
Plan sponsor (name): Northwestern Steel & Wire (plan A);
Total number of participants: 3,576;
Number of participants with recoupments: 450;
Percentage of total number of participants with recoupments (22,623):
2.0%;
Cumulative percentage: 65.3%.
Source: GAO analysis of PBGC data and documents as of February 2009 for
participants in plans terminated and trusteed during fiscal years 2000
through 2008.
[End of table]
[End of section]
Appendix VII: Sample Benefit Determination Letter:
To illustrate the complexity of some benefit calculations, this
appendix provides a sample benefit determination letter that PBGC sent
to an RTI (USWA) participant, as well as a 5-page benefit statement and
a recoupment summary that were provided as attachments.
Pension Benefit Guaranty Corporation:
U.S. Government Agency:
PBGC/Insurance Operations Department:
Box 151750:
Alexandria VA 22315-1750:
November 04, 2008:
PBGC Case Number:
Plan Name: Republic Technologies International LLC - USWA Defined
Benefit Plan:
Dear:
This letter is about changes in the amount of your monthly payment from
the Pension Benefit Guaranty Corporation ("PBGC").
As you know, the PBGC assumed responsibility for the Republic
Technologies International LLC - USWA Defined Benefit Plan. Because
federal law limits the benefits the PBGC can pay, we are required to
conduct a comprehensive review of all benefits payable under the
pension plan. While this review is underway, the benefit amounts paid
by the PBGC represent our best estimate of the final benefit amount. We
have been paying you an estimated amount of $548.89. The PBGC calls
this the estimated benefit amount.
The review process is now complete. Under your pension plan and the
limits set by law, the monthly benefit amount that you are entitled to
receive from PBGC is $397.85. This is less than the estimated benefit
amount you have been receiving. In accordance with the provisions of
the plan your monthly benefit will change to $342.16 beginning
06/01/2012. PBGC is required to apply the legal limits to all benefits
you received after June 14, 2002, the date as of which your plan was
terminated. As a result, there is a $11932.16 overpayment balance that
PBGC must now recover. The PBGC charges no interest on this
overpayment. Please see the enclosed Recoupment Summary sheet for a
detailed history of your pension payments.
PBGC will recover this overpayment balance by making deductions in
future benefit payments. The new monthly amount of $358.06 that you
will receive until 05/01/2012 reflects a deduction of $39.79 from your
final benefit amount. The monthly amount of $307.94 that you will
receive from 06/01/2012 until 07/01/2037 reflects a deduction of $34.22
from your final benefit amount. After that date, monthly payments made
to you will be increased to your final benefit amount of $342.16. This
final benefit is the amount that the PBGC is legally allowed to pay
you. It was calculated by determining the benefit you are entitled to
under your plan and then applying the limits spelled out in federal
pension law.
Your benefit is paid in the form of a Joint and 50% Survivor Annuity
with a Surviving Spouse Benefit. The Joint and 50% Survivor Annuity
provides you with a reduced monthly benefit for the rest of your life.
Thereafter, your surviving beneficiary will receive 50% of your
benefit. The Surviving Spouse Benefit is payable if your spouse meets
the eligibility requirements defined by the pension plan.
The Surviving Spouse Benefit will provide your spouse with a benefit
equal to 50% of your basic benefit until age 60. After age 60, the
pension plan requires us to recalculate the benefit and reduce the
benefit 50% of the Social Security Widow's benefit.
The enclosed Benefit Statement and Question and Answer sheet offer more
detail on these calculations. If you have questions about your final
benefit or overpayment amount, please contact us. The PBGC benefit
specialists will be happy to answer any questions that you may have.
We understand that this is a complicated situation and you may have
questions. You may call our Customer Contact Center at 1 (800) 400-
7242, Monday through Friday, 8:00 a.m. - 7:00 p. ET. If you use a
TTY/TDD, call 1 (800)-877-8339, and ask the relay operator to call our
telephone number. Or, you may write to: PBGC/Benefits Administration
and Payment Department, P.O. Box 151750, Alexandria. VA 22315-1750.
Please include your Social Security number, PBGC case number, I and a
daytime telephone number.
This is PBGC's formal determination of your benefit. If you have a
question about how your benefit was calculated, please call us for an
explanation. If you still disagree with the benefit determination, you
have the right to appeal. Your appeal must state the specific reason
you believe the determination is wrong, and must be made in writing
within 45 days of the date of this letter. The enclosed pamphlet, Your
Right to Appeal, explains more about filing an appeal.
If you do not appeal this determination, we anticipate that your
benefit will be reduced as of 02/01/2009. If you choose to appeal, your
benefit will not change until the PBGC Appeals Board reaches a
decision.
Please keep this letter in your records for future reference.
Sincerely,
FBA Pension Benefit Analyst:
Field Benefit Administration:
Enclosure:
Your Right to Appeal:
PACS Calculation Sheet:
(Manual Insert) Benefit Statement:
[End of letter]
Privacy Act Data:
Benefit Statement:
10/31/2008 11:38 AM:
Page 1 of 5:
Republic Technologies International LLC-USWA DB PL:
PBGC Case Number:
Date of Plan Termination: June 14,2002.
