This is the accessible text file for GAO report number GAO-08-223 
entitled 'State and Local Government Retiree Benefits: Current Funded 
Status of Pension and Health Benefits' which was released on February 
28, 2008.

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 the Committee on Finance, U.S. Senate: 

United States Government Accountability Office:
GAO: 

January 2008: 

State and Local Government Retiree Benefits: 

Current Funded Status of Pension and Health Benefits: 

GAO-08-223: 

GAO Highlights: 

Highlights of GAO-08-223, a report to the Committee on Finance, U.S. 
Senate. 

Why GAO Did This Study: 

Pension and other retiree benefits for state and local government 
employees represent liabilities for state and local governments and 
ultimately a burden for state and local taxpayers. Since 1986, 
accounting standards have required state and local governments to 
report their unfunded pension liabilities. Recently, however, standards 
changed and now call for governments also to report retiree health 
liabilities. The extent of these liabilities nationwide is not yet 
known, but some predict they will be very large, possibly exceeding a 
trillion dollars in present value terms. 

The federal government has an interest in assuring that all Americans 
have a secure retirement, as reflected in the federal tax deferral for 
contributions to both public and private pension plans. Consequently, 
the GAO was asked to examine: 1) the key measures of the funded status 
of retiree benefits and 2) the current funded status of retiree 
benefits. GAO analyzed data on public pensions, reviewed current 
literature, and interviewed a range of experts on public retiree 
benefits, actuarial science, and accounting. 

What GAO Found: 

Three key measures help to understand different aspects of the funded 
status of state and local government pension and other retiree 
benefits. First, governments’ annual contributions indicate the extent 
to which governments are keeping up with the benefits as they are 
accumulating. Second, the funded ratio indicates the percentage of 
actuarially accrued benefit liabilities covered by the actuarial value 
of assets. Third, unfunded actuarial accrued liabilities indicate the 
excess, if any, of liabilities over assets in dollars. Governments have 
been reporting these three measures for pensions for years, but new 
accounting standards will also require governments to report the same 
for retiree health benefits. Because a variety of methods and actuarial 
assumptions are used to calculate the funded status, different plans 
cannot be easily compared. 

Currently, most state and local government pension plans have enough 
invested resources set aside to keep up with the benefits they are 
scheduled to pay over the next several decades, but governments 
offering retiree health benefits generally have large unfunded 
liabilities. Many experts consider a funded ratio of about 80 percent 
or better to be sound for government pensions. We found that 58 percent 
of 65 large pension plans were funded to that level in 2006, a decrease 
since 2000. Low funded ratios would eventually require the government 
employer to improve funding, for example, by reducing benefits or by 
increasing contributions. However, pension benefits are generally not 
at risk in the near term because current assets and new contributions 
may be sufficient to pay benefits for several years. Still, many 
governments have often contributed less than the amount needed to 
improve or maintain funded ratios. Low contributions raise concerns 
about the future funded status. For retiree health benefits, studies 
estimate that the total unfunded actuarial accrued liability for state 
and local governments lies between $600 billion and $1.6 trillion in 
present value terms. The unfunded liabilities are large because 
governments typically have not set aside any funds for the future 
payment of retiree health benefits as they have for pensions. 

Figure: Percentage of State and Local Pension Plans with Funded Ratios 
above or below 80 Percent: 

[See PDF for image] 

This figure is a stacked bar graph depicting the following data: 

Fiscal year: 1994; 
Funded ratio 80 percent or more: approximately 60%; 
Funded ratio less than 80 percent: approximately 40%. 

Fiscal year: 1996; 
Funded ratio 80 percent or more: approximately 68%; 
Funded ratio less than 80 percent: approximately 32%. 

Fiscal year: 2000; 
Funded ratio 80 percent or more: approximately 95%; 
Funded ratio less than 80 percent: approximately 5%. 

Fiscal year: 2001; 
Funded ratio 80 percent or more: approximately 93%; 
Funded ratio less than 80 percent: approximately 7%. 

Fiscal year: 2002; 
Funded ratio 80 percent or more: approximately 88%; 
Funded ratio less than 80 percent: approximately 12%. 

Fiscal year: 2003; 
Funded ratio 80 percent or more: approximately 82%; 
Funded ratio less than 80 percent: approximately 18%. 

Fiscal year: 2004; 
Funded ratio 80 percent or more: approximately 80%; 
Funded ratio less than 80 percent: approximately 20%. 

Fiscal year: 2005; 
Funded ratio 80 percent or more: approximately 68%; 
Funded ratio less than 80 percent: approximately 32%. 

Fiscal year: 2006; 
Funded ratio 80 percent or more: approximately 62%; 
Funded ratio less than 80 percent: approximately 38%. 

Source: GAO analysis of PFS, PENDAT data. 

[End of figure] 

What GAO Recommends: 

GAO is not making recommendations in this report. Experts on public 
benefits funding provided technical clarifications, which were 
incorporated as appropriate. 

To view the full product, including the scope and methodology, click on 
[hyperlink, http://www.GAO-08-223]. For more information, contact 
Barbara Bovbjerg at (202) 512-7215 or bovbjergb@gao.gov. 

[End of section] 

Contents: 

Letter: 

Results in Brief: 

Background: 

Key Measures of the Funded Status of Retiree Benefits Are 
Contributions, Funded Ratios, and Unfunded Liabilities of Individual 
Plans over Time: 

Most Public Pensions Have Assets to Pay Benefits over Several Decades, 
Though Contributions Vary, While Unfunded Liabilities for Retiree 
Health Are Significant: 

Concluding Observations: 

Agency Comments: 

Appendix I: Objectives, Scope, and Methodology: 

Related GAO Products: 

Tables: 

Table 1: Effective Dates for GASB Statements 43 and 45, Requiring 
Public Employers to Estimate Health Care Liabilities: 

Table 2: Normal Cost Calculations for Three Most Commonly Used 
Actuarial Cost Methods: 

Figures: 

Figure 1: Relationship among the Key Measures of the Funded Status: 

Figure 2: Division of the Current Value of Future Benefits among Time 
Periods: 

Figure 3: Percentage of State and Local Government Pension Plans with 
Funded Ratios above or below 80 Percent, by Fiscal Year: 

Figure 4: Percentage of State and Local Government Pension Plans for 
which Governments Contributed More or Less Than 100 Percent of the ARC, 
by Fiscal Year: 

Abbreviations: 

ARC: annual required contribution: 

AAL: actuarial accrued liability: 

ERISA: Employee Retirement Income Security Act: 

GASB: Governmental Accounting Standards Board: 

NASRA: National Association of State Retirement Administrators: 

PFS: Public Fund Survey: 

PBGC: Pension Benefit Guaranty Corporation: 

PPCC: Public Pension Coordinating Council: 

[End of section] 

United States Government Accountability Office:
Washington, DC 20548: 

January 29, 2008: 

The Honorable Max Baucus: 
Chairman: 
The Honorable Charles E. Grassley: 
Ranking Member: 
Committee on Finance: 
United States Senate: 

Nearly 20 million employees and 7 million retirees and dependents of 
state and local governments--including school teachers, police, 
firefighters, and other public servants--are promised pensions, and 
many are promised retiree health benefits. Many of these benefits are 
guaranteed by state law or contract and represent actuarial accrued 
liabilities[Footnote 1] for state and local governments and ultimately 
the taxpayer. Typically, pension benefits are paid from a fund made up 
of assets from employers' and employees' annual contributions and the 
investment earnings from those contributions. Such a fund has an 
unfunded liability when the actuarial value of assets is less than 
actuarial accrued liabilities. Accounting standards have called for 
state and local governments to report their unfunded pension 
liabilities since 1986. But accounting standards have only recently 
been established that call for reporting the size of unfunded retiree 
health liabilities. While few state and local governments have as yet 
officially reported these unfunded liabilities, some studies have 
estimated that they may exceed $1 trillion dollars nationwide in 
present value terms. Such estimates raise concerns about the fiscal 
challenges that state and local governments will face in the coming 
decades. As discussion of the unfunded liabilities of state and local 
governments has increased, questions have been raised by some about how 
to understand these amounts. 

State and local retiree benefits are not subject, for the most part, to 
the federal funding requirements that apply to pensions sponsored by 
private employers. Nevertheless, the federal government has an interest 
in assuring that all Americans have a secure retirement, as reflected 
in the federal tax deferral for contributions to both public and 
private pension plans. Given the concerns about unfunded liabilities 
for state and local retiree benefits, we are reporting on: 1) the key 
measures of the funded status of retiree benefits and 2) the current 
funded status of retiree benefits. 

To address these objectives, we reviewed literature and interviewed a 
range of experts and stakeholders, including national associations of 
state and local officials, labor unions, bond raters, and actuarial and 
accounting professionals, among others. To describe the funded status 
of state and local pension plans, we analyzed self-reported data from 
the Public Fund Survey (PFS) as well as surveys by the Public Pension 
Coordinating Council (PPCC).[Footnote 2] This report represents one of 
two recent reports on state and local government retiree benefits. The 
other report, State and Local Government Retiree Benefits: Current 
Status of Benefit Structures, Protections, and Fiscal Outlook for 
Funding Future Costs (GAO-07-1156), provides a descriptive overview of 
such benefits. 

We conducted our work in Washington, D.C.; New York; and Connecticut 
from July 2006 to January 2008 in accordance with generally accepted 
government auditing standards. 

Results in Brief: 

Three key measures help to understand different aspects of the funded 
status of state and local government retiree benefits. First, 
governments' annual contributions indicate the extent to which they are 
keeping up with the value of benefits as they are accumulating. Second, 
the funded ratio indicates the percentage of a plan's liabilities 
covered by its assets. Third, unfunded liabilities indicate the excess, 
if any, of liabilities over assets in dollars. Low funded ratios 
correspond to high unfunded liabilities and require larger future 
contributions to pay benefits, which may create future budget problems 
and means future generations will bear more of the cost. Governments 
have been reporting these funded status measures for pensions for 
years. However, new accounting rules will also call on governments to 
report the funded status of retiree health benefits in a similar 
manner, even though many have not made any contributions to build 
assets to cover liabilities. These funded status measures should be 
reviewed using several years of data because in some years fiscal 
pressures may encourage governments to choose other budget priorities. 
Also, the value of assets can fluctuate from year to year with changes 
in investment returns, so examining a single year of funding data can 
be misleading. Because governments use a variety of methods and 
actuarial assumptions to calculate the funded status, different plans 
cannot be easily compared. 

Currently, most state and local government pension plans have enough 
invested resources set aside to pay for the benefits they are scheduled 
to pay over the next several decades, but governments that offer 
retiree health benefits generally have large unfunded liabilities. Many 
experts consider a funded ratio of about 80 percent or better to be 
sound for state and local government pensions. According to the self-
reported PFS data, 58 percent of 65 large public pension plans were 
funded to that level in 2006, a decrease since 2000 when about 90 
percent of plans were so funded. While most plans' funding may be 
sound, a few plans have persistently reported low funded ratios. Low 
funded ratios will eventually require the government employer to 
improve funding, for example, by reducing benefits or by increasing 
contributions. Increasing contributions may require revenue increases 
or reductions in non-benefit spending. However, even for many plans 
with lower funded ratios, benefits are generally not at risk in the 
near term because current assets and new contributions may be 
sufficient to pay benefits for several years. Still, many governments 
have often contributed less than the amount needed to improve or 
maintain funded ratios. Low contributions raise concerns about the 
future funded status, and may shift costs to future generations. For 
retiree health benefits, various studies estimate that the total 
unfunded liability for state and local governments lies between $600 
billion and $1.6 trillion although the estimates are based on samples 
of governments that are not necessarily representative. The unfunded 
liabilities are large because state and local governments typically 
have not set aside any funds for future retiree health benefits in the 
way they have for pensions. Instead, their practice has been to pay for 
the retiree health benefits due in a given year from the revenues for 
that year, like many private employers. This financing approach can 
leave little flexibility for governments, and therefore may stress 
future budgets. As a result, as health care costs increase, governments 
may face even greater pressure to reduce benefits or increase revenues. 
However, our analysis shows that the annual amount paid for retiree 
health benefits is currently low compared to pensions, but growth of 
health costs will be faster and less predictable. 

The Internal Revenue Service and experts in the field provided 
technical comments, which we incorporated as appropriate. 

Background: 

State and local governments will likely face daunting fiscal challenges 
in the next few years, driven in large part by the growth in health-
related costs.[Footnote 3] Medicaid and health insurance for state and 
local employees and retirees make up a large share of such costs. In 
contrast, our analysis shows that state and local governments on 
average would need to increase pension contribution rates to 9.3 
percent of salaries--less than .5 percent more than the 9.0 percent 
contribution rate in 2006 to achieve healthy funding on an ongoing 
basis. 

With few exceptions, defined benefit pension plans still provide the 
primary pension benefit for most state and local workers. About 
90 percent of full-time state and local employees participated in 
defined benefit pension plans as of 1998.[Footnote 4] A defined benefit 
plan determines benefit amounts by a formula that is generally based on 
such factors as years of employment, age at retirement, and salary 
level.[Footnote 5] A few states offer defined contribution or other 
types of plans as the primary retirement instrument.[Footnote 6] In 
fiscal year 2006, state and local government pension systems covered 
18.4 million members and made periodic payments to 7.3 million 
beneficiaries, paying out $151.7 billion in benefits. 

Many state and local governments also offer retirees health care 
benefits--in addition to Medicare benefits provided by the federal 
government--the costs of which have been growing rapidly. One study 
estimated that state and local governments paid $20.7 billion in fiscal 
year 2004 for retiree health benefits. For retirees who are under age 
65 (that is, not yet Medicare-eligible), many state and local employers 
provide access to group health coverage with varying levels of employer 
contributions. As of 2006, 14 states did not contribute to the premium 
for this coverage, while 14 states picked up the entire cost, and the 
remainder fell somewhere in between. For virtually all state and local 
retirees age 65 or older, Medicare provides the primary coverage. Most 
state and local government employers provide supplemental coverage for 
Medicare-eligible retirees that covers prescription drugs.[Footnote 7] 

Financing of State and Local Retiree Benefits: 

Both government employers and employees generally make contributions to 
fund state and local pension benefits. States follow statutes 
specifying contribution amounts or determine the contribution amount 
each legislative session. However many state and local governments are 
statutorily required to make yearly contributions based either on 
actuarial calculations or according to a statutorily specified amount. 
For plans in which employees are covered by Social Security, the median 
contribution rate in fiscal year 2006 was 8.5 percent of payroll for 
employers and 5 percent of pay for employees, in addition to 6.2 
percent of payroll from both employers and employees to Social 
Security. For plans in which employees are not covered by Social 
Security, the median contribution rate was 11.5 percent of payroll for 
employers and 8 percent of pay for employees. 

Actuaries estimate the amount that will be needed to pay future 
benefits. The benefits that are attributable to past service are called 
the "actuarial accrued liabilities." (In this report, the actuarial 
accrued liabilities are referred to as "liabilities.") Actuaries 
calculate liabilities based on an actuarial cost method and a number of 
assumptions including discount rates and worker and retiree mortality. 
Actuaries also estimate the "actuarial value of assets" that fund a 
plan (in this report, the actuarial value of assets is referred to as 
"assets"). The excess of actuarial accrued liabilities over the 
actuarial value of assets is referred to as the "unfunded actuarial 
accrued liability" or "unfunded liability." Under accounting standards, 
such information is disclosed in financial statements. In contrast, the 
liability that is recognized on the balance sheet is the cumulative 
excess of annual benefit costs over contributions to the plan. Certain 
amounts included in the actuarial accrued liability are not yet 
recognized as annual benefit costs under accounting standards, as they 
are amortized over several years. 

In a typical defined benefit pension plan, employer and employee 
contributions are made to a specific fund from which benefits will be 
paid. The yearly contributions from employers and employees are 
invested in the stock market, bonds, and other investments. Unlike most 
pension plans, retiree health benefits have generally been financed on 
a pay-as-you-go basis. Pay-as-you-go financing means that state and 
local governments have not set aside funds in a trust reserved for 
future retiree health costs. Instead, governments pay for each year's 
retiree health benefits from the current year's budget. 

Oversight of State and Local Retiree Benefits: 

The federal government has an interest in the funded status of state 
and local government retiree pensions and health care, even though it 
has not imposed the same funding and reporting requirements as it has 
on private sector pension plans. State and local government pension 
plans are not covered by most of the substantive requirements, or the 
insurance program operated by the Pension Benefit Guaranty Corporation 
(PBGC), under the Employee Retirement Income Security Act of 1974 
(ERISA), which apply to most private employer benefit plans. Federal 
law generally does not require state and local governments to prefund 
or report on the funded status of pension plans or health care 
benefits.[Footnote 8] However, in order to receive preferential tax 
treatment, state and local pensions must comply with requirements of 
the Internal Revenue Code. In addition, the retirement income security 
of all Americans is an ongoing concern of the federal government. 

All states have legal protections for their pensions. The majority of 
states have constitutional provisions prescribing how pension trusts 
are to be funded, protected, managed, or governed. The remaining states 
have pension protections in their statutes or recognize legal 
protections under common law. Legal protections usually apply to 
benefits for existing workers or benefits that have already accrued; 
thus, state and local governments generally can change the benefits for 
new hires.[Footnote 9] In contrast to pensions, retiree health benefits 
generally do not have the same constitutional or statutory protections. 
Instead, to the extent retiree health benefits are legally protected, 
it is generally because they have been collectively bargained and are 
subject to current labor contracts. 

Since the 1980s, the Governmental Accounting Standards Board (GASB) has 
maintained standards for accounting and financial reporting for state 
and local governments. GASB operates independently and has no authority 
to enforce the use of its standards. Still, many state laws require 
local governments to follow GASB standards, and bond raters do consider 
whether GASB standards are followed. Also, to receive a "clean" audit 
opinion under generally accepted accounting principles, state and local 
governments are required to follow GASB standards. These standards 
require reporting financial information on pensions, such as 
contributions and the ratio of assets to liabilities. In contrast to 
pensions, the financial status of retiree health care benefits has 
generally not been reported or even estimated actuarially until 
recently. However, new GASB standards (Statements 43 and 45) call for 
employers to quantify and report on the size of retiree health care 
benefit liabilities. The new health care reporting standards are being 
phased in over time to give more time to smaller state and local 
government sponsors to generate estimates. Table 1 shows the respective 
GASB 43 and 45 effective dates, as well as to what type of entity each 
statement applies. 

Table 1: Effective Dates for GASB Statements 43 and 45, Requiring 
Public Employers to Estimate Health Care Liabilities: 

GASB 43: 
Applies to: Plans administered as trusts and multiple-employer plans 
that are not administered as trusts; 
Total annual revenues as of 1999: $100,000,000 or more: Applies for 
periods beginning after 12/15/05; 
Total annual revenues as of 1999: $10,000,000 - $99,999,999: Applies 
for periods beginning after 12/15/06; 
Total annual revenues as of 1999: Less than $10,000,000: Applies for 
periods beginning after 12/15/07. 

GASB 45: 
Applies to: All employers that provide retiree health benefits. 
Total annual revenues as of 1999: $100,000,000 or more: Applies for 
periods beginning after 12/15/05; 
Total annual revenues as of 1999: $10,000,000 - $99,999,999: Applies 
for periods beginning after 12/15/06; 
Total annual revenues as of 1999: Less than $10,000,000: Applies for 
periods beginning after 12/15/07. 

Source: GASB. 

[End of table] 

Key Measures of the Funded Status of Retiree Benefits Are 
Contributions, Funded Ratios, and Unfunded Liabilities of Individual 
Plans over Time: 

Understanding the financial health of pension plans can be confusing. 
To help clarify, we found that three measures are key to understanding 
pension plans' funded status. GASB standards require reporting all 
three of these measures. First, one can look at yearly contributions 
governments are making to their plans. Actuaries calculate yearly 
contribution amounts needed to maintain or improve the funded status of 
plans over time. Comparing this amount to the amount governments 
actually contribute indicates how well governments are keeping up with 
yearly funding needs. Two other measures, funded ratios and unfunded 
liabilities, both suggest the extent to which current assets can cover 
accrued benefits. These three measures should be viewed together and 
over time to get a complete picture of the funded status. The funded 
status measures of different plans cannot be compared to one another 
easily because different governments use different actuarial funding 
methods and assumptions to estimate them. 

Three Measures, Viewed in Relation to One Another over Time, Describe 
Funded Status: 

Some officials we interviewed expressed confusion about how to 
understand the funded status of public pension plans. State and local 
governments report a significant amount of information on funding, 
required by GASB standards. The media often report various measures of 
the funded status without explaining the meaning of the terms or 
without enough context. In addition, governments have been reporting 
these funded status measures for pensions for years. However, the new 
accounting rules will also call on governments to report the funded 
status of retiree health benefits in a similar manner, even though many 
have not made any contributions to build assets to cover liabilities. 

We identified three key measures to help explain plans' funded status: 
contributions, funded ratios, and unfunded liabilities. According to 
experts we interviewed, any single measure at a point in time may give 
a dimension of a plan's funded status, but it does not give a complete 
picture. Instead, the measures should be reviewed collectively over 
time to understand how the funded status is improving or worsening. For 
example, a strong funded status means that, over time, the amount of 
assets, along with future scheduled contributions, comes close to 
matching a plan's liabilities. 

Comparing governments' actual contributions to the "annual required 
contribution" (ARC) helps in evaluating the funded status of each plan. 
Each year, plan actuaries calculate a contribution amount that, if paid 
in full, would normally maintain or improve the funded status.[Footnote 
10] This amount is referred to as the ARC, although the use of the word 
"required" can be misleading because governments can choose to pay more 
or less than this amount.[Footnote 11] If the actuarial assumptions are 
consistent with the plans' future experience, paying the full ARC each 
year provides reasonable assurance that sufficient money is being set 
aside to cover currently accruing benefits as well as a portion of any 
unfunded accrued benefits left over from previous years, instead of 
leaving those costs for the future. In other words, when a government 
consistently pays the ARC, the benefits accrued by employees are paid 
for by the taxpayers who receive the employees' services. When the ARC 
is not paid in full each year, future generations must make up for the 
costs of benefits that accrued to employees in the past. In addition, 
the ARC can be compared to the government's yearly budget to understand 
the financial burden of the benefits, according to officials. This 
comparison indicates how affordable the plan is to the government in a 
given year. A high ARC relative to a government's budget may indicate 
that the costs of benefits are relatively high or that payments have 
been deferred from previous years. 

The funded ratio is the ratio of assets to liabilities. Liabilities are 
the amount governments owe in benefits to current employees who have 
already accrued benefits they will collect in the future. The funded 
ratio indicates the extent to which a plan has enough funds set aside 
to pay accrued benefits. If a plan has a funded ratio of 80 percent, 
the plan has enough assets to pay for 80 percent of all accrued 
benefits. A rising funded ratio over time indicates that the government 
is accumulating the assets needed to make future payments for benefits 
accrued to date. A low or declining funded ratio over time may raise 
concerns that the government will not have the assets set aside to pay 
for benefits. 

While the funded ratio equals the ratio of assets to liabilities, 
unfunded liabilities equal the difference between liabilities and 
assets in dollars. Thus, unfunded liabilities indicate the amount of 
benefits accrued for which no money is set aside. Assets may fall short 
of liabilities, for example, when governments do not contribute the 
full ARC, when they increase benefits retroactively, or when returns on 
investments are lower than assumed. Additionally, because all these 
financial calculations involve estimates of future payments, they are 
based on a number of assumptions about the future. Unfunded liabilities 
can grow if actuaries' assumptions do not hold true. For example, if 
beneficiaries live longer than anticipated, they will receive more 
benefits than predicted, even if the government has been paying the ARC 
consistently. Unfunded liabilities will eventually require the 
government employer to increase revenue, reduce benefits or other 
government spending, or do some combination of these. Revenue increases 
could include higher taxes, returns on investments, or employee 
contributions. Nevertheless, we found that unfunded liabilities do not 
necessarily imply that pension benefits are at risk in the near term. 
Current funds and new contributions may be sufficient to pay benefits 
for several years, even when funded ratios are relatively low. 

As described in figure 1, unfunded liabilities are calculated as 
intermediate steps in the process of calculating the ARC. After 
calculating the unfunded liabilities, actuaries usually determine an 
amount to fund the unfunded liabilities over several years or 
"amortize" the cost of the liability. That amortized portion is added 
to the cost of benefits that employees accrued in the current year to 
determine the ARC. If a government pays the ARC, then a portion of the 
unfunded liabilities is paid off each year. When no more unfunded 
liabilities exist, the funded ratio is 100 percent, and the plan has 
"fully funded" all the benefits that its current employees have accrued 
under the plan's actuarial cost method. However, a fully funded plan 
still requires yearly contributions to maintain full funding because as 
employees perform additional service, they accrue additional benefits. 

Figure 1: Figure 1. Relationship among the Key Measures of the Funded 
Status: 

[See PDF for image] 

This figure is an illustration of the relationship among the key 
measures of the Funded Status. The illustration depicts the following 
information: 

Funded ratio: 
Assets divided by liabilities (which include unfunded liabilities); 

Portion of unfunded liabilities to be paid off this year: 
ARC: cost of benefits accrued this year; 
Actual contribution: May be greater or less than ARC. 

Assets: = sum of past contributions from the state and local government 
plan sponsors, employees, and investment earnings that have not been 
paid out in benefits or administrative expenses. 

Liabilities = current cost of all future benefits that have been 
accrued to date. 

Source: GAO analysis; images partially by Art Explosion. 

[End of figure] 

The funded status measures should be reviewed over time because several 
factors can affect them. In particular, the money set aside is invested 
and returns can fluctuate. If a plan's invested assets grow at a rate 
significantly above or below the rate assumed for funding purposes in a 
given year, it can change the funded status measures, regardless of the 
government's contributions. Granting retroactive benefits also 
increases liabilities and increases unfunded liabilities, even if a 
government has been contributing the full ARC each year. Funded ratios 
and unfunded liabilities also can reflect changes in assumptions about 
member characteristics. For example, as plan members are projected to 
live in retirement longer, the estimated amount expected to be paid for 
future benefits rises. 

Comparing the Funded Status of Different Plans Is Difficult: 

Under GASB reporting standards, the funded status of different pension 
plans cannot be compared easily because governments use different 
actuarial approaches such as different actuarial cost methods, 
assumptions, amortization periods, and "smoothing" mechanisms. 

Actuarial Cost Methods: 

Most public pension plans use one of three "actuarial cost methods," 
out of the six GASB approves. Actuarial cost methods differ in several 
ways. First, each uses a different approach to calculate the "normal 
cost," the portion of future benefits that the cost method allocates to 
a specific year, resulting in different funding patterns for each, as 
described in Table 2. 

Table 2: Normal Cost Calculations for Three Most Commonly Used 
Actuarial Cost Methods: 

Actuarial cost Method: Projected unit credit; 
Description: Projected benefits of each employee covered by the plan 
are allocated by a consistent formula to valuation years; 
How the method calculates the normal cost for the current year: Equal 
to the current value of the future benefit that each employee earned 
this year, using the employee's projected salary at retirement as a 
base. 

Actuarial cost Method: Entry age normal; 
Description: The current value of future benefits of each employee is 
allocated on a level basis over the earnings or service of the employee 
between entry age and assumed exit age; 
How the method calculates the normal cost for the current year: Equal 
to the level percentage of payroll that would exactly fund each 
employee's prospective benefits if contributed from the member's date 
of eligibility until retirement. 

Actuarial cost Method: Aggregate; 
Description: The excess of the value of future benefits of all 
employees over the current value of assets is allocated on a level 
basis over the earnings or service of the group between the valuation 
date and assumed exit. This allocation is performed for the group as a 
whole, not as a sum of individual allocations; 
How the method calculates the normal cost for the current year: The 
percentage of payroll equal to the current value of future benefits 
minus assets, divided by the current value of future salaries. 

Sources: Actuarial Standards Board, Government Accountants Journal, 
Organization for Economic Cooperation and Development, American Academy 
of Actuaries. 

[End of table] 

The Aggregate Cost Method: 

Some news reports have expressed uncertainty about the use of the 
aggregate actuarial cost method, but experts indicated that the 
aggregate method is as sound as the other methods. Experts explained 
that under the aggregate method, unfunded liabilities are allocated as 
future normal costs instead of being amortized and added to the normal 
cost. As a result, no unfunded liabilities are reported, and the funded 
ratio is often reported as 100 percent and year-to-year payments may be 
more volatile. Relatively few plans actually employ the aggregate 
method. For those plans, GASB recently began to require governments to 
report the funded ratio using the entry age normal method. 

Actuarial cost methods are used to allocate the current value of future 
benefits into amounts attributable to the past, to the current year, 
and to future years, as shown in figure 2. The cost of future benefits 
that are attributable to past years under the actuarial cost method is 
called the actuarial accrued liability (AAL), while the cost of 
benefits accrued under the cost method in the current year is known as 
the normal cost. 

Figure 2: Division of the Current Value of Future Benefits among Time 
Periods: 

[See PDF for image] 

This figure is a chart that depicts the following information: 

Current value of future benefits: 
Can be divided into: 
* Actuarial accrued liability (AAL), benefits accrued in past years; 
* Normal cost, benefits accrued in the current year; 
* Future normal costs, benefits that will accrue in future years. 

Source: Paul Angelo, Fellow of the Society of Actuaries, and GAO. 

[End of figure] 

The funded status of plans using different cost methods differs because 
each has a different approach to dividing up the value of future 
benefits. Different cost methods are designed for plans to accrue 
liabilities at different rates, so the normal cost and the AAL vary 
according to the cost method. For example, under some cost methods, 
governments accrue more liabilities in the early part of employees' 
career rather than later. As a result, two identical plans, using 
identical actuarial assumptions but different cost methods, would 
report a different funded status.[Footnote 12] 

Some Call for Assuming Risk-Free Investment Returns: 

Some in the pension community have been advocating an alternative 
approach to measuring the funded status of public plans. Proponents of 
this approach point to certain implications of the field of "financial 
economics" that suggest that using the expected rate of return to 
project future fund earnings does not adequately take into account the 
risk inherent in some investments. They believe it is preferable, for 
disclosure purposes, that a plan's assets and liabilities be "marked to 
market." In particular, plan liabilities should be measured, 
independent of the actuarial cost method used for funding, as the cost 
of closing out the plan's accrued benefit obligations based on service 
to date. This implies using the cost of annuities or discounting the 
expected cash flows using a risk-free rate of return and would likely 
result in much less favorable funded status estimates. Further, they 
believe that using a "smoothed" value of assets rather than the market 
value of assets obscures the plan's risk profile and may have 
operational consequences as well. 

Most governments do not use risk-free return assumptions to calculate 
funded status. Most public plan actuaries believe that using this 
approach is inappropriate because their plans do invest in diversified 
portfolios with higher rates of returns than risk-free rates. Those 
higher returns are reasonable to expect, they feel, based on past 
experience and will decrease the contributions that would be required 
if assumed returns were lower. Their current practice, they argue, 
produces estimates of contributions that best reflect what will 
actually be required on average over the long term. Using a risk-free 
return assumption would result in higher current contribution rates, 
requiring current taxpayers to pay more for the cost of future 
benefits. 

Assumptions: 

In addition to the cost methods, differences in assumptions used to 
calculate the funded status can result in significant differences among 
plans that make comparisons difficult. One key assumption is the rate 
at which governments assume their invested assets will grow. If 
governments assume a high growth rate, their calculations will indicate 
that they do not have to pay as much today, because the assets set 
aside will grow more rapidly. In 2006, 70 percent of state and local 
government pension plans assumed a return of 8.0 to 8.5 percent, while 
30 percent assumed a lower rate of return (7 percent at the lowest). If 
a plan's assets fail to grow at the assumed rate of return, then the 
shortfall becomes part of the unfunded liabilities. However, in other 
years, assets may earn more than the assumed rate of return, reducing 
unfunded liabilities. 

Amortization Periods for Unfunded Liabilities: 

In addition to actuarial cost methods and assumptions, differences in 
amortization periods make it difficult to compare the funded status of 
different plans. Governments amortize unfunded liabilities to reduce 
the volatility of contributions from year to year. Governments can 
choose shorter or longer periods over which to amortize unfunded 
liabilities. GASB standards allow governments to amortize unfunded 
liabilities over a period of up to 30 years.[Footnote 13] State and 
local governments can amortize their benefits because there is little 
chance that they will cease to exist. 

Smoothing Periods: 

Finally, actuaries for many plans calculate the value of current assets 
based on an average value of past years. As a result, if the value of 
assets fluctuates significantly from year to year, the "smoothed" value 
of assets changes less dramatically. GASB does not limit the number of 
years governments may use to smooth the value of assets, but in 2006, 
most governments averaged the value of current assets with those of the 
last zero to 5 years. Comparing the funded status of plans that use 
different smoothing periods can be confusing because the value of the 
different plans' assets reflects a different number of years. Given 
fluctuations in the stock market from year to year, the reported value 
of assets for plans that use different numbers of years for smoothing 
calculations could reflect significantly different market returns. 

Most Public Pensions Have Assets to Pay Benefits over Several Decades, 
Though Contributions Vary, While Unfunded Liabilities for Retiree 
Health Are Significant: 

More than half of public pension plans reported that they have put 
enough assets aside in advance to pay for benefits over the next 
several decades, while governments providing retiree health benefits 
generally have significant unfunded liabilities. The percentage of 
pension plans with funded ratios below 80 percent, a level viewed by 
many experts as sound, has increased in recent years, and a few plans 
are persistently underfunded. Although members of these plans may not 
be at risk of losing benefits in the near term, the unfunded 
liabilities will have to be made up in the future. In addition, a 
number of governments reported not contributing enough to reduce 
unfunded liabilities, which can shift costs to future generations. For 
state and local governments' retiree health benefits, studies have 
estimated unfunded liabilities nationwide to be between $600 million 
and $1.6 trillion, although the amounts for individual governments vary 
widely. Even though annual costs for retiree health benefits are 
currently low compared to pensions, continuing to pay for current 
benefits with current revenues can put stress on government budgets 
because health care costs are increasing rapidly. 

Most Public Pension Plans Have Enough Funds to Pay for Benefits over 
the Long-Term: 

Most public pension plans report having sufficient assets to pay for 
retiree benefits over the next several decades. Many experts and 
officials to whom we spoke consider a funded ratio of 80 percent to be 
sufficient for public plans for a couple of reasons.[Footnote 14] 
First, it is unlikely that public entities will go bankrupt as can 
happen with private sector employers, and state and local governments 
can spread the costs of unfunded liabilities over up to 30 years under 
current GASB standards. In addition, several commented that it can be 
politically unwise for a plan to be overfunded; that is, to have a 
funded ratio over 100 percent. The contributions made to funds with 
"excess" assets can become a target for lawmakers with other priorities 
or for those wishing to increase retiree benefits. 

More than half of state and local governments' plans reviewed by the 
Public Fund Survey (PFS) had a funded ratio of 80 percent or better in 
fiscal year 2006, but the percentage of plans with a funded ratio of 80 
percent or better has decreased since 2000, as shown in figure 
3.[Footnote 15] Our analysis of the PFS data on 65 self-reported state 
and local government pension plans showed that 38 (58 percent) had a 
funded ratio of 80 percent or more, while 27 had a funded ratio of less 
than 80 percent. In the early 2000s, according to one study, the funded 
ratio of 114 state and local government pension plans together reached 
about 100 percent; it has since declined.[Footnote 16] In fiscal year 
2006, the aggregate funded ratio was about 86 percent. Some officials 
attribute the decline in funded ratios since the late 1990s to the 
decline of the stock market, which reduced the value of assets. This 
sharp decline would likely affect funded ratios for several years 
because most plans use smoothing techniques to average out the value of 
assets over several years. Our analysis of several factors affecting 
the funded ratio showed that changes in investment returns had the most 
significant impact on the funded ratio between 1988 and 2005, followed 
by changes in liabilities.[Footnote 17] 

Figure 3: Percentage of State and Local Government Pension Plans with 
Funded Ratios above or below 80 Percent, by Fiscal Year: 

[See PDF for image] 

Fiscal year: 1994; 
Funded ratio 80 percent or more: approximately 60%; 
Funded ratio less than 80 percent: approximately 40%. 

Fiscal year: 1996; 
Funded ratio 80 percent or more: approximately 68%; 
Funded ratio less than 80 percent: approximately 32%. 

Fiscal year: 2000; 
Funded ratio 80 percent or more: approximately 95%; 
Funded ratio less than 80 percent: approximately 5%. 

Fiscal year: 2001; 
Funded ratio 80 percent or more: approximately 93%; 
Funded ratio less than 80 percent: approximately 7%. 

Fiscal year: 2002; 
Funded ratio 80 percent or more: approximately 88%; 
Funded ratio less than 80 percent: approximately 12%. 

Fiscal year: 2003; 
Funded ratio 80 percent or more: approximately 82%; 
Funded ratio less than 80 percent: approximately 18%. 

Fiscal year: 2004; 
Funded ratio 80 percent or more: approximately 80%; 
Funded ratio less than 80 percent: approximately 20%. 

Fiscal year: 2005; 
Funded ratio 80 percent or more: approximately 68%; 
Funded ratio less than 80 percent: approximately 32%. 

Fiscal year: 2006; 
Funded ratio 80 percent or more: approximately 62%; 
Funded ratio less than 80 percent: approximately 38%. 

Source: GAO analysis of PFS, PENDAT data. 

[End of figure] 

Although most plans report being soundly funded in 2006, a few have 
been persistently underfunded, and some plans have seen funded ratio 
declines in recent years.[Footnote 18] We found that several plans in 
our data set had funded ratios below 80 percent in each of the years 
for which data is available. Of 70 plans in our data set, 6 had funded 
ratios below 80 percent for 9 years between 1994 and 2006. Two plans 
had funded ratios below 50 percent for the same time period. In 
addition, of the 27 plans that had funded ratios below 80 percent in 
2006, 15 had lower funded ratios in 2006 than in 1994. The sponsors of 
these plans may be at risk in the future of increased budget pressures. 

By themselves, lower funded ratios and unfunded liabilities do not 
necessarily indicate that benefits for current plan members are at 
risk, according to experts we interviewed. Unfunded liabilities are 
generally not paid off in a single year, so it can be misleading to 
review total unfunded liabilities without knowing the length of the 
period over which the government plans to pay them off. Large unfunded 
liabilities may represent a fiscal challenge, particularly if the 
period to pay them off is short. But all unfunded liabilities shift the 
responsibility for paying for benefits accrued in past years to the 
future. 

Some Pension Sponsors Do Not Contribute Enough to Improve Funding 
Status: 

A number of governments reported not contributing enough to keep up 
with yearly costs. Governments need to contribute the full ARC yearly 
to maintain the funded ratio of a fully funded plan or improve the 
funded ratio of a plan with unfunded liabilities. In fiscal year 2006, 
the sponsors of 46 percent of the 70 plans in our data set contributed 
less than 100 percent of the ARC, as shown in figure 4, including 39 
percent that contributed less than 90 percent of the ARC. In fact, the 
percentage of governments contributing less than the full ARC has risen 
in recent years. This continues a trend in recent years of about half 
of governments making full contributions. 

Figure 4: Percentage of State and Local Government Pension Plans for 
which Governments Contributed More or Less Than 100 Percent of the ARC, 
by Fiscal Year: 

[See PDF for image] 

This figure is a stacked bar graph depicting the following data: 

Fiscal year: 1994; 
Percentage of plans, 100 percent or more of the ARC: approximately 14%; 
Percentage of plans, Less than 100 percent of the ARC: approximately 
86%. 

Fiscal year: 1996; 
Percentage of plans, 100 percent or more of the ARC: approximately 82%; 
Percentage of plans, Less than 100 percent of the ARC: approximately 
18%. 

Fiscal year: 2000; 
Percentage of plans, 100 percent or more of the ARC: approximately 76%; 
Percentage of plans, Less than 100 percent of the ARC: approximately 
24%. 

Fiscal year: 2001; 
Percentage of plans, 100 percent or more of the ARC: approximately 80%; 
Percentage of plans, Less than 100 percent of the ARC: approximately 
20%. 

Fiscal year: 2002; 
Percentage of plans, 100 percent or more of the ARC: approximately 64%; 
Percentage of plans, Less than 100 percent of the ARC: approximately 
36%. 

Fiscal year: 2003; 
Percentage of plans, 100 percent or more of the ARC: approximately 60%; 
Percentage of plans, Less than 100 percent of the ARC: approximately 
40%. 

Fiscal year: 2004; 
Percentage of plans, 100 percent or more of the ARC: approximately 55%; 
Percentage of plans, Less than 100 percent of the ARC: approximately 
45%. 

Fiscal year: 2005%; 
Percentage of plans, 100 percent or more of the ARC: approximately 55%; 
Percentage of plans, Less than 100 percent of the ARC: approximately 
45%. 

Fiscal year: 2006; 
Percentage of plans, 100 percent or more of the ARC: 58%; 
Percentage of plans, Less than 100 percent of the ARC: 42%. 

Source: GAO analysis of PFS, PENDAT data. 

[End of figure] 

In particular, some of the governments that did not contribute the full 
ARC in multiple years were sponsors of plans with lower funded ratios. 
In 2006, almost two-thirds of plans with funded ratios below 80 percent 
in 2006 did not contribute the full ARC in multiple years. Of the 32 
plans that in 2006 had funded ratios below 80 percent, 20 did not 
contribute the full ARC in more than half of the 9 years for which data 
is available. In addition, 17 of these governments did not contribute 
more than 90 percent of the full ARC in more than half the years. 

State and local government pension representatives told us that 
governments may not contribute the full ARC each year for a number of 
reasons. First, when state and local governments are under fiscal 
pressure, they may have to make difficult choices about paying for 
competing interests. State and local governments will likely face 
increasing fiscal challenges in the next several years as the cost of 
health care continues to rise. In light of this stress, the ability of 
some governments to continue to pay the ARC may be questioned. Second, 
changes in the value of assets can affect governments' expectations 
about how much they will have to contribute. Because a high proportion 
of plan assets are invested in the stock market, the decline in the 
early 2000s decreased funded ratios and increased the unfunded 
liabilities of many plans. Such a marked decline in asset values was 
not typical in the experience of public pension funds, according to one 
expert. Reflecting the need to keep up with the increase in unfunded 
liabilities, ARCs increased, challenging many governments to make full 
contributions after they had grown accustomed to lower ARCs in the late 
1990s. Moreover, some plans have contribution rates that are fixed by 
constitution, statute, or practice and do not change in response to 
changes in the ARC. Even when the contribution rate is not fixed, the 
political process may take time to recognize and act on the need for 
increased contributions. Nonetheless, many states have been increasing 
their contribution rates in recent years, according to information 
compiled by the National Conference of State Legislatures. Third, some 
governments may not contribute the full ARC because they are not 
committed to pre-funding their pension plans and instead have other 
priorities, regardless of fiscal conditions. 

When a government contributes less than the full ARC, the funded ratio 
can decline and unfunded liabilities can rise, if all other assumptions 
are met about the change in assets and liabilities.[Footnote 19] 
Increased unfunded liabilities will require larger contributions in the 
future to keep pace with the liabilities that accrue each year and to 
make up for liabilities that accrued in the past. As a result, costs 
are shifted from current to future generations. 

Unfunded Retiree Health Liabilities Are Large for Many State and Local 
Governments: 

Our review of studies estimating the total retiree health benefits for 
all state and local governments showed that liabilities are between 
$600 billion and $1.6 trillion.[Footnote 20] The studies noted that, 
like many private employers, few governments have set aside any assets 
to pay for these obligations. The projected unfunded liabilities do not 
have to be paid all at once, but can be paid over many years. Some 
governments do not pay for any retiree health benefits and therefore do 
not have any unfunded liabilities. Others may have large unfunded 
liabilities. For example, California has estimated its unfunded retiree 
health benefits liabilities at $70 billion, while the state of Utah 
estimates $749 million. 

Estimates of unfunded liabilities for retiree health benefits are 
subject to change substantially because projecting future costs of 
health care is difficult. Compared to the future payments for pension 
benefits, payments for health care benefits are significantly more 
unpredictable. Pension calculations generally use salaries as a base 
for calculations and result in a predictable benefit amount per year. 
But the cost of providing health care benefits varies with the changing 
cost of health care as well as with each individual's usage. In 
addition, state and local governments usually have the ability to 
reduce or eliminate benefits. 

Unfunded liabilities for retiree health benefits are high because 
unlike pension plans, nearly all state and local government retiree 
health benefits have been financed on a pay-as-you-go basis. In other 
words, most governments have not set aside funds in a trust dedicated 
for future retiree health benefit payments. As a result, governments do 
not pay a yearly ARC, but rather pay for retiree health benefits as 
they become due from annual funds. However, the new GASB accounting 
standards will require state and local governments to report their 
funding status on an accrual basis. In other words, for the first time, 
most governments will begin to calculate and report their funding 
status in a manner similar to the way they report pensions' funding 
status, whether or not they are prefunded. 

Officials told us that state and local governments have not prefunded 
retiree health benefits for several reasons. First, for many 
governments, retiree health benefits began as an extension of employee 
health care benefits, which are usually paid for from general funds. 
Governments did not view retiree health as a separate stream of 
payments. Second, retiree health benefits were established at a time 
when health care costs were more affordable, so paying for the benefits 
as a yearly expense was less burdensome. Third, the inflation rate for 
health care is less predictable than for pensions, so calculating the 
current funding status is difficult. Fourth, given that specific 
retiree health benefits are generally not guaranteed by law, employers 
are freer to modify benefits; as a result, state and local governments 
are reluctant to commit funds to an obligation that may be reduced or 
eliminated in the future. Finally, changes in national health care 
policy and health insurance markets can affect what benefits state and 
local governments cover, so state and local governments may have 
resisted locking in their commitment to pay for future retiree health 
benefits by prefunding, and instead preferred to finance on a pay-as-
you-go basis. 

Although the unfunded liabilities for retiree health benefits are 
generally much higher than for pensions, their current annual payments 
are considerably lower. According to our analysis presented in our 
recent report on this topic,[Footnote 21] in 2006, the aggregate state 
and local contribution rate for pensions was about 9 percent of 
salaries, and the pay-as-you-go expense for retiree health benefits was 
about 2 percent of salaries. However, if retiree health continues to be 
financed on a pay-as-you-go basis, the pay-as-you-go amount is 
estimated to more than double to 5 percent of salaries by 2050 to keep 
up with the growth in health costs, adding to budgetary stress. Pay-as-
you-go financing also leaves less budgetary flexibility because state 
and local governments must pay the full costs of each year's benefits. 
In contrast, under pre-funding, benefits are paid from a fund that 
already exists, so government contributions can be reduced when fiscal 
pressures are great. As a result, governments may face even greater 
pressure to reduce benefits or shift the costs of benefits to 
beneficiaries, for example, by restricting eligibility, reducing 
coverage, or increasing premiums. Still, pre-funding retiree health 
benefits would require significantly higher contributions in the short 
term than pay-as-you-go financing would require. 

Concluding Observations: 

Understanding the funded status of state and local government retiree 
benefits requires examining, on a plan-by-plan basis, whether funding 
levels are improving over time and whether governments are making the 
contributions recommended by the plan's actuary each year. The variety 
of actuarial funding methods and assumptions makes it difficult to 
compare funded status across different pension plans. However, funded 
status information is not intended to help compare plans, but rather to 
determine contributions that will achieve full funding over time and to 
assess a given plan's funded status over time. 

The funded status of state and local government pensions overall is 
reasonably sound, though recent deterioration underscores the 
importance of keeping up with contributions, especially in light of 
anticipated fiscal and economic challenges. Since the stock market 
downturn in the early 2000s, the funded ratios of some governments have 
declined. Governments can gradually recover from these losses. However, 
the failure of some to consistently make the annual required 
contributions undermines that progress and is cause for concern, 
particularly as state and local governments will likely face increasing 
fiscal pressure in the coming decades. While unfunded liabilities do 
not generally put benefits at risk in the near-term, they do shift 
costs and risks to the future. 

In the case of retiree health benefits, pay-as-you-go financing has 
been the norm up to the present day. The initial estimates of the 
unfunded liabilities will be daunting. But that is a natural 
consequence of pay-as-you-go financing. Just as the unfunded 
liabilities did not accumulate overnight, it may be unrealistic to 
expect them to be paid for overnight. Rather, state and local 
governments need to find strategies for dealing with unfunded 
liabilities, and such strategies will take time, will require difficult 
choices, and could be affected by changes in national health policy. 

Agency Comments: 

We provided officials from the Internal Revenue Service, GASB staff, 
and other external reviewers knowledgeable about the subject area a 
copy of this report for their review. They provided us with technical 
comments that we incorporated, where appropriate. 

As agreed with your offices, unless you publicly announce the contents 
of this report earlier, we plan no further distribution until 30 days 
from the report date. At that time, we will send copies of this report 
to relevant congressional committees, the Acting Commissioner of 
Internal Revenue, and other interested parties. Copies will also be 
made available to others upon request. In addition, the report will be 
available at no charge on the GAO Web site at [hypertext, 
http://www.gao.gov]. Please contact me at (202) 512-7215, if you have 
any questions about this report. Other major contributors include 
Tamara Cross, Assistant Director; Ken Stockbridge; Anna Bonelli; Temeca 
Simpson; Amy Abramowitz; Joseph Applebaum; Rick Krashevski; Jeremy 
Schwartz; Walter Vance; Charles Willson; and Craig Winslow. 

Signed by: 

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

[End of section] 

Appendix I: Objectives, Scope, and Methodology: 

The objectives of this report were to examine 1) the key measures of 
the funded status of retiree benefits and 2) the current funded status 
of state and local pension and retiree health benefits. 

To describe the key measures of the funded status of retiree benefits, 
we interviewed experts on state and local government pension and 
retiree health benefits such as national organizations, bond rating 
agencies, and representatives from one local government retiree benefit 
system. We also spoke with experts on actuarial science such as the 
Actuarial Standards Board, the American Academy of Actuaries, and 
independent actuaries. We spoke to staff of the Governmental Accounting 
Standards Board to understand accounting practices and principles. We 
also reviewed actuarial literature and attended conferences. In 
addition, we conducted the following analysis: 

* To understand the impact of various economic factors on the funding 
ratio of public pension plans, we developed a simple model of the 
determinants of the funding ratio and conducted "counterfactuals" 
holding rates of return on investments constant. To do this, we used 
the following data sources: 

* funding ratio data from the Public Fund Survey (PFS) for years 2001 
to 2005 and the Survey of State and Local Pensions for years 1988 to 
2000; 

* market value of pension assets from the Federal Reserve's Flow of 
Funds Accounts; 

* contributions and benefits data from the Bureau of Economic 
Analysis's National Income and Product Accounts database; and: 

* data on returns on pension fund portfolios by analyzing market data. 

* Our methodology and data sources for this analysis include some 
limitations. First, annual data are not available in the Survey of 
State and Local Pensions for 5 years during the period. For those 
years, values were imputed by using the average growth between the two 
closest values. In addition, the funding ratios are available on a 
fiscal year basis and were subsequently adjusted to a calendar year 
period. Second, assumptions may not be representative of all pension 
plans, such as the assumptions based on smoothing functions and the 
real expected returns on investments. Last, counterfactuals do not 
include policy adjustments that may occur because of different rates of 
return. 

To describe the funded status of state and local governments' pensions, 
in addition to a literature review, we analyzed pension funding data 
provided by the National Association of State Retirement Administrators 
(NASRA). The data come from two different databases. The first database 
is the PFS and is sponsored by NASRA and the National Council on 
Teacher Retirement (NCTR). Data from years 2001 to 2006 were available. 
PFS data are gathered by reviewing publicly available financial 
documents from the state and local government plans. The second 
database is called the PENDAT database and was sponsored by the Public 
Pension Coordinating Council. PENDAT data are available in fiscal years 
1992, 1994, 1996, 1998, and 2000.[Footnote 22] PENDAT data were 
collected via a survey sent to the administrators of a sample of plans 
nationwide. 

* The PFS and PENDAT databases do not include all of the same entries. 
We matched individual entries from PENDAT to PFS, resulting in a sample 
with between 63 and 71 plans that had data across each of the available 
years from 1994 to 2006. In fiscal year 2005, these plans represented 
58 percent of plan assets nationwide, and 72 percent of state and local 
government pension plan members. 

* We reviewed the PFS and PENDAT data and found them to be reliable for 
our purposes. To do this, we reviewed all entries of key data points in 
the PFS data using publicly available sources from the state and local 
government plan sponsors and made adjustments to the data as needed. 
The corrections made to the PFS data were not material. To review the 
PENDAT database, we reviewed the methodology used to collect the data 
and verified the data of 23 percent of entries using external sources. 
The corrections were not found to be material. 

* The information contained in the PFS and PENDAT databases have 
limitations: 1) surveys, including PENDAT, are subject to several kinds 
of error such as the failure to include all members of the population 
in the sample, nonresponse error, and data processing error; 2) the 
funding ratio and other funding indicators represent the financial 
status for the fiscal year with the most recent actuarial valuation, 
and thus do not all represent the same fiscal year's financial status; 
3) the plans included in the analysis are not necessarily 
representative of all state and local government pension plans 
nationwide; and 4) data for every plan is not available in each year. 

To obtain information on the funded status of retiree health benefits, 
we interviewed experts on retiree health benefits funding from national 
organizations, bond rating agencies, and one local government retiree 
benefits system. We also reviewed studies conducted by various 
organizations estimating the funded status. These organizations each 
obtained information about retiree health benefits liabilities from a 
number of different state and local governments and then extrapolated 
these figures to generate a nationwide estimate of all state and local 
governments. We reviewed the following studies: 

* Credit Suisse, You Dropped a Bomb on Me, GASB, 2007. Limitations of 
this study include: only states in the analysis, not local 
jurisdictions, are included; assumes that those government entities for 
which Credit Suisse was able to find estimates of future retiree health 
benefit obligations were representative of governments overall in terms 
of age distribution and funding levels; and does not consider the 
variation in actuarial assumptions and methods between the different 
plans. 

* Cato Institute, Unfunded State and Local Health Costs: $1.4 Trillion, 
2006. Limitations of this study include: includes states only in the 
analysis, not local jurisdictions; assumes that those government 
entities for which Cato was able to find estimates of future retiree 
health benefit obligations were representative of governments overall 
in terms of age distribution and funding levels; does not consider the 
variation in actuarial assumptions and methods between the different 
plans; it is not clear how many employees were covered by the sample 
because there were so many localities; and figures on the percentage of 
employees covered by health care plans in state and local government 
jurisdictions may not be precise. 

* OPEB for Public Entities: GASB 45 and Other Challenges, JP Morgan, 
2005. Limitations of this study include: assumes that those government 
entities for which they were able to find estimates of future retiree 
health benefit obligations were representative of governments overall 
in terms of age distribution and funding levels; and does not consider 
the variation in assumptions and methods between the different plans. 

We conducted our work in Washington, D.C.; New York; and Connecticut, 
from July 2006 to January 2008 in accordance with generally accepted 
government auditing standards. 

[End of section] 

Related GAO Products: 

State and Local Government Retiree Benefits: Current Status of Benefit 
Structures, Protections, and Fiscal Outlook for Funding Future Costs. 
GAO-07-1156. Washington, D.C.: September 24, 2007. 

State and Local Governments: Persistent Fiscal Challenges Will Likely 
Emerge within the Next Decade. GAO-07-1080SP. Washington, D.C.: July 
18, 2007. 

Retiree Health Benefits: Majority of Sponsors Continued to Offer 
Prescription Drug Coverage and Chose the Retiree Drug Subsidy. GAO-07-
572. Washington, D.C.: May 31, 2007. 

Employer-Sponsored Health and Retirement Benefits: Efforts to Control 
Employer Costs and the Implications for Workers. GAO-07-355. 
Washington, D.C.: March 30, 2007. 

State Pension Plans: Similarities and Differences Between Federal and 
State Designs. GAO/GGD-99-45. Washington, D.C.: March 19, 1999. 

Public Pensions: Section 457 Plans Pose Greater Risk than Other 
Supplemental Plans. GAO/HEHS-96-38. Washington, D.C.: April 30, 1996. 

Public Pensions: State and Local Government Contributions to 
Underfunded Plans. GAO/HEHS-96-56. Washington, D.C.: March 14, 1996. 

[End of section] 

Footnotes: 

[1] Actuarial accrued liabilities, referred to in this report as 
"liabilities," are the portion of the present value of future benefits 
that is attributable to employee services in past periods, under the 
actuarial cost method utilized. 

[2] The PFS is sponsored by the National Association of State 
Retirement Administrators and the National Council on Teacher 
Retirement. In 2005, the PFS data we used represented 58 percent of 
total assets invested in public pension plans nationwide, and 72 
percent of total members. PFS data covered years beginning with 2001. 
PPCC data covered years 1994, 1996, and 2000. 

[3] GAO, State and Local Governments: Persistent Fiscal Challenges Will 
Likely Emerge within the Next Decade, GAO-07-1080SP (Washington, D.C.: 
July 18, 2007). 

[4] The last year for which the Bureau of Labor Statistics published 
these data was 1998. U.S. Department of Labor, Bureau of Labor 
Statistics, Employee Benefits in State and Local Governments, 1998 
(Washington, D.C.: 2000). 

[5] In contrast, for defined contribution plans, the key determinants 
of the benefit amount are the employee's and employer's contribution 
rates and the rate of return achieved on plan assets (made up of the 
amounts contributed to an individual's account over time). Defined 
contribution plans include 401(k)s. 

[6] Two states (Alaska and Michigan) and the District of Columbia offer 
defined contribution plans as their primary plan for general public 
employees. Two states (Indiana and Oregon) offer primary plans with 
both defined benefit and defined contribution components; and one state 
(Nebraska) offers a cash balance defined benefit plan as its primary 
plan. 

[7] States also typically offer other retiree benefits such as vision, 
dental, long-term care, and life insurance, but these are generally 
funded entirely by retirees. For more information on the range and 
types of benefits provided, see GAO, State and Local Government Retiree 
Benefits: Current Status of Benefit Structures, Protections, and Fiscal 
Outlook for Funding Future Costs, GAO-07-1156 (Washington, D.C.: Sept. 
24, 2007). 

[8] Similarly, ERISA generally does not include funding and reporting 
requirements for private companies' health benefits. 

[9] For more information on the protections for state and local retiree 
benefits, see GAO, State and Local Government Retiree Benefits: Current 
Status of Benefit Structures, Protections, and Fiscal Outlook for 
Funding Future Costs, GAO-07-1156 (Washington, D.C.: Sept. 24, 2007). 

[10] The ARC is made up of the amount of future benefits promised to 
plan participants that accumulated in the current year, plus a portion 
of any unfunded liabilities. 

[11] Contributions from both sponsors and employees, combined with 
investment earnings on plan assets, must cover both future benefit 
payments and the administrative expenses associated with the plan. 

[12] Even if a single method were required for financial reporting 
purposes, government sponsors could still use a different method for 
funding purposes, since financial reporting standards do not dictate 
the fiscal policies used to fund the plans. 

[13] Under GASB standards, sponsors can also re-amortize unfunded 
liabilities each year, known as "open amortization." Under such an 
approach, for example, each year sponsors can pay the annual cost for a 
30-year amortization of that year's unfunded liabilities; the following 
year, the sponsor can re-amortize the remaining unfunded liabilities 
over an additional 30 years, and so on. 

[14] The Pension Protection Act of 2006 provided that large private 
sector pension plans will be considered at risk of defaulting on their 
liabilities if they have less than 80 percent funded ratios under 
standard actuarial assumptions and less than 70 percent funded ratios 
under certain additional 'worst-case' actuarial assumptions. When 
private sector plans default on their liabilities, the Pension Benefit 
Guaranty Corporation becomes liable for benefits. These funding 
standards will be phased in, becoming fully effective in 2011, and at-
risk plans are required to use stricter actuarial assumptions that will 
result in them having to make larger plan contributions. Pub. L. No. 
109-280, sec. 112(a), § 430(i), 120 Stat. 780, 839-42. 

[15] In this section, we refer to our analysis of the Public Fund 
Survey (PFS) and the PENDAT database. The PFS is sponsored by the 
National Association of State Retirement Administrators and the 
National Council on Teacher Retirement. These sources contain self-
reported data on state and local government pension plans in years 
1994, 1996, and 2000 to 2006. Each year, between 62 and 72 plans were 
represented in our dataset. In 2005, the 70 plans represented 58 
percent of total assets invested in public pension plans nationwide in 
2005, and 72 percent of total members. 

[16] K. Brainard, Public Fund Survey Summary of Findings for FY 2006, 
National Association of State Retirement Administrators, (Georgetown, 
Tex.: October 2007). 

[17] These findings may be unique to the time period examined (1988-
2005). In other periods, other factors, such as changes to benefits, 
may account for more of the change in the funded ratio than the rates 
of return on the investment portfolio. 

[18] Reports estimate total unfunded liabilities for public pension 
plans nationwide between $307 and $385 billion, but the estimates do 
not cover all state and local government plans. One study by the 
National Association of State Retirement Administrators reviewed the 
funding status of 125 of the nation's large public pension plans in 
fiscal year 2006 and found total unfunded liabilities to be more than 
$385 billion. Another study reviewed state-only pension plans and found 
that in 2005, the most recent year for which substantially complete 
data was available, total unfunded liabilities for 108 plans were about 
$307 billion. Neither study is a random sample of state and local 
government pension plans that represents all public plans nationwide. 
NASRA Public Fund Survey (2006). This estimate represents 85 percent of 
public plan assets nationwide. Wilshire Consulting, 2007 Wilshire 
Report on State Retirement Systems: Funding Levels and Asset Allocation 
(2007). This study includes only state plans, not local plans. 

[19] When a government does not contribute at least the normal cost 
plus interest on the unfunded liability (which is an amount less than 
the full ARC), unfunded liabilities will increase. 

[20] Chris Edwards and Jagadeesh Gokhale. "Unfunded State and Local 
Health Costs: $1.4 Trillion." Tax and Budget Bulletin, no. 40 (Cato 
Institute: 2006); David Zion and Amit Varshney, "You Dropped a Bomb on 
Me, GASB: Uncovering $1.5 Trillion in Hidden OPEB Liabilities for State 
and Local Governments," Equity Research, Accounting and Tax (Credit 
Suisse: 2007); Brian Whitworth, Igor Balevich, and Jim Kelly OPEB for 
Public Entities: GASB 45 and other Challenges (J.P. Morgan: 2005). 
These estimates of health care liabilities are limited by their 
methodologies. For example, the reports generalize about all state and 
local governments' liabilities from a non-representative sample, and 
the reports did not consider the variation in actuarial methods and 
assumptions in the calculations (See app. I). The studies base their 
estimates on actuarial valuations and public reports that have been 
performed by some state and local governments in advance of the 
deadlines for the new GASB standards. The studies then extrapolate the 
findings to calculate a nationwide total for all state and local 
governments. 

[21] GAO-07-1156, pp. 27-30. As noted in that report, the simulations 
of future contribution rates are very sensitive to assumptions about 
the growth rate of health care costs and to assumptions about the rate 
of return on investments. 

[22] Few entries were available from 1998, so we did not use any data 
from this year. 

[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 "Subscribe to Updates." 

Order by Mail or Phone: 

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

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

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

To Report Fraud, Waste, and Abuse in Federal Programs: 

Contact: 

Web site: [hyperlink, http://www.gao.gov/fraudnet/fraudnet.htm]: 
E-mail: fraudnet@gao.gov: 
Automated answering system: (800) 424-5454 or (202) 512-7470: 

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: