# 2000-11
August 2000
THE IMPACT OF MANDATORY
SOCIAL SECURITY COVERAGE OF
STATE AND LOCAL WORKERS:
A MULTI-STATE REVIEW
by
Alicia H. Munnell
Boston College
Laurel Beedon and Jules Lichtenstein, PPI Project Managers
The Public Policy Institute, formed in 1985, is part of the Research
Group of AARP. One of the missions of the Institute is to foster
research and analysis on public policy issues of interest to older
Americans. This paper represents part of that effort.
The views expressed herein are for information, debate, and
discussion, and do not necessarily represent formal policies of AARP.
 2000, AARP
Reprinting with Permission Only.
Acknowledgements
Peter A. Diamond helped straighten out some of the conceptual issues in the report. Robert M.
Ball and Annika Sunden provided valuable comments. Yogesh Singh, Ye Feng, and Catherine
Taylor, Boston College graduate students in Finance and Economics, respectively, and Michael
Hough, an undergraduate in the School of Management, provided very helpful research
assistance.
Alicia H. Munnell
i
ii
Foreword
Retirement income security is often described as a three-tiered system with Social
Security, pensions and individual savings making up the tiers. Some American workers,
however, are employed in jobs that are not covered by Social Security. The majority of these
non-covered individuals is employed by state and local governments that provide their own
pensions financed by a combination of employee and government contributions plus earnings on
pension assets. As the discussion of Social Security’s long-term solvency becomes more
prominent, so does the debate about the coverage of state and local workers.
Inclusion of all government workers in Social Security is not a new idea. Amendments to
the Social Security Act in 1979 required a study of the implications of mandating coverage for all
public employees. In 1983, Congress extended mandatory coverage to new federal employees
and in 1990 coverage was extended to state and local workers who had no other pension plans.
Unable to agree upon a single plan that would eliminate Social Security’s long-term
deficit, the 1994-1996 Advisory Council on Social Security members did agree on several points,
one of which was universal coverage. The Council stated “To the extent feasible, everyone who
works for pay should be covered by the Social Security program. Every occupational group
contains substantial numbers of people who at one time or another will need the protection of the
program.” The Council also observed, “Moreover, all Americans have an obligation to
participate, since an effective Social Security program helps to reduce public costs for relief and
assistance, which, in turn, means lower general taxes.”
Uncovered workers and their state and local government employers are not so sure. They
have strong objections to mandatory coverage – even if it is restricted to newly hired workers.
Public employers assert that mandatory coverage will impose financial burdens on states and
localities for which there will be little or no return. Government workers are concerned that
changes in the structure of their pension systems could result in a move from secure defined
benefit pension plans to more risky defined contribution plans. They fear that plans designed for
specific categories of workers, such as fire fighters and police could be lost. And, they voice
concerns about the possibility that government pensions could be cut back and even eliminated
altogether, leaving Social Security to act not as the base retirement plan, but as the only plan.
Additionally, employers and administrators argue that mandatory coverage would lower public
pension benefits and thus create morale, labor relations and recruitment problems.
In this paper Alicia Munnell describes and analyzes the major arguments for and against
mandating Social Security coverage for newly hired state and local employees. She notes that
while the arguments against coverage have most recently centered on financing and cost issues,
the arguments supporting mandatory coverage are the “more complex and subtle issues of
equity.” It is observed that while mandatory coverage may make benefits more equitable and
improve benefits coverage for workers, it will likely raise costs by 5 to 6 percent of payroll. Part
of the increase reflects improved real benefits, the costs of which, will probably be borne by the
worker; part reflects the tax to cover Social Security’s unfunded liability, the cost of which, will
be borne by the taxpayer. Specifically, the study acknowledges that about half the cost associated
iii
with mandatory coverage will probably be borne by the taxpayers of the non-participating states
and localities. The cost increase associated with new benefits such as portability, disability and
dependent and survivor protection, and full cost-of-living adjustments will most likely be borne
by the employees. In addition, states and localities are likely to explicitly adjust their existing
pensions to reflect Social Security’s benefits and costs.
The most compelling arguments for mandatory coverage, according to the authors, have
little to do with the long-run deficit for the Social Security system. Rather, they conclude that the
case for mandatory coverage rests largely on issues of equity as well as the adequacy of state and
local benefits.
Specifically, the equity issues are whether excluded state and local government
employees and employers are paying their share of the nation’s effort, through the Social Security
program, to redistribute income from workers with higher lifetime earnings to workers with
lower lifetime earnings as well as paying their share of financing the unfunded liability
associated with the startup of the Social Security program. The benefit adequacy issues relate to
ensuring that all state and local workers have a full range of protections not currently provided
under public plans.
The study arrives at its conclusions with caution. It suggests paying close attention to the
lessons learned from the extension of Medicare coverage for employees hired after March 1986.
It also suggests that the legal issues involved in mandatory coverage, while beyond the scope of
the study, remain important. In concluding, the authors suggest that transition relief is important
and any changes should be made prudently to ensure economic efficiency and fairness.
Laurel Beedon and Jules Lichtenstein
Senior Policy Advisors
AARP Public Policy Institute
iv
Executive Summary
Introduction
The Social Security Act of 1935 excluded state and local workers from mandatory
coverage due to constitutional concerns about whether the federal government could impose
taxes on state governments. As Congress expanded coverage to new groups of private sector
workers, it also passed legislation in the 1950s that allowed states to elect voluntary coverage for
their employees. The question of mandatory – as opposed to voluntary – coverage for state and
local workers surfaced in the 1960s and 1970s. Today, about 30 percent of the state and local
workforce – roughly 4 million workers – still are not covered by Social Security. The 1994-96
Advisory Council on Social Security and many other groups have proposed extending Social
Security coverage to all state and local workers. Uncovered workers and their employers,
however, strongly oppose being forced to participate in Social Security.
Purpose
This study explores the major arguments for and against mandating Social Security
coverage for state and local employees.
Methodology
The approach of this study is a reasoned analysis of published sources, and an integration
of recent studies of Social Security’s unfunded liability into the standard arguments for and
against mandating coverage for state and local workers.
Principal Findings
Although the rationale for mandating Social Security coverage has recently focused on
the financial considerations, the strongest arguments for mandatory coverage are complex and
somewhat subtle issues of equity. Excluded state and local government employees or taxpayers
in their jurisdictions are not paying their share of income redistribution. Excluded state and local
workers and their employers also are not paying their share of financing the unfunded liability
associated with the startup of the Social Security program. The parents and grandparents of these
state/local workers – like those of covered workers – have received Social Security benefits far in
excess of their contributions to the system. This has created a debt that covered employees
finance with part of their payroll tax payment. By not participating in the system, state and local
government employees and their employers escape this “tax,” which is estimated to equal
roughly 3 percent of the 12.4 percent payroll tax.
Another argument for mandatory Social Security coverage is that it would bring excluded
workers three types of benefit protection that they do not have currently. They would enjoy full
portability of benefits as they move from job to job, even if they leave government employment.
Second, they would have full inflation protection for their retirement and disability benefits.
v
Third, they would be eligible for dependent and survivor benefits not offered by public plans. In
addition to these benefits, state and local workers would not face gaps in insurance – particularly
for disability.
Mandating coverage for all state and local employees will help Social Security. In the
short run, the impact will be modest because most proposals involve covering only new
employees hired after a specified date. Over the 75-year projection period, however, coverage
would improve the financial outlook for two reasons. First, the currently excluded state and local
workers would be relatively low-cost participants in the Social Security system. Second, part of
the payroll tax contribution from these uncovered workers would go to pay off Social Security’s
unfunded liability, as discussed below.
The strongest argument against mandatory coverage of state and local workers is that
even if plan sponsors cut back on their own pensions in an attempt to maintain a constant level of
benefits they will face an increase in costs. Some of the cost increase will reflect new protections
for state and local government employees as discussed above; some will reflect a roughly 3percent “tax” component of Social Security to cover the unfunded liability that arose from paying
benefits to older generations; some will reflect state and local workers’ contribution to income
redistribution. The extent of the cost increase depends on how states and localities adjust their
own plans in response to Social Security coverage.
About half of the cost increase associated with mandatory coverage will probably be
borne by the taxpayers of the affected states and localities, because current compensation, even in
jurisdictions without Social Security, already reflects the “tax” to cover Social Security’s
unfunded liability and government employers are unlikely to be able to shift much of this cost
onto their employees. The cost increase associated with new benefit protections most likely will
be borne by the employees.
Conclusion
The legal issues involved in mandatory Social Security coverage for state and local
employees are beyond the scope of this study, but they remain important. The GAO (1998b)
concluded that “mandatory coverage is likely to be upheld under the current U.S. Supreme Court
decisions.” "Further, in a recent case, Reno v. Condon, No. 98-1468 (U.S.S.Ct., January 12,
2000), the Supreme Court has signaled that, while it is respectful of state sovereignty, it will rule
cautiously when reviewing federal-state relationships. Thus, how the Court will rule regarding
mandatory Social Security coverage appears uncertain today."
While mandatory Social Security coverage may be equitable and improve benefits, it will
raise costs by 5 to 6 percent of payroll. Part of the increase would reflect improved real benefits,
the cost of which would be borne by the workers; part would reflect a tax to cover Social
Security’s unfunded liability, the cost of which would probably be borne by the taxpayer. Since
mandating coverage does change the rules for a significant number of taxpayers, it might be
worth considering some transition relief to state and local governments as they bring their
vi
employees into Social Security. It is also important to pay attention to the lessons learned from
the extension of Medicare coverage for employees hired after March 1986. For example, one
sensitive area is the definition of a new employee. States object to classifying a teacher who
moves from one town to another but remains within the teachers retirement system, as a new
employee. Every effort should be made to use reasonable definitions to ensure both economic
efficiency and fairness. Flexibility and good will are required on all sides to solve this problem.
vii
viii
TABLE OF CONTENTS
Acknowledgments… ….…………………………………………………………………………..i
Foreword…………….………………………………………………………………………..….iii
Executive Summary…………………………………………………………………………..…..v
Table of Tables………….……………...………………………………………………………....x
Table of Figures…….………………………………………………………………………….…xi
Introduction…………………………………………………………………………………….…1
An Overview of State and Local Pension Plans……………………………………………….….4
Social Security and State and Local Plans…….………………………………………………....10
The Case for Mandatory Coverage of State and Local Workers……………………………..….15
The Costs of Mandating Coverage for All State and Local Employees…………………………19
Conclusion……………………………………………………………………………………….25
References………………………………………………………………………………………..27
Tables……………………………………………………………………………………………34
Figures………….………………………………………………………………………………..47
Appendix A………………………………………………..……………………………………A-1
Appendix B………………………………………………………..……………………………B-1
Appendix C……………………………………………………………………..………………C-1
ix
TABLE OF TABLES
Table 1. State and Local Pension Systems, 1996-1997…………..……………………………..34
Table 2. State and Local Government Employees Not Covered by Social Security, 1996..……35
Table 3. Prevalence of State and Local Government Employees Not Covered by Social Security,
1996……………………………………………………………………………………36
Table 4. State and Local Government Employee Coverage by Type of Retirement Plan….…...37
Table 5. Percentage Distribution of State and Local Pension Plans by Type of Cost-of-Living
Adjustments (COLA) and Size of Plan, 1996……..………...…………………………38
Table 6. Cost-of-Living Protection in Major State Plans with Employees Not Covered by
Social Security………………………..……………………………………………….39
Table 7. Funding Status of State and Local Pension Plans, 1996……….………………………40
Table 8. Funding Status of Major Plans Without Social Security Coverage, 1998…….……….41
Table 9. Comparison of Pension Benefits for a Four-Job Worker and a One-Job Worker…..…42
Table 10. Normal Cost for State and Local Plans Before and After Mandatory Social Security
Coverage from Actuarial Education and Research Fund Study, 1980…….…..………43
Table 11. Normal Cost for State and Local Plans Before and After Mandatory Social Security
Coverage from the Urban Institute Study, 1980…….…………………………………44
Table 12. Ratio of Required State/Local Benefits to Maintain Benefit Levels with Inflation
Relative to the Amount Required with Zero Inflation..………….……………………45
Table 13. Five-Year Mandatory Social Security Coverage as a Percent of State Expenditures
For Selected States and for the Nation, 2000-2004………...…………………………46
x
TABLE OF FIGURES
Figure 1. Date of Retirement Plan Establishment or Major Restructuring for Large State and
Local Systems…………………………………………………………………………47
Figure 2. Average Benefit Percentage per Year of Service for State and Local Plans by Type of
Covered Employee, 1996……………………..……………………………………….48
Figure 3. Employee and Government Contributions as a Percent of Payroll by Type of Covered
Employees…………….……………………………………………………………….49
Figure 4. Ratio of Assets to Benefit Payments for Total State and Local Pension Plans…….....50
xi
Introduction
The Social Security Act of 1935 excluded state and local workers from mandatory
coverage due to constitutional concerns about whether the federal government could impose
taxes on state governments.1 As Congress expanded coverage to new groups of private sector
workers, it also passed legislation in the 1950s that allowed states to elect voluntary coverage for
their employees. The same legislation permitted states to withdraw from Social Security after a
specified period, but Congress eliminated this option in 1983.
The question of mandatory – as opposed to voluntary – coverage for state and local
workers surfaced in the 1960s and 1970s. In response to increasing interest in such a change, the
1977 Amendments to the Social Security Act required a study of the desirability and feasibility of
covering all public employees. In the wake of that study, Congress extended mandatory coverage
to new federal employees in the 1983 Social Security Amendments and, in 1990, to state and
local workers who had no other pension plan. About 30 percent of the state and local workforce
– roughly 4 million workers – remain without Social Security coverage. The bulk of uncovered
workers (75 percent) reside in seven states: California, Colorado, Illinois, Louisiana,
Massachusetts, Ohio, and Texas. In Colorado, Louisiana, Massachusetts, and Ohio virtually no
government workers are covered by Social Security in their government employment. In
California, Illinois, and Texas, uncovered state and local workers constitute 49 percent, 62
percent, and 55 percent of the total, respectively.
The 1994-96 Advisory Council on Social Security and many other groups have proposed
extending Social Security coverage to all state and local workers. Uncovered workers and their
employers, however, strongly oppose mandatory participation in Social Security. They say that
such requirements would impose a financial burden on states and localities, for which public
employees would receive little (Segal Company, 1999). Additionally, at the 1998 White House
Conference on Social Security, representatives of uncovered state and local workers and their
employers raised other concerns (White House Conference on Social Security). A spokesperson
for the American Federation of State, County, and Municipal Employees (AFSCME) suggested
that restructuring pension systems to integrate Social Security would create an opportunity for
pension plan sponsors to change defined benefit plans to defined contribution plans. The
International Association of Fire Fighters worried particularly that mandatory coverage could
endanger pension plans that had been tailored to the unique needs of its members. James E.
Burton, CEO of California’s Public Employees Retirement System (CalPERS), argued that
requiring new hires to participate in Social Security and thus providing them with lower public
pension benefits would create morale, labor relations, and recruitment problems.
1
In hearings on S. 1130 before the Senate Committee on Finance January 22 to February 20, 1935, Ed Witte, the
Executive Director of the Committee for Economic Security, said "In the contributory annuity plan, we exclude all
public employees, for the reason that not only do they very often have their own systems, but also that the federal
government cannot impose a tax on state governments." Witte’s comment about public employees having their own
system applied primarily to federal employees since coverage for state and local employees did not become
widespread until the 1930s and 1940s.
1
This study explores the major arguments for and against mandating Social Security
coverage for state and local employees. Section II provides an overview of the types of
retirement, disability, and survivor protection that state and local workers enjoy today. These
descriptions draw heavily on detailed information about a sample of large state plans that Social
Security does not cover (see Appendix A).2 Section III explores the history of the relationship
between Social Security and state and local workers, and looks at the impact of mandatory
coverage on Social Security. Section IV presents the case for mandatory coverage, focusing on
the equity issues and the increased protections that Social Security would provide to workers.
Section V examines the costs of extending mandatory coverage and considers who would bear
the burden of these costs. The following conclusions emerge from this study:
•
State and local plans collectively provide substantial retirement income, and they are in much
better financial health than they were 20, or even 10, years ago. Those plans not covered by
Social Security generally provide somewhat greater benefit accrual rates per year of coverage
and somewhat more inflation protection than the group as a whole. Nevertheless, state and
local employees without Social Security coverage lack the benefit protection for dependents
and survivors found in Social Security. Moreover, COLAs provided by public pension plans
are frequently capped at 3 percent, which means that inflation can erode the purchasing
power of benefits. In addition, state and local employees without Social Security who leave
public employment for other jobs either forfeit accrued benefits if they are not vested, or face
the erosion in the value of benefits between termination and retirement that is characteristic
of defined benefit plans. Finally, mobile workers may experience gaps in insurance
protection, particularly in case of disability.
•
Mandating coverage for all state and local employees would help Social Security. In the
short run, the impact would be modest because most proposals involve covering only new
employees hired after a specified date. Over the 75-year projection period, however,
coverage would improve the financial outlook for two reasons. First, the currently excluded
state and local workers would be relatively low-cost participants in the Social Security
system, since state and local employees tend to earn a higher than average wage. Moreover, a
large number of the state and local employees without coverage are teachers, who are
predominately women. Women are less likely to have spouses who receive dependents’ and
survivors’ benefits based on their earnings record. Second, part of the payroll tax
contribution from these uncovered workers would go to pay off Social Security’s unfunded
liability, as discussed below.
•
Although the rationale for mandating Social Security coverage has recently focused on the
financial considerations, the strongest arguments for mandatory coverage are complex and
somewhat subtle issues of equity.
2
The state and local documents used to gather specific information about the plans, such as handbooks and summary
plan descriptions, are listed in the Reference section.
2
◊
Excluded state and local government employees or taxpayers in their jurisdictions do
not pay their share of income redistribution. Social Security redistributes income
toward workers with low lifetime earnings. As generally higher paid workers, state
and local employees as a group would be net contributors to this redistribution. State
and local workers, like their private and public sector counterparts, benefit from the
reduced need for taxes to assist older workers, and from living in a society with fewer
poor elderly. Proponents of mandatory coverage believe that uncovered state and
local workers should bear a greater share of this societal burden.
◊
Excluded state and local workers and their employers do not pay their share of
financing the unfunded liability associated with the startup of the Social Security
program. The parents and grandparents of these state and local workers – like those
of covered workers – have received Social Security benefits far in excess of their
contributions to the system. This has created a debt that covered employees finance
with part of their payroll tax payment. By not participating in the system, state and
local government employees and their employers escape this “tax,” which is estimated
to equal roughly 3 percent of the 12.4 percent payroll tax.
•
Another argument for mandatory Social Security coverage is that it would bring excluded
workers three types of protection that they do not have currently. They would enjoy full
portability of benefits as they move from job to job, even if they were to leave government
employment. Second, they would have full inflation protection for their retirement and
disability benefits. Third, they would be eligible for dependent and survivor benefits not
offered by many public plans. In addition to these benefits, state and local workers would not
face gaps in insurance, particularly for disability.
•
The strongest argument against mandatory coverage of state and local workers is that, even if
plan sponsors cut back on their own pensions in an attempt to maintain a constant level of
benefits, they would face an increase in costs. These costs would likely amount to between 5
and 6 percent of payroll. Some of the cost increase would reflect new protections for state
and local government employees, as discussed above; some would reflect a roughly 3
percent “tax” component of Social Security to cover the unfunded liability that arose from
paying benefits to older generations; some would reflect state and local workers’
contribution to income redistribution.
◊
The extent of the cost increase would depend on how states and localities adjusted
their own plans in response to Social Security coverage. Most studies assume that
they would try to maintain constant benefits, and measure constant benefits in terms
of equalizing first-year retirement benefits for the average unmarried worker. This
approach exaggerates the cost increase. A more accurate calculation would equalize
lifetime, not first-year, benefits, which would reveal the impact of Social Security’s
full inflation protection over the duration of retirement. Even if done perfectly,
however, costs would increase.
3
◊
•
Taxpayers of the affected states and localities would probably bear roughly half of a
cost increase associated with mandatory coverage. Current compensation, even in
jurisdictions without Social Security, already reflects the “tax” to cover Social
Security’s unfunded liability, and government employers are unlikely to be able to
shift this cost onto their employees. The cost increase associated with new benefit
protections most likely would be borne by the employees.
Although costs would increase gradually (four years to revise pension provisions after which
provisions apply only to new employees), mandating coverage would change the rules of the
game for employees and taxpayers of the affected jurisdictions. Some of the plans in
jurisdictions currently without Social Security are not fully funded. Consequently, some
taxpayers would find themselves contributing to cover both Social Security’s unfunded
liability as well as the unfunded liability of their state or local pension. Hence, it might be
worth considering some transition relief for the affected governments. In addition, it is also
important to remember the lessons learned from the extension of Medicare coverage for
employees hired after March 1986.
An Overview of State and Local Pension Plans
State and local pension plans are a major component of the nation’s retirement system.
They cover roughly 12.8 million active employees, pay benefits to more than 5 million
beneficiaries and had assets of $1.5 trillion, as of the last Census of Governments for 1996-97
(see Table 1). Today, because of the run-up in the stock market, the Federal Reserve, which uses
a slightly different definition of assets, shows state and local plan holdings of $3.0 trillion, an
amount equal to 65 percent of the $4.6 trillion held in private plans (Board of Governors of the
Federal Reserve System).3
The first state or local plan dates from 1857, when New York City provided lump-sum
benefits to policemen injured in the line of duty (Bleakney). Many municipalities created plans
during the last half of the nineteenth century, including a number of systems for teachers. In
1911, Massachusetts developed the first state system to cover its general government employees,
but the major expansion of coverage came in the wake of the 1935 Social Security legislation.
During the 1930s and 1940s, nearly half of the large state and local plans were established or
significantly restructured (see Figure 1). By the early 1960s, most states and localities had
established their pension systems, and those government employers who wanted to join Social
Security had done so.
3 Called
the Flow of Funds Accounts of the United States: Flows and Outstandings, this document reports total state
and local plan assets of $1,464.6 billion at the end of calendar year 1995 and $1,789.6 billion at the end of calendar
year 1996. Interpolating between the two figures would put assets at $1.6 trillion in mid-1996 compared with the
$1.4 trillion figure reported in the Census of Governments. Most of the increase between 1996 and the present,
therefore, is attributable to the fact that state and local plans held nearly 60 percent of their portfolio in stocks in
1996, and stocks have increased substantially in value.
4
State and local government employment roughly doubled between the early 1960s and the
mid-1970s, resulting in enormous growth in the population covered by state and local pension
plans. This growth, combined with interest in private plan reform that culminated in the 1974
passage of the Employee Retirement Income Security Act (ERISA), focused attention on public
pensions. ERISA mandated a study to determine whether public plans needed further regulation.
The 1978 Pension Task Force Report on Public Employee Retirement Systems (hereafter Pension
Task Force Report; see U.S. Congress, House of Representatives, Committee on Education and
Labor, 1978) did not result in the extension of federal regulation to governmental plans, but it did
document the status of public systems.
In addition to the 1978 Pension Task Force Report discussed above, we obtained
information about state and local plans from four major sources. One is government reports,
including the Bureau of Labor Statistics’ surveys of state and local pension plans (U.S.
Department of Labor 1999, 1996) and the Bureau of the Census reports on employee retirement
systems (U.S. Department of Commerce 1997, 1995, 1978). Another is the so-called PENDAT
files created from the Surveys of State and Local Employee Retirement Systems for Members of
the Public Pension Coordinating Council (hereafter PENDAT; see Zorn, 1997,1995, 1993, 1991).
The most recent survey includes coverage information for 1996.4 Third, very recent data on
pension assets are available from the Flow of Funds Accounts produced quarterly by the Federal
Reserve Board of Governors. Finally, drawing from financial reports, actuarial valuations, plan
handbooks, and direct contact, we collected data on the 14 largest state plans that do not
participate in Social Security. Detailed information about these plans, which cover 2.6 million
workers, appears in Appendix A.
State and Local Workers Not Covered by Social Security
The most elementary questions of how many workers are not covered by Social Security
and where they reside are difficult to answer. Several estimates are available, but in many cases
they provide very different pictures of the employment and coverage status.
Both the Census of Governments and the Social Security Administration (SSA) produced
estimates for 1992 of total state and local government employment and of those not covered by
Social Security; more recent government estimates are unavailable. These estimates are not only
dated, but also inconsistent. For example, for New York, the Census shows 592,855 without
coverage, while SSA reports 120,000. Unfortunately, it is not possible to explain the relative
magnitudes of these numbers by differences in definitions. One would expect SSA, which
counts people who had a job at any time during the year, to produce a higher number than
Census, which counts people during the last month of the fiscal year.
Two sources provide estimates for 1996. The PENDAT survey, described above, shows
total employment and the number not covered by Social Security for the plans included in the
survey. The problem is that the survey does not include all major plans. For example, PENDAT
4 In
1997, 261 public retirement systems responded to the survey, representing 379 retirement plans. These plans
covered 81 percent of the 13.6 million active members reported by the Bureau of the Census and held 81 percent of
the $1.6 trillion in assets reported by the Federal Reserve.
5
does not include Massachusetts Public Employee or Teachers Retirement Systems, and
Massachusetts is the fifth largest state in terms of the number of uncovered workers. The second
source of 1996 estimates is a recent study by the Segal Company for the AFSCME, AFL-CIO,
and the Coalition to Preserve Retirement Security; the Segal study includes estimates of the
number not covered but not of total employment (Segal, 1999). The Segal numbers are
reportedly based on SSA estimates for 1992, augmented by data in state reports.
In any event, the information from the 1996 sources is inconsistent in many cases with
that from 1992. For example, for New York, PENDAT shows no workers without Social
Security coverage, and Segal reports 90,680, compared with 1992 estimates from Census and
SSA of 592,855 and 120,000, respectively. In cases like New York, we called administrators of
the state plans in an attempt to reconcile the various numbers. Our conclusion for New York was
that virtually all state and local employees are covered, so we report the PENDAT number.
Appendix B provides extensive detail about the various estimates and the decision rules we used
to produce Table 2. While it would be preferable to have a clear explanation for the differences
in the various estimates, the published reports do not provide enough information for such an
assessment.
Table 2 shows that in 1996, roughly 30 percent of state and local workers – almost 4
million out of 12.8 million workers – were not covered by Social Security. Although almost
every state has some workers without coverage, 75 percent of state and local government
employees not covered by Social Security are concentrated in seven states. Listed in order of
number of employees not covered, these states are: Ohio, California, Texas, Illinois,
Massachusetts, Louisiana, and Colorado (see Table 3). Employees in teacher systems constitute
a disproportionate share of those employees not covered by Social Security. For example, in
California and Texas, the system for general public employees provides coverage, while the
teachers’ retirement system does not.
Characteristics of State and Local Retirement Systems
State and local retirement systems are not homogeneous because they cover several very
different groups of workers – general government employees, teachers, and public safety
personnel – each of which have very different career paths. For example, police and firefighters,
with physically demanding jobs, have plans that allow retirement at earlier ages and offer more
extensive disability protection. Most workers in state and local plans (57 percent) are covered
under the major state retirement systems; teachers and school employee systems cover another
30 percent; local systems, while numerous, cover only 10 percent; and systems for police and
firefighters cover the remaining 3 percent (see Table 4).
While most private pension plans provide for vesting after 5 years, vesting for public
plans is bimodal (State of Wisconsin Retirement Research Committee, 1996 and U.S.
Department of Labor, 1996). That is, roughly 45 percent of plans have 5-year vesting and
another 45 percent vest after 10 years. Vesting requirements for police and fire vary widely,
from a low of 3 years to a high of 20 years. Employees who leave with less than the required
6
service generally receive their contributions with some interest, but forfeit rights to earned
pension credits.
Most state and local pensions (92 percent) are defined benefit plans, 4 percent are defined
contribution, and 4 percent are a combination of the two (Zorn, 1997).5 Under the typical
defined benefit plan, the employee’s retirement benefit is calculated as a percentage of final
average salary for each year of service. For example, an employee who works for 30 years at a
benefit rate of 2 percent would earn benefits equal to 60 percent of final average pay upon
retirement.
Figure 2 shows the average annual benefit percentage that workers accrue per year of
service in the 1997 PENDAT survey by type of employee and by Social Security coverage. The
fact that plans covered by Social Security have a somewhat lower benefit accrual rate indicates
that even though most public plans do not explicitly integrate their plans with Social Security,
they do implicitly design their benefit formulas to take account of Social Security benefits.6 For
example, teacher plans covered by Social Security have an accrual rate of 1.77 percent per year
compared with 2.09 percent for plans not covered by Social Security. The difference in accrual
rates covers only a portion of the lack of Social Security, however.
When calculating benefits, respondents to the survey generally defined final average pay
as the salary earned during the highest or last 3 to 5 years. In general public employee and
teacher plans, employees can retire at age 60 or 65 with 5 or 10 years of service. Early retirement
with reduced benefits is generally available at age 50 or 55 with 5 or 10 years of service, or at any
age with 25 years.
Plans for police and firefighters are also defined benefit plans, and they tend to have
higher annual benefit accrual rates than those for other systems. This probably reflects the fact
that only 20 percent of police and fire plans are covered by Social Security, compared with 75
percent of plans for general employees and 59 percent for teachers (Zorn, 1997).7 Some police
and firefighter plans require 20 years of service for vesting purposes, and these plans often
provide a flat percentage of annual pay at retirement (usually 50 percent). Police and firefighters
can generally retire at age 50 or 55 with 20 years of service or sometimes at any age with 20
years.
In addition to retirement, most plans provide disability benefits in case of permanent and
total service-related disability. Because disability benefits are based on age and salary, benefits
can be quite low if disability occurs early in a worker’s career, although many states establish a
5 This
information is found on p. 31 of Paul Zorn’s research.
to the 1997 PENDAT survey, only 10 percent of plans with Social Security coverage explicitly integrate
their benefits with Social Security (Zorn, 1997, pp. 33-34). Approximately half of the plans that do integrate use the
“offset” approach, and the other half use the “step-rate” approach. Under the offset approach, the plan sponsor
reduces recipients’ benefits by a fixed percentage (for example 50 percent) of their primary Social Security benefit.
Under the step-rate approach, the plan sponsor applies a higher benefit accrual rate above the integration level
(usually set at or below the maximum taxable wage base) than below it
7 This information is found in Table IV-7, p. 114 of Zorn’s 1997 survey report.
6 According
7
floor of 25 to 33 percent of salary. Most of the 14 plans that we examined require at least 5 years
of service, and some require 10, before the worker is eligible for disability benefits (see
Appendix A). Plans for police and firefighters tend to be more generous; most pay 50 percent
and some pay 75 percent for service-related disability.
State and local systems offer only limited survivor benefits. Prior to retirement, the
benefit is either a refund of employees’ contributions or a lump sum, whichever is greater. After
retirement, survivor benefits are available if the employee selects a joint-and-survivor option for
his annuity. These annuities offer reduced benefits while the worker is alive in exchange for
providing benefits to the widow or widower after the worker’s death. Systems without Social
Security coverage do not appear to provide any more extensive survivor protection than those
systems with coverage.
In order to mitigate the effect of inflation on retirement income, many public plans
provide retirees with some post-retirement cost-of-living adjustments (COLAs). About a third of
plans provide automatic increases linked to the Consumer Price Index (CPI), although these
increases are generally capped at 3 percent. A somewhat smaller percentage provides automatic
adjustments at a fixed rate specified by the plan, and others provide ad hoc adjustments. Among
small plans, 31 percent provide no COLA, but this figure drops to 11 percent for medium-sized
plans and to only 5 percent for large plans (see Table 5). The sample of uncovered plans we
examined appears to have more generous cost-of-living provisions than average; most provide an
automatic COLA linked to the CPI and a cap of 3 percent (see Table 6).
Public pension benefits are taxable under the federal personal income tax, to the extent
that they exceed employee after-tax contributions. The taxability of public benefits at the state
level was dramatically affected by the 1989 U.S. Supreme Court decision in Davis v. Michigan.8
This decision noted that federal law requires federal and military retirees be treated at least as
favorably as state and local retirees under state law. Facing a substantial loss of revenues, states
became less willing to exempt state and local pensions from taxation. One survey showed that,
whereas 45 of 85 plans in 1986 were totally exempt from state income taxes, this number had
dropped to 21 plans by 1996 (State of Wisconsin Retirement Research Committee, 1996).
State and local pensions are financed by a combination of employee and government
contributions plus earnings on pension assets. This differs markedly from the private sector,
where employees rarely contribute to defined benefit plans.9 Roughly 9 percent of annual
receipts come from employees, 20 percent from government employers, and the bulk (71 percent)
from investment earnings (see Table 1). As a percentage of payroll, contributions vary by type of
8 Davis
v. Michigan Department of the Treasury, 489 U.S. 803 (1989).
relatively recently, employees made their contributions to state and local plans from after tax dollars, so they
did not enjoy the benefits of deferral generally associated with contributions to employer-sponsored pensions. A
noticeable trend in the last fifteen years, however, has been the adoption of Internal Revenue Code 414(h)
provisions. Under these provisions, employers may “pick up” employees’ contributions, presumably in lieu of a
salary increase, or employees can continue to make contributions, but on a tax-sheltered basis (State of Wisconsin
Retirement Research Committee,1996).
9 Until
8
employee covered (see Figure 3). According to PENDAT, for general government, employee
contributions were 4.6 percent of payroll and government contributions were 11.1 percent; the
comparable numbers for police and fire were 7.2 and 16.4 percent, respectively.
In summary, state and local plans provide relatively generous pensions, although workers
bear a substantial portion of the cost. Plans not covered by Social Security appear to have higher
benefit accrual rates and more generous cost-of-living provisions than the group as a whole,
although none provide the full inflation protection offered by Social Security. State and local
plans also provide little in the way of dependent and survivor benefits, except the option of jointlife annuities. Disability protection is available, but only after 5 or 10 years of service, and
benefits are not fully indexed for inflation.
The Funding Status of State and Local Pension Plans
While the characteristics of public pension plans are important for determining what
benefit protections uncovered state and local workers currently lack, the funding status of public
plans has implications for the financial impact of mandating coverage. If public plans were
financed on a pay-as-you-go basis, plan sponsors – even if they reduced public plan benefits to
reflect Social Security coverage – could not expect to see any reduction in costs until the newly
covered workers started to retire. Only then would employers reap the benefit of having already
paid for the Social Security component of the combined retirement pension. The story is
different if public plans are advance funded. If employers reduced benefits under the state or
local plan, the new lower accrual rate would reduce employers’ public plan costs immediately,
because contributions to an advanced funded system are designed to cover benefits as they are
earned. Thus, the extent to which state and local plans are advance funded is crucial to assessing
the financial burden of mandatory coverage during the transition phase.
The funding of state and local pension plans has improved substantially over the last two
decades. One development in the late 1970s and early 1980s was a move to expand investment
options by substituting a general standard of prudence for the more restrictive “legal lists,” which
favored fixed investments, such as bonds, and often prohibited investments in stocks altogether.
The movement to a prudence standard allowed public plans to increase their holdings of equities.
The expansion of investment options was accompanied by renewed interest in prefunding
accruing liabilities.10 The combined effect of improved funding and expanded investment
options positioned public plans to take advantage of the remarkable stock market returns over the
last 20 years. The result has been a startling growth in the assets of public plans and a significant
improvement in the health of public plans.
10 For
example, in Massachusetts, the last state to defend unfunded or pay-as-you-go financing, the Massachusetts
Retirement Law Commission recommended in 1976 to fund all the state’s pension plans on an actuarial basis,
reversing the policy set in 1945 “to place all contributory retirement systems ... on a non-reserve basis.” The
Massachusetts legislature passed the first bill allowing systems to put aside reserves in 1978 and adopted a formal
funding schedule, moving ad hoc reserves into the structure of the retirement system, in 1987.
9
The improvement is especially striking when compared with the dire situation reported in
the 1978 Pension Task ForceReport. At that time, 23 percent of state plans and 16 percent of
local plans were financed on a pay-as-you-go basis. Only 45 percent of state plans and 26
percent of local plans were funded on an actuarial basis, whereby contributions covered normal
costs and amortized the accrued unfunded liability over 40 years or less.11 In fact, only 70
percent of state plans and 45 percent of local plans even produced an actuarial report at least
every three years. Among plans that did have recent actuarial valuations, the mean and median
ratio of assets to liabilities were both 51 percent, and this was an optimistic figure, since many
systems assumed no salary growth when projecting their liabilities.
Today, almost all public plans have regular valuations, and funding ratios have improved
substantially. For example, plans in the 1997 PENDAT data set that provided information on
both liability and assets had an actuarial accrued liability of $1,248.5 billion and assets of
$1,107.0 billion, yielding an aggregate funding ratio of 88.7 percent (see Table 7). In other
words, state and local plans had enough assets on hand to pay off 88.7 percent of their benefit
promises. The 88.7 percent figure in the 1997 survey reflects an upward trend in funding from
82.0 percent in the 1993 survey and 84.9 percent in the 1995 survey.12 Our survey of the 14
major plans not covered by Social Security shows slightly better funding ratios: 91 percent for
the group as whole (see Table 8). Given that state and local plan assets increased by 67 percent
between the end of 1996 and the second quarter of 1999, considerably faster than liabilities, total
plan assets now may well exceed liabilities.13
Another measure of the improvement in the funding status of public plans, when
comparable data are not available for early years, is the ratio of assets to benefit payments. As
shown in Figure 4, this ratio has a clear upward trend. In 1978, assets were 14.9 times annual
benefits, which – with benefits increasing at about 15 percent per year -- implied enough money
to cover 8 or 9 years. By 1997, the ratio had increased to 21.3, which – with benefits growing at
10 percent – was enough to cover between 11 and 12 years of benefit payments. Our estimate for
1999 shows an asset-to-benefit ratio of 27.2, which would be enough to cover benefit payments
for 13 to 14 years. Thus, regardless of the measure used, the funding status of state and local
pensions has improved enormously.
Social Security and State and Local Plans
This section shifts the focus from state and local plans to Social Security. It provides a
brief history of the relationship between Social Security and state and local plans, and then
explores the impact of mandatory coverage on Social Security.
11
This information is found in Table G. I., p. 151 of the Pension Task Force Report.
Limiting the comparison to matched cases, the funding ratio went from 83.9 percent in the 1995 survey to 87.4
percent in the 1997 survey (Zorn 1997, Table I-5, p. 90).
13 The Flow of Funds account data show state and local pension assets totaling $1,789.6 billion for year-end 1996
and $2,994.5 billion for the second quarter of 1999, a 67.3 percent increase.
12
10
A Brief History
When Congress enacted Social Security in 1935, it excluded many categories of workers
from the program, among them public employees. It omitted federal employees because most of
them were already protected under the federal Civil Service Retirement System. Although most
state and local employees were not similarly situated, Congress excluded them because it was
assumed that the U.S. Constitution prevented the federal government from taxing states and
municipalities. However, as Congress expanded Social Security coverage to include virtually all
private-sector employees, it also opened coverage to many public sector workers. Specifically,
amendments to the Social Security Act in 1950, 1954, and 1956 allowed states, with the consent
of employees in the pension plan, to elect Social Security coverage through agreements with the
Social Security Administration (making their taxation voluntary). The amendments also allowed
states to withdraw from the program after meeting certain conditions.
Policymakers have addressed the question of mandatory, as opposed to voluntary,
coverage of state and local workers on several occasions, but the major initiative came in
response to the financial problems faced by the system in the early 1970s. The 1977 Social
Security Amendments directed the Secretary of Health, Education, and Welfare (now Health and
Human Services) to undertake a thorough study of the extent of Social Security coverage and of
the desirability and feasibility of covering all government workers, including state and local
government employees. Although the Universal Social Security Coverage Study Group
(hereafter the “1980 Study Group”), which was established by the Secretary of Health,
Education, and Welfare, did not draw conclusions about the desirability of the various options,
Joseph Bartlett, the Chair of the Study Group, stated in the cover letter of the group’s report that,
in his judgment, “state and local workers should be brought into Social Security”(U.S.
Department of Health, Education, and Welfare, 1980).14
Although Congress has not extended mandatory coverage to state and local workers, it
has taken steps to eliminate some of the inequities that can arise from so-called “double dipping.”
The 1977 and the 1983 amendments both introduced windfall provisions so that spouses and
surviving spouses who had worked in uncovered employment could not claim full Social
Security spousal or survivor benefits (“Government Pension Offset”) or state and local workers
who gained minimum coverage under Social Security through second jobs could not profit from
the progressive benefit formula (“Windfall Elimination Provision”). These provisions will be
discussed in more detail in the next section.
The 1983 amendments also eliminated the option for states and localities that had
voluntarily elected Social Security coverage to withdraw from the program. And, in 1990,
Congress mandated coverage for state and local workers not covered by a pension plan with
benefits comparable to those provided by Social Security. The following is a brief summary of
the major legislative changes:
14 This
letter is found on page XIII of the report.
11
1935: Social Security Act is passed prohibiting participation by states and localities.
1950: Social Security Act is amended to permit coverage on an optional basis for state and local
employees not already covered by a public employee retirement system. Legislation also
permits state and local government groups to withdraw from Social Security two years
after submitting notice of intent.15
1954: Social Security becomes available to state and local employees already covered by a
pension system when the majority of the group elects coverage and the state agrees.
1956: Congress creates a divided option. If all new employees are to be covered by Social
Security, current workers can choose to be covered by Social Security and the public plan
or only by the public plan. Legislation also allows five states to cover police and fire
fighters, two groups that had been excluded from Social Security.
1977: Social Security Act Amendments introduce a “Government Pension Offset” to reduce
benefits to spouses and surviving spouses who worked in employment not covered by
Social Security.
1983: Social Security Act is amended to eliminate the ability of states and localities to withdraw
from the system16 and to enact the “Windfall Elimination Provision” to prevent higher
income earners from taking advantage of the benefit formula that was designed to provide
higher replacement rates for lower income earners.
1985: Consolidated Omnibus Budget Reconciliation Act requires all state and local government
new hires after March 1986 to participate in Medicare.
1990: Omnibus Budget Reconciliation Act requires coverage of all state and local government
employees not covered by a state and local plan that provides benefits comparable to
Social Security.
The Impact of Mandatory Coverage on Social Security
Social Security provides retirement, disability, and survivor protection to insured workers
and their families. Insured workers are eligible for full retirement benefits at age 65 and reduced
benefits at age 62. The 1983 Social Security Amendments gradually increased the age for full
benefits from 65 to 67 over the period 1999-2022. Benefits are based on career earnings,
indexed to reflect the growth in average wages over time. The benefit structure is progressive in
the sense that it replaces a higher percentage of the pre-retirement earnings of a low-wage earner
15 State
and local organizations were required to participate for at least five years before submitting a notice of intent
to withdraw. Any withdrawal was permanent.
16 By 1979, 674 jurisdictions had terminated coverage for 112,000 employees. An additional 222 notices were
pending, affecting an additional 98,100 employees.
12
than of a high earner. In 1999, the replacement rate for a single individual retiring at age 65 with
a history of average earnings ($28,861 in 1998) was 43.0 percent, for the low earner (45 percent
of the average national wage) 57.8 percent, and for the person earning at the taxable maximum
($68,400 in 1998) 25.4 percent (U.S. House of Representatives, Committee on Ways and Means,
1998). Benefits are adjusted fully for changes in the cost-of-living after age 62.
Social Security provides additional benefits for spouses and for survivors, who are
usually women. For a woman who spends her career taking care of her family, Social Security
provides a retirement benefit equal to 50 percent of her spouse’s benefit when her husband is
living. For a homemaker who becomes divorced after at least 10 years of marriage, Social
Security provides a retirement benefit based on 50 percent of her former spouse’s benefits if he is
still living. For an older woman whose husband (or former husband of at least ten years) dies,
Social Security provides widow’s benefits equal to 100 percent of her husband’s benefits. If
children are getting survivors’ benefits, younger widows also receive benefits based on their
husbands’ earnings while they are caring for the children and not working in paid employment.
Social Security pays disability benefits to workers whose disability is sufficient to prevent
them from doing any substantial gainful work and is expected to last at least 12 months or end in
death. The benefits are based on workers’ average indexed monthly earnings at the time of
disability and are adjusted annually to reflect changes in the cost of living thereafter. In addition,
benefits are available to the worker’s spouse and eligible children, subject to a family maximum.
Workers are eligible for disability insurance (DI) benefits if they have earnings in covered
employment for half the quarters since age 21.17 Workers are permanently and fully insured –
that is, their insurance will not lapse if they leave covered employment – once they have
accumulated 40 quarters of coverage.
For a long time, Social Security benefits were not subject to the federal personal income
tax. In 1983, however, Congress made them partially taxable and increased their taxation further
in 1993. Under current law, up to 50 percent of Social Security benefits are taxable if recognized
income falls within specified levels – $25,000 to $34,000 for single individuals and $32,000 to
$44,000 for married couples. If income exceeds these levels, then 85 percent of benefits are
taxable.18 At the state level, a 1992 survey showed that 26 states allowed a full exemption of
Social Security benefits from the income tax; 14 states allowed a partial exemption; and 10 states
had a very limited or no income tax (State of Wisconsin Retirement Research Committee, 1996).
17 Under
current law, workers over age 31 are eligible for Social Security disability benefits if they have covered
earnings in 20 of the 40 quarters immediately preceding disability. Younger workers can qualify if they had covered
wages in one half the quarters they worked since age 21. A minimum of 6 quarters is required. Workers are fully
insured when they have achieved one quarter of coverage for each year since 1950 or their 21st birthday, whichever
comes later. Workers are permanently and fully insured once they have forty quarters of coverage. Workers are
currently insured with at least 6 quarters of coverage during the last 13 quarters. Many state and local plans require
at least 5 years of service before a worker becomes eligible for disability benefits, and some require a significantly
longer period of employment.
18 The 85 percent figure is still short of the standard used for taxation of private and public pensions generally –
namely, full taxation of benefits in excess of those attributable to employee after-tax contributions.
13
The program has been financed primarily on a pay-as-you-go basis. In calendar 1999,
Social Security had revenues of $526.6 billion and expenditures of $392.9 billion. The bulk of
Social Security’s revenues comes from the payroll tax, which is 12.4 percent of payrolls up to an
established maximum. That maximum, which is adjusted each year to reflect the increase in
average earnings in covered employment, was $68,400 in 1998, $72,600 in 1999, and $76,200 in
2000. The Social Security trust funds held reserves of $737.9 billion at the end of 1999, equal to
2.3 times annual outlays.
Every year, the Social Security Trustees publish an actuarial report that includes three sets
of projections based on alternative assumptions about future mortality, fertility, economic
growth, and interest rates. The intermediate projection from the 2000 report shows that over the
next 75 years, Social Security has a long-run deficit equal to 1.87 percent of covered payroll
earnings.19 The Social Security actuaries estimate that extending coverage to all new state and
local workers would reduce the 75-year deficit by about 10 percent (0.22 percent of taxable
payrolls) and would extend the trust fund solvency by about 2 years (Advisory Council on Social
Security, 1997).20
The favorable impact on long-run costs comes from two factors.21 The first is the nature
of the workers not covered by Social Security, who tend in large part to be teachers. The average
salary for teachers not covered by Social Security is higher than the average salary for the nation
as a whole: $40,790 in the California Teachers Retirement System, $41,210 in the Teachers’
Retirement System of Ohio, and $42,393 in the Massachusetts Teachers Retirement System,
compared with average national earnings of $28,861 in 1998 (see Appendix A). In addition,
teachers tend to be disproportionately women. For example, in the major teachers plan, women
account for 62 percent of the total in California, 73 percent in Kentucky, 83 percent in Louisiana,
and 76 percent in Missouri. Most likely, their spouses have worked full time and therefore
would be unlikely to receive the spouses’ and survivors’ benefits provided by Social Security.
The combination of higher salaries and lower supplementary benefits would make teachers
relatively inexpensive new participants for the Social Security system.
The second way in which mandatory coverage affects the long-run financing of Social
Security involves covering the costs associated with the startup of the pay-as-you-go system.
When Social Security started, the president and Congress decided to pay benefits to the first
generation of retirees whose lives had been disrupted by the Great Depression. The workers who
had paid taxes only for a short period received benefits far in excess of their contributions. These
19 Between
now and 2015 the Social Security system will bring in more tax revenues than it pays out. From 2015 to
2024, the system will have to draw on the interest on trust fund assets to cover benefit payments, and thereafter it will
have to draw down trust fund assets until the funds are exhausted in 2037. If no tax or benefit changes were made,
current payroll tax rates and benefit taxation would provide enough money to cover roughly 70 percent of benefits
after 2037.
20 The data are found in Vol. 1, Table 1.1A, p.182.
21 In addition to the factors described below, extending coverage affects costs because the Social Security
projections run for only 75 years. Thus, workers who pay taxes within the 75-year horizon but receive at least part of
their benefits after the horizon help the actuarial calculation, even if they do not contribute to Social Security from a
lifetime perspective.
14
net transfers continued for many years, and totaled roughly $9 trillion. While early generations
enjoyed the benefits, later generations have to pay the bill. One way to describe the magnitude of
this unfunded liability relative to taxable earnings is to consider the option that every generation
gives up the same percentage of its earnings. Based on that assumption and the existing $9
trillion unfunded liability, Geanakoplos, Mitchell, and Zeldes (1998) estimate that roughly 3
percentage points of the current 12.4 percent payroll tax goes towards covering the startup costs.
In other words, 25 percent of the Social Security tax goes to cover the implicit interest costs of
the unfunded liability and improves the long-run financial outlook.
As a result of additional workers to share the burden of the unfunded liability and the
relatively low cost of the excluded workers, extending coverage to all new state and local
workers would improve Social Security’s long-run financial situation. Some opponents of
mandatory coverage have characterized this improvement as trivial: “…the resulting extensions
of the Trust Funds’ solvency for two years (to calendar year 2036) would be nearly meaningless
in terms of Social Security’s long-term projected deficit” (Segal, 1999). Perhaps a more
meaningful way to view the financial implications of the coverage issue is to consider its impact
today. With 30 percent of state and local workers not paying into Social Security, covered
workers and their employers face a combined employer-employee payroll tax that is 0.22 percent
higher than it needs to be (Advisory Council on Social Security, 1997). This means that the
worker with average annual earnings of roughly $30,000 currently pays an estimated additional
$66 in tax each year. All else equal, bringing in uncovered state and local workers would reduce
the burden on covered workers.
The Case for Mandatory Coverage of State and Local Workers
The most compelling arguments for mandatory coverage, however, have little to do with
the projected long-run deficit for the Social Security system. Rather the case rests on issues of
equity and ensuring that all state and local workers have protections not currently provided under
public plans.
Equity Considerations
All three factions of the 1994–1996 Advisory Council on Social Security recommended
mandatory coverage for newly hired state and local employees. The Advisory Council report
stated that: “…an effective Social Security program helps to reduce the public costs for relief and
assistance, which, in turn, means lower general taxes. There is an element of unfairness in a
situation where practically all contribute to Social Security, while a few benefit both directly and
indirectly, but are excused from contributing to the program” (Advisory Council on Social
Security 1997, Vol. I, p.19). The Council’s concern about equity reflects the fact that excluded
state and local government employees and their employers are not paying their share of income
redistribution, and they are not paying their share of financing the unfunded liability associated
with the startup of the Social Security program.
15
The first equity issue is straightforward. Social Security redistributes income from
workers with higher lifetime earnings towards workers with lower lifetime earnings. To the
extent that higher paid people do not participate in this national program of redistribution, it has
been argued that they place an extra burden on the rest of the population.22 Moreover, the
excluded state and local workers benefit from this redistribution; they have to pay fewer taxes
for means-tested programs for the elderly, and they live in a society with fewer poor elderly than
would exist in the absence of Social Security. Advocates of universal coverage contend that it is
only fair that state and local government workers and their employers, like their private and
public sector counterparts, participate in this national endeavor.
The second equity issue relates to paying for the unfunded liability associated with the
startup of the program that was discussed above in connection with the impact of mandatory
coverage on long-run Social Security costs. The early benefits during the 1940s, 1950s, and
1960s went to the parents, grandparents, and great grandparents of today’s uncovered state and
local workers. Those who support mandatory coverage believe that state and local workers
should pay their share of the inherited unfunded liability.
The exclusion of some state and local workers from Social Security used to raise further
equity problems, but these were largely addressed by the 1977 and 1983 amendments. State and
local workers who are not covered by Social Security in their government work can easily gain
coverage as a result of a second career or moonlighting. Since a worker’s monthly earnings for
purposes of benefit calculation are averaged over a typical working lifetime rather than over the
years actually spent in covered employment, a high-wage earner with a short period of time in
covered employment cannot be distinguished from an individual who worked a lifetime in
covered employment at an exceptionally low wage. Thus, a worker who was entitled to a state
and local pension and to Social Security could qualify for the subsidized benefits associated with
the progressive benefit formula. Similarly, a spouse who had a full career in uncovered
employment – and worked in covered employment for only a short time or not at all – would be
eligible for the spouse’s and survivor’s benefits.
In 1977, Congress established the Government Pension Offset (GPO) and in 1983
introduced the Windfall Elimination Provision (WEP) to reduce the unfair advantage enjoyed by
these “double dippers,” their spouses, or both.23 The provisions appear to have ameliorated the
22
Concerns about redistribution were much less pressing in 1935 when the decision was initially made to exclude
state and local government workers. The 1935 Social Security Act actually set up a system that more closely
resembled a defined contribution arrangement than does Social Security as we know it today. The initial legislation
planned for the accumulation of a trust fund and stressed the principle of a fair return for each worker, which made it
far less important that some workers were excluded from contributing. The 1939 amendments to the Social Security
Act introduced more redistributional elements into the system.
23 Under the GPO, SSA reduces Social Security benefits to those beneficiaries whose entitlement is based on
another person’s (usually their spouse’s) Social Security coverage. Social Security spouse and survivor benefits
were intended to provide protection to spouses with no earnings of their own, not to those with a career in uncovered
employment. The GPO reduces spouse or survivor Social Security benefits by two-thirds of the amount of a
government pension. Under the WEP, SSA uses a modified formula to calculate the benefit that applicants will
receive when they have had a limited career in covered employment.
16
problem. A brief survey of the literature produced no articles characterizing the offsets as
inadequate. In fact, the only reports were that several members of Congress have introduced
legislation to mitigate the offsets’ effect.24 On a purely administrative level, however, the
General Accounting Office (GAO) suggests that the offsets have not worked perfectly. The
GAO (1998a) reported that the Social Security Administration cannot always determine whether
an applicant should be subject to WEP or GPO, and this failure has led to some overpayments.
The GAO estimated total overpayments for the period 1978 to 1995 were between $160 million
and $335 million.
These amounts are relatively small, however, and play only a minor role in the arguments
for mandatory Social Security coverage for state and local workers. Even if the WEP and GPO
worked perfectly and Social Security were not facing a long-term deficit, many argue that equity
demands all state and local workers, just like all workers in the private sector, should participate
in Social Security. Exempting some state and local workers from paying into a redistributive
program and from paying their share of the unfunded liability increases the burden on the rest of
the population.
Improved Protection for State and Local Employees and Their Families
Equity considerations are only half the story, however. The second major argument for
mandatory Social Security coverage is that it would provide important protections for state and
local workers that they do not have now. Uncovered state and local workers face two types of
gap in protection. The first is a gap in insurance protection for those workers who move between
covered and uncovered employment or between two separate uncovered jobs. The second is a
gap in benefit protection for those workers who spend their careers in uncovered employment
and for their families, because Social Security provides some benefits not available in most state
and local retirement plans.
Gaps in insurance protection, particularly in the event of disability, occur because both
Social Security and state and local plans require a relatively long period before workers are
eligible for some types of benefits. Workers moving between jobs that are covered and jobs that
are not covered by Social Security may experience long periods without disability and survivor
protection. When young workers leave covered employment and go to work for a state or local
government not covered by Social Security, their insured status under Social Security may lapse
if they have not accumulated 40 quarters of coverage. Since it may take 5 years, or even 10
years, to become insured under the public plan, they will have a significant period with no
protection at all should they become disabled or for their survivors. Similarly, a worker who
leaves a noncovered state or local plan may have to wait five years before gaining insured status
24
Senator Barbara Mikulski (D-MD) and Congressman William Jefferson (D-LA) have introduced legislation that
would remove the spousal offset for anyone whose combined public pension benefit and spousal Social Security
benefit is less than $1,200 per month. Congressman Barney Frank (D-MA) has introduced legislation to limit the
Windfall Elimination Provision.
17
under Social Security.25 Although young, such workers are losing substantial value – that is,
insurance against catastrophic risk. Should an unfortunate event occur, the costs would be great,
because the workers face a very long period with little earnings.
Gaps in benefit protection arise because state and local plans do not always provide the same
benefits as those offered by Social Security. Benefit protections often missing from state and
local pensions are portability, dependents’ and survivors’ benefits, and full cost-of-living
adjustments.
Portability. Defined benefit plans have the advantage of offering a predictable benefit,
expressed as a percentage of final pay for each year of service. A problem arises, however, in the
case of mobile employees, since portability of state and local defined benefit pensions is usually
limited to employment within state government. The problem would arise even if all state and
local plans and private employers had identical plans and immediate vesting. Mobile employees
receive significantly lower benefits as a result of changing jobs than they would receive from
continuous coverage under a single plan, because a measure of final earnings levels usually
determines pension benefits in defined benefit plans. A worker who remains with a plan receives
benefits related to earnings for a period, e.g., 5 years, just before retirement. The benefits for
mobile employees, however, are based on earnings at the time they terminate employment. A
simple example indicates that if wages increased 4 percent annually, the pension of a worker who
held four jobs would equal 61 percent of the pension of a worker who remained continuously
employed by one firm (see Table 9). The more wages rise with productivity and inflation, the
relatively lower the benefits received by the mobile employee, because defined benefit plans –
that is, most public plans – do not account for real wage growth and inflation between
termination and retirement. In contrast, Social Security allows employees to build on previous
earnings as they move from job to job, so mobility does not reduce benefits.
Dependents’ and survivors’ benefits. As noted above, most state and local plans provide
little in the way of dependents’ benefits. They provide nothing for the spouse of a retired worker
when the worker is alive; Social Security offers a benefit equal to 50 percent of the employee’s
for spouses without significant earnings records and provides similar benefits to divorced
spouses who have been married at least 10 years. State and local plans generally offer only
modest survivor benefits: before retirement the benefit is often either a refund of employees’
contributions or a lump sum, whichever is greater. After retirement, survivor benefits are
available only if the employee selects a joint-and-survivor option for his or her annuity. The
extent to which workers select joint-and-survivor options is problematic.26 Social Security
25
A similar problem occurs for workers moving between uncovered plans in different states.
left on their own, tend to select single-life as opposed to joint-and-survivor annuities, which would
continue payment to a worker’s spouse after the worker’s death. Single-life annuities were very popular in
employer-provided retirement plans prior to ERISA (1974). This pattern began to change when ERISA provided for
a 50-percent joint-life annuity as the default provision. Holden (1997) estimates that 48 percent of men with
pensions beginning before 1974 had joint-and-survivor pensions, while 64 percent of those beginning after 1974 did
so. Similarly, the U.S. GAO (1992) estimates that the percentage selecting a single-life annuity dropped by 15
26 Many workers,
18
provides a worker’s widow with benefits equal to 100 percent of the worker’s pension and also
provides benefits to young widows.
Cost-of-living adjustments. State and local pension plans are generally much better than
private sector plans at providing post-retirement cost-of-living adjustments, and those plans
under which workers do not have Social Security coverage tend to provide more COLA
protection than the average. Almost all the plans that we examine closely in this study provide
an automatic adjustment (see Table 6), compared with roughly half of large state and local plans
generally (see Table 5). Nevertheless, the COLA adjustments were generally capped at 3
percent, and it is unclear what the sponsors of these plans would do if inflation started to
increase. Social Security, on the other hand, provides full cost-of-living adjustments no matter
how rapid inflation might be, ensuring that beneficiaries do not see the value of their pension and
insurance benefits eroded.
The guarantees of portability, full inflation protection, and generous ancillary benefits,
which are offered only by Social Security, mean that extending mandatory coverage to state and
local employees would bring some real gains in benefits to certain groups of workers and their
spouses. Proponents argue that such guarantees would also eliminate gaps in disability
protection that occur as workers move between covered and noncovered employment.
The Costs of Mandating Coverage for All State and Local Employees
The case for mandatory coverage, therefore, rests on equity considerations and offering
excluded state and local workers important protections that they do not have currently. The
arguments against mandatory coverage center on the issue of costs. Opponents of mandating
Social Security coverage for state and local workers also raise concerns about destabilizing
existing plans for current employees and about morale issues associated with having new
employees receive different pension benefits than current workers. Others raise the issue of
administrative burden.
Almost all proposals for mandating Social Security coverage are limited to new
employees only (Advisory Council on Social Security, 1997 and Moynihan, 1999). Focusing on
new employees avoids personnel problems and legal challenges associated with benefit
reductions for existing personnel; it requires no “hold harmless” provisions; it means each
employee deals with only one benefit formula; and it allows a gradual phase-in of the required
tax increase (U.S. Department of Health, Education, and Welfare, 1980).
Although covering only new employees would ease the transition, costs would increase.
The ultimate cost of the combined Social Security/public pension system would depend crucially
on how plan sponsors would respond to the introduction of Social Security. As previous
analyses of mandatory coverage have pointed out, it is plausible that states and localities would
change their retirement and disability plans for new employees and not simply add Social
Security on top of existing provisions; if not, the resulting package would be too expensive and
percentage points after the 1984 Retirement Equity Act (REA), which required notarized spousal approval before
selecting a single-life annuity. Thus, the degree of joint-life annuitization is very sensitive to system design.
19
produce unduly high replacement rates. More likely, they would explicitly adjust their existing
pension provisions to reflect Social Security’s benefits and costs. Indeed, as noted above, benefit
accrual rates are lower for plans covered by Social Security than for those plans not covered.
Although the work of the 1980 Study Group is now 20 years old and many things have
changed – including the Social Security tax rate, the taxation of benefits, and the coverage of all
newly hired state and local workers under Medicare – it still provides the most extensive analysis
of possible cost increases. For this reason, recent studies, such as the 1998 GAO report, have
built on the 1980 work. The Study Group examined two sources of data. One source was the
Actuarial Education and Research Fund (AERF) study of 25 representative uncovered systems
(Actuarial Education and Research Fund, 1979). The second source was the Urban Institute
study, produced in collaboration with the actuarial firm of Howard Winklevoss and Associates,
Inc., of 22 of the 50 largest uncovered plans. The difference between these two studies is that the
AERF relied on the local actuaries to design the revised benefit formulas, while the Urban
Institute specified the revisions so that the results for different plans were directly comparable.
Both studies employed the actuarial and economic assumptions actually used by the various state
and local plans.
The 1980 Study Group analyzed three different responses to Social Security by state and
local plans. The first was the reproduction of existing benefits or the “constant benefit”
response. Here, the AERF attempted to preserve first-year benefits for all age and salary
combinations, while the Urban Institute selected a single target group – an unmarried employee
with average age, salary, and service characteristics.27 (The problem of equalizing first-year –
rather than lifetime— benefits will be discussed below.) The second analysis, by AERF, asked
the actuaries to estimate the costs of the plans that they thought “most likely” to be adopted. The
third analysis, by The Urban Institute, reproduced “typical benefits” for employees already
covered by Social Security.
As shown in Table 10, the AERF study calculated that the normal cost of maintaining
“constant benefits” in large state and local plans would be roughly halved (from 14.37 percent of
payroll to 7.19 percent of payroll) and reduced by a third (from 23.67 to 16.13 percent of payroll)
for small plans, which cover primarily police and fire. To calculate total costs, however, it is
necessary to add the current Social Security payroll tax of 12.4 percent of taxable payrolls to the
reduced state and local costs. In terms of total cost, the net increase would be about 5 percent of
payroll, according to the AERF calculations, to maintain constant benefits. The Urban Institute
produced comparable results (see Table 11). If plans adopted the actuaries’ “most likely”
scenario, the increase would be about 6 percent of payroll (see Table 10). If the plans adopted
the “typical plan” of those pensions with Social Security coverage, the cost increase would be
between 6 and 11 percent of payroll (see Table 11).
27
Constant benefits were estimated on an after-tax basis to reflect the fact that state and local pensions were taxable
under the federal personal income tax, while Social Security benefits were not. This situation changed with the 1983
Social Security Amendments, which made Social Security benefits partially taxable. Thus, the estimates of “constant
benefits” are understated to the extent that Social Security benefits no longer have a full tax advantage.
20
Colorado, Illinois, and Ohio have recently undertaken studies to determine the cost to
their systems of mandatory Social Security coverage and found costs to be in the range estimated
by the Universal Coverage Study Group. For example, a 1997 study for an Ohio plan found that
providing new employees with retirement benefits that approximated those for current employees
would require an increase in contributions of between 6 and 7 percent of new employee payrolls.
Similarly, a study for the State University Retirement System of Illinois concluded that it would
cost an additional 6 percent for the combined Social Security/state system to provide benefits
comparable to the current state plan. These studies also equalized first-year, rather than lifetime,
benefits.
While costs would almost certainly increase with mandatory coverage, existing estimates
of the cost of maintaining “constant benefits” are misleading. They represent the costs of
duplicating first-year benefits for an unmarried worker in the middle of the age, service, and pay
ranges. As the 1980 Study Group and some other reports point out, participants would actually
get more in lifetime benefits under the combined Social Security/public system than they get
under the public plan alone for two major reasons. First, Social Security provides full
cost-of-living adjustment, while public pension plans typically provide either ad hoc adjustments
or COLAs capped at 3 percent. Second, Social Security also provides generous spouse and
survivor benefits for qualifying married and divorced couples.
If the goal were to equalize lifetime benefits, public plans could be cut back significantly
more than assumed in most calculations. For example, consider the issue of the COLA and how
much less the state and local plans could provide to achieve equal real lifetime benefits.28 As
shown in Table 12, the answer depends on the rate of inflation. If the inflation rate is zero, then
the reduced state and local benefit used in the calculations of the impact of mandatory coverage
is correct. Even with an inflation rate of 3 percent, however, the full COLA provided by Social
Security becomes important. For example, the non-Social-Security-covered plans in California,
Louisiana, Massachusetts, and Texas, which guarantee less than 3 percent, could provide 78 to
88 percent of the reduced state and local benefit and still have their employees end up with equal
lifetime benefits. At a 6 percent inflation rate, most of the plans could provide 50 percent or less
of reduced state and local benefits. While not a plausible long-term rate, it is worth noting that at
a 9 percent inflation rate, in most cases, Social Security alone would provide as much or more in
real lifetime benefits as was provided under the original state and local plan.
Nevertheless, costs for state and local plans would increase with mandatory Social
Security coverage. This would be true even if the excluded state and local plans provided the
same type of benefits granted by Social Security, and even if it were possible to design revised
state and local benefit provisions that produced constant lifetime benefits. The reasons for the
28
Similarly, if benefits are held constant for a single worker, married workers will come out ahead because Social
Security provides spouse and survivor benefits. It is somewhat less clear how to adjust for this factor, however.
Unlike the case of the COLA where a constant adjustment could be made for all participants, workers vary in the
extent to which they are married and the extent to which the earnings of their spouses will entitle them to spouse and
survivor benefits.
21
inescapable increase are the costs associated with the unfunded liability and the redistribution to
lower paid workers. The extent to which costs increase beyond this minimum depends to a large
extent on the nature of the public pension, because this affects the sponsor’s ability to preserve
the existing level of benefits in the combined Social Security/public plan. The potential to
produce a combined plan with the same average benefits is much greater if the public plan is
quite similar to Social Security. To the extent that the plan differs, merging the plans has the
potential of creating winners and losers, and the desire to avoid imposing losses may end up
producing a more expensive package.
This factor helps explain why analysts project significant cost increases for state and local
plans, although extending Social Security coverage to federal workers had little impact on federal
pension costs.29 The similarity between Social Security and the Civil Service Retirement System
(CSRS) greatly exceeds that between Social Security and public plans. For example, both CSRS
and Social Security fully adjust benefits for inflation after retirement. This is not true generally
for state and local plans. Second, CSRS has much lower turnover rates than state and local plans,
and turnover saves money in defined benefit plans.30 Finally, the CSRS package of ancillary
benefits for survivors closely parallels that provided by Social Security. Most state and local
plans provide little in the way of ancillary benefits. Thus, the larger cost increases for state and
local plans would be due to strengthening the COLA, reducing benefit forfeitures when workers
move among jobs, and providing substantial benefits to dependents and survivors. Since these
would be real gains from Social Security participation, they would add to total cost.
When considering the impact of higher pension costs on public plans, it is necessary to
consider who would bear the burden in the long run. If contributions to finance pension benefits
were to increase by 5 to 6 percent of payrolls for all workers in the economy, economists almost
universally agree that the burden of the tax would shift back onto the worker. Private sector
employers establish how much they will pay in total compensation, and if they face an increase in
one component, such as retirement costs, they will reduce another component, such as other
fringe benefits or cash wages.31
29 The
GAO (GAO/T-GGD-98-27) reported that the normal cost for CSRS was 25.14 percent of payrolls (7 percent
paid by the employee and 18.14 percent paid by the employer) in 1996, while the federal government’s cost for the
Federal Employee Retirement System (FERS) was 20.9 percent. The FERS cost consisted of 10.7 percent for the
defined benefit, 6.2 percent for Social Security, and 4 percent for the employer match in the Thrift saving Plan. To
get the total cost of the new system, it is necessary to add the 6.2 percent paid by employees for Social Security.
Thus, the combined employee and employer costs under FERS are very close to the total costs of CSRS.
30 State and local plans – like all defined benefit plans – gain financially from the fact that the value of the mobile
employee’s terminated benefit does not keep pace with wage growth between termination and retirement. In fact, the
1980 Study Group estimated that for a teacher plan, even with the return to employees of their own contributions,
mobile employees save the pension plan up to 3.9 percent of payrolls. For a given level of expenditure, a change to
include Social Security will benefit mobile workers at the expense of career workers. If the desire is to keep career
employees whole, total costs have to increase. If a plan has few mobile workers (CSRS), the cost increase will be
modest; if it has many (state and local), the required cost increase would be greater.
31
A sizeable literature has attempted to document empirically the tradeoff between pension benefits and wages.
Early studies (Schiller and Weiss, 1980 and Smith and Ehrenberg, 1983) found little evidence of such a tradeoff.
However, each of these studies employed a spot market model that examined only period-by-period tradeoffs of
wages for pensions. More recent studies have corrected for this problem by computing lifetime compensating
22
Extending coverage raises different issues, however. First, in the present case, it pertains
to the public, not the private, sector. Second, roughly half the cost increase would involve
extending a tax that 95 percent of workers in the country already pay.
The literature suggests that the wage setting process in the public sector differs
substantially from that in a profit-maximizing private sector firm (Ehrenberg, 1980). At the same
time, recent surveys from the Bureau of Labor Statistics indicate that for any individual job class,
the compensation in the two sectors is quite comparable (Braden and Hyland, 1993). That is, a
secretary in the government sector appears to receive roughly the same compensation as a
secretary in the private sector. This equivalence provides a basis for analyzing the incidence of
extending coverage.
The second step is to separate the portion of the cost increase associated with the
unfunded liability from that required to finance real benefit improvements. Taxpayers would
likely pay the tax component, since the compensation of employees in plans without Social
Security most probably has already been reduced to reflect this component of the payroll tax.
Compensation for workers, such as secretaries, is set in the rest of the economy, where this
Social Security tax already exists. Thus, state and local government employers without Social
Security coverage – a very small fraction of the total – pay market rate for secretaries. The
government employer would not be able to further reduce secretary compensation and remain
competitive in the labor market, and therefore would have to cover the unfunded liability cost
with tax revenues.
The incidence of the cost increase associated with real benefit improvements is different.
In this case, the compensation in plans without Social Security coverage includes compensation
to match the protection offered by Social Security. As a result, if Social Security coverage were
introduced, the higher costs associated with increased benefit protection would decrease nonSocial-Security compensation in order to keep the total compensation package in line with that in
the rest of the economy. Thus, both the taxpayer and the employee would bear some of the cost
increase.
The Segal Company (1999) study estimates that the total costs of mandatory Social
Security coverage for the first five years after the mandate took effect would be $1.6 billion, $3.3
billion, $5.1 billion, $7.0 billion, and $9.0 billion, for a cumulative total of $26.0 billion.32 As
differentials that examine the tradeoff over several periods. Using the 1983 Survey of Consumer Finances,
Montgomery et al. (1992) find a substantial compensating differential between the expected present value of
pensions and wages. The magnitude of this differential suggests a pension-wage tradeoff of dollar for dollar over the
worker’s lifetime. Gunderson et. al (1992) find similar evidence of a tradeoff. In another study, Montgomery, Shaw,
and Benedict (1997) examine the magnitude of compensating differentials across job sectors and find that the
pension-wage tradeoff is greater in small, nonunion firms than in large, unionized firms. Montgomery, Shaw, and
Benedict suggest that workers in the large, unionized sector pay a lower implicit price for their pensions because
pensions in large firms either provide productivity-heightening benefits to the firm or constitute economic rents that
the firm prefers to pay to the worker in the form of a pension.
32 These estimates assume that payrolls grow at 3 percent and that 9 percent of employees turn over each year. The
statistics presented on page 12 of the Segal Company study are mislabeled. For, example, the “Fifth year employee
23
just discussed, taxpayers would not bear all of these costs, because some reflect real benefit
improvements. Nevertheless, to get some understanding of the magnitudes involved, it is useful
to consider the Segal numbers and place them in the context of total state expenditures.
Assuming, conservatively, that state expenditures will grow at 3 percent in nominal terms over
the next five years, the total increase in costs due to mandatory Social Security coverage would
amount to 0.6 percent of state expenditures over the five-year period. Of course, this is an
average, and the burden would not be distributed evenly; those states with the largest shares of
employees not covered by Social Security would see the greatest increase in costs. As shown in
Table 13, the costs of coverage in Ohio for the first five years would amount to about 2.1 percent
of total state expenditures if the taxpayer were to bear the entire burden, which is unlikely given
that some of the increase reflects real benefit improvements.
To the extent that taxpayers would bear some of the costs, states are in a better position
than they have been in the past to take on additional fiscal burden. State government budgets
reflect the strong economy in the form of soaring revenues and diminished expenditures for needrelated programs (National Governors’ Association, 1999). Governors are proposing moderate
tax cuts for fiscal year 2000 of $3.8 billion. Fiscal year 2000 is the sixth consecutive year that
states have reduced taxes and fees, for a total decrease of $25.9 billion over the six-year period.
That means in the last six years, states have reduced taxes by roughly the amount that mandatory
coverage would cost in the first five years. Moreover, during the past several years, states have
been building up “rainy day” fund balances and ending balances. These balances, which amount
to 7.4 percent of expenditures for fiscal year 1999, should prevent major disruptions in services
in the event the economy slows from its current rapid pace of growth.
Although state finances are healthy, taxpayers in some states where public employees are
not covered by Social Security are already contributing to pay off the unfunded liability of their
own public pension plan. For example, despite the great improvement in funding status, plan
assets are only 67 percent of accrued liability in the Illinois teachers system, 69 and 72 percent,
respectively, in Louisiana’s teachers and state employee systems, and 78 and 87 percent in the
two major Massachusetts systems (see Table 8). The taxpayers in these states are paying for
benefits for workers who provided services to their parents and grandparents. In some sense,
asking this generation of taxpayers to help cover Social Security’s unfunded liability is asking
them to pay twice for the lack of funding in the past. For this reason, it might be worth
considering some transition relief to state and local governments if they were to bring their
employees into Social Security.
In addition to concern about higher costs, some opponents suggest that mandatory
coverage could destabilize existing systems and create morale problems when state and local
workers with different levels of seniority have different pension plans. The morale issue does
not appear to have been a problem at the federal level, where the federal government introduced
a new pension system in 1986. Different pension plans are much less likely to affect morale than
different wages, particularly if the government sets the combined Social Security/public plan for
and employer Social Security cost” is shown as $26,021,562,351, when, in fact, that is the cumulative total over the
five-year period, not a single year amount.
24
new workers to roughly match the plan for current employees and carefully explains the changes
to new hires. Redesigning the plan would, however, require opening up pensions for negotiation
in the case of unionized workers, which raises the possibility of ending up with a very different
type of pension arrangement. A dramatic change in pension design seems unlikely, however,
since the vast majority of state and local systems with Social Security coverage remain defined
benefit. Police and fire fighters should have the least to fear, since localities would have to retain
most of their existing pension provisions to accommodate the early retirement needs of these
workers in physically demanding and potentially dangerous jobs.
Some argue that extending Social Security coverage to the 30 percent of state and local
workers currently not covered would create a serious administrative burden. It is important to
remember, however, that all employees and employers in the private sector and 70 percent in the
public sector now participate in Social Security. The states would require some time to adapt
their plans to mandatory coverage provisions. The federal government required three years to
enact a new federal pension plan after Congress mandated Social Security coverage for federal
employees in 1983. The Report of the 1980 Study Group (U.S. Department of Health, Education
and Welfare) estimated that it would take four years for states and localities to redesign pension
formulas, legislate changes, and educate employees. The GAO reports that their discussions with
employers, employees, and pension officials also confirm that four years would be required to
complete the complex and time-consuming task of negotiating with state legislatures and
employee representatives. Once negotiated, however, the administrative requirements should not
be significant.
In short, the major argument against mandating coverage of all state and local workers
under Social Security is not administrative burden, morale, or destabilizing public plans, but
rather costs. The cost increases would be significant, and to the extent that they would fall on
taxpayers may merit consideration of some transition relief.
Conclusion
Proponents of mandatory Social Security coverage contend that universal participation is
the only equitable way to run a national social insurance plan that redistributes income from
those with high lifetime earnings to those with low lifetime earnings. Most high earners would
choose not to participate if they had the option; if they could withdraw, the system would become
unsustainable. Universal coverage is also the only way to ensure that all workers pay their share
of the unfunded liability associated with the startup of the program. Today, 30 percent of state
and local workers benefit from the system – albeit indirectly – without contributing to the
program’s support. At the same time, these excluded workers lack some important protections –
portability, full inflation protection, and survivors’ benefits – that only Social Security now
offers. These workers also face gaps in disability and survivor protection, to the extent that they
move between covered and uncovered employment.
The legal issues involved in mandatory Social Security coverage for state and local
employees are beyond the scope of this study, but they remain important. The GAO (1998b)
25
concluded that “mandatory coverage is likely to be upheld under the current U.S. Supreme Court
decisions.” Further, in a recent case, Reno v. Condon, No. 98-1468 (U.S.S.Ct., January 12,
2000), the Supreme Court signaled that, while it is respectful of state sovereignty, it will rule
cautiously when reviewing federal-state relationships. Thus, how the Court will rule regarding
mandatory Social Security coverage appears uncertain today.
While mandatory Social Security coverage may be equitable and improve benefits,
estimates indicate it would likely raise costs by 5 to 6 percent of payroll. Part of the increase
would reflect improved real benefits, the cost of which would probably be borne by the workers;
part would reflect a tax to cover Social Security’s unfunded liability, the cost of which would
probably be borne by the taxpayer. Since mandating coverage would change the rules for a
significant number of taxpayers, any proposal to mandate coverage should probably consider
some transition relief to state and local governments as they bring their employees into Social
Security. It is also important to pay attention to the lessons learned from the extension of
Medicare coverage for employees hired after March 1986. For example, one sensitive area is the
definition of a “new employee.” States object to classifying a teacher who moves from one town
to another but remains within the teachers retirement system as a new employee. Every effort
should be made to use reasonable definitions to ensure both economic efficiency and fairness.
Obviously, flexibility and goodwill would be required on all sides to solve this problem.
26
REFERENCES
Actuarial Education and Research Fund (AERF). 1979. Report on the Study of the Feasibility of
Mandating Universal Coverage under Social Security for the Participants of State and
Local Public Employee Retirement Systems, Deliverable under contract #600- 79-0052 to
The Universal Coverage Study Group of the Department of Health, Education and
Welfare, Social Security Administration (November 30).
Advisory Council on Social Security. 1997. Report of the 1994-1996 Advisory Council on Social
Security. Washington, DC: U.S. Government Printing Office.
Advisory Council on Social Security. 1980. Report of the 1979 Advisory Council on Social
Security. Washington, DC: U.S. Government Printing Office.
Annual Report of the Board of Trustees of the Federal Old-Age and Survivors Insurance and
Disability Insurance Trust Funds. Washington, DC: U.S. Government Printing Office,
2000.
AFL-CIO Public Employee Department. 1979. “Testimony submitted to the Study Group.”
Washington, DC (April 12).
Ball, Robert. 1978. Social Security Today and Tomorrow. New York: Columbia University
Press.
Bleakney, Thomas P., FSA. 1972. Retirement Systems for Public Employees. Philadelphia, PA:
Pension Research Council.
Board of Governors of the Federal Reserve System. 1996. Flow of Funds Accounts of the United
States: Flows and Outstandings. Washington, DC.
Braden, Bradley R. and Stephanie L. Hyland. 1993. “Cost of Employee Compensation in the
Public and Private Sectors,” Monthly Labor Review (May).
California State Teachers’ Retirement Board. 1999. “Mandatory Social Security Impact on
California State Teachers’ Retirement System.”
Colorado State University. 1998. “Mandatory Social Security Impact on New Hire State
Employees at Colorado State University.”
Denver Public School Employees' Pension and Benefit Association. 1999. “Mandatory Social
Security Impact for New Hires on Denver Public Schools.”
Ehrenberg, Ronald G. 1980. “Retirement System Characteristics and Compensating Wage
Differentials in the Public Sector,” Industrial and Labor Relations, 33(4), 470-484.
27
Fire and Police Pension Association of Colorado. 1999. “Study of the Impact of Mandatory
Social Security Coverage.”
Fox, Alan. 1979. “Earnings Replacement Rates of Retired Couples: Findings from the
Retirement History Study.” Social Security Bulletin. 42(1).
Geanakoplos, John, Olivia S. Mitchell, and Stephen P. Zeldes. 1998. “Would a Privatized Social
Security System Really Pay a Higher Return?” in Framing the Social Security Debate:
Values, Politics and Economics, edited by A. Douglas Arnold, Michael J. Graetz, and
Alicia H. Munnell. Washington, DC: Brookings Institution Press for the National
Academy of Social Insurance.
Grad, Susan. 1996. “Income of the Population 55 or Older, 1994.” Washington, DC: Social
Security Administration, Office of Research and Statistics.
Greenwich Research Associates. 1977. Public Pension Funds Trustees Report.
Gunderson, Morley, Douglas Hyatt, and James Pesando. 1992. “Wage-Pension Trade-Offs in
Collective Bargaining Agreements.” Industrial and Labor Relations Review 46 (1): 14660.
Holden, Karen. 1999. “Women as Widows under a Reformed Social Security System,” in
Prospects for Social Security Reform, edited by Olivia S. Mitchell, Robert Myers, and
Howard Young. Pension Research Council Publication: University of Pennsylvania Press.
Louisiana School Employees’ Retirement System. 1998. Actuarial Valuation.
Mitchell, Olivia, David McCarthy, Stanley C. Wisniewski, and Paul Zorn. 1999. “Developments
in State and Local Pension Plans,” Pension Research Council Working Paper 99-4.
Montgomery, Edward, Kathryn Shaw and Mary Ellen Benedict. 1992. “Pensions and Wages: An
Hedonic Price Theory Approach.” International Economic Review 33 (1): 111-28.
Montgomery, Edward, Kathryn Shaw and Mary Ellen Benedict. 1997. “Pensions and Wage
Premia.” Economic Inquiry 35 (3): 510-22.
Moynihan, Daniel Patrick. 1999. “Social Security Solvency Act of 1999: Brief Description of
Provisions.” (www.senate.gov/~moynihan/summary.htm).
Munnell, Alicia H. 1977. The Future of Social Security. Washington, DC: The Brookings
Institution.
28
Munnell, Alicia H. and Annika Sunden with the Assistance of Cynthia Perry and Ryan Kling.
1999 forthcoming. “Investment Practices of State and Local Pensions.” Pensions in the
Public Sector. Editors: Olivia Mitchell and Edwin Hustead. Pension Research Council
Publication: University of Pennsylvania Press.
Munnell, Alicia H. 1979. Pensions for Public Employees, Washington, DC: National Planning
Association.
Myers, Robert J. 1976. Social Security. Homewood, Illinois: Richard D. Irwin.
National Education Association. 1998. Characteristics of 100 Large Public Pension Plans.
Research Division (September).
National Governors’ Association and National Association of State Budget Officers. 1999. The
Fiscal Survey of States (June).
Price, Daniel. 1979. “Experience of Federal Annuitants under OASDHI: Age and Sex.” Social
Security Bulletin, 42(7).
Price, Daniel and Andrea Novotny. 1975. “Federal Civil-Service Annuitants and Social
Security.” Social Security Bulletin, 40(11).
Public Employee Retirement Administration Commission of Massachusetts. 1998. Actuarial
Valuation.
Public Employee Retirement Administration Commission of Massachusetts. 1998. Letter to
GAO from Massachusetts Public Employee Retirement Administration Commission.
Public Employee Retirement Administration Commission of Massachusetts. 1999.
Massachusetts Public Employee Retirement Guide.
Public Employees’ Retirement Association of Colorado. 1998. Actuarial Valuation.
Public Employees’ Retirement Association of Colorado. 1999. “Comparison of Colorado PERA
and Social Security.”
Public Employees’ Retirement Association of Colorado. 1999. “Mandatory Social Security
Impact on Colorado PERA and Its Covered Members and Employers.”
Public Employees’ Retirement Association of Colorado. 1998. Member Handbook.
Public Employee Retirement System of Nevada. 1998. Comprehensive Annual Financial Report.
Public Employee Retirement System of Nevada. 1998. “Mandatory Social Security Impact on
Nevada”.
29
Public Employee Retirement System of Nevada. 1998. Member Handbook.
Public Employee Retirement System of Ohio. 1997. “Actuarial Considerations of Proposed
Mandatory Social Security Coverage of New Hires.”
Public Employee Retirement System of Ohio. 1998. Comprehensive Annual Financial Report.
Public Employee Retirement System of Ohio. 1997. “Mandatory Social Security and its Impact
on Ohio Public Employees and Employers.”
Public Employee Retirement System of Ohio. 1999. Member Handbook.
Public School Retirement System of Missouri. 1997. Comprehensive Annual Financial Report.
Public School Retirement System of Missouri. 1997. Member Handbook.
Schiller, Bradley and Randall Weiss. 1980. “Pensions and Wages: A Test for Equalizing
Differences.” Review of Economics and Statistics 62 (4): 529-38.
Segal Company. 1999. “The Cost Impact of Mandating Social Security for State and Local
Workers” for the AFSCME, AFL-CIO, and Coalition to Preserve Retirement Security
(May).
Smith, R.S. and Ronald Ehrenberg. 1983. “Estimating Wage-Fringe Tradeoffs: Some Data
Problems.” In Measurement of Labor Cost, edited by Jack E. Triplett. University of
Chicago Press.
Social Security Administration. 1999. Facts & Figures about Social Security 1999.
Social Security Administration Office of the Inspector General. 1996.“Social Security Coverage
of State and Local Government Employees” (A-04-95-06013, Dec 13).
State of Wisconsin Retirement Research Committee. 1996. Staff Report No.82, Comparative
Study of Major Public Employee Retirement Systems (December, 1996).
State Teachers’ Retirement System of California. 1999. Actuarial Valuation.
State Teachers’ Retirement System of California. 1999. Member Handbook.
State Teachers’ Retirement System of California. 1999. “Mandatory Social Security Impact on
California State Teachers’ Retirement System.”
State Teachers Retirement System of Ohio. 1998. Benefit Summary.
30
State Teachers Retirement System of Ohio. 1999. “Mandatory Social Security Impact on the
State Teachers Retirement System of Ohio.”
State Teachers Retirement System of Ohio. 1998. Actuarial Valuation.
State Teachers Retirement System of Ohio. 1998. Member Handbook.
State Universities Retirement System in Illinois. 1998. “SURS Alternative under Universal
Social Security Coverage,” (June).
State Universities Retirement System in Illinois. 1999. Actuarial Valuation.
State Universities Retirement System in Illinois. 1998. Comprehensive Annual Financial Report.
Teachers’ Retirement System of Georgia. 1999. Comprehensive Annual Financial Report.
Teachers’ Retirement System of Georgia. 1999. Member Handbook.
Teachers’ Retirement System of Illinois. 1998. Comprehensive Annual Financial Report.
Teachers’ Retirement System of Illinois. 1999. Member Handbook.
Teachers’ Retirement System of Kentucky. 1997. Comprehensive Annual Financial Report.
Teachers’ Retirement System of Kentucky. 1997. “Mandatory Social Security Impact on
Kentucky.”
Teachers’ Retirement System of Kentucky. 1997. Member Handbook.
Teachers’ Retirement System of Louisiana. 1998. Actuarial Valuation.
Teachers’ Retirement System of Louisiana. 1998. Comprehensive Annual Financial Report.
Teachers’ Retirement System of Louisiana. 1998. “Effect of Mandatory Social Security.”
Teachers’ Retirement System of Louisiana. 1998. Member Handbook.
Teachers’ Retirement System of Massachusetts. 1999. “Mandatory Social Security: An Unfunded
Federal Mandate.”
Teachers’ Retirement System of Massachusetts. 1999. Member Handbook.
Teachers’ Retirement System of Texas. 1998. Actuarial Valuation.
Teachers’ Retirement System of Texas. 1998. Member Handbook.
31
Tilove, Robert. 1976. Public Employee Pension Funds. New York: Columbia University Press.
U.S. Congress, House of Representatives, Committee on Education and Labor. 1978. Pension
Task Force Report on Public Employee Retirement Systems. Washington, DC: U.S.
Government Printing Office.
U.S. Congress, House of Representatives, Committee on Ways and Means. 1998. 1998 Green
Book. Washington, DC: U.S. Government Printing Office.
U.S. Congress, House of Representatives, Committee on Ways and Means and Committee on
Post Office and Civil Service. 1980. Report of the Universal Social Security Coverage
Study Group, Joint Committee Print, WMCP: 96-54, March 27.33
U.S. Congress, House of Representatives, Committee on Ways and Means, Subcommittee on
Social Security. 1976. “Social Security Coverage and Government Employers and
Employees of Non-Profit Organizations.”
U.S. Department of Commerce, Bureau of the Census. 1978. 1977 Census of Governments:
Employee-Retirement Systems of State and Local Governments. Topical Studies, 6(1).
U.S. Department of Commerce, Bureau of the Census. 1995. 1992 Census of Governments, 6(4).
U.S. Department of Commerce, Bureau of the Census. 1997. Employee Retirement Systems of
State and Local Governments.
U.S. Department of Health, Education, and Welfare. 1980. Report of the Universal Social
Security Coverage Study Group. Washington, DC: U.S. Government Printing Office.34
U.S. Department of Labor, Bureau of Labor Statistics. 1999. Employee Benefits in Medium and
Large Private Establishments, 1997. Washington, DC: Department of Labor
(stats.bls.gov/ebshome.htm).
U.S. Department of Labor, Bureau of Labor Statistics. 1996. Employee Benefits in State and
Local Governments, 1994. Bulletin 2477. Washington, DC.
U.S. General Accounting Office. 1998a. “Social Security: Better Payment Controls for Benefit
Reduction Provisions Could Save Millions.” GAO/HEHS-98-76, April 30.
Both the U.S. Department of Health and Human Services and the U.S. Congress released this
report in 1980. Thus, it is referenced under both names.
33
34
See footnote #33.
32
U.S. General Accounting Office. 1998b. “Social Security: Implications of Extending Mandatory
Coverage for State and Local Employees.” GAO/HEHS-98-196, August.
U.S. Office of Personnel Management, Office of the Actuary. 1979. “An Alternative to
Extending Social Security Coverage to Federal Employees.”
White House Conference on Social Security. 1998. Statements from Participants. Washington,
DC: The White House, December 8–9.
Zorn, Paul. 1997. Survey Report: 1997 Survey of State and Local Government Employee
Retirement Systems. Chicago, Illinois: Government Finance Officers Association.
Zorn, Paul. 1995. Survey Report: 1995 Survey of State and Local Government Employee
Retirement Systems. Chicago, Illinois: Government Finance Officers Association.
Zorn, Paul. 1993. Survey Report: 1993 Survey of State and Local Government Employee
Retirement Systems. Chicago, Illinois: Government Finance Officers Association.
Zorn, Paul. 1991. Survey Report: 1991 Survey of State and Local Government Employee
Retirement Systems. Chicago, Illinois: Government Finance Officers Association.
33
Table 1. State and Local Pension Systems, 1996–97
Pension Systems
State
Local
Total
Number of Systems
212
2,064
2,276
Members a
Active
Inactive
13.5
11.2
2.3
1.7
1.6
0.1
15.2
12.8
2.4
Beneficiaries a
4.3
1.1
5.4
Contributions b
188.4
36.3
226.6
53.7
15.7
69.5
--
--
1,480
9.2%
19.7%
71.1%
100.0%
9.2%
21.5%
69.1%
100.0%
9.4%
19.9%
70.9%
100.0%
Benefits b
Assets b
Addendum:
Contributions from:
Employees
Government
Investment earnings
Total*
*Sum may not add to total due to rounding.
a
b
In millions.
In billions of dollars.
Source: U.S. Bureau of the Census, Employee Retirement Systems of State and Local
Governments, 1997.
34
Table 2. State and Local Government Employees not Covered by Social Security, 1996
State
Alabama
Alaska
Arizona
Arkansas
California
Colorado
Connecticut
Delaware
Florida
Georgia
Hawaii
Idaho
Illinois
Indiana
Iowa
Kansas
Kentucky
Louisiana
Maine
Maryland
Massachusetts
Michigan
Minnesota
Mississippi
Missouri
Montana
Nebraska
Nevada
New Hamphire
New Jersey
New Mexico
New York
North Carolina
North Dakota
Ohio
Oklahoma
Oregon
Pennsylvania
Rhode Island
South Carolina
South Dakota
Tennessee
Texas
Utah
Vermont
Virginia
Washington
West Virginia
Wisconsin
Wyoming
Total
Total
Not Covered
210,412
45,579
203,484
105,661
1,368,736
174,817
122,890
37,740
638,072
338,090
57,044
57,295
589,341
239,790
153,563
144,755
177,264
226,357
47,494
202,635
288,708
456,930
281,123
146,825
239,127
50,833
65,952
70,789
41,253
401,893
107,682
1,067,339
404,261
27,487
674,816
144,869
150,614
431,530
36,962
211,150
33,506
213,461
1,049,057
85,637
21,873
326,215
269,984
68,701
258,299
33,953
12,801,848
1,647
33,325
11,908
5,922
669,085
166,535
63,835
5,234 **
51,442
102,120
4,559
831
363,546
22,787
3,543
3,663
53,355
221,912
37,838
325
288,708
95,862
58,662
17,722
106,844
12,666
10,294
70,789
4,523
40,274
3,939
0
37,034
921
674,816
1,579
5,859
42,004
9,852
420
309
10,384
575,707
4,090
8,000
1,303
14,913
455
46,579
2,315
3,970,235
Source*
p
s
s
p
p
p
c
p
c
s
p
p
c
c
p
p
p
c
s
p
c
c
c
s
c
s
s
c
p
p
t
p
s
p
c
p
p
s
s
p
p
p
p
t
t
p
p
p
s
p
Percentage Not
Covered
0.8
73.1
5.9
5.6
48.9
95.3
51.9
13.9
8.1
30.2
8.0
1.5
61.7
9.5
2.3
2.5
30.1
98.0
79.7
0.2
100.0
21.0
20.9
12.1
44.7
24.9
15.6
100.0
11.0
10.0
3.7
0.0
9.2
3.4
100.0
1.1
3.9
9.7
26.7
0.2
0.9
4.9
54.9
4.8
36.6
0.4
5.5
0.7
18.0
6.8
31.0
Source: Author's estimates based on Appendix B Table B-1.
* Sources for "Not Covered" estimates: s= Segal (1999), p=PENDAT, c=U.S. Bureau of the Census,
t=Telephone Call
** Volunteer Fire Fighters who are not paid a salary are considered "not covered"
35
Table 3. Prevalence of State and Local Government Employees Not by
Covered Social Security, 1996
Rank
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Rank
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
State
Number Not Covered
Ohio
California
Texas
Illinois
Massachusetts
Louisiana
Colorado
Missouri
Georgia
Michigan
Nevada
Connecticut
Minnesota
Kentucky
Florida
674,816
669,085
575,707
363,546
288,708
221,912
166,535
106,844
102,120
95,862
70,789
63,835
58,662
53,355
51,442
Percentage Not
Covered
100.0
100.0
100.0
98.0
95.3
79.7
73.1
61.7
54.9
51.9
48.9
44.7
36.6
30.2
30.1
State
Ohio
Massachusetts
Nevada
Louisiana
Colorado
Maine
Alaska
Illinois
Texas
Connecticut
California
Missouri
Vermont
Georgia
Kentucky
Source: Author's estimates based on Appendix B Table B-1.
36
Table 4. State and Local Government Employee Coverage by Type of Retirement Plan,
1996
Type of System
State Employees
Teachers and School Employees
Police and Firefighters
Local Government Employees
Total
Active Members
Number
Percentage
(Thousands)
of Total
Plans
Number
Percentage
of Total
6,320
3,276
327
1,105
57.3
29.7
3.0
10.0
72
40
126
140
19.0
10.6
33.3
37.0
11,028
100.0
378
100.0
Source: Mitchell et al. 1999. The data are derived from the 1997 PENDAT database.
37
Table 5. Percentage Distribution of State and Local Pension Plans by Type of Cost-ofLiving Adjustment (COLA) and Size of Plan, 1996
Type of Adjustment
No COLA
Automatic at Fixed Rate
CPI, generally capped at 3%
Ad Hoc
Combination and Other
Total
Addendum:
Total Responding Plans
Small
(under 1,000)
30.9%
15.8
22.3
18.7
12.2
Plan Size
Medium
(1,000-50,000)
10.6%
16.9
43.7
10.6
18.3
Large
(over 50,000)
5.2%
22.8
29.8
15.8
26.3
100.0
100.0
100.0
139
142
57
Source: Zorn 1997.
38
Table 6. Cost-of-Living Protection in Major State Plans with Employees Not Covered by
Social Security
State
Plan
CA
CO
GA
IL
KY
LA
MA
MO
NV
STRS
PERA
TRS
TRS
TRS
SERS/TRS
SERS/TRS
PSRS
PERS
OH
TX
PERS/STRS
TRS
Note:
PERA=
PERS =
PSRS =
SERS =
STRS =
TRS =
Cost-of-Living Protection
Automatic 2% annually plus ad-hoc to supplement.
Automatic based on CPI, capped at 3.5%.
Automatic based on CPI, capped at 1.5% every six months.
Automatic 3%.
1.5% COLA annually plus additional ad-hoc increases.
Beginning 1999, automatic based on CPI, capped at 2%.
Ad-hoc based on CPI, capped at 3% on first $12,000.
Automatic based on CPI, capped at 5%.
Automatic based on CPI, capped at 2% to 5% depending on
years in retirement.
Automatic based on CPI, capped at 3%.
Ad-hoc. Calculations based on 63.5% purchasing power
maintenance.
Public Employees’ Retirement Association
Public Employees’ Retirement System
Public School Retirement System
State Employees’ Retirement System
State Teachers’ Retirement System
Teachers’ Retirement System
Source: Appendix A, “Detailed Description on Major Plans with Uncovered Workers.”
39
Table 7. Funding Status of State and Local Pension Plans, 1996
System
Total
Actuarial
Accrued
Liability
($ in billions)
Actuarial Value of
Assets
($ in billions)
Ratio of
Assets to
Liabilities
$1,248.5
$1,107.0
88.7
State Employees
713.9
675.6
94.6
Teachers/ School Employees
439.4
340.9
77.6
85.9
75.7
88.1
9.3
7.4
79.1
Police/Firefighters
Other
Note: The actuarial value of assets was estimated for the 150 valid cases in the PENDAT
sample by multiplying the funding ratio reported in Table VI-8 by the actuarial accrued
liability reported in Table VI-1. The assets for the subgroups may not add to the total
because of rounding.
Source: Zorn 1997, Tables VI-1 and VI-8, pp.125 and 132.
40
Table 8. Funding Status of Major Plans without Social Security Coverage, 1998
State
Plan
CA
CO
GA
IL
KY
LA
LA
MA
MA
MO
NV
OH
OH
TX
STRS
PERA
TRS
TRS
TRS
TRS
SERS
SERS
TRS
PSRS
PERS
PERS
STRS
TRS
All
a
b
Actuarial Accrued
Liability
($ in billions)
Actuarial Value of
Assets
($ in billions)
7.4
23.9
24.9a
29.9
9.9a
13.2
7.0
11.4
13.1
16.3b
9.9
37.7
54.8b
57.9
317.3
Ratio of Assets
To Liabilities
7.7
23.1
22.5
19.9
8.8
9.1
5.0
10.2
9.9
17.2
7.9
38.4
49.1
60.4
1.041
0.965
0.904
0.666
0.888
0.687
0.715
0.778
0.868
1.051
0.801
1.020
0.896
1.043
289.2
0.911
1997
1999
Note:
PERA=Public Employees’ Retirement Association
PERS=Public Employees’ Retirement System
PSRS=Public School Retirement System
SERS=State Employees’ Retirement System
STRS=State Teachers’ Retirement System
TRS=Teachers’ Retirement System
Source: Appendix A, “Detailed Description on Major Plans with Uncovered Workers.”
41
Table 9. Comparison of Pension Benefits for a Four-Job Worker and a One-Job Worker
Category
Wage growth: 0 percent
Four-job worker
Job 1
Job 2
Job 3
Job 4
Total
One-job worker
Wage growth: 4 percent
Four-job worker
Job 1
Job 2
Job 3
Job 4
Total
One-job worker
Wage growth: 8 percent
Four-job worker
Job 1
Job 2
Job 3
Job 4
Total
One-job worker
Final Pay
Benefits
Percent of
Amount
Salary
$10,000
10,000
10,000
10,000
10
10
10
10
40
$1,000
1,000
1,000
1,000
4,000
10,000
40
4,000
14,802
21,911
32,434
48,010
10
10
10
10
40
1,480
2,191
3,243
4,801
11,715
48,010
40
19,204
21,589
46,609
100,626
217,243
10
10
10
10
40
2,159
4,661
10,063
21,724
38,607
217,243
40
86,897
Ratio of Benefits
Four-job/One-job
Worker
1.00
0.61
0.44
Note: Benefits are calculated on earnings in last year of employment. Annual benefits accrue
at 1 percent a year. Assumes worker stays at each job for 10 years.
Source: Author’s calculations.
42
Table 10. Normal Cost, As a Percentage of Payroll, for State and Local Plans before and
after Mandatory Social Security Coverage, from Actuarial Education and Research Fund
Study, 1980
Plan Size
(# of members)
Large
Current
Formula
“Constant
Benefit”
“Most
Likely”
“Constant
Benefit” Plus
Social
Security
“Most
Likely” Plus
Social
Security
14.37
7.19
7.74
19.59
20.14
20.08
12.10
12.68
24.50
25.08
23.67
16.13
18.60
28.53
31.00
(1,000+)
Medium
(100-999)
Small
(<100)
Note: “Constant benefit” represents the costs associated with preserving the current
benefits for all age and salary combinations. “Most likely” represents the costs associated
with the plan the actuaries thought would most likely result from mandating coverage.
“Constant benefit” plus Social Security is the costs from “constant benefit” plus the 12.4
percent Social Security tax and similarly for “most likely” plus Social Security.
Source: U.S. House, 1980. Tables 6-23 and 6-27 and, pp. 194 and 199.
43
Table 11. Normal Cost, As a Percentage of Payroll, for State and Local Plans before and
after Mandatory Social Security Coverage, from The Urban Institute Study, 1980
Type of Employee
Current
Formula
“Constant
Benefit”
“Typical
Plan”
“Constant
Plan” Plus
Social
Security
“Typical
Plan” Plus
Social
Security
General Employees
13.4
5.4
9.7
17.8
22.1
Teachers
11.6
6.6
10.6
19.0
23.0
Police and Fire
23.4
16.8
17.4
29.2
29.8
Note: For “constant benefit,” The Urban Institute computes the costs of keeping the
current benefits for a single target group—an unmarried employee with average age,
salary, and service characteristics. “Typical plan” represents the costs associated with
reproducing “typical benefits” of employees who are already covered by Social Security.
“Constant benefit” plus Social Security is the cost from “constant benefit” plus the 12.4
percent Social Security tax and similarly for “typical plan” plus Social Security.
Source: U.S. House, 1980. Tables 6-22 and 6-30, pp. 192 and 202.
44
Table 12. Ratio of Required State/Local Benefits to Maintain Benefit Levels with
Inflation, Relative to the Amount Required with Zero Inflation
State Plan
0%
CA
CO
GA
IL
KY
LA
MA
MO
NV
OH
TX
a
STRS
b
PERA
c
TRS
d
TRS
e
TRS
f
SERS/TRS
g
SERS/TRS
h
PSRS
i
PERS
j
PERS/STRS
k
TRS
Potential Inflation Rate
3%
6%
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
0.79
1.00
1.00
1.00
1.00
0.88
0.78
1.00
1.00
1.00
0.83
0.12
0.71
0.36
0.22
0.36
0.51
0.44
0.89
0.71
0.45
0.65
9%
*
0.31
*
*
*
0.09
0.06
0.52
0.31
*
0.48
*Social Security alone would provide as much or more in real lifetime benefits as was
provided under the original state/local plan.
Note:
PERA=Public Employees’ Retirement Association
PERS=Public Employees’ Retirement System
PSRS=Public School Retirement System
SERS =State Employees’ Retirement System
STRS =State Teachers’ Retirement System
TRS=Teachers’ Retirement System
Source: Author’s calculations.
a
Automatic 2% annually plus ad-hoc to supplement. In addition, supplementary benefits available for
participants whose allowances have fallen below 75% of purchasing power. COLA has averaged 2%
between 1992 to 1996. Calculations based on automatic 2%.
b Automatic based on CPI, capped at 3.5%. COLA has averaged 2.4% between 1992 to1996.
c Automatic based on CPI, capped at 1.5% every six months.
d Automatic 3%, COLA has averaged 3% between 1992 and 1996.
e 1.5% COLA annually plus additional ad-hoc increases. COLA has averaged 3% between 1992 and 1996.
f Beginning 1999, automatic based on CPI, capped at 2%.
g Ad-hoc based on CPI, capped at 3% on first $12,000.
h Automatic based on CPI, capped at 5%.
i 2% (yrs. 4-6 after benefits commence), 3% (7-9), 3.5% (10-12), 4% (13-14), 5% thereafter.
j Automatic based on CPI, capped at 3%.
k Ad-hoc. Calculations based on 63.5% purchasing power maintenance.
45
Table 13. Five-Year Cost of Mandatory Social Security Coverage as a Percentage of
State Expenditures for Selected States and for the Nation, 2000–2004
State
5-year Cost of
Mandatory Social
Security Coverage
(millions)
Projected State
Spending for 20002004
(millions)
Cost as a Percentage
of Total State
Spending
(%)
California
Colorado
Illinois
Louisiana
Massachusetts
Ohio
Texas
$4,103
1,123
2,479
959
2,021
3,975
2,647
$558,394
53,756
163,386
85,473
121,494
190,486
231,217
0.7
1.0
1.5
1.1
1.7
2.1
1.1
Nation
26,022
4,541,451
0.6
Note: The costs of mandatory coverage are from Appendix A of the Segal Company
study. Five-year projections of state spending were derived by growing the 1997
expenditure figures in the Statistical Abstract by 3 percent per year.
Sources: Segal Company 1999. U.S. Bureau of the Census 1998, Table 513, p.316.
46
Figure 1. Date of Retirement Establishment or Major Restructuring for Large State and Local
Systems
35
30
47
Percent of Systems
25
20
15
10
5
0
Unknown
before
1901
1901-10
1911-20
1921-30
1931-40
1941-50
1951-60
1961-70
1971-76
Period
Source: U.S. House of Representatives, 95 Congress, Session 2. 1978. Pension Task Force Report on Public Employee Retirement Systems (March 15).
Appendix 1, Table 2, p.202.
Figure 2. Average Benefit Percentage per Year of Service for State and Local Plans by Type of
Covered Employee, 1996
3.5
3.2
Covered by Social Security
3
48
Percent of Final Average Salary
Not Covered by Social Security
2.37
2.5
2.4
2.47
2.24
2.09
2
1.84
1.77
1.5
1
0.5
0
General
Teachers/School
Police/Fire
Other
Type of Covered Em ployees
Source: Paul Zorn 1997. Survey of State and Local Government Systems. Exhibit IV-2, p.37. Illinois: Public Pension Coordinating Council.
(PENDAT)
Figure 3. Employee and Government Contributions as a Percent of Payroll by Type of Covered Employees
20
17.52
18
16.4
Employee
16
Government
Percent of Payroll
14
11.13
12
10.1
10
7.16
8
6.32
6
4.6
3.82
49
4
2
0
General
Teacher/School
Police/Fire
Others
Type of Covered Employees
Source: Paul Zorn 1997. Survey of State and Local Government Systems. Chicago, Illinois: Government Finance Officers Association, Table VII-3 and
Table VII-7, pp. 135 and 139. (PENDAT)
Figure 4. Ratio of Assets to Benefit Payments for Total State and Local Pension Plans
30
25
15
10
5
98
19
96
19
94
19
92
19
90
19
88
19
86
19
84
19
82
19
80
19
78
19
76
19
74
19
72
19
70
19
68
19
66
19
64
19
62
19
60
19
58
19
56
19
54
19
52
0
19
50
Ratio
20
Year
Note: Ratios for 1998 and 1999 are estimates.
Sources: Ratios for 1952-1975 are from Alicia H. Munnell and Ann M. Connolly. 1976. Funding Pensions: Issues and Implications for Financial Markets. Chicago, Illinois: Federal Reserve Bank of Boston. P.84.
For 1978-1981, ratios are from the U.S. Department of Commerce, Bureau of the Census. 1982. Census of Governments: Employee Retirement Systems of State and Local Governments. Vol. 6: Table 1. Ratios
for 1977 and 1982-1987 are from the U.S. Department of Commerce, Bureau of the Census. 1987. Census of Governments: Employee Retirement Systems of State and Local Governments. Vol. 4: Table 1. Ratios
for 1988-1996 are from the U.S. Department of Commerce, Bureau of the Census. State and Local Government Finance Estimates by State. Ratio for 1997 is from the U.S. Department of Commerce, Bureau of
the Census. 1997. Census of Governments :Employee Retirement Systems of State and Local Governments. Table 1.
APPENDIX A
DETAILED DESCRIPTION OF MAJOR PLANS WITH UNCOVERED WORKERS
California State Teachers’ Retirement System
Benefit Provision
Administration Cost
Normal Retirement Provisions
Early Retirement Provisions
Benefits Reduction
Vesting Period
Final Average Salary (FAS)
Benefit Formula
Cost-of-Living Adjustment
State Taxation of Benefits
Disability Benefits
Death Benefits
Survivor Benefits
Portability of Benefits
Interest on Contributions
Other Pension Benefits
Benefit Options
0.25% of annual payroll
60/5
55/5, 50/30
6% per year; plus 3% per year if less than age 55
5 years
Average of highest 3 consecutive years
2% * FAS * years plus career bonus
Automatic 2% annually plus ad hoc to supplement
Taxable, 90% of tax liability is withheld or paid.
Members before 10/16/92: Maximum age 60, 5+ y.o.s. to be eligible.
50%*FAS (more than 10 years of service); 5%*FAS*years (less than
10 years of service); plus 10% per child - up to 40% FAS. Members
after 10/16/92: No age restriction, 5+ y.o.s. for eligibility, 50% of FAS
plus 10% for up to 4 eligible children, may provide for an option
beneficiary.
$5,598 lump-sum
(Coverage A-prior to 10/16/92): $5,598 lump sum & surviving spouse
with eligible children: 40% spouse plus 10% per child up to 50%;
spouse with no children: return of accumulated contributions and
interest in member's account at the time of death less all monthly
allowances paid. (Coverage B-after 10/16/92): lump sum of $22,394,
joint survivor option, return of member contributions. and interest to
surviving spouse, or survivor allowance plus 10% per child up to 50%
for eligible children.
All contributions made by member inclusive of credited interest until
date are refunded upon written request. This includes tax-deferred
member contributions. Partial refunds are not allowed and neither are
state and employer contribution refunds.
Yes, at the rate of 2 year Treasury bill
403(b), cash balance benefit program
7 options with various amounts to survivor or beneficiary
Actuarial Information
Actuarial Method
Members
Active
Average Age
Average Pay
Percentage Female
Annuitants
Are Benefits Provided to :
Half-time workers?
Less than Half-time?
Annual Payroll
Contribution (% of annual payroll)
Employees
Employers
State
Normal Cost Rate
For Year Ending 6/30/98:
Actuarial Value of Plan Assets
Accrued Actuarial Liability (AAL)
Unfunded AAL (UAAL)
Funding Ratio (Assets/AAL)
Entry age normal
385,530
44.2
$40,790
61.7%
157,747
Yes
Alternative CalSTRS Cash Balance Benefit Program
$15.7 billion
8.00%
8.25%
3.10%
17%
$7.7 billion
$7.4 billion
$(0.3) billion
104.1%
Source: Benefit information can be found in the 1999 California STRS Member Handbook . Actuarial information and assumptions are
taken from the 1999 California STRS Actuarial Valuation and direct inquiry of CALSTRS EAPD Branch.
A-1
Public Employees’ Retirement Association of Colorado
Benefit Provision
Administration Cost
Normal Retirement Provisions
Early Retirement Provisions
Benefits Reduction
Vesting Period
Final Average Salary (FAS)
Benefit Formula
Cost-of-Living Adjustment
State Taxation of Benefits
Disability Benefits
Death Benefits
Survivor Benefits
Portability of Benefits
Interest on Contributions
Other Pension Benefits
Benefit Options
0.30% of annual payroll
50/30, 60/20, 65/5 (age/years of service)
50/25, 55/20, 60/5
3% per year from 55 to 60 and 6% per year before 55
5 years
Average of 3 highest consecutive years
2.5% * FAS * Years of Service (40 years max)
Automatic based on CPI, capped at 3.5%
Taxable (exempt to $20,000/yr/person for retiree age >55)
Eligibility: 5 or more years of service credit, 6 months of which have
been in the most recent period of PERA membership, not withdrawn
membership contribution account, not be eligible for service
retirement. Formula: for more than 20 years of service credit at
disability - benefits based on actual service credit at disability, else
based on actual service credit + service credit projected to age 65
but not to exceed 20 years of service credit.
Same as survivors benefits
Qualified children under age 23: 40%*FAS for one child, an equal
share of 50%FAS if there are two or more children. Spouse: if no
qualified children a.) less than 10 years of service credit, 25% FAS,
benefits begin at age 60 b.) 10 or more years of service credit, the
greater of 25%FAS or the benefit which would have been payable
as a 100% joint and survivor option if deceased member had been
eligible for service retirement and retired on the date of death,
benefits begin immediately.
If member leaves service for reasons other than death or retirement,
employee contribution accounts including interest plus matching
employer contributions with interest are refunded upon request.
7% compounded annually
401(k) voluntary investment plan, tax-deferred member
contributions, long term care insurance program, two voluntary term
life insurance plans, health care coverage for annuitants (single and
two others)
3 options with various amount to survivor or beneficiary
Actuarial Information
Actuarial Method
Members
Active
Average Age
Average Pay
Percentage Female
Annuitants
Are Benefits Provided to :
Half-time workers?
Less than Half-time?
Annual Payroll
Contribution (% of annual payroll)
Employees
Employers
Normal Cost Rate
For Year Ending 12/31/98:
Actuarial Value of Plan Assets
Accrued Actuarial Liability (AAL)
Unfunded AAL (UAAL)
Funding Ratio (Assets/AAL)
Entry Age Normal
154,235
43.4
$28,680
65%
49,808
Yes. Varies depending on years of service credit
Yes. Varies depending on years of service credit
$4.5 billion
8.00%
11.40%
14.8%
$23.1 billion
$23.9 billion
$0.8 billion
96.50%
Source: Benefit information can be found in the 1998 Colorado PERA Member Handbook . Actuarial information and assumptions are
taken from the 1998 Colorado PERA Actuarial Valuation.
A-2
Teachers Retirement System of Georgia
Benefit Provision
Administration Cost
Normal Retirement Provisions
Early Retirement Provisions
Benefits Reduction
Vesting Period
Final Average Salary (FAS)
Benefit Formula
Cost-of-Living Adjustment
State Taxation of Benefits
Disability Benefits
Death Benefits
0.11% of annual payroll
Any/30 ; 60/10 (age/years of service)
Any/25
1/12 of 7% for each month the member is below age 60 or 7% for
each year or fraction thereof the member has less than 30 years of
creditable service.
10 years
Average of highest 2 consecutive years
2% * FAS * years of service (40 years max)
Based on CPI
Taxable
Calculated using 2%*years of creditable service*FAS. No age
requirement with 9.5 y.o.s.
Surviving spouse is entitled to receive annually an amount calculated
as: 2%*FAS*years of creditable service (minimum 10 years)
Portability of Benefits
Less than 10 years: lump sum refund of employee contribution and
interest. More than 10 years: lump sum refund of employee
contribution and interest or monthly benefit for life.
Withdrawal allowed, however not after the date of reacceptance of
any TRS covered employment. Refund from TRS is subject to federal
and Georgia tax regulations. In addition, if withdrawal is before age 591/2, taxable portion of refund is subject to a 10% federal excise tax.
Interest on Contributions
Other Pension Benefits
Benefit Options
4.5%
State health and dental coverage plans
7 options with various amounts to survivor or beneficiary
Survivor Benefits
Actuarial Information
Actuarial Method
Members
Active
Average Age
Average Pay
Percentage Female
Annuitants
Are Benefits Provided to :
Half-time workers?
Less than Half-time?
Annual Payroll
Contribution (% of annual payroll)
Employees
Employers
Normal Cost Rate
For Year Ending 6/30/97:
Actuarial Value of Plan Assets
Accrued Actuarial Liability (AAL)
Unfunded AAL (UAAL)
Funding Ratio (Assets/AAL)
Entry age normal
173,599
43.5
$31,497
79%
39,278
Yes
Ineligible for membership
$5.5 billion
5.00%
11.81%
17.1%
$22.5 billion
$24.9 billion
$2.4 billion
90.4%
Source: Benefit information can be found in the 1999 Georgia TRS Member Handbook . Actuarial information and assumptions are
taken from the 1999 Georgia TRS Comprehensive Annual Financial Report .
A-3
Illinois Teachers’ Retirement System
Benefit Provision
Administration Cost
Normal Retirement Provisions
Early Retirement Provisions
Benefits Reduction
Vesting Period
Final Average Salary (FAS)
Benefit Formula
0.18% of annual payroll
62/5, 60/10, 55/35 (age/years of service)
55/20.This option avoids annuity discounting if both member and
employer make a one time contribution.
6% per year.
5 years
Average of highest 4 consecutive years
2.2%*FAS*years of service beginning at 6/30/98; before 6/30/98:
1.67%*FAS*years(1-10), 1.9%*FAS*years(11-20),
2.1%*FAS*years(21-30), 2.3%*FAS*year(30+); Limit: 75%*FAS
Cost-of-Living Adjustment
State Taxation of Benefits
Disability Benefits
Automatic 3% annually
Exempt
Non-occupational: 40% of the greater of contract salary rate in
effect at the time the benefit becomes payable or the annual
contract rate on the date the disability commenced. Occupational:
60% with same provisions as under non-occupational.
Death Benefits
Return of member contributions includes the 6.5% contribution with
interest and the 1/2% paid toward annual increases in annuity. Also
survivor benefits are paid.
Nondependent: Lump-sum = 1/6*FAS*years, for survivors of an
annuitant, lump sum benefit is the greater of 1/6*FAS*years or the
annuitant's survivor benefit contributions or $3,000; Dependent:
$1,000+monthly income as follows-1child: min(30% avg. monthly
salary or $400), spouse & 1 child min(60% avg. salary or $600,
spouse & 2 or more children: min(80% avg. salary or $600)
Survivor Benefits
Portability of Benefits
Interest on Contributions
Other Pension Benefits
Benefit Options
Actuarial Information
Actuarial Method
Members
Active
Average Age
Average Pay
Percentage Female
Annuitants
Are Benefits Provided to:
Half-time workers?
Less than Half-time?
Annual Payroll
Contribution (% of annual payroll)
Employees
Employers
Normal Cost Rate
For Year Ending 6/30/98:
Actuarial Value of Plan Assets
Accrued Actuarial Liability (AAL)
Unfunded AAL (UAAL)
Funding Ratio (Assets/AAL)
After a 4-month waiting period from the date last taught, a member
ceasing covered employment may withdraw all contributions, except
the 1% death benefit. When accepting a refund, the member forfeits
all service credit and benefit rights.
6% p.a. on member accumulated contributions; no interest on
Post-retirement insurance (health coverage and medical plan)
Non-discounted annuity, early retirement option and discounted
it
Projected unit credit
115,116
43.5
$44,769
73.2%
51,472
Yes
Yes
$5.3 billion
8.00%; 9.00% beginning 7/1/98, latest 9.50%
9.00%
8.6%
$19.9 billion
$29.9 billion
$10.0 billion
66.60%
Source: Benefit information can be found in the 1999 Illinois TRS Member Handbook . Actuarial information and assumptions are
taken from the 1998 Illinois TRS Comprehensive Annual Financial Report .
A-4
Kentucky Teachers’ Retirement System
Benefit Provision
Administration Cost
Normal Retirement Provisions
Early Retirement Provisions
Benefits Reduction
Vesting Period
Final Average Salary (FAS)
Benefit Formula
Cost-of-Living Adjustment
State Taxation of Benefits
Disability Benefits
Death Benefits
Survivor Benefits
Portability of Benefits
Interest on Contributions
Other Pension Benefits
Benefit Options
0.18% of annual payroll
Any/27, 60/5 (age/years of service)
55/5
5% per year under age 60
5 years
Average of highest 5 consecutive years
2% * FAS * years
1.5% COLA annually plus additional ad hoc increases.
Taxable
60% FAS if 5 or more years of service and applied for disability
retirement within 1 year of last contributing service. Active member
with less than 5 years would get 50% of current annual contract
salary.
covered by $2,000 death benefit from first working day with KTRS ad
till member remains in active contributing status. If death occurs with
27 or more years of service or he or she is 55 with 5 years of service,
the surviving beneficiary may receive an annuity or a refund of the
member's account.
Eligibility: member with 10 or more years of service. Spouse: basic
monthly-$180 ; $240 (income less than $550). These monthly
payments are guaranteed for spouse's life unless he/she remarries.
Dependent children under age 18 get monthly benefits irrespective of
parents income or marital status. They may apply for additional
benefits between ages 18 and 23 if they are full-time students.
Refund issued within 30 days of refund application. However, refund
beneficiary must meet IRS early distribution exceptions, else he/she
has to pay an additional 10% excise tax on the taxable portion of
refund.
3% per annum after first year
403(b), medical insurance, purchase of service credit
5 options: Straight life annuity with refundable balance; 10 years
certain and life thereafter; joint-survivor annuity; joint-survivor
annuity, one-half benefit to beneficiary and other payments
Actuarial Information
Actuarial Method
Members
Active
Average Age
Average Pay
Percentage Female
Annuitants
Are Benefits Provided to:
Half-time workers?
Less than Half-time?
Annual Payroll
Contribution (% of annual payroll)
Employees
Employers
Normal Cost Rate
For Year Ending 6/30/97:
Actuarial Value of Plan Assets
Accrued Actuarial Liability (AAL)
Unfunded AAL (UAAL)
Funding Ratio (Assets/AAL)
Unit credit actuarial cost method
50,445
43.5
$38,907
72.90%
26,283
Yes
Yes
$1.9 billion
9.86%
13.11%
24.2%
$8.8 billion
$9.9 billion
$1.1 billion
88.8%
Source: Benefit information can be found in the 1997 Kentucky TRS Member Handbook . Actuarial information and assumptions are
taken from the 1997 Kentucky TRS Comprehensive Annual Financial Report .
A-5
Teachers’ Retirement System of Louisiana
Benefit Provision
Administration Cost
Normal Retirement Provisions
Early Retirement Provisions
Benefits Reduction
Vesting Period
Final Average Salary (FAS)
Benefit Formula
Cost-of-Living Adjustment
State Taxation of Benefits
Disability Benefits
Death Benefits
Survivor Benefits
Portability of Benefits
Interest on Contributions
Other Pension Benefits
Benefit Options
0.24% of annual payroll
Regular 2% plan: 60/10, Any/20. Regular 2.5% plan: 65/20, 55/25,
Any/30, Plan A 3%: 60/10, 55/25, Any/30, plan B 2%: 60/10, 55/30
(age/years of service)
60/10, Any/20.
Multiplier reduced to 2% *FAS from 2.5%*FAS. If joined after 7/1/99
no longer have 2.0 provision, but the Any/20 benefit reduced based
on member's age after retirement.
10 years
Highest consecutive 3 years average
2.5% * FAS * Years of Service (55/25, 65/20, Any/30); 2%*FAS*Years
(60/10, Any/20)
Beginning 1999, capped at 2%, automatic based on CPI
Exempt
Regular Plan: 2.5%*FAS *years of service, not exceeding 75% of FAS
and not below 40% of the stated minimum salary for a beginning
teacher. Eligibility: 5 or more years of service. Additional 50% of
member's benefits payable if minor child, but family benefit limited to
75% FAS
See survivor benefits
Greater of 50% member's benefit with 5 years of creditable service (2
years immediately prior to death) and $300 per month, for spouse with
minor children; Greater of 50% benefit with 10 years of creditable
service and $300 per month for spouse without children.
Refund of employee contributions excluding interest. Upon reemployment, a member may reinstate the credit forfeited through
termination of previous membership by repaying the refunded
contributions plus interest.
No
Purchase of additional service credits
Reduced benefits payable in the form of joint and survivor option,
reduced benefit with a lump sum payment which cannot exceed 36
months benefit payments. Optional retirement plan for higher
education members desiring a defined contribution plan.
Actuarial Information
Actuarial Method
Members
Active
Average Age
Average Pay
Percentage Female
Annuitants
Are Benefits Provided to:
Half-time workers?
Less than Half-time?
Annual Payroll
Contribution (% of annual payroll)
Employees
Employers
Normal Cost Rate
For Year Ending 6/30/98
Actuarial Value of Plan Assets
Accrued Actuarial Liability (AAL)
Unfunded AAL (UAAL)
Funding Ratio (Assets/AAL)
Projected unit credit actuarial cost method
87,193
43
$28,695
82.9%
42,445
No
No
$2.5 billion
8.00% (regular), 9.10% (Plan A), 5.00% (Plan B)
16.4% (determined annually)
15.4%
$9.07 billion
$13.2 billion
$4.1 billion
68.70%
Source: Benefit information can be found in the 1998 Louisiana TRS Member Handbook and Actuarial Valuation . Actuarial information
and assumptions are taken from the 1998 Louisiana TRS Actuarial Valuation .
A-6
Louisiana State Employees’ Retirement System
Benefit Provision
Administration Cost
Normal Retirement Provisions
Early Retirement Provisions
Benefits Reduction
Vesting Period
Final Average Salary (FAS)
Benefit Formula
Cost-of-Living Adjustment
State Taxation of Benefits
Disability Benefits
Death Benefits
Survivor Benefits
Portability of Benefits
Interest on Contributions
Other Pension Benefits
Benefit Options
0.41% of annual payroll
Any/30, 55/25, 60/10 (age/years of service)
Any/20 (actuarially reduced benefits)
Multiplier is reduced for each year under normal retirement age
10 years
Average of highest 3 consecutive years
2.5% * FAS * years, plus $300 if joined before 7/1/86
Automatic based on CPI, capped at 2%
Exempt
2.5%*FAS*years under maximum disability retirement benefit,
choice of optional plans as for regular retirees is available, 10 years
of service required
See survivor benefits
With children: greater of 75% FAS or $300/month (if surviving
spouse is entitled to a benefit, he/she receives on pro rata basis);
without children: greater of 50% FAS or $200/month
Can apply for refund without interest or investment earnings, when
leaving state services. Transfers from one state agency to another
do not qualify for refunds. Interest not paid on refunds.
Yes
Purchase of additional service credits
4 options and deferred retirement benefits
Actuarial Information
Actuarial Method
Members
Active
Average Age
Average Pay
Percentage Female
Annuitants
Are Benefits Provided to:
Half-time workers?
Less than Half-time?
Annual Payroll
Contribution (% of annual payroll)
Employees
Employers
Normal Cost Rate
For Year Ending 6/30/98:
Actuarial Value of Plan Assets
Accrued Actuarial Liability (AAL)
Unfunded AAL (UAAL)
Funding Ratio (Assets/AAL)
Unit credit actuarial cost method
69,949
43
$24,303
Not available
29,420
No
No
$1.7 billion
7.67%
12.40%
21.1%
$5.0 billion
$7.0 billion
$2.0 billion
71.50%
Source: Benefit and actuarial information, as well as actuarial assumptions are taken from the 1998 LASERS 1997 Actuarial
Valuation .
A-7
Massachusetts State Employees’ Contributory Retirement System
Benefit Provision
Administration Cost
Assumed to be paid separately and are not included in the appropriation.
Normal Retirement Provisions
Any/20, if hired prior to 1978-attainment of age 55, else 55/10 (age/years of
service)
No mandatory retirement age, barring certain public safety and state police
officers.
0.1% reduction for each year of age under 65.
10 years
Average of highest 3 consecutive years
Benefit rate* FAS *years (capped at 80% FAS). 2.5% benefit rate for
general employees who retire at or after age 65.
Ad hoc based on CPI, capped at 3% on first $12,000
Exempt
10 y.o.s. requirement for eligibility if injury is non-job-related. Ordinary:
50% last pay plus annuity based on accumulated member contributions
with interest. Accident: Lump-sum 72%* last pay plus annuity based on
contributions with interest, not exceeding 75%.
If member dies as a result of a work-related injury or if he was retired for
accidental disability and death was the natural result of that injury, there is
immediate payment to named beneficiary of accumulated
deductions+pension of 72% current salary+supplemental $312 per year per
child until child attains age 18 or 21(if full-time student).
If death occurs in active service, an allowance of what would have been
payable had the member retired. Death prior to member's superannuation
retirement age results in use of age 55 benefit rate. Minimum allowance to
surviving spouse is $3,000+$1,440 per year for first child and $1,080 per
year for each additional child.
If member is under age 55, member contributions may be withdrawn.
Employees who first become members on or after 1/1/84, may receive only
limited interest on their contributions, if they voluntarily terminate their
services. Those who leave with less than 5 years service get no interest
and those who leave with more than 5 but less than 10years of service get
50% of interest.
5.5% per year
Health and life insurance sponsored
Three options for receiving retirement allowance. Total payable in monthly
installments, reduced annual allowance with balance paid as lump sum to
designated beneficiary, reduced allowance with 2/3 payable at retired
employees death, to designated beneficiary.
Early Retirement Provisions
Benefits Reduction
Vesting Period
Final Average Salary (FAS)
Benefit Formula
Cost-of-Living Adjustment
State Taxation of Benefits
Disability Benefits
Death Benefits
Survivor Benefits
Portability of Benefits
Interest on Contributions
Other Pension Benefits
Benefit Options
Actuarial Information
Actuarial Method
Members
Active
Average Age
Average Pay
Percentage Female
Annuitants
Are Benefits Provided to :
Half-time workers?
Less than Half-time?
Annual Payroll
Contribution (% of annual payroll)
Employees
Employers
Entry age normal
82,631
44.3
$37,649
52.70%
43,144
Jurisdiction of retirement board in determining eligibility for membership.
Jurisdiction of retirement board in determining eligibility for membership.
$3.1 billion
Varies depending on the most recent date of membership. For 7/1/96 to
present - 9% of regular compensation and 1979 to present - an additional
2% of regular compensation in excess of $30,000
8.25%
Normal Cost Rate
For Year Ending 1/1/98:
Actuarial Value of Plan Assets
13.1%
Accrued Actuarial Liability (AAL)
$11.4 billion
Unfunded AAL (UAAL)
Funding Ratio (Assets/AAL)
$1.5 billion
86.8%
$9.9 billion
Source: Benefit information can be found in the 1999 Mass. Public Employee Retirement Guide . Actuarial information and assumptions are taken
from 1998 Mass. Public Employee Retirement Administration Commission Actuarial Valuation.
A-8
Massachusetts Teachers’ Retirement System
Benefit Provision
Administration Cost
Normal Retirement Provisions
Early Retirement Provisions
Benefits Reduction
Vesting Period
Final Average Salary (FAS)
Benefit Formula
Cost-of-Living Adjustment
State Taxation of Benefits
Disability Benefits
Death Benefits
Survivor Benefits
Assumed to be paid separately and are not included in the
appropriation.
Any/20, 55/and if hired before 1978, 55/10 (age/years of service)
No mandatory retirement age
0.1% per year under age 65
10 years
Average of highest 3 consecutive years
2.5%*FAS*years (capped at 80%FAS)
Ad hoc based on CPI, capped at 3% on first $12,000
Exempt
10 y.o.s. requirement for eligibility if injury is non-job-related.
Ordinary: 50% last pay plus annuity based on accumulated
member contributions with interest. Accident: lump-sum 72%*last
pay plus annuity based on contributions with interest, not
exceeding 75%.
If member dies as a result of a work-related injury or if he/she was
retired for accidental disability and death was the natural result of
that injury, there is immediate payment to named beneficiary of
accumulated deductions+pension of 72% current
salary+supplemental $312 per year per child until child attains age
18 or 21(if full-time student).
If death occurs in active service, an allowance of what would have
been payable had the member retired. Death prior to member's
superannuation retirement age results in use of age 55 benefit
rate. Minimum allowance to surviving spouse is $3,000+$1,440 per
year for first child and $1,080 per year for each additional child.
Portability of Benefits
If member is under age 55, member contributions may be
withdrawn. Employees who first became members on or after
1/1/84, may receive only limited interest on their contributions, if
they voluntarily terminate their services. Those who leave with less
than 5 years service get no interest and those who leave with more
than 5 but less than 10 years of service get 50% of interest.
Interest on Contributions
Other Pension Benefits
Benefit Options
5.5% per year
Optional group medical insurance plan
3 options for receiving retirement allowance. Total payable in
monthly installments, reduced annual allowance with balance paid
as lump sum to designated beneficiary, reduced allowance with
2/3 payable at retired employees death, to designated beneficiary.
Actuarial Information
Actuarial Method
Members
Active
Average Age
Average Pay
Percentage Female
Annuitants
Are Benefits Provided to:
Half-time workers?
Less than Half-time?
Annual Payroll
Contribution (% of annual payroll)
Employees
Employers
Normal Cost Rate
For Year Ending 1/1/98:
Actuarial Value of Plan Assets
Accrued Actuarial Liability (AAL)
Unfunded AAL (UAAL)
Funding Ratio (Assets/AAL)
Entry age normal
74,899
43.5
$42,393
71.60%
30,499
Yes
No
$3.2 billion
Varies depending on the most recent date of membership. For
7/1/96 to present - 9% of regular compensation and 1979 to
present - an additional 2% of regular compensation in excess of
$30,000
8.25%
10.5%
$10.2 billion
$13.1 billion
$2.9 billion
77.80%
Source: Benefit information can be found in the 1999 Mass. Public Employee Retirement Guide . Actuarial information and assumptions are
taken from 1998 Mass. Public Employee Retirement Administration Commission Actuarial Valuation .
A-9
Public School Retirement System of Missouri
Benefit Provision
Administration Cost
Normal Retirement Provisions
Early Retirement Provisions
Benefits Reduction
Vesting Period
Final Average Salary (FAS)
Benefit Formula
Cost-of-Living Adjustment
State Taxation of Benefits
Disability Benefits
Death Benefits
Survivor Benefits
Portability of Benefits
Interest on Contributions
Other Pension Benefits
Benefit Options
0.16% of annual payroll
60/5, 55/25, any/30, or the sum of age and years of service exceeds 80
(age/years of service)
55/5, any/25
Benefits reduced from 9.23% to 73.91% based on years under age 60
5 years
Average of highest 3 consecutive years
2.5% * FAS * Years. See HB p41 for replacement rate for 55/25.
Automatic based on CPI, not exceeding 5%
Taxable
90% of service retirement benefit payable as though you were age 60, or
50% of your salary for your last full year of credit, whichever is greater
Lump-sum refund (contribution plus payment due to reinstatement and
other purchases plus interest credited to account as of date of death),
$5000 one-time lump sum
% of applicable salary for spouse-20%, spouse and dependent children20%+10% per child, dependent children-16 2/3% each, and dependent
parents - 16 2/3% each
Refund of accrued contribution with interest through preceding June 30
Yes, set by Board of Trustees to be comparable to average savings or
short term CD accounts
Health care
6 options with various amounts to survivor or beneficiary
Actuarial Information
Actuarial Method
Members
Active
Average Age
Average Pay
Percentage Female
Annuitants
Are Benefits Provided to :
Half-time workers?
Less than Half-time?
Annual Payroll
Contribution (% of annual payroll)
Employees
Employers
Normal Cost Rate
For Year Ending 6/30/99:
Actuarial Value of Plan Assets
Accrued Actuarial Liability (AAL)
Unfunded AAL (UAAL)
Funding Ratio (Assets/AAL)
Entry age normal
70,092
42.5
$38,383
76%
23,292
Yes, certified part-time employees/teachers employed for 20 hours/week
on a regular basis for at least one month are automatically PSERS
employees.
No
$2.69 billion
10.50%
10.5% (neither exceeding 11.5%, as set by board of trustees)
18.2%
$17.2 billion
$16.3 billion
($0.9 billion)
105.1%
Source: Benefit information can be found in the 1997 Missouri PSRS Member Handbook . Actuarial information and assumptions are taken
from 1997 Missouri PSRS Comprehensive Annual Financial Report, and direct inquiry of the Missouri PSRS.
A-10
Public Employees’ Retirement System of Nevada – General Employees
Benefit Provision
Administration Cost
Normal Retirement Provisions
Early Retirement Provisions
Benefits Reduction
Vesting Period
Final Average Salary (FAS)
Benefit Formula
Cost-of-Living Adjustment
State Taxation of Benefits
Disability Benefits
Death Benefits
Survivor Benefits
Portability of Benefits
Interest on Contributions
Other Pension Benefits
Benefit Options
0.15% of annual payroll
65/5, 60/10, any/30 (age/years of service)
Any/5
4% per year for each year under early retirement
5 years
Average of 3 highest consecutive years
2.5%*FAS*years, no exceeding 75% FAS
2% (yrs. 4-6 after benefits commence), 3% (7-9), 3.5% (10-12), 4%
(13,14), 5% (thereafter)
No state income tax
Members with 5 or more years of service who become disabled are
eligible to receive normal retirement benefits without reduction
None
Spouse: $450 if less than 10 years of service; greater of $450 or 50%
FAS between 10 to 15 years; greater of $450 or 100% FAS if more
than 15 years. $400 per dependent child. $400 dependent parent.
Lump-sum: refund of employee contribution and half of employer
contribution.
Members who have contributed or who were mandated under the
Employee Pay Contribution Plan, may withdraw employee
contributions and/or receive a distribution of the employee portion of
the mandatory employer-pay contributions, if terminating all
employment for which contribution is required. A refund cancels
membership.
Yes at actuarially determined rate
No health insurance
7 options with various amounts to survivor or beneficiary.
Actuarial Information
Actuarial Method
Members
Active
Average Age
Average Pay
Percentage Female
Annuitants
Are Benefits Provided to :
Half-time workers?
Less than Half-time?
Annual Payroll
Contribution (% of annual payroll)
Employees
Employers
Normal Cost Rate
For Year Ending 6/30/98:
Actuarial Value of Plan Assets
Accrued Actuarial Liability (AAL)
Unfunded AAL (UAAL)
Funding Ratio (Assets/AAL)
Entry age normal
66,374
43.5
$32,714
Not available
14,822
Yes, if employed in a position which is normally considered to be half
time or more according to full-time work schedule
No
$2.4 billion
10.00% (under employer/employee contribution plan)
10.00% (under employer/employee contribution plan); 18.75% (under
employer contribution plan)
15.1%
$7.9 billion
$9.9 billion
$2.0 billion
80.1%
Source: Benefit and actuarial information, as well as actuarial assumptions are taken from 1998 Nevada PERS Comprehensive Annual
Financial Report .
A-11
Public Employees’ Retirement System of Ohio-State and Local Division
Benefit Provision
Administration Cost
Normal Retirement Provisions
Early Retirement Provisions
Benefits Reduction
Vesting Period
Final Average Salary (FAS)
Benefit Formula
Cost-of-Living Adjustment
State Taxation of Benefits
Disability Benefits
0.297% of annual payroll
60/5, any/30 (age/years of service)
55/25 (if 30 y.o.s. no benefit reduction)
Benefits reduced up to 25% between age 55-65.
5 years
Average of 3 highest years
2.1%*FAS*years + 2.5%*FAS*years (30+)
Automatic based on CPI, capped at 3%
Taxable
5 y.o.s. eligibility requirement. Original plan - benefits depend on
FAS, years of service with PERS and the length of time between the
effective date of disability and age 60. This benefit can't exceed 75%
or be less than 30% of the member's FAS. Revised plan - benefits
based on FAS, years of service with PERS with no early retirement
reductions, but can't be less than 45% or exceed 60% FAS.
Death Benefits
$500 if less than 10 years of service; $1,000 between 10 to 15
years; $1,500 between 15 to 20 years; $2,000 between 20 to 25
years; $2,500 maximum
Greater of: 1 person: 25% FAS or $96 per month; 2 persons: 40%
FAS or $186; 3 persons: 50% FAS or $236; 4 persons: 55% FAS or
$236; 5 persons: 60% FAS or $236
Refund of accrued savings
Yes
Health care
5 options with various amounts to survivor or beneficiary
Survivor Benefits
Portability of Benefits
Interest on Contributions
Other Pension Benefits
Benefit Options
Actuarial Information
Actuarial Method
Members
Active
Average Age
Average Pay
Percentage Female
Annuitants
Are Benefits Provided to :
Half-time workers?
Less than Half-time?
Annual Payroll
Contribution (% of annual payroll)
Employees
Employers
Normal Cost Rate
For Year Ending 12/31/98:
Actuarial Value of Plan Assets
Accrued Actuarial Liability (AAL)
Unfunded AAL (UAAL)
Funding Ratio (Assets/AAL)
Entry age normal
354,431
41.4
$24,479
51%
124,258
Yes
Yes
$9.0 billion
8.50%
13.31% (state), 13.55% (local)
14.7%
$38.4 billion
$37.7 billion
$(0.7) billion
102.0%
Source: Benefit information can be found in the 1999 Ohio PERS Member Handbook . Actuarial information and assumptions are
taken from the 1998 Ohio PERS Comprehensive Annual Financial Report and through direct inquiry of the Ohio PERS.
A-12
State Teachers’ Retirement System of Ohio
Benefit Provision
Administration Cost
Normal Retirement Provisions
Early Retirement Provisions
Benefits Reduction
Vesting Period
Final Average Salary (FAS)
Benefit Formula
Cost-of-Living Adjustment
State Taxation of Benefits
Disability Benefits
Death Benefits
Survivor Benefits
Portability of Benefits
Interest on Contributions
Other Pension Benefits
Benefit Options
0.62% of annual payroll
Any/30 (age/years of service)
60/5, 55/25
Varies with ages and years of service
5 years
Average of highest 3 consecutive years
2.1%*FAS*years of service (0-30) + 2.5%*FAS*Y.O.S. (31+), not
exceeding 100% FAS. This adjusted percentage is based on
number of y.o.s. beyond 30.
Automatic based on CPI, capped at 3%
Taxable
5 years of service eligibility requirement. 2.1%*FAS*years. Joined
before 7/30/92: 2%*FAS *years plus 60 minus age. Capped at 75%
of FAS.
$1,000 automatic; additional $1,000 or $2,000 optional.
Only spouse: refund of accrued contribution or annuity of 25% FAS.
Spouse with children: 1 dependent: annuity of 25% FAS; 2: 40%
FAS; 3: 50%; 4: 55% FAS; 5 or more: 60% FAS.
Refund of accrued contribution with interest. Interest rates vary with
years of service.
Yes
Health care
5 options with various amount to survivors or beneficiary
Actuarial Information
Actuarial Method
Members
Active
Average Age
Average Pay
Percentage Female
Annuitants
Are Benefits Provided to :
Half-time workers?
Less than Half-time?
Annual Payroll
Contribution (% of annual payroll)
Employees
Employers
Normal Cost Rate
For Year Ending 7/1/99:
Actuarial Value of Plan Assets
Accrued Actuarial Liability (AAL)
Unfunded AAL (UAAL)
Funding Ratio (Assets/AAL)
Entry age normal
170,854
43.6
$41,210
69.00%
95,796
Yes. Percent-based salaried service is determined by the percent of
full time employment
$7.4 billion
9.30%
14.00%
15.3%
$49.1 billion
$54.8 billion
$5.7 billion
89.6%
Source: Benefit information can be found in the 1998 Ohio STRS Member Handbook . Actuarial information and assumptions are taken
from the 1999 Ohio STRS Actuarial Valuation .
A-13
Teacher Retirement System of Texas
Benefit Provision
Administration Cost
Normal Retirement Provisions
Early Retirement Provisions
Benefits Reduction
Vesting Period
Final Average Salary (FAS)
Benefit Formula
Cost-of-Living Adjustment
State Taxation of Benefits
Disability Benefits
Death Benefits
Survivor Benefits
Portability of Benefits
Interest on Contributions
Other Pension Benefits
Benefit Options
0.13% of annual payroll
65/5, 60/20, (age/years of service) or sum of age and years of
service equal to 80.
55/5, any/30
Reduced up to 53% depending on age and years of service
5 years
Average of highest 3 years
2.2%*FAS*years (2% prior to 9/1/99)
Ad hoc
No state income tax
If less than 10 years of service, $1,800 per year. If more than 10
years of service, the greater of accrued retirement benefit or
$78*years per years, no less than $1,800 per year.
Lump-sum: 2 * last pay, not exceeding $80,000. Annuity: 60 monthly
payments of accrued normal annuity or life annuity as if the member
had retired on the last day of the month preceding death. Refund of
employee contribution. Also three other options.
Lump-sum: $10,000. Or lump-sum $2,500 plus $200 per month for
only spouse, commencing at age 65; $300 for spouse and children
until the youngest child reaches 18 or 21 if a full- time student; $200
for only one child; $300 for two or more
Refund of employee contribution plus interest
Yes
Deferred Retirement Option Plan (DROP), excess benefit
arrangement, partial lump sum option
5 options with various amount to survivor or beneficiary
Actuarial Information
Actuarial Method
Members
Active
Average Age
Average Pay
Percentage Female
Annuitants
Are Benefits Provided to:
Half-time workers?
Less than Half-time?
Annual Payroll
Contribution (% of annual payroll)
Employees
Employers
Normal Cost Rate
For Year Ending 8/31/98:
Actuarial Value of Plan Assets
Accrued Actuarial Liability (AAL)
Unfunded AAL (UAAL)
Funding Ratio (Assets/AAL)
Entry age normal
705,447
42.91
$26,533
75.00%
170,351
Yes
No
$18.3 billion
6.40%
6.00%
11.4%
$60.4 billion
$57.9 billion
$(2.5) billion
104.30%
Source: Benefit information can be found in the 1998 Texas TRS Member Handbook . Actuarial information and assumptions are taken
from the 1998 Texas TRS Comprehensive Annual Financial Report .
A-14
APPENDIX B
STATE AND LOCAL WORKERS NOT COVERED BY SOCIAL SECURITY
The most elementary questions of how many workers are not covered by Social
Security and where they reside are not as straightforward to answer as one would think.
Several estimates are available and in many cases they provide very different pictures of the
employment and coverage situation. Appendix Table B-1 summarizes the alternative
estimates.
The first source of information is the 1992 Census of Governments. It reports the
number active members covered by Social Security for plans with more than $20 million in
cash and investment holdings. This is the last government survey to address the issue of
Social Security coverage; the coverage picture could have changed since 1992.
The last data from the Social Security Administration (SSA) for total employment
and coverage by state is also from 1992 (1998 Green Book). 1 In addition to being dated,
these numbers tend to be considerably higher than those from other sources. The reason is
that SSA counts the number of people who had a job at any time during the calendar year,
while the Census data report active members of each significant pension plan during the last
month of the plan’s fiscal year. When SSA splits employment between those covered and
those not covered by Social Security, the employment count grows even further. A person
can have both a covered and a non-covered job during the calendar year; for example, a
teacher whose job is not covered by Social Security takes a covered local government job
during the summer. In such a case, SSA will count that person twice. Despite this double
counting, the percent covered figure provides some information about the coverage ratio for
the state.
In addition to the two data sources for 1992, three estimates are available for 1996.
The first one we construct by applying the Census coverage ratio by state in 1992 to total
state and local government employment in 1996 taken from Employee Retirement Systems of
State and Local Governments 1997 by the U.S. Bureau of the Census. 2 This report covers
fiscal years that ended between July 1, 1996 and June 30, 1997. To the extent that the
coverage ratios are applicable, these estimates represent somewhat more up-to-date numbers.
The second source of data for 1996 is the so-called PENDAT files created from the
Surveys of State and Local Employee Retirement Systems for Members of the Public
Pension Coordinating Council (Zorn, 1997). The most recent survey includes coverage
information for 1996. The problem with this survey is that it does not include all major plans
1
While SSA has not published state-by-state numbers for later years, it estimated that in 1996 22.3 million
individuals worked sometime during the year for state or local government, and the wages of 25 percent of
these individuals were not covered by Social Security (1998 Green Book, p. 11).
2
The 1998 version of Employee Retirement Systems of State and Local Governments is recently available.
However, in order to directly compare the estimates from each source, we keep the year of the estimates
constant among sources.
B-1
and suffers from survey non-response. The following table shows that state by state, the
PENDAT number of total active members is usually either very close to the census number
of below it. For example, the survey does not include Massachusetts Public Employee or
Teachers Retirement Systems, and Massachusetts is the fifth largest state in terms of the
number of uncovered state and local workers. Nevertheless, the PENDAT files are a useful
additional source of information.
The final source of data on the number of employees not covered is a recent (May
1999) study “The Cost Impact of Mandating Social Security for State and Local Workers” by
the Segal Company for the AFSCME, AFL-CIO, and Coalition to Preserve Retirement
Security. This study includes 1996 estimates of uncovered workers. The authors state the
estimates are based on unpublished SSA 1992 estimates of non-covered state and local
government workers which are augmented with data from state reports. No further
explanation is provided, so it is difficult to determine the relationship between these
estimates and the others.
As noted above, Appendix A, Table 1 summarizes all these estimates. Although we
are able to define why the total number of active members differs from source to source, it is
not so easy to determine why the number of non-covered members differs between sources.
Based on the arrayed numbers, we make the following selections. For the number of total
active members, we use the 1996−97 Census data throughout. For the number of state and
local workers not covered by Social Security, we use the PENDAT figure when the total
number of active members reported by PENDAT is within 5 percent of the total number from
1996 Census. We are wary that PENDAT does not contain data from all of the pension
plans in some states, but when the totals of PENDAT and Census match, we are assured that
PENDAT has indeed collected data from all the plans in the state. When PENDAT is not
within 5 percent of the total from 1996 Census but the other sources match, we use Census’
number of workers not covered by Social Security. If the other sources do not match, we use
information gathered from calls to the plans within the state with over 5,000 members and
then use the estimate of the source that most closely matches the information gathered by the
phone calls. When phone calls do not provide the needed information, we use the source
whose percent not covered is closest to the SSA numbers. Finally, if the call indicates
something very different from all published sources, we use the call information. This
procedure led us to use PENDAT for 25 states, Segal for 11, Census for 11, and telephone
information for 3. Although we believe that we have been as thorough as possible, these
numbers again differ from those published elsewhere.
Table 2 in the text summarizes our best estimate of total state and local government
employment and Social Security coverage for 1996. The total uncovered shown in Table 2 in
the text is 4.0 million; Segal Company reported 4.8 million uncovered and Census suggests a
number closer to 6.6 million. Table 3 in the text presents the states with the highest number
of uncovered workers and the states with the largest percentage of workers not covered by
Social Security.
B-2
Appendix Table B-1. Alternative Estimates of State and Local Government Employees Not
Covered under Social Security
State
Alabama
Alaska
Arizona
Arkansas
California
Colorado
Year
Source
Active
Members
Members Not
Covered
Percent Not
Covered
1992
Census
SSA
192,626
360,000
72,801
36,000
37.8
10.0
1996
Census
PENDAT
Segal
210,412
200,840
--
79,523
1,647
27,911
37.8
0.8
--
1992
Census
SSA
49,796
82,000
49,796
48,000
100.0
58.5
1996
Census
PENDAT
Segal
45,579
46,074
--
45,579
46,074
33,325
100.0
100.0
--
1992
Census
SSA
179,499
340,000
34,489
16,000
19.2
4.7
1996
Census
PENDAT
Segal
203,484
178,280
--
39,097
6,247
11,908
19.2
3.5
--
1992
Census
SSA
121,635
191,000
11,636
19,000
9.6
9.9
1996
Census
PENDAT
Segal
105,661
104,884
--
10,106
5,922
15,902
9.6
5.6
--
1992
Census
SSA
1,492,572
2,198,000
876,679
1,129,000
58.7
51.4
1996
Census
PENDAT
Segal
1,368,736
1,429,862
--
803,943
669,085
903,027
58.7
46.8
--
1992
Census
SSA
172,056
330,000
160,561
208,000
93.3
63.0
1996
Census
PENDAT
Segal
174,817
169,260
--
163,138
166,535
180,076
93.3
98.4
--
B-3
Appendix Table B-1—continued. Alternative Estimates of State and Local Government
Employees Not Covered under Social Security
State
Connecticut
Delaware
Florida
Georgia
Hawaii
Idaho
Year
Source
Active
Members
Members Not
Covered
Percent Not
Covered
1992
Census
SSA
126,022
255,000
65,462
81,000
51.9
31.8
1996
Census
PENDAT
Segal
122,890
98,307
--
63,835
42,276
61,228
51.9
43.0
--
1992
Census
SSA
33,058
65,000
31,989
5,000
96.8
7.7
1996
Census
PENDAT
Segal
37,740
34,292
--
36,520
5,234
3,632
96.8
15.3
--
1992
Census
SSA
603,999
1,003,000
48,695
76,000
8.1
7.6
1996
Census
PENDAT
Segal
638,072
597,930
--
51,442
6,125
61,817
8.1
1.0
--
1992
Census
SSA
323,345
580,000
312,820
119,000
96.7
20.5
1996
Census
PENDAT
Segal
338,090
283,121
--
327,085
54,452
102,120
96.7
19.2
--
1992
Census
SSA
57,401
107,000
6,469
19,000
11.3
17.8
1996
Census
PENDAT
Segal
57,044
56,985
--
6,429
4,559
14,209
11.3
8.0
--
1992
Census
SSA
1996
Census
PENDAT
Segal
52,170
113,000
52,171
5,000
100.0
4.4
57,295
56,730
--
57,295
831
3,741
100.0
1.5
--
B-4
Appendix Table B-1—continued. Alternative Estimates of State and Local Government
Employees Not Covered under Social Security
State
Illinois
Indiana
Iowa
Kansas
Kentucky
Louisiana
Year
Source
Active
Members
Members Not
Covered
Percent Not
Covered
1992
Census
SSA
511,125
985,000
315,297
470,000
61.7
47.7
1996
Census
PENDAT
Segal
589,341
518,568
--
363,546
302,871
351,845
61.7
58.4
--
1992
Census
SSA
230,420
436,000
21,897
58,000
9.5
13.3
1996
Census
PENDAT
Segal
239,790
78,917
--
22,787
51
48,530
9.5
0.1
--
1992
Census
SSA
157,294
270,000
22,809
28,000
14.5
10.4
1996
Census
PENDAT
Segal
153,563
151,046
--
22,268
3,543
21,702
14.5
2.3
--
1992
Census
SSA
124,077
257,000
11,768
24,000
9.5
9.3
1996
Census
PENDAT
Segal
144,755
144,115
--
13,729
3,663
19,512
9.5
2.5
--
1992
Census
SSA
191,681
325,000
89,332
84,000
46.6
25.8
1996
Census
PENDAT
Segal
177,264
178,044
--
82,613
53,355
64,120
46.6
30.0
--
1992
Census
SSA
282,663
396,000
277,112
282,000
98.0
71.2
1996
Census
PENDAT
Segal
226,357
164,603
--
221,912
164,603
220,876
98.0
100.0
--
B-5
Appendix Table B-1—continued. Alternative Estimates of State and Local Government
Employees Not Covered under Social Security
State
Maine
Maryland
Massachusetts
Michigan
Minnesota
Mississippi
Year
Source
1992
Census
SSA
1996
Census
PENDAT
Segal
Active
Members
Members Not
Covered
Percent Not
Covered
49,899
110,000
49,899
59,000
100.0
53.6
47,494
---
47,494
37,838
100.0
---
1992
Census
SSA
200,754
396,000
189,352
39,000
94.3
9.8
1996
Census
PENDAT
Segal
202,635
180,933
--
191,116
325
28,126
94.3
0.2
--
1992
Census
SSA
275,027
325,000
275,027
279,000
100.0
85.8
1996
Census
PENDAT
Segal
288,708
11,174
--
288,708
11,174
301,520
100.0
100.0
--
1992
Census
SSA
510,097
790,000
107,016
116,000
21.0
14.7
1996
Census
PENDAT
Segal
456,930
96,751
--
95,862
7,578
85,808
21.0
7.8
--
1992
Census
SSA
288,704
422,000
60,244
--
20.9
--
1996
Census
PENDAT
Segal
281,123
169,122
--
58,662
11,173
72,948
20.9
6.6
--
1992
Census
SSA
189,389
222,000
51,293
20,000
27.1
9.0
1996
Census
PENDAT
Segal
146,825
144,003
--
39,765
17,722
27.1
0.0
--
B-6
Appendix Table B-1. Alternative Estimates of State & Local Government Employees not Covered
under Social Security
State
Missouri
Montana
Nebraska
Nevada
New Hamphire
New Jersey
Year
Source
Active
Members
Members Not
Covered
Percent Not
Covered
1992
Census
SSA
195,932
385,000
87,544
72,000
44.7
18.7
1996
Census
PENDAT
Segal
239,127
206,014
--
106,844
68,615
59,992
44.7
33.3
--
1992
Census
SSA
56,985
93,000
40,348
16,000
70.8
17.2
1996
Census
PENDAT
Segal
50,833
18,332
--
35,992
12,666
70.8
0.0
--
1992
Census
SSA
51,032
165,000
15,680
13,000
30.7
7.9
1996
Census
PENDAT
Segal
65,952
---
20,264
-10,294
30.7
---
1992
Census
SSA
106,437
93,000
106,437
32,000
100.0
34.4
1996
Census
PENDAT
Segal
70,789
68,115
--
70,789
68,115
75,000
100.0
100.0
--
1992
Census
SSA
39,838
88,000
19,576
14,000
49.1
15.9
1996
Census
PENDAT
Segal
41,253
38,899
--
20,271
4,523
11,039
49.1
11.6
--
1992
Census
SSA
435,039
591,000
433,783
35,000
99.7
5.9
1996
Census
PENDAT
Segal
401,893
421,125
--
400,733
40,274
39,264
99.7
9.6
--
B-7
Appendix Table B-1—continued. Alternative Estimates of State and Local Government
Employees Not Covered under Social Security
State
New Mexico
New York
N. Carolina
N. Dakota
Ohio
Oklahoma
Year
Source
Active
Members
Members Not
Covered
Percent Not
Covered
1992
Census
SSA
112,508
175,000
79,777
30,000
70.9
17.1
1996
Census
PENDAT
Segal
107,682
96,277
--
76,355
40,495
22,269
70.9
42.1
--
1992
Census
SSA
1,237,779
1,673,000
592,855
120,000
47.9
7.2
1996
Census
PENDAT
Segal
1,067,339
1,030,731
--
511,220
90,680
47.9
0.0
--
1992
Census
SSA
400,435
579,000
82,423
47,000
20.6
8.1
1996
Census
PENDAT
Segal
404,261
258,087
--
83,211
3,751
37,034
20.6
1.5
--
1992
Census
SSA
26,872
70,000
12,048
9,000
44.8
12.9
1996
Census
PENDAT
Segal
27,487
26,101
--
12,324
921
7,831
44.8
3.5
--
1992
Census
SSA
812,899
800,000
812,899
739,000
100.0
92.4
1996
Census
PENDAT
Segal
674,816
665,771
--
674,816
665,771
921,404
100.0
100.0
--
1992
Census
SSA
146,608
267,000
28,181
17,000
19.2
6.4
1996
Census
PENDAT
Segal
144,869
139,217
--
27,857
1,579
13,633
19.2
1.1
--
B-8
Appendix Table B-1—continued. Alternative Estimates of State and Local Government
Employees Not Covered under Social Security
State
Oregon
Pennsylvania
Rhode Island
S. Carolina
S. Dakota
Tennessee
Year
Source
Active
Members
Members Not
Covered
Percent Not
Covered
1992
Census
SSA
161,617
264,000
25,484
18,000
15.8
6.8
1996
Census
PENDAT
Segal
150,614
148,111
--
23,749
5,859
14,288
15.8
4.0
--
1992
Census
SSA
431,070
740,000
230,641
50,000
53.5
6.8
1996
Census
PENDAT
Segal
431,530
329,428
--
230,887
11,494
42,004
53.5
3.5
--
1992
Census
SSA
34,854
74,000
32,046
13,000
91.9
17.6
1996
Census
PENDAT
Segal
36,962
32,661
--
33,984
6,477
9,852
91.9
19.8
--
1992
Census
SSA
311,036
310,000
311,036
30,000
100.0
9.7
1996
Census
PENDAT
Segal
211,150
220,660
--
211,150
420
23,175
100.0
0.2
--
1992
Census
SSA
35,913
75,000
3,649
3,000
10.2
4.0
1996
Census
PENDAT
Segal
33,506
33,508
--
3,404
309
2,266
10.2
0.9
--
1992
Census
SSA
198,749
409,000
146,754
56,000
73.8
13.7
1996
Census
PENDAT
Segal
213,461
185,935
--
157,617
10,384
41,549
73.8
5.6
--
B-9
Appendix Table B-1—continued. Alternative Estimates of State and Local Government
Employees Not Covered under Social Security
State
Texas
Utah
Vermont
Virginia
Washington
W. Virginia
Year
Source
Active
Members
Members Not
Covered
Percent Not
Covered
1992
Census
SSA
795,364
1,355,000
126,198
562,000
15.9
41.5
1996
Census
PENDAT
Segal
1,049,057
907,052
--
166,451
575,707
515,751
15.9
63.5
--
1992
Census
SSA
104,003
165,000
104,003
8,000
100.0
4.8
1996
Census
PENDAT
Segal
85,637
85,637
--
85,637
14,559
100.0
0.0
--
1992
Census
SSA
21,498
52,000
21,498
2,000
100.0
3.8
1996
Census
PENDAT
Segal
21,873
653
--
21,873
162
1,664
100.0
24.8
--
1992
Census
SSA
358,000
518,000
66,926
47,000
18.7
9.1
1996
Census
PENDAT
Segal
326,215
290,171
--
60,984
1,303
38,517
18.7
0.4
--
1992
Census
SSA
282,458
437,000
278,656
63,000
98.7
14.4
1996
Census
PENDAT
Segal
269,984
262,870
--
266,350
14,913
49,868
98.7
5.7
--
1992
Census
SSA
72,385
154,000
33,593
9,000
46.4
5.8
1996
Census
PENDAT
Segal
68,701
68,135
--
31,883
455
6,220
46.4
0.7
--
B-10
Appendix Table B-1—continued. Alternative Estimates of State and Local Government
Employees Not Covered under Social Security
State
Wisconsin
Wyoming
Year
Source
Active
Members
Members Not
Covered
Percent Not
Covered
1992
Census
SSA
313,878
464,000
88,378
65,000
28.2
14.0
1996
Census
PENDAT
Segal
258,299
246,830
--
72,729
2,396
46,579
28.2
1.0
--
1992
Census
SSA
31,803
66,000
31,803
10,000
100.0
15.2
Census
PENDAT
Segal
33,953
54,569
--
33,953
2,315
7,035
100.0
4.2
--
Sources: Zorn, 1996. U.S. Congress, House of Representatives, Committee on Ways and Means,
1998. U.S. Department of Commerce, Bureau of the Census, 1997.
(www.census.gov/govs/per/re97t5.txt) U.S. Department of Commerce, Bureau of the Census, 1995.
Segal Company, 1999.
B-11
APPENDIX C
DISCUSSION OF GPO AND WEP
As noted in the body of this report, the Government Pension Offset (GPO) of 1977 and
the Windfall Elimination Provision (WEP) of 1983 are amendments of the Social Security Act
which deal specifically with the treatment of people who have a history of noncovered
employment.
Two issues regarding these provisions arise when evaluating the impact of
mandatory coverage of state and local workers. The first issue is whether these provisions are
fair to state and local workers. That is, do these provisions create a penalty for noncovered
workers that would disappear if workers entered Social Security? If so, coverage inclusion in
Social Security may look attractive to state and local workers.
If these provisions treat
noncovered workers better than covered workers, fairness concerns speak for including these
workers to eliminate preferential treatment. The second issue is how well these provisions are
administered. If it is difficult and costly to administer these provisions, then bringing all state
and local workers into the system would eliminate the need for special provisions and thereby
eliminate the costs of administering them.
The GPO reduces the Social Security benefits to a person who earns a government
pension through noncovered employment and who also has a Social Security entitlement through
their spouse. The reason for the offset is that the spouse’s benefit is intended to provide income
for men and women who are financially dependent on their spouse. Spouses who earn a pension
through their own work history are not considered financially dependent on their spouses. One
can view the government as willing to ensure that every person is entitled to at least half of the
Social Security benefit of their spouse. Thus, the government is willing to provide the difference
between the earned benefit and half of the spouse’s benefit.
The GPO reduction is 2/3 of
amount of government pension. 1 The reason the offset is not dollar for dollar is that the pension
provided by noncovered government employment acts in part as a substitute for Social Security
and in part as an employer provided pension similar to those in the private sector. Thus, just as
Social Security does not reduce the Social Security benefit for private pensions, it should not
subtract comparable element of the government pensions from the Social Security benefit. Of
course, the choice of 2/3 is an ad hoc adjustment factor. 2 CRS (1990) investigate whether the
1
Take for example, a noncovered state worker whose pension is $600 and whose spouse has a Social
Security benefit of $1,000. His Social Security benefit is $100 = (1/2 * $1,000 ) – ( 2/3 * $600).
2
CRS (1990) investigated whether the 2/3 adjustment is appropriate for federal workers who are covered by
CSRS. They find that, in general, WEP over-penalizes lower-paid workers with shorter work careers and
under-penalizes higher-paid workers with long careers.
C-1
2/3 adjustment is appropriate for federal workers who are covered by CSRS. They find that in
general WEP over penalizes lower-paid workers with shorter work careers and underpenalizes
higher-paid workers with long careers. Inclusion of state and local workers in Social Security
would remove the arbitrary nature of this adjustment.
WEP, the other special provision for noncovered workers, institutes a modified benefit
formula for people who qualify for a Social Security benefit based on a brief work history in
covered employment but who earn a pension from noncovered employment. Whereas the regular
benefit provides 90% of the first $505 of a person’s average indexed monthly earnings (AIME),
the revised formula provides 40% if the person has 20 or less years of covered employment. For
each year of covered employment up to 30 years, an additional 5 percent is added to the 40
percent. For 30 or more years of covered employment, it remains at 90 percent. 3 WEP also
provides protection for workers with low pensions; the Social Security benefit under WEP plus
½ the pension must be greater than the Social Security benefit under the regular treatment.
Thus, Social Security treats people who have short period of covered employment due to leaving
the workforce, perhaps because of disability or for child rearing, better than people who have a
short period of covered employment due to switching to noncovered employment.
This discrimination occurs because Social Security aims to be redistributive, i.e. people
with lower earnings should get a higher return than people with high earnings. Thus, someone
who has a full working career should receive a smaller percent return than someone who was
only able to work a portion of their potential career because it is reasonable to assume the person
who works longer has more wealth.
Figure C-1 illustrates how the percent of AIME replaced changes by years of covered
employment for an illustrative worker who earns the indexed equivalent of $30,000 a year.
Under regular treatment, the replacement percent increases. Under WEP, replacement percent
hovers around 40 percent, which is near the replacement rate for someone who earns indexed
equivalent of $30,000 a year for their entire working career in covered employment. Thus, the
WEP appears to attain its objective of ensuring that people with limited working careers in
covered employment due to employment in noncovered jobs are treated the same with respect to
redistribution as people who spend their entire working career in covered employment.
Another issue of the GPO and WEP is administration of the amendments. The 1998
General Accounting Office (GAO) report “Better Payment Controls for Benefit Reduction
3
For example, a worker who has 24 years of covered employment receives 60 percent of their first $505 of
their AIME.
C-2
Provisions Could Save Millions” describes the procedure SSA uses for identifying people with a
history on noncovered employment. SSA staff asks applicants for SSA benefits whether they
currently or will in the future receive a pension from noncovered employment. Also, SSA staff
investigates the work history of the applicant for significant periods of no covered employment
and then asks the applicant the reason for the gap. The appropriate adjustments to the SSA
benefit are then applied. The GAO investigators estimate that by failing to identify noncovered
workers, SSA overpaid between $109 and $274 million related to GPO. They also estimate that
between $52 and $81 million in WEP related overpayments were made from 1986 to 1995.
These overpayments are made chiefly to state and local workers since information on the
pensions of retired federal employees (who at one time were not included in Social Security) can
be attained from Office of Personnel Management (OPM). The report notes that this process
could benefit by verifying the reports of the applicants with a second source of information on
pension payments such as the individual pension plans or the IRS. They recommend working
with IRS to build a method for receiving the needed information from income tax returns. SSA
is developing a proposal that would permit IRS to identify persons receiving pensions from
noncovered employment. Of course, if state and local workers are included in Social Security
the need for this type of investigation will eventually be eliminated since SSA will have the
necessary information on all employment histories.
In short, GPO and WEP appear to treat noncovered workers neither advantageously nor
disadvantageously, the administration of the provisions appears to be effective, and there are
plans for improvement. Thus, the affects of these provisions do not seem to lend support for
either side of the argument concerning mandatory coverage for state and local workers.
Additionally, although there are currently bills being considered to reduce the GPO from 2/3,
there does not seem to be reason for this action.
References:
Kollman, Geoffry C. 1990. “What Amount of a Civil Service Retirement System
(CSRS) Pension is Equivalent to a Social Security Benefit?” Congressional Research
Service.
C-3