Participant's Name:
Participant's Information:
Social Security Number:
Gender:
Date of Birth:
Credit Service Date: 02/03/1969;
Date of Termination of Employment: 02/28/2002;
Normal Retirement Date (NRD): 06/01/2014;
Summary of Participant's Benefit:
Actual Retirement Date (ARD): 03/01/2002;
Current Monthly Benefit: $548.89;
PBGC Monthly Benefit Before 06/01/2012: $397.85;
PBGC Monthly Benefit After 06/01/2012: $342.16;
PBGC Form of Annuity: 5-Year Certain Then Joint and 50% Survivor
Annuity;
PBGC Monthly Surviving Spouse Benefit Before Spouse Age 60: $198.92;
PBGC Monthly Surviving Spouse Benefit After Spouse Age 60: $97.31.
Your Current Monthly Benefit of $548.89 will change to benefits listed
above due to pension law limitations that apply to your benefit See the
Benefit Calculation section below for details.
Beneficiary's Information:
Beneficiary Name:
Social Security Number:
Date of Birth: 07/01/1954;
Participant's Pension Information for Benefit Calculation:
Years of Credited Service CS): 33.0833;
Benefit Rate (BR): 35.0000;
5-Year Certain End Date: 02/28/2007;
New Plan Temporary Monthly Supplement (SUPP): $200.00;
New Plan Supplement End Date: 05/31/2012;
Plant Code and Plant Location: 151-Massillon Hot Roll;
Retirement Type: 30 Year;
Benefit under Provisions Effective June 1, 1993 (The Old Plan):
(1) Net LTV DBP Benefit Payable at NRD as a Straight Life Annuity:
$381.57;
(2) RESI DCP Monthly Benefit Offset: $203.98;
(3) Plan Monthly Benefit:
$333334 x CS, Maximum 30 = $33.3334 x 30.0000 = $1,000.00;
(4) Net Plan Monthly Benefit at NRD as a 5-Year Certain Then Straight
Life Annuity: (3) - (2) - (1) = $1,000.00 - $203.98 - $381.57 =
$414.45;
(5) Adjustment Factor for Early Retirement: 0.3827;
(6) Adjustment Factor for Joint and Survivor Annuity: 0.8600;
(7) Plan Monthly Benefit at ARD as a Joint and 50% Survivor Annuity:
(4) x (5) a (6) = $414.45 x 0.3827 x 0.8600 = $136.40.
Because your benefit provisions under the Old Plan did not include the
New Plan's Surviving Spouse Benefit and 5 Year Certain and Continuous
features, we have to adjust this benefit to the same survivor benefit
and form of benefit in order to make correct comparisons (the
normalized benefit).
(8) Normalization factor for New Plan's Surviving Spouse Benefit and 5
Year Certain and Continuous Feature: 0.9256;
(9) Normalized Monthly Benefit at ARD: (7) x (8) = $136.40 a 0.9256 =
$126.25;
Benefit under Provisions Effective September 8, 1998 (The New Plan):
(10) Plan Monthly Benefit: BR x CS = 535.0000 x 33.0833 =
$1,157.92.
(11) Net Plan Monthly Benefit at NRD as a 5-Year Certain Then Straight
Life Annuity: (10)-(2)-(1)= $1,157.92 - $203.98 - $381.57 = $572.37.
(12) Adjustment Factor for Joint and Survivor Annuity: 0.8600.
(13) Plan Monthly Benefit at ARD: (11) x (12) = $572.37 x 0.8600 =
$492.24.
(14) Plan Monthly Benefit Before 03/012007: (11) + SUPP = $572.37 +
$200.00 = $772.37.
(15) Plan Monthly Benefit After 03/01/2007 and Before 06/01/2011:
(13) + SUPP = $492.24 + $200.00 = $692.24.
(16) Plan Monthly Benefit After 06/01/2011 and Before 06/01/2012:
(13) + SUPP = $492.24 + $200.00 = $692.24.
(17) Plan Monthly Benefit After 06/01/2012: Same as (13) = $492.24.
When plans have provisions for supplemental benefits before age 65, the
benefit is limited to the monthly benefit payable to you as a straight
life annuity at NRD (Age 65). This is called the PBGC Accrued-at-Normal
Limit. See the next line for application of this limit.
(18) PBGC Accrued-at-Normal Limit: Same as (11): $572.37.
(19) PBGC Monthly Benefit Before 03/01/2007: The Lesser of (18) and
(14) = The Lesser of $572.37 and $772.37 = $572.37.
(20) PBGC Monthly Benefit After 03/01/2007 and Before 06/01/2011: [The
Lesser of (18) and (15)] = (The Lesser of $572.37 and $692.24] =
$572.37.
(21) PBGC Monthly Benefit After 06/01/2011 and Before 06/01/2012: [The
Lesser of (18) and (16)] = [The Lesser of $572.37 and $692.24] =
$572.37.
(22) PBGC Monthly Benefit After 06/01/2012: Same as (17) = $492.24.
(23) PBGC Levelizing Factor for Benefit Payable After 06/14/2002 but
Before 06/01/2012: 0.4844.
Because your supplement is not payable for your lifetime, it is
necessary for the PBGC to convert the supplement to an equivalent
lifetime benefit (your level-life benefit). This conversion is
necessary to determined limitations on your benefit. Once all-
limitations have been applied to your level-life benefit, your PBGC
guaranteed benefit is then converted back to a lifetime benefit and
supplement.
(24) Levelized PBGC Monthly Benefit: [(21) - (22)] x (23) + (22) =
($572.37 - $492.241 x 0.4844 +$492.24] = $531.06.
Calculation of PBGC Benefit:
Your plan increased benefits within the 5 years before the plan
terminated, but PBGC only guarantees a part of the increase. For each
FULL year that the benefit increase was In effect, PBGC guarantees 20%
of the increase or $20, whichever is greater.
(25) Benefit Increase under Provisions Effective September 8, 1998:
(24) - (9) = $531.06 - $126.25 = $404.81.
(26) Phased-In Benefit Increase: The Lesser of (25) or [The Greater of
$60.00 or 60% x (25)] = The Lesser of $404.81 or [The Greater of $60.00
or $242.89] = $242.89/
(27) Leveled Monthly Benefit after Phase-in Limit: (9) + (26) = $126.25
+ $242.89 = $369.14.
(28) Guarantee Ratio: (27)/(24) = $369.14/$531.06 = 0.695100.
(29) PBGC Monthly Benefit Before 03/01/2007: (19) x (28) - $572.37 x
0.695100 = $397.85.
(30) PBGC Monthly Benefit After 03/01/2007 and Before 06/01/2011: (20)
x (28) = $572.37 x 0.695100 = $397.85.
(31) PBGC Monthly Benefit After 06/01/2011 and Before 06/01/2012: (21)
x (28) = $57237 x 0.695100 = $397.85.
(32) PBGC Monthly Benefit After 06/01/2012: (22) x (28) = $492.24 x
0.695100 = $342.16.
Privacy Act Data:
Recoupment Summary:
Plan Name: Republic Technologies International LLC - USWA Defined
Benefit Plan:
Case Number:
Participant:
Prepared by:
Social Security Number:
Date Prepared: November 4, 2008:
Cust Id:
IPS Document Type: Benefit Calculations:
Date of Plan Termination: 06/14/2002:
Overpayment Accrual Commencement Date: 06/14/2002:
Target Adjustment Date: 02/01/2009:
Present Value of Termination Benefit: $67,217.00:
Maximum Guaranteeable Monthly Benefit: $3,579.55:
Total Number of Months for Calculation: 421:
Recoupment Data:
Total Amount of Money Subject to Recoupment: $11,932.16:
Expected Number of Months over which Recoupment will Occur: 342:
Recoupment End Date: 07/01/2037:
Correct Entitlements Entered:
07/01/2002-05/01/2012: $397.85;
06/01/2012-12/01/2099: $342.16.
Actual Pay Entered/Used:
07/01/2002-01/01/2009: $548.89.
Recoupment Percent and Amount Information:
Pay Dates: 02/01/2009-03/01/2012;
Recoup. %: 10%;
Recoup. Amount: $39.79;
Reduced Pay Amount: $358.06.
Pay Dates: 06/01/2012-07/01/2037;
Recoup. %: 10%;
Recoup. Amount: $34.22;
Reduced Pay Amount: $307.94.
This calculation uses interest rotes that were updated more than 45
days before the Target Payment Date for this transaction. Per PBGC
regulation, this result is valid for payment.
System: (PACS) Payment Adjustment Calculation System Version 3.0.
[End of section]
Appendix VIII: Comments from the Pension Benefit Guaranty Corporation:
Pension Benefit Guaranty Corporation:
Office of the Director:
1200 K Street, NW:
Washington, DC 20005-4026:
July 31, 2009:
Barbara D. Bovbjerg, Director:
Education, Workforce and Income Security Issues:
U.S. Government Accountability Office:
Washington, DC 20548:
Dear Ms. Bovbjerg:
Thank you for the opportunity to comment on the draft GAO report
entitled, Pension Benefit Guaranty Corporation: More Strategic Approach
Needed for Processing Complex Plans Prone to Delays and Overpayments
(GAO-09-716). GAO's efforts in performing this important work are
certainly appreciated. We welcome the report's acknowledgement of the
many positive comments about PBGC made by plan participants, and its
conclusion that PBGC is able to timely complete and issue benefit
determination letters within three years for most plans. We agree that
more can be done when PBGC terminates large, complex plans, and are
committed to doing so. PBGC is in general agreement with the report's
recommendations, and we look forward to updating your office as we make
progress toward their implementation.
PBGC guarantees the basic pension benefits covering 44 million
individuals in more than 30,000 private-sector defined benefit plans.
Since 1974, when Congress created PBGC to guarantee payment of defined
benefit pensions, more than 1.1 million workers and retirees in 3,860
terminated pension plans, and more than 120,000 participants in
multiemployer plans receiving financial assistance, have relied on PBGC
for their retirement income. In Fiscal Year 2008 alone, PBGC paid about
$4.3 billion to more than 64,000 participants in terminated pension
plans.
PBGC pays benefits according to the provisions of each individual
pension plan within statutory limits and ensures that benefits remain
uninterrupted for those already receiving benefits when a plan is
terminated. PBGC makes every effort to ensure that participants receive
what they are entitled to under their plan, and most participants of
plans taken over by PROC receive the full benefit due under their
respective plans. As the report notes, most plans are completed timely;
however, PBGC has terminated some plans that are particularly complex,
such as plans with multiple benefit formulas due to mergers, or which
have data-quality issues, and some that involve more than 100,000
participants. In the time period studied by GAO, there have been 10
such large plans. In order to better address issues associated with
these plans, PBGC recently created a Business Process Management Unit
reporting directly to the Chief Operating Officer. This group has
coordinated the performance of business-process engineering efforts,
leveraging Lean principles, to streamline our processes. Multiple
processes have been reengineered in both BAPD and the Appeals Division
since 2008. We look forward to further improving our existing
processes, so that PBGC is able to maintain its strong rating from the
American Customer Satisfaction Index. For FY 2008, PBGC received a
score of 88 in customer satisfaction of retirees receiving benefits
from PBGC - one of the highest scores received by a federal government
entity.
In the paragraphs below, we provide our specific responses to the
report's recommendations, which we expect to implement over the next
fiscal year:
Recommendation:
PBGC should set goals for timeliness and monitor the progress made in
finalizing benefit determinations for large complex plans separately
drum other plans.
Response:
We agree. PBGC has already started to implement steps for tracking and
monitoring the tasks associated with case-processing activities for
large, complex plans. For example, PBGC today uses detailed project-
management software to track timeliness and monitor progress for large,
complex plans.
Recommendation:
To reduce the number and size of overpayments in large complex plans,
PBGC should prioritize the calculation of estimated benefits For
retirees subject to statutory limits and adjust estimates,
as needed, throughout the benefit determination process. To reduce
increased overpayments due to appeals, PBGC should prioritize the
processing of appeals for those already receiving benefits and should
consider implementing the final benefit determination for retirees
during the appeals process.
Response:
We agree. PBGC generally does already identity and prioritize those
cases where adjustment may be likely, and makes estimated adjustments
accordingly. Given the special nature of these large, complex plans,
PBGC is intent on focusing keenly on them as we move forward.
Certainly, we will continue to look for ways to improve our processes
in order to minimize the impact of overpayments on participants
affected by the statutory limitations. As to the second part of this
recommendation, we have legal concerns about the effect that
implementing the final benefit determination before the completion of
the administrative appeals process would have on our requirement that
participants exhaust PBGC's administrative review procedure before
resorting to Court. These concerns arise from Section 704 of the
Administrative Procedure Act, which permits an agency to require
exhaustion of its administrative appeal procedures as a prerequisite to
a lawsuit only where the action that is subject to administrative
appeal "meanwhile is inoperative." Nonetheless, as noted above, we are
willing to explore options for making earlier adjustments in
appropriate cases, while seeking to avoid subjecting participants to
multiple revisions of their benefit payment. Finally, we note that, as
two thirds of our appeals are from participants already receiving
benefits, we are substantially in compliance with the balance of the
recommendation.
Recommendation:
PBGC should develop improved procedures for adapting and reviewing
letters to participants in large complex plans, such as by (1)
providing more specific information in letters to participants who
receive benefit reductions describing which limits were applied and
why; (20 ensuring all letters to participants involving benefit
reductions are reviewed for accuracy and coherence before being sent;
and (3) establishing processes to more frequently communicate with
participants who are experiencing delays in receiving final benefit
determinations.
Response:
We agree. Regarding the first part of your recommendation, PBGC has
just completed a six month effort to improve our letters and is near
complete in revising the guidelines for our benefit statements to
better communicate the complexities of PBGC benefits and to better
manage expectations of plan participants. On the second part of your
recommendation, we will evaluate and make necessary modifications to
our process for reviewing benefit determination letters before issuance
to ensure that letters to participants involving benefit reductions are
accurate and coherent. On the third part of your recommendation, PBGC
is committed to researching and establishing cost-effective processes
to more frequently--and more clearly--communicate with participants who
may be experiencing delays in receiving final benefit determinations.
Recommendation:
PBGC should provide information or resources to help participants in
large complex plans better understand their benefit calculations and
also to avoid any unnecessary appeals. Specifically, PBGC's benefit
determination letters should provide information, such as how
participants can obtain additional information by using the Freedom of
Information Act or other resources.
Response:
We agree. PBGC has already established working groups to examine
several aspects of our communications with participants, and we
certainly want to help them avoid unnecessary appeals. While we are
unsure that encouraging participants to obtain the wealth of technical
information in their files will clarify their understanding of their
benefit determination letter, and may create unnecessary Freedom of
Information Act (FOIA) requests, we are willing to be more proactive in
informing participants about the availability of FOIA in connection
with review of their benefit determinations. Accordingly, we will amend
the appeals brochure to include information about the FOIA process and
a link to our FOIA page on our web site.
PBGC continues to press forward in addressing the issues related to our
most complex plans. By way of update, as of June 30, 2009 (the end of
the third quarter of FY09), there were 376 RTI-USWA cases with appeals.
Of these 376 appeals opened, only 160 appeals remain pending, and 216
of the appeals are now closed.
Your efforts and those of your staff in preparing this important review
are valued. Again, thank you for the opportunity to comment.
Sincerely,
Signed by:
Vincent K. Snowbarger:
Acting Director:
[End of section]
Appendix IX: GAO Contact and Staff Acknowledgments:
GAO Contact:
Barbara D. Bovbjerg, (202) 512-7215 or bovbjergb@gao.gov:
Staff Acknowledgments:
In addition to the contact named above, Blake L. Ainsworth, Assistant
Director; Margie K. Shields, Analyst-in-Charge; Kristen W. Jones; and
Wayne Turowski made significant contributions to this report. Joseph A.
Applebaum, Jeffrey L. Bernstein, Susan C. Bernstein, Jena Y. Sinkfield,
Walter K. Vance, and Craig H. Winslow also made important
contributions.
[End of section]
Related GAO Products:
Pension Benefit Guaranty Corporation: Financial Challenges Highlight
Need for Improved Governance and Management, [hyperlink,
http://www.gao.gov/products/GAO-09-702T] (Washington, D.C.: May 20,
2009).
Auto Industry: Summary of Government Efforts and Automakers'
Restructuring to Date, [hyperlink,
http://www.gao.gov/products/GAO-09-553] (Washington, D.C.: April 23,
2009).
Defined Benefit Pensions: Survey Results of the Nation's Largest
Private Defined Benefit Plan Sponsors, [hyperlink,
http://www.gao.gov/products/GAO-09-291] (Washington, D.C.: March 30,
2009).
Pension Benefit Guaranty Corporation: Improvements Needed to Address
Financial and Management Challenges, [hyperlink,
http://www.gao.gov/products/GAO-08-1162T] (Washington, D.C.: September
24, 2008).
Pension Benefit Guaranty Corporation: Some Steps Have Been Taken to
Improve Contracting, but a More Strategic Approach Is Needed,
[hyperlink, http://www.gao.gov/products/GAO-08-871] (Washington, D.C.:
August 18, 2008).
Defined Benefit Pensions: Plan Freezes Affect Millions of Participants
and May Pose Retirement Income Challenges, [hyperlink,
http://www.gao.gov/products/GAO-08-817] (Washington, D.C.: July 21,
2008).
Pension Benefit Guaranty Corporation: A More Strategic Approach Could
Improve Human Capital Management, [hyperlink,
http://www.gao.gov/products/GAO-08-624] (Washington, D.C.: June 12,
2008).
Pension Benefit Guaranty Corporation: Governance Structure Needs
Improvements to Ensure Policy Direction and Oversight, [hyperlink,
http://www.gao.gov/products/GAO-07-808] (Washington, D.C.: July 6,
2007).
Private Pensions: Timely and Accurate Information Is Needed to Identify
and Track Frozen Defined Benefit Plans, [hyperlink,
http://www.gao.gov/products/GAO-04-200R] (Washington, D.C.: December
17, 2003).
Pension Benefit Guaranty Corporation Single-Employer Insurance Program:
Long-Term Vulnerabilities Warrant "High Risk" Designation, [hyperlink,
http://www.gao.gov/products/GAO-03-1050SP] (Washington, D.C.: July 23,
2003).
Pension Plans: Benefits Lost When Plans Terminate, [hyperlink,
http://www.gao.gov/products/T-HRD-92-58] (Washington, D.C.: September
24, 1992).
[End of section]
Footnotes:
[1] PBGC Office of Inspector General, The Length of Time It Has Taken
PBGC To Issue Initial Determination Letters, 99-3/23128-2, (Washington,
D.C.: March 1999); PBGC Office of Inspector General, Improvements Are
Needed To Achieve Better Efficiency And Effectiveness In PBGC's Benefit
Determination Process, 99-2/23128-1 (Washington, D.C.: March 1999);
PBGC Office of Inspector General, Pension Plan Participants Impacted by
Delays In Initial Determination Letter Issuance, 99-1/23128-3
(Washington, D.C.: March 1999); and PBGC Office of Inspector General,
Audit of PBGC's Response to Certain Questions Concerning Appeals of
PBGC Initial Determinations of Pension Benefits, 98-10/23131
(Washington, D.C.: September 1998).
[2] PBGC administers two separate insurance programs: a single-employer
program and a multiemployer program. This report focuses solely on
plans in PBGC's single-employer program, as PBGC will provide
assistance, but will not trustee multiemployer plans that are unable to
pay guaranteed benefits when due.
[3] Pub. L. No. 93-406, 88 Stat. 829 (codified, as amended, at 29
U.S.C. §§ 1001-1461).
[4] GAO, Pension Benefit Guaranty Corporation Single-Employer Insurance
Program: Long-Term Vulnerabilities Warrant "High Risk" Designation,
[hyperlink, http://www.gao.gov/products/GAO-03-1050SP] (Washington,
D.C.: July 23, 2003).
[5] GAO, Pension Benefit Guaranty Corporation: Governance Structure
Needs Improvements to Ensure Policy Direction and Oversight,
[hyperlink, http://www.gao.gov/products/GAO-07-808] (Washington, D.C.:
July 6, 2007).
[6] GAO, Pension Benefit Guaranty Corporation: Some Steps Have Been
Taken to Improve Contracting, but a More Strategic Approach Is Needed,
[hyperlink, http://www.gao.gov/products/GAO-08-871] (Washington, D.C.:
August 18, 2008); and GAO, Pension Benefit Guaranty Corporation: A More
Strategic Approach Could Improve Human Capital Management, [hyperlink,
http://www.gao.gov/products/GAO-08-624] (Washington, D.C.: June 12,
2008).
[7] At least one of the following criteria must be met in order for
PBGC to approve a distress termination filing: (1) liquidation in
bankruptcy (Chapter 7) or insolvency proceedings; (2) reorganization in
bankruptcy (Chapter 11); (3) a company will be unable to continue to
stay in business unless its plan is terminated; or (4) unreasonable,
burdensome pension costs caused solely by a decline in workforce. 29
U.S.C. § 1341(c)(2)(B).
[8] PBGC may initiate involuntary terminations for several reasons,
including if PBGC's loss from that plan may be expected to increase
unreasonably if the plan is not terminated. 29 U.S.C. § 1342(a).
[9] The process for determining how the plan's assets are distributed
among the plan's participants is detailed in ERISA. For a description
of the allocation process, see appendix III.
[10] For a description of guarantee limits from a participant's
perspective, see appendix IV.
[11] The termination date is set by either the plan sponsor or PBGC,
depending on which party initiates the plan termination, except in
cases where the plan sponsor has filed for bankruptcy. In cases of
Chapter 7 or Chapter 11 bankruptcy, the Pension Protection Act of 2006
required that the date on which a plan sponsor files for bankruptcy be
used as the date of plan termination for purposes of determining the
amount of guaranteed benefits and the allocation of assets. Pub. L. No.
109-280, § 404(a), 120 Stat. 280, 928 (codified at 29 U.S.C. §
1322(g)). If the plan sponsor and PBGC disagree over the date of
termination, a court determines the date.
[12] GAO has issued several reports on the related topic of "plan
freezes," which limit some or all future pension accruals for some or
all plan participants. For example, see GAO, Defined Benefit Pensions:
Survey Results of the Nation's Largest Private Defined Benefit Plan
Sponsors, [hyperlink, http://www.gao.gov/products/GAO-09-291]
(Washington, D.C.: Mar. 30, 2009); GAO, Defined Benefit Pensions: Plan
Freezes Affect Millions of Participants and May Pose Retirement Income
Challenges, [hyperlink, http://www.gao.gov/products/GAO-08-817]
(Washington, D.C.: July 21, 2008); GAO, Private Pensions: Timely and
Accurate Information Is Needed to Identify and Track Frozen Defined
Benefit Plans, [hyperlink, http://www.gao.gov/products/GAO-04-200R]
(Washington, D.C.: Dec. 17, 2003); and GAO, Pension Plans: Benefits
Lost When Plans Terminate, [hyperlink,
http://www.gao.gov/products/T-HRD-92-58] (Washington, D.C.: Sept. 24,
1992).
[13] Throughout this report, the term "participant" refers to the
primary participant (that is, the individual who earned the pension),
or if deceased, the beneficiary. For a more detailed explanation of our
methodology, see appendix I.
[14] PBGC Office of Inspector General, Pension Plan Participants
Impacted by Delays In Initial Determination Letter Issuance, 99-1/
23128-3 (Washington, D.C.: March 1999).
[15] For a description of the allocation process, see appendix III.
[16] A controlled group is a group of businesses under common control
that is treated as a single employer for the purposes of determining
the extent of employer liability.
[17] As noted previously (see background section), PBGC's 2006 study of
these limits found that participants in steel and airline industry
plans are often subject to the guarantee limits because these plans
provide generous benefits and allow their participants to retire at
relatively young ages.
[18] In a more recent plan termination, PBGC did not provide shutdown
benefits for two RTI plans (USWA and USS/KOBE). These plans'
termination dates--which were ultimately decided in court and set prior
to the shutdown date--precluded some participants from qualifying for
shutdown benefits. (For a more detailed discussion of the RTI plan
terminations, see appendix V). However, in accordance with the Pension
Protection Act of 2006, PBGC's guarantee of shutdown benefits is now
phased in from the date of plan shutdown. If this requirement had been
in place at the time of the RTI case, the date of plan termination
would not have been significant for determining benefits. § 403, 120
Stat. 928.
[19] Participants may be subject to more than one type of limitation,
so numbers cannot be added together to determine a total.
[20] 29 U.S.C. §§ 1302, 1303, and 1310.
[21] Of the cases with the lengthiest processing times, the RTI plan
was the only one with more than 1 year between termination and
trusteeship. PBGC officials regarded the circumstances of this case--
specifically, the protracted legal battle over the plan's termination
date--to be rare. Moreover, the circumstances giving rise to this
debate are unlikely to be repeated due to changes made by the Pension
Protection Act of 2006, which provides for the phase-in of PBGC's
guarantee from the date of a plant shutdown, and for the date on which
a sponsor files for bankruptcy to be used as the date of plan
termination for the purposes of determining the amount of guaranteed
benefits and the allocation of assets. §§ 403 and 404, 120 Stat. 928
and 928.
[22] Data reliability issues prevented us from conducting a more
definitive analysis of total overpayment amounts. For a more detailed
discussion of these data limitations, see appendix I.
[23] PBGC regulations generally limit benefit reductions to the greater
of (a) 10 percent of the participant's monthly benefit, or (b) the
amount in excess of the participant's "maximum guaranteeable benefit."
29 C.F.R. § 4022.82 (2009). In addition, the regulation provides that
PBGC may use its discretion to recoup overpayments by other methods. 29
C.F.R. § 4022.81(a) (2009). According to the PBGC Operating Policy
Manual, these other methods include situations when there is a delay in
implementing the recoupment; when the overpayment was due to (a) the
participant providing false information, (b) a failure to pay an
alternate payee under a qualified domestic relations order, or (c) a
change in disability benefits; or when the recoupment was initiated by
the plan sponsor prior to termination. In these situations, the amount
of the reduction may be increased to 25 percent, 40 percent, or some
other percentage of the benefit, and, in certain situations, the entire
benefit may be suspended.
[24] Four of these 10 plans also were among the 10 plans with the
greatest number of benefit determinations taking 4 years or longer to
process (see figure 6).
[25] Bethlehem Steel could not be included in the table because the
valuation of that plan was not complete as of February 2009, and the
number of participants subject to guarantee limits had not yet been
documented.
[26] For a description of the allocation process, see appendix III.
[27] 29 U.S.C. § 1341(c)(3)(D)(ii)(IV). For example, although our
analysis of plan data suggests that several airline plans were among
those most affected by guarantee limits, no airline plans were among
those most affected by overpayments (see appendix VI). This is likely
due to the fact that these airlines plans were terminated at the
sponsor's request. For example, according to agency officials, the
Delta Air Lines plan and the US Airways pilots' plan were both distress
terminations at the sponsor's request, and both reduced participants'
benefits as of the date of plan termination, as required by law. Thus,
we found that although nearly 60 percent of the individuals associated
with the US Airways pilots' plan were affected by one or more of the
guarantee limits (see table 2), US Airways was not on the list of 10
plans most affected by overpayments. (For a more detailed discussion of
the US Airways plan termination, see appendix V.)
[28] The accrued-at-normal limit requires guaranteed benefits to be
calculated based on retirement at the plan's normal retirement age,
with no survivor benefits. (For more details, see the background
section, table 1.)
[29] Phase-in limits apply to benefit changes made during the 5 years
before plan termination. The term "$20 or 20 percent" means the larger
of either $20 per month or 20 percent of the benefit increase for each
full year the increase was in effect prior to termination. (For more
details, see the background section, table 1.)
[30] 29 U.S.C. § 1322(b).
[31] Data reliability issues prevented us from conducting a more
definitive analysis of total overpayment amounts. For a more detailed
discussion of these data limitations, see appendix I.
[32] Based on the Social Security Administration's Cohort Life
Expectancy Tables for the cohort age 65, as of January 1, 2009, males
are projected to live another 16.8 to 18.1 years on average, and
females another 19.2 to 20.6 years on average. PBGC uses different
tables that project slightly greater life expectancies. In 2009, these
tables projected healthy males to live another 19.3 years, and healthy
females to live another 21.8 years.
[33] PBGC has reserved the right to use its discretion to recoup
overpayments by methods other than those specifically set out in its
applicable regulation, and may decide not to recoup net overpayments
that it determines to be de minimis. 29 C.F.R. §§ 4022.81(a) and
4022.82(a)(4) (2009).
[34] In addition, in the last month that benefits are to be reduced to
repay an overpayment, PBGC policy allows the final monthly reduction
amount to be waived if the remaining balance due is less than the
normal monthly reduction amount. 29 C.F.R. § 4022.82(a)(5) (2009).
[35] If no benefits are being paid, SSA can recover the overpayment
from federal income tax refunds or from the person's wages and report
delinquencies to credit bureaus.
[36] According to SSA's Performance and Accountability Reports, in
fiscal year 2007, SSA detected $5.1 billion in overpayments and
collected $2.5 billion.
[37] 5 U.S.C. § 8346(b).
[38] According to OPM's annual report, in fiscal year 2008, OPM
identified $194.1 million in overpayments and recovered $165.9 million
(including amounts due from prior years).
[39] Prior to termination, plan sponsors are required to notify
participants if the plan is significantly underfunded and warn them
that if the plan is terminated, their benefits must be cut back to
ERISA levels as of the plan termination date. 29 U.S.C. § 1021.
[40] PBGC produces an annual newsletter for retirees and a biannual
newsletter for future retirees.
[41] For a list of these groups, see appendix II.
[42] However, if a participant applies to start benefit payments during
this time, communications would be exchanged between PBGC and the
participant about the participant's current status, eligibility, and
benefit amount, based on the requested retirement date.
[43] For a description of our random selection process, see appendix I.
[44] We reviewed letters only to determine if they accurately conveyed
information documented elsewhere in the files. We did not attempt to
verify PBGC calculations of benefit amounts.
[45] This triage approach was later formalized in PBGC's Rules for
Administrative Review of Agency Decisions. 29 C.F.R. § 4003 (2009).
[46] 29 C.F.R. §§ 4003.52 and 4003.54 (2009).
[47] More specifically, 27 appeals were closed because the appeal was
not submitted within the required time frame; 7 were closed because the
appellant failed to identify a specific error; and 3 were closed
because the appellant attempted to appeal an issue that cannot be
appealed through this process.
[48] When PBGC identifies an error through the appeals process in how
it interpreted plan provisions, it also typically adjusts the benefits
of all other similarly-situated participants accordingly; but,
generally only does so if it results in a higher benefit amount.
[49] The Administration on Aging currently funds six regional
counseling projects that provide core services in 27 states for
individuals who need help in understanding and enforcing their pension
and retirement savings plan rights.
[50] 5 U.S.C. § 552.
[51] Formerly the United Steelworkers of America (USWA); in 2005, the
union's name was officially changed to United Steel, Paper and
Forestry, Rubber, Manufacturing, Energy, Allied Industrial and Service
Workers International Union, but is commonly known as United
Steelworkers.
[52] 29 U.S.C. 1344.
[53] 29 U.S.C. 1322(c).
[54] Before 2006, for any plan that had due and unpaid employer
contributions, PBGC would value the collectible portion of these
contributions and include that value as part of the plan assets, as
determined through adjudication (for example, in bankruptcy) or
settlement. The Pension Protection Act of 2006 created a new mechanism,
similar to the recovery ratio used for purposes of section 4022(c),
that allows PBGC to assign a value to the collectible portion of these
contributions in order to complete the valuation of the plan more
quickly. Pub. L. No. 109-280, § 408(b)(2), 120 Stat. 780, 931-32
(codified at 29 U.S.C. § 1344(f)).
[55] Based on provisions in the Pension Protection Act of 2006, the
date that the sponsor files for bankruptcy is treated as the plan
termination date for purposes of determining the amount of guaranteed
benefits and allocating assets, if that date is after September 16,
2006 and the sponsor is still in bankruptcy when the plan actually
terminates. § 404, 120 Stat. 928 (codified at 29 U.S.C. §§ 1322(g) and
1344(e)).
[56] Omnibus Budget Reconciliation Act of 1987, Pub. L. No. 100-203, §
9312(b)(3)(A)(ii), 111 Stat. 1330, 1330-362--1330-363 (codified at 29
U.S.C. § 1322(c)).
[57] This average is known as the Small Plan Average Recovery Ratio
(SPARR). Before passage of the Pension Protection Act of 2006, this 5-
year period consisted of the 5 fiscal years before the fiscal year in
which termination was initiated for the plan in which benefits are
being determined. The Pension Protection Act of 2006 backed up the 5-
year period for plans with unfunded nonguaranteed benefits not
exceeding $20 million. Specifically, the 5-year period was changed to
consist of the 5 full fiscal years ending with the third fiscal year
before the fiscal year in which termination was initiated for the plan
in which benefits are being determined. § 408(a), 120 Stat. 931
(codified at 29 U.S.C. § 1322(c)(3)(B)(ii). This change allowed
processing for these plans to be expedited.
[58] US Airways (pilots) was 2nd on the list of plans most affected by
guarantee limits, but 18th on the list of plans most affected by
processing delays, and 27th on the list of plans most affected by
overpayments.
[End of section]
GAO's Mission:
The Government Accountability Office, the audit, evaluation and
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 GAO's Web site [hyperlink, http://www.gao.gov]. Each
weekday, GAO posts newly released reports, testimony, and
correspondence on its Web site. To have GAO e-mail you a list of newly
posted products every afternoon, go to [hyperlink, http://www.gao.gov]
and select "E-mail Updates."
Order by Phone:
The price of each GAO publication reflects GAO’s actual cost of
production and distribution and depends on the number of pages in the
publication and whether the publication is printed in color or black and
white. Pricing and ordering information is posted on GAO’s Web site,
[hyperlink, http://www.gao.gov/ordering.htm].
Place orders by calling (202) 512-6000, toll free (866) 801-7077, or
TDD (202) 512-2537.
Orders may be paid for using American Express, Discover Card,
MasterCard, Visa, check, or money order. Call for additional
information.
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:
Congressional Relations:
Ralph Dawn, Managing Director, dawnr@gao.gov:
(202) 512-4400:
U.S. Government Accountability Office:
441 G Street NW, Room 7125:
Washington, D.C. 20548:
Public Affairs:
Chuck Young, Managing Director, youngc1@gao.gov:
(202) 512-4800:
U.S. Government Accountability Office:
441 G Street NW, Room 7149:
Washington, D.C. 20548: