Social Security: Implications of Extending Mandatory Coverage to State
and Local Employees (Letter Report, 08/18/98, GAO/HEHS-98-196).

Pursuant to a congressional request, GAO examined the implications of
extending mandatory social security coverage to all newly hired state
and local employees, focusing on: (1) the implications of mandatory
coverage for the social security program and for public employers,
employees, and pension plans; and (2) potential legal or administrative
problems associated with implementing mandatory coverage.

GAO noted that: (1) the Social Security Administration (SSA) estimates
that extending mandatory social security coverage to all newly hired
state and local government employees would reduce the program's
long-term actuarial deficit by about 10 percent and would extend the
trust funds' solvency by about 2 years; (2) in addition to helping to
some extent resolve the solvency problem, mandatory coverage would
broaden participation in an important national program and simplify
program administration; (3) the impact on public employers, employees,
and pension plans would depend on how state and local governments with
noncovered employees respond to the additional costs and benefits
associated with social security coverage; (4) social security retirement
benefits are fully protected from inflation and are weighted in favor of
families and low-income employees; (5) many public pension plans, on the
other hand, permit employees to retire earlier and provide a higher
retirement income benefit than social security; (6) those states and
localities that decide to maintain benefit levels for new employees
consistent with the earlier retirement age and enhanced retirement
income benefit would experience increased costs; (7) however, those
employees would also have the additional family and other protection
provided by social security; (8) alternatively, states and localities
that choose to maintain level retirement spending might need to reduce
some retirement benefits for newly hired employees; (9) several
employer, employee, and plan representatives stated that mandating
social security coverage for all new state and local government
employees would raise constitutional issues and would be challenged in
court; (10) however, GAO believes that mandatory coverage is likely to
be upheld under current Supreme Court decisions; (11) mandatory coverage
would also present administrative issues for implementing state and
local governments; and (12) up to 4 years could be required for states
and localities to develop, legislate, and implement pension plans that
are coordinated with social security.

--------------------------- Indexing Terms -----------------------------

 REPORTNUM:  HEHS-98-196
     TITLE:  Social Security: Implications of Extending Mandatory 
             Coverage to State and Local Employees
      DATE:  08/18/98
   SUBJECT:  Employee retirement plans
             Social security benefits
             Government retirement benefits
             State employees
             Retirement pensions
IDENTIFIER:  Social Security Program
             California
             Colorado
             Illinois
             Louisiana
             Massachusetts
             Ohio
             Texas
             
******************************************************************
** This file contains an ASCII representation of the text of a  **
** GAO report.  Delineations within the text indicating chapter **
** titles, headings, and bullets are preserved.  Major          **
** divisions and subdivisions of the text, such as Chapters,    **
** Sections, and Appendixes, are identified by double and       **
** single lines.  The numbers on the right end of these lines   **
** indicate the position of each of the subsections in the      **
** document outline.  These numbers do NOT correspond with the  **
** page numbers of the printed product.                         **
**                                                              **
** No attempt has been made to display graphic images, although **
** figure captions are reproduced.  Tables are included, but    **
** may not resemble those in the printed version.               **
**                                                              **
** Please see the PDF (Portable Document Format) file, when     **
** available, for a complete electronic file of the printed     **
** document's contents.                                         **
**                                                              **
** A printed copy of this report may be obtained from the GAO   **
** Document Distribution Center.  For further details, please   **
** send an e-mail message to:                                   **
**                                                              **
**                                            **
**                                                              **
** with the message 'info' in the body.                         **
******************************************************************


Cover
================================================================ COVER


Report to the Chairman, Subcommittee on Social Security, Committee on
Ways and Means, House of Representatives

August 1998

SOCIAL SECURITY - IMPLICATIONS OF
EXTENDING MANDATORY COVERAGE TO
STATE AND LOCAL EMPLOYEES

GAO/HEHS-98-196

Mandatory Social Security Coverage

(207012)


Abbreviations
=============================================================== ABBREV

  BLS - Bureau of Labor Statistics
  GPO - Government Pension Offset
  IRS - Internal Revenue Service
  PPCC - Public Pension Coordinating Council
  SSA - Social Security Administration
  WEP - Windfall Elimination Provision

Letter
=============================================================== LETTER


B-278999

August 18, 1998

The Honorable Jim Bunning
Chairman, Subcommittee on Social Security
Committee on Ways and Means
House of Representatives

Dear Mr.  Chairman: 

The Social Security Act of 1935 excluded state and local government
employees from coverage because there was concern over the question
of the federal government's right to impose a tax on state
governments and because many state and local employees were already
protected by public pension plans.  Since 1935, the Congress has
extended mandatory Social Security coverage to state and local
employees not covered by a public pension plan and voluntary coverage
to other state and local government employees.\1

The Social Security Administration (SSA) estimates that about 5
million state and local government employees, with annual salaries
totaling about $132.5 million, are currently occupying positions not
covered by Social Security. 

Under current estimates, Social Security revenues will fall short of
expenditures and the Trust Funds will be exhausted by 2032.  The
1994-1996 Social Security Advisory Council\2

could not agree on how to resolve Social Security's financial
shortfall, but groups of council members coalesced around three
proposals.  A common element of the three proposals was to extend
mandatory coverage to all newly hired state and local government
workers. 

You asked us to examine the implications of extending mandatory
coverage to all newly hired state and local employees.  Specifically,
you asked us to examine the implications of mandatory coverage for
the Social Security program and for public employers, employees, and
pension plans.  You also asked us to identify potential legal or
administrative problems associated with implementing mandatory
coverage.  We presented preliminary findings in testimony before this
Subcommittee on May 21, 1998.\3

In response to your request, we examined SSA's estimates of the
impact of mandatory coverage for Social Security revenues,
expenditures, and trust fund balances.  We discussed the implications
of mandatory coverage with state and local government employer,
employee, and pension plan representatives in the seven states that
account for over 75 percent of the noncovered payroll.  We also
examined relevant studies provided by these representatives.  We
reviewed a 1980 study of the feasibility of extending mandatory
coverage to state and local employees prepared at the request of the
Congress and two 1994 surveys of state and local pension plan costs
and benefits.  We discussed the financial implications of mandatory
coverage for public pension plans with actuaries and representatives
of state and local pension plans for covered and noncovered
employees.  Our methodology is described in more detail in appendix
1. 


--------------------
\1 Coverage is generally at the option of the state, and employees
who are members of a retirement system have a choice as a group with
respect to coverage. 

\2 The Social Security Act, as amended, required the appointment of
an advisory council every 4 years to examine issues related to the
Social Security and Medicare programs.  Under provisions of the
Social Security Independence and Program Improvements Act of 1994
(P.L.  103-296), the 1994-1996 Advisory Council was the last advisory
council to be appointed. 

\3 Social Security:  Mandating Coverage for State and Local Employees
(GAO/T-HEHS-98-127, May 21, 1998). 


   RESULTS IN BRIEF
------------------------------------------------------------ Letter :1

SSA estimates that extending mandatory Social Security coverage to
all newly hired state and local government employees would reduce the
program's long-term actuarial deficit by about 10 percent and would
extend the trust funds' solvency by about 2 years.  In addition to
helping to some extent resolve the solvency problem, mandatory
coverage would broaden participation in an important national program
and simplify program administration. 

The effect on public employers, employees, and pension plans would
depend on how state and local governments with noncovered employees
respond to the additional costs and benefits associated with Social
Security coverage.  Social Security retirement benefits are fully
protected from inflation and are weighted in favor of families and
low-income employees.  Many public pension plans, on the other hand,
permit employees to retire earlier and provide a higher retirement
income benefit than Social Security.  Those states and localities
that decide to maintain benefit levels for new employees consistent
with the earlier retirement age and enhanced retirement income
benefit would experience increased costs.  However, those employees
would also have the additional family and other protection provided
by Social Security.  Alternatively, states and localities that choose
to maintain level retirement spending might need to reduce some
retirement benefits for newly hired employees. 

Several employer, employee, and plan representatives stated that
mandating Social Security coverage for all new state and local
government employees would raise constitutional issues and would be
challenged in court.  However, we believe that mandatory coverage is
likely to be upheld under current U.S.  Supreme Court decisions. 
Mandatory coverage would also present administrative issues for
implementing state and local governments.  Up to 4 years could be
required for states and localities to develop, legislate, and
implement pension plans that are coordinated with Social Security. 


   BACKGROUND
------------------------------------------------------------ Letter :2

The Social Security Act of 1935 required most workers in commerce and
industry, then about 60 percent of the workforce, to be covered. 
Amendments to the act in 1950, 1954, and 1956 allowed states,
generally acting for their employees, to voluntarily elect Social
Security coverage through agreements with SSA.  The amendments also
permitted states and localities that elected coverage to withdraw
from the program after meeting certain conditions. 

Policymakers have addressed the issue of extending mandatory Social
Security coverage for state and local government employees on several
occasions.  In response to financial problems the Social Security
system faced in the early 1970s, for example, the 1977 Social
Security amendments directed that a study be made of the desirability
and feasibility of extending mandatory coverage to employees at all
levels of government, including state and local governments.  The
Secretary of the Department of Health, Education, and Welfare--now
the Departments of Health and Human Services and
Education--established the Universal Social Security Coverage Study
Group to develop options for mandatory coverage and analyze the
fiscal effects of each option. 

Recognizing the diversity of state and local systems, the study group
selected representative plans for analysis.  Two data sources were
developed and analyzed.  First, the Actuarial Education and Research
Fund, sponsored by six professional actuarial organizations,
established a task force of plan actuaries to study 25 representative
large and small noncovered retirement systems.  Second, the Urban
Institute, under a grant from several government agencies, used an
actuarial firm to obtain data on 22 of the largest 50 noncovered
employee retirement systems.  The study group report, issued in 1980,
provided information on the costs and benefits of various options but
did not draw conclusions about their relative desirability.\4

In 1983, the Congress removed authority for states and localities
that had voluntarily elected Social Security coverage to withdraw
from the program, which effectively made coverage mandatory for many
state and local employees.  Additionally, in 1990, the Congress
mandated coverage for state and local employees not covered by public
pension plans.  SSA estimates that 96 percent of the workforce,
including 70 percent of the state and local government workforce, is
now covered by Social Security. 

During 1997, Social Security had $457.7 billion in revenues and
$369.1 billion in expenditures.  About 89 percent of Social
Security's revenues came from payroll taxes.  The Social Security
payroll tax is 6.2 percent of pay each for employers and employees,
up to an established maximum.  Maximum earnings subject to Social
Security payroll taxes were $65,400 in 1997 and are $68,400 in 1998. 

Social Security provides retirement, disability, and survivor
benefits to insured workers and their families.  Insured workers are
eligible for full retirement benefits at age 65 and reduced benefits
at age 62.  The retirement age was increased by the 1983 Social
Security amendments.  Beginning with those born in 1938, the age at
which full benefits are payable will increase in gradual steps from
age 65 to age 67. 

Benefit amounts are based on a worker's age and career earnings, are
fully indexed for inflation, and as shown in table 1, replace a
relatively higher proportion of the final year's wages for low
earners. 



                          Table 1
          
             Estimated Average Social Security
            Earnings Replacement Rates Based on
            Intermediate Assumptions in the 1998
                      Trustees Report

Year of     Retirement         Low     Average     Maximum
retirement         age    earner\a      earner    earner\b
----------  ----------  ----------  ----------  ----------
2000                65        55.4        41.2        24.2
2015                66        56.7        42.2        27.7
2040                67        56.5        42.1        27.7
----------------------------------------------------------
\a Equal to 45 percent of the average wage. 

\b Equal to maximum earnings subject to Social Security payroll
taxes. 

Source:  Office of the Chief Actuary, SSA. 

Social Security provides additional benefits for eligible family
members, including spouses aged 62 or older--or younger spouses if a
child meeting certain requirements is in their care--and children up
to age 18--or older if they are disabled.  The amount of a spouse's
or child's benefit is one-half the insured worker's age-65 benefit
amount.  A spouse's benefit is reduced if taken earlier than age 65,
unless the spouse has a child in his or her care. 

SSA estimates that about 5 million state and local government
employees, excluding students and election workers, occupy positions
not covered by Social Security.  SSA also estimates that the
noncovered employees have annual salaries totaling about $132.5
billion.  Seven states--California, Colorado, Illinois, Louisiana,
Massachusetts, Ohio, and Texas--account for over 75 percent of the
noncovered payroll.  Based on a 1995 survey of public pension
plans,\5 the Public Pension Coordinating Council (PPCC) estimates
that police, firefighters, and teachers are more likely to occupy
noncovered positions than other employees are. 

According to a 1994 Bureau of Labor Statistics (BLS) survey,\6 most
full-time state and local employees participate in defined benefit
pension plans.  Table 2 shows membership and contribution rates for
nine defined benefit state and local pension plans that we studied as
part of the review.  For the most part, active members in the nine
plans occupy positions that are not covered by Social Security. 



                                         Table 2
                         
                           Membership, Contribution Rates, and
                         Assets for Selected Public Pension Plans

                                                   Contribution rate\a
                                            ---------------------------------
Public pension          Active     Benefit                                     Net assets
plan                   members  recipients    Employer    Employee      Total  (billions)
------------------  ----------  ----------  ----------  ----------  ---------  ----------
California State       364,000     154,000       12.5%        8.0%      20.5%       $74.8
 Teachers'
 Retirement System
Public Employees'      148,000      46,000        11.6         8.0       19.6        19.9
 Retirement
 Association of
 Colorado
Teachers'              137,000      59,000         7.9         8.0       15.9        17.4
 Retirement System
 of the State of
 Illinois
Louisiana State         70,000      27,000        12.0         7.5       19.5         4.3
 Employees'
 Retirement System
Massachusetts           83,000      42,000        14.5         9.0       23.5         9.6
 State Retirement
 System
Massachusetts           69,000      29,000        14.0         9.0       23.0         9.9
 Teachers'
 Contributory
 Retirement System
State Teachers         169,000      89,000        14.0         9.3       23.3        42.4
 Retirement System
 of Ohio
Public Employees'      345,000     146,000        13.3         8.5       21.8        39.8
 Retirement System
 of Ohio
Teacher Retirement     695,000     158,000         6.0         6.4       12.4        62.2
 System of Texas
=========================================================================================
Total                2,080,000     750,000                                         $280.3
-----------------------------------------------------------------------------------------
\a The employer rate includes contributions toward the plan's
unfunded liability.  Several plans have multiple employee
contribution rates.  The rate provided is for state employees,
excluding special-rate groups, such as state police. 

Source:  State and pension plan financial reports. 

Defined benefit plans promise a specific level of benefits to their
members when they retire.  Minimum retirement age and benefits vary;
however, the BLS and PPCC surveys indicate that many public employees
can retire with full benefits at age 55 or earlier with 30 years of
service.  The surveys also indicate that plan members typically have
a benefit formula that calculates retirement income on the basis of
specified benefit rates for each year of service and the members'
average salary over a specified time period--usually the final 3
years. 

For example, the benefit rates for members of the Colorado Public
Employees' Retirement Association are 2.5 percent of highest average
salary per year over a 3-year period for the first 20 years of
service and 1.5 percent of highest average salary per year for each
additional year of service.  Full retirement benefits are available
at any age with 35 years of service, at age 55 with 30 years of
service, age 60 with 20 years of service, or at age 65 with 5 years
of service.  Therefore, plan members who retire at age 55 with 30
years of service receive annual retirement income amounting to 65
percent of their highest average salary.  Reduced retirement benefits
are available, for example, at age 55 with 20 years of service. 

In addition to retirement income benefits, most public pension plans
provide other benefits, such as disability or survivor benefits.  For
example, BLS reported that of defined benefit plan members, 91
percent were provided with disability benefits, all have a survivor
annuity option, and 62 percent receive some cost-of-living increases
after retirement. 

Public pension plan coverage for part-time, seasonal, and temporary
employees varies.  In Ohio, for example, part-time and temporary
state employees participate in a defined benefit plan.  In
California, the 16,000 part-time, seasonal, and temporary state
employees have a defined contribution plan.  Plan benefits are based
on plan contributions, which consist of 7.5 percent of the employees'
gross pay deducted from their pay and returns on plan investments. 


--------------------
\4 Universal Social Security Coverage Study Group, The Desirability
and Feasibility of Social Security Coverage for Employees of Federal,
State, and Local Governments and Private, Nonprofit Organizations
(Mar.  1980). 

\5 Paul Zorn, Survey Report:  1995 Survey of State and Local
Government Employee Retirement Systems (Washington, D.C.:  PPCC, July
1996). 

\6 BLS, Employee Benefits in State and Local Governments, 1994,
Bulletin 2477 (Washington, D.C.:  Department of Labor, May 1996). 


   MANDATORY COVERAGE WOULD
   BENEFIT THE SOCIAL SECURITY
   PROGRAM
------------------------------------------------------------ Letter :3

SSA estimates that extending mandatory Social Security coverage to
all newly hired state and local employees would reduce the trust
funds' 75-year actuarial deficit by about 10 percent.\7 The surplus
payroll tax revenues associated with mandatory coverage and interest
on that surplus would extend the trust funds' solvency by about 2
years.  Extending mandatory coverage to newly hired employees would
also increase program participation and, in the long run, simplify
program administration. 


--------------------
\7 SSA uses a period of 75 years for evaluating the program's
long-term actuarial status to obtain the full range of financial
commitments that will be incurred on behalf of current program
participants. 


      TRUST FUNDS' DEFICIT WOULD
      BE REDUCED
---------------------------------------------------------- Letter :3.1

Table 3 shows SSA's analysis of the present discounted value of
revenues and expenditures with and without mandatory coverage over
the 75-year period beginning January 1, 1998.  The analysis indicates
that extending mandatory coverage to all state and local employees
hired beginning January 1, 2000, would reduce the program's long-term
actuarial deficit by 10 percent, from about 2.19 percent of payroll
to 1.97 percent of payroll. 



                          Table 3
          
              Present Value of Social Security
          Revenues and Expenditures and Actuarial
           Balance Over 75 Years With and Without
                     Mandatory Coverage

                           Without        With
                         mandatory   mandatory
                          coverage    coverage      Change
                        (billions)  (billions)  (billions)
----------------------  ----------  ----------  ----------
Beginning trust fund        $655.5      $655.5        $0.0
 balance
Present value of total    18,413.4    18,934.6       521.2
 revenues
Present value of total    21,983.0    22,274.7       291.7
 expenditures
Revenue minus            (3,569.6)   (3,340.1)       229.5
 expenditures
Target trust fund            185.4       192.7         7.3
 balance\a
Actuarial balance        (3,099.5)   (2,877.3)       222.2
Present value of         141,779.0   145,878.9     4,099.9
 payroll
Actuarial balance as a      (2.19)      (1.97)        0.22
 percent of payroll
----------------------------------------------------------
\a The target trust fund balance is an amount equal to the following
year's projected expenditures. 

Source:  SSA, Office of the Chief Actuary. 

Figure 1 shows that SSA's analysis indicates that extending mandatory
coverage to new state and local employees would extend the trust
funds' solvency by about 2 years, from 2032 to 2034.  As with most
other elements of the reform proposals put forward by the 1994-1996
Social Security Advisory Council, extending mandatory coverage to
newly hired state and local employees would contribute to the
resolution of--but not fully resolve--the trust funds' solvency
problem.  A combination of adjustments will be needed to extend the
program's solvency over the entire 75-year period. 

   Figure 1:  Comparison of
   Projected End-of-Year Trust
   Fund Balances With and Without
   Mandatory Coverage, From 1998
   to 2048

   (See figure in printed
   edition.)

Note:  SSA data were based on intermediate assumptions in the 1998
Board of Trustees' report.  SSA assumed that mandatory coverage would
be effective beginning January 1, 2000. 

Source:  GAO analysis of SSA data. 

SSA's analysis indicates that revenues resulting from an extension of
mandatory coverage, including payroll taxes and interest on surplus
revenues, would substantially exceed additional expenditures
throughout the 75-year period.  SSA assumes that payroll tax
collections for new employees would accelerate early in the 75-year
period, while benefits for those employees would not accelerate until
later in the period.  For example, annual revenues from payroll taxes
collected from the newly covered employees and their employers are
expected to exceed expenditures for benefits to those employees until
2050.  In that year, however, revenues resulting from an extension of
mandatory coverage, including interest on cumulative surplus
revenues, are projected to exceed expenditures on those employees by
over 300 percent. 


      MANDATORY COVERAGE WOULD
      HAVE OTHER BENEFICIAL
      EFFECTS
---------------------------------------------------------- Letter :3.2

While Social Security's solvency problems triggered the analysis of
the effect of mandatory coverage on program revenues and
expenditures, the inclusion of such coverage in a comprehensive
reform package would likely be grounded in other considerations as
well, such as broadening Social Security's coverage and simplifying
program administration. 

According to SSA, about 91 percent of elderly households received
Social Security benefits in 1994.\8 Social Security contributes
substantially to reducing poverty in these households.  For example,
in 1994, 11.7 percent of persons age 65 or over were poor.  Excluding
Social Security benefits, however, the incomes of about 54 percent of
persons age 65 or older would have been below the poverty
threshold.\9 In recommending that mandatory coverage be included in
its reform proposals, the advisory council stated that mandatory
coverage is basically "an issue of fairness." The advisory council
report stated that

     an effective Social Security program helps to reduce 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. 

According to SSA, one important way that noncovered employees benefit
from, without contributing to, Social Security is that their parents,
grandparents, or other relatives receive Social Security's
retirement, disability, or survivor benefits.  Social Security is
designed as a national intergenerational transfer program where the
taxes of current workers fund the benefits of current beneficiaries. 
SSA stated that those not contributing to the program still receive
the benefits of this transfer. 

Extending mandatory Social Security coverage to all newly hired state
and local employees would also simplify program administration by
eliminating, over time, the need to administer and enforce special
rules for noncovered state and local employees.  For example, SSA's
Office of Research, Evaluation, and Statistics estimates that 95
percent of state and local employees occupying noncovered positions
become entitled to Social Security as either workers or dependents. 
Additionally, the Office of the Chief Actuary estimates that 50 to 60
percent of state and local employees in noncovered positions will be
fully insured by age 62 from other, covered employment. 

The Congress established the Windfall Elimination Provision (WEP) and
Government Pension Offset (GPO) to reduce the unfair advantage that
workers eligible for pension benefits on the basis of noncovered
employment may have when they apply for Social Security benefits. 
The earnings history for workers with noncovered earnings may appear
to qualify them for increased Social Security benefits as low-income
wage earners--or for additional benefits for a nonworking
spouse--when in fact they have had substantial income from noncovered
employment.  With a few exceptions, WEP and GPO require SSA to use
revised formulas to calculate benefits for workers with noncovered
employment. 

In April 1998, we reported that SSA is often unable to determine
whether applicants should be subject to WEP or GPO and this has led
to overpayments.\10 We estimated total overpayments to be between
$160 million and $355 million over the period 1978 to 1995.  In
response, SSA plans to perform additional computer matches with the
Office of Personnel Management and the Internal Revenue Service (IRS)
to obtain noncovered pension data and ensure WEP and GPO are
correctly applied.  Mandatory coverage would reduce required WEP and
GPO adjustments to benefits by gradually reducing the number of
employees in noncovered employment.  Eventually, all state and local
employees--with the exception of a few categories of workers, such as
students and election workers--would be in covered employment, and
adjustments would be unnecessary. 

In 1995, SSA asked its Office of the Inspector General to review
state and local government employers' compliance with Social Security
coverage provisions.  In December 1996, the Inspector General
reported that Social Security provisions related to coverage of state
and local employees are complex and difficult to administer.\11 The
report stated that few resources were devoted to training state and
local officials and ensuring that administration and enforcement
roles and responsibilities are clearly defined.  The report concluded
that there is a significant risk of sizeable noncompliance with state
and local coverage provisions.  In response, SSA and IRS have
initiated an effort to educate employers and ensure compliance with
legal requirements for withholding Social Security payroll taxes. 
Extending coverage to all newly hired state and local government
employees would eventually eliminate this problem. 

SSA stated that the time needed to fully phase in mandatory coverage
could be 20 to 30 years, if it followed estimates of the time needed
to phase in Medicare coverage, which was mandated for newly hired
state and local employees starting in 1986.  SSA also stated that
mandatory Social Security coverage for new hires would possibly
create another tier in the payroll reporting process resulting in
additional compliance issues in the near term.  Additionally, payroll
practitioners would need to account for Social Security covered and
noncovered government employment--along with Medicare covered and
noncovered employment--and, as a result, they would face additional
reporting burdens in the near term as they extended Social Security
coverage to new employees. 


--------------------
\8 Susan Grad, Income of the Population 55 or Older, 1994
(Washington, D.C.:  SSA, Office of Research and Statistics, 1996). 

\9 However, without Social Security, people may save more or continue
working. 

\10 Social Security:  Better Payment Controls for Benefit Reduction
Provisions Could Save Millions (GAO/HEHS-98-76, Apr.  30, 1998). 

\11 SSA Office of the Inspector General, Social Security Coverage of
State and Local Government Employees (A-04-95-06013, Dec.  13, 1996). 


   EFFECT OF MANDATORY COVERAGE
   FOR EMPLOYERS, EMPLOYEES, AND
   THEIR PUBLIC PENSION PLANS
   WOULD VARY
------------------------------------------------------------ Letter :4

If Social Security becomes mandatory, all newly hired state and local
employees would be provided with the minimum income protection
afforded by Social Security.  Also, they and their employers would
pay Social Security's combined 12.4-percent payroll tax.  Each state
and locality with noncovered employees would then decide how to
respond to the increase in benefits and costs.  Possible responses
range from the government's absorbing the added costs and leaving
current pension plans unchanged to entirely eliminating state and
local pension plan benefits for newly hired employees. 

From discussions with state and local representatives, however,
noncovered employers would likely adjust their pension plans to
reflect Social Security's benefits and costs.  To illustrate the
implications of mandatory coverage for public employers and
employees, we examined three possible responses: 

  -- States and localities could maintain similar total retirement
     benefits for current and newly hired employees.  For example,
     employees who retire before age 62 would be paid supplemental
     retirement benefits until they become eligible for Social
     Security benefits.  This response would likely result in an
     increase in total retirement costs and some additional family
     and other benefits for many newly hired employees. 

  -- States and localities could examine other pension plans that are
     already coordinated with Social Security and provide newly hired
     employees with similar benefits.  For example, employees who
     retire before age 62 would receive, on average, a smaller
     initial retirement benefit than current noncovered employees. 
     This response would also likely result in an increase in total
     retirement costs and some additional family and other benefits
     for newly hired employees. 

  -- States and localities could maintain level retirement costs. 
     This response would likely require a reduction in pension
     benefits from the government's plans for many newly hired
     employees, but the new employees would also have Social Security
     benefits. 

According to pension plan representatives, the changes to current
pension plans in response to mandatory coverage could result in
reduced contributions to those plans, which could affect their
long-term financing. 


      MAINTAINING BENEFITS OF
      NONCOVERED EMPLOYEES FOR
      NEWLY HIRED EMPLOYEES WOULD
      LIKELY INCREASE COSTS
---------------------------------------------------------- Letter :4.1

States and localities with noncovered employees could decide to
provide newly hired employees with pension benefits at retirement,
which, when combined with Social Security benefits, approximate the
pension benefits of current employees.  Studies indicate that such a
decision would likely result in an increase in retirement costs.  The
amount of increase would vary depending on a number of factors;
however, studies indicate the increase could be about 7 percent of
new-employee payroll. 

The 1980 Universal Social Security Coverage Study Group report
estimated that total retirement costs, including Social Security
payroll taxes and pension plan contributions, would need to increase
an average of 5 to 10 percent of payroll to maintain level benefits
for current and newly hired employees.  However, the estimated
increase included the 2.9 percent of payroll Medicare tax that was
mandated for all new state and local employees in 1986--6 years after
the study was completed.  Deducting the Medicare tax reduces the
estimate of additional costs to between 2 and 7 percent of payroll. 

The 1980 study group assumed that most newly hired employees would
have salary replacement percentages in their first year of retirement
that would be comparable to the salary replacement percentages
provided to current employees.  For example, employees retiring
before age 62 would receive a temporary supplemental pension benefit
to more closely maintain the benefits of the current plan.  Since
Social Security benefits are weighted in favor of families and lower
income employees--and because Social Security benefits are fully
indexed for inflation, while many pension plans provide limited or no
cost-of-living protection--total lifetime benefits for some new
employees would be greater than those provided to current employees. 

More recent studies by pension plan actuaries in Colorado, Illinois,
and Ohio also indicate the cost increase would be in the same range. 
For example, a December 1997 study for a plan in Ohio indicated that
providing retirement benefits for new employees that, when added to
Social Security benefits, approximate retirement benefits for current
employees would require an increase in contributions of 6 to 7
percent of new-employee payroll. 

A 1997 study for a pension plan in Illinois indicated the increased
payments necessary to maintain similar total retirement benefits for
current and new employees would be about 6.5 percent of new-employee
payroll.  Since it would be limited to new employees, the cost
increase would be phased in over several years.  For example, the
cost increase would be about 0.25 percent of total payroll starting
the first year, 2.83 percent of total payroll in 10 years, and 6.54
percent of total payroll after all current employees have been
replaced. 

The 1980 study group report stated that the causes of the cost
increase cannot be ascribed directly to specific Social Security or
pension plan provisions.  According to the study, however, among the
most important factors contributing to the cost increase are Social
Security's

  -- strengthening of cost-of-living protection,

  -- provision of substantial additional benefits to some families,
     and

  -- reduction in pension benefit forfeitures occurring when
     employees move between jobs. 

The study stated that another contributing factor would be the need
for pension plans to provide supplemental benefits to employees,
especially police and firefighters, who retire before they begin
receiving Social Security benefits at age 62.  The study also found
that the magnitude of the cost increase would depend on the pension
plan's current benefits.  Cost increases would be less for plans that
already provide benefits similar to those provided by Social Security
because those plans would be able to eliminate duplicate benefits. 

Maintaining level benefits for noncovered and newly hired employees
would require states and localities in redesigning plans for the
newly hired employees to adopt benefit formulas that explicitly
integrate pension and Social Security benefits.  For example,
affected states and localities could adopt a benefit formula that
offsets a portion of the member's pension benefit with a specified
percentage of the member's Social Security benefit.  This approach is
more common in the private sector--where a 1995 BLS survey of large
and medium establishments found that about 51 percent of full-time
employees had benefits integrated with Social Security\12 --than the
public sector, where a survey found that only about 4 percent of
full-time employees had pension benefits integrated with Social
Security.  In the public sector, pension plans for covered employees
generally recognize Social Security benefits implicitly by providing
their members with lower benefit rates than are provided to
noncovered employees. 


--------------------
\12 BLS, Employee Benefits in Medium and Large Private
Establishments, 1995, Bulletin 2496 (Washington, D.C.:  Department of
Labor, Apr.  1998). 


      PROVIDING BENEFITS OF
      CURRENTLY COVERED EMPLOYEES
      WOULD LIKELY INCREASE COSTS
---------------------------------------------------------- Letter :4.2

SSA estimates that about 70 percent of the state and local workforce
is already covered by Social Security.  The 1980 study group examined
the impact on retirement costs if states and localities with
noncovered employees provide newly hired employees with pension
benefits that are similar to the benefits provided to employees who
are already covered by Social Security.  The study group concluded
that implementing such formulas would increase overall retirement
costs by 6 to 14 percent of payroll--or about 3 to 11 percent of
payroll after deducting the Medicare tax.  The study also concluded
that for most pension plans, the present value of lifetime benefits
for new employees covered by Social Security would be greater than
the value of benefits of current noncovered employees. 

As shown in table 4, our analysis of 1995 PPCC data also indicates
that total retirement costs for states and localities covered by
Social Security are higher than the costs for noncovered states and
localities. 



                          Table 4
          
             Average Employee Contribution and
           Employer Benefit Costs for Covered and
                  Noncovered Pension Plans

                                  Noncovered       Covered
                                       plans       plans\a
------------------------------  ------------  ------------
Employee contribution rate                8%            9%
Employer cost rate\b                      8%           12%
----------------------------------------------------------
\a Includes Social Security payroll tax. 

\b Excludes administrative costs and the cost of amortizing the
plan's accrued unfunded liability. 

Source:  GAO analysis of 1995 PPCC survey data.  Our methodology is
described in appendix 1. 

PPCC data also indicate that many employees, especially police and
firefighters, retire before age 62, when they would first be eligible
for Social Security retirement benefits.  The data indicate, for
example, that police and firefighters in noncovered plans retired, on
average, at age 54.  The average retirement age of other employees in
noncovered plans was age 60.  In covered plans, the average
retirement age for police and firefighters and other employees was
somewhat higher at ages 55 and 62, respectively. 

Analyses indicate that, initially, the percentage of salary that is
replaced by retirement income is smaller for covered employees who
retire before they are eligible for Social Security benefits than for
noncovered employees.  Our analysis of PPCC data indicates, for
example, that public pension plans replace about 65 percent of the
final average salary of members who retired with 30 years of service
and were not covered by Social Security.  For members who retired
with 30 years of service and were covered by both a pension plan and
Social Security, the PPCC data indicate that pension plans replace
only about 53 percent of their members' final average salary.  After
Social Security benefits begin, however, covered employees generally
have higher salary replacement rates.  For example, the average
salary replacement rates in 1994 were higher for covered state and
local employees than for noncovered employees, after they reach age
62 at all salary levels between $15,000 and $65,000.  (See table 5.)



                          Table 5
          
          Average 1994 Salary Replacement Rates in
            Defined Benefit Plans for Full-Time
           State and Local Employees Retiring at
              Age 62 With 30 Years of Service

                            Employees covered by Social
                             Security and pension plan
                          --------------------------------
               Employees
                 covered
                 only by
                 pension     Pension      Social
Final salary        plan        plan  Security\a     Total
------------  ----------  ----------  ----------  --------
$15,000              62%         50%         37%       87%
$25,000               62          50          31        81
$35,000               62          50          27        77
$45,000               62          50          23        73
$55,000               62          50          20        70
$65,000               62          50          17        67
----------------------------------------------------------
\a Excludes Social Security spousal and dependent benefits. 

Source:  BLS, Employee Benefits in State and Local Governments, 1994. 

We did not compare the expected value of total lifetime benefits for
covered and noncovered employees because amounts would vary depending
on the benefits offered by each plan. 

The extent to which the experience of states and localities with
covered employees can be generalized to those with noncovered
employees is limited.  According to the 1980 study group report, most
public pension plans that coordinated with Social Security did so in
the 1950s and 1960s when Social Security benefits and payroll taxes
were much smaller.  As Social Security benefits grew, pension plan
benefits remained basically unchanged.  The study stated that,
starting in the 1970s, however, rising pension costs caused several
large state systems to consider reducing their relatively liberal
pension benefits.  In the 1980s, for example, California created an
alternative set of reduced benefits for general employees to, among
other things, reduce the state's retirement costs.  Initially,
general employees were permitted to select between the higher costs
and benefits of the original plan and the lower costs and benefits of
the revised plan.  Subsequently, however, newly hired general
employees were limited to the reduced benefits.  Regardless, the
circumstances surrounding the experiences of states with covered
employees make it difficult to predict what changes would occur from
further extension of coverage. 


      LEVEL RETIREMENT SPENDING
      WOULD MEAN REDUCED BENEFITS
---------------------------------------------------------- Letter :4.3

Several employer, employee, and pension plan representatives with
whom we spoke stated that spending increases necessary to maintain
level retirement income and other benefits would be difficult to
achieve.  State and pension plan officials noted that spending for
retirement benefits must compete for funds with spending for
education, law enforcement, and other areas that cannot be readily
reduced.  For example, Ohio officials noted that the state is having
difficulty finding the additional funds for education needed to
comply with court ordered changes in school financing.  A
representative of local government officials in Ohio stated that
payroll represents 75 to 80 percent of county budgets, and there is
little chance that voters would approve revenue increases needed to
maintain level retirement benefits.  He stated the more likely
options for responding to increased retirement costs were to decrease
the number of employees or reduce benefits under state and local
pension plans. 

If states and localities decide to maintain level spending for
retirement, they might need to reduce pension benefits under public
pension plans for many employees.  For example, a June 1997 actuarial
evaluation of an Ohio pension plan examined the impact on benefits of
mandating Social Security coverage for all employees, assuming no
increase in total retirement costs.  The study concluded that level
spending could be maintained if

  -- service retirement benefits were reduced (for example, salary
     replacement rates for employees retiring with 30 years of
     service would be reduced from 60.3 percent to 44.1 percent);

  -- retiree health benefits were eliminated for both current and
     future employees; and

  -- the funding period of the plan's unfunded accrued liability was
     extended from 27 years to 40 years. 

The study also stated that additional benefit reductions might be
needed to maintain level spending if additional investment income was
not available to subsidize pension benefits for newly hired
employees. 


      EFFECT ON PENSION PLAN
      FINANCES IS UNCERTAIN
---------------------------------------------------------- Letter :4.4

States and localities typically use a "reserve funding" approach to
finance their pension plans.  Under this approach, employers--and
frequently employees--make systematic contributions toward funding
the benefits earned by active employees.  These contributions,
together with investment income, are intended to accumulate
sufficient assets to cover promised benefits by the time employees
retire. 

However, many public pension plans have unfunded liabilities.  The
nine plans that we examined, for example, have unfunded accrued
liabilities ranging from less than 1 percent to over 30 percent of
total liabilities.  Unfunded liabilities occur for a number of
reasons.  For example, public plans generally use actuarial methods
and assumptions to calculate required contribution rates.  Unfunded
liabilities can occur if a plan's actuarial assumptions do not
accurately predict reality.  Additionally, retroactive increases in
plan benefits can create unfunded liabilities.  Unlike private
pension plans, the unfunded liabilities of public pension plans are
not regulated by the federal government.  States or localities
determine how and when unfunded liabilities will be financed. 

Mandatory coverage and the resulting pension plan modifications would
likely result in reduced contributions to public pension plans.  This
would occur because pension plan contributions are directly tied to
benefit levels and plan contributions would be reduced to the extent
plan benefits are reduced and replaced by Social Security benefits. 

The impact of reduced contributions on plan finances would depend on
the actuarial method and assumptions used by each plan, the adequacy
of current plan funding, and other factors.  For example, some plan
representatives are concerned that efforts to provide adequate
retirement income benefits for newly hired employees would affect
employers' willingness or ability to continue amortizing their
current plans' unfunded accrued liabilities at current rates. 

Actuaries also believe that reducing contributions to current pension
plans could adversely affect the liquidity of some plans.  In 1997,
for example, an Arizona state legislative committee considered
closing the state's defined benefit pension plan to new members and
implementing a defined contribution plan.  Arizona state employees
are already covered by Social Security; however, states and
localities faced with mandatory coverage might consider making a
similar change to their pension plans.  A March 1997 analysis of the
proposed change stated that as the number of employees covered by the
plan decreased, the amount of contributions flowing into the plan
would also decrease.  At the same time, the number of members
approaching retirement age was increasing and benefit payments were
expected to increase.  As a result, external cash flow would become
increasingly negative over time.  The analysis estimated that about
10 years after the plan was closed to new members, benefit payments
would exceed contributions by over $1 billion each year.  In another
10 years, the annual shortfall would increase to $2 billion. 

The analysis stated that the large negative external cash flow would
require that greater proportions of investment income be used to meet
benefit payment requirements.  In turn, this would require the
pension plan to hold larger proportions of plan assets in cash or
lower yielding short-term assets.  Once this change in asset
allocation occurs, the plan would find it increasingly difficult to
achieve the investment returns assumed in current actuarial analyses
and employer costs would increase. 


   LEGAL AND OTHER CONSIDERATIONS
------------------------------------------------------------ Letter :5

Mandatory coverage presents several legal and administrative issues,
and states and localities with noncovered employees would require
several years to design, legislate, and implement changes to current
pension plans. 


      LEGAL ISSUES
---------------------------------------------------------- Letter :5.1

Although mandating Social Security coverage for state and local
employees could elicit a constitutional challenge, mandatory coverage
is likely to be upheld under current U.S.  Supreme Court decisions. 

Several employer, employee, and plan representatives with whom we
spoke stated that they believe mandatory Social Security coverage
would be unconstitutional and should be challenged in court. 
However, recent Supreme Court cases have affirmed the authority of
the federal government to enact taxes that affect the states and to
impose federal requirements governing the states' relations with
their employees. 

A plan representative suggested that the Supreme Court might now come
to a different conclusion.  He pointed out that a case upholding
federal authority to apply minimum wage and overtime requirements to
the states was a 5 to 4 decision and that until then, the Supreme
Court had clearly said that applying such requirements to the states
was unconstitutional.  States and localities also point to several
recent Supreme Court decisions that they see as sympathetic to the
concept of state sovereignty.  However, the facts of these cases are
generally distinguishable from the situation that would be presented
by mandatory Social Security coverage. 

Unless the Supreme Court were to reverse itself, which it seldom
does, mandatory Social Security coverage of state and local employees
is likely to be upheld.  Current decisions indicate that mandating
such coverage is within the authority of the federal government. 


      ADMINISTRATIVE ISSUES
---------------------------------------------------------- Letter :5.2

The states would require some time to adjust to a mandatory coverage
requirement.  The federal government required approximately 3 years
to enact legislation to implement a new federal employee pension plan
after Social Security coverage was mandated for federal employees. 
The 1980 study group estimated that 4 years would be required for
states and localities to redesign pension formulas, legislate
changes, adjust budgets, and disseminate information to employers and
employees.  Our discussions with employer, employee, and pension plan
representatives also indicate that up to 4 years would be needed to
implement a mandatory coverage decision.  They indicated, for
example, that developing revised benefit formulas for each affected
pension plan would require complex and time-consuming negotiations
among state legislatures, state and local budget and personnel
offices, and employee representatives. 

Additionally, constitutional provisions or statutes in some states
may prevent employers from reducing benefits for employees once they
are hired.  Those states would need to immediately enact legislation
that would establish a demarcation between current and future
employees until decisions were made concerning benefit formulas for
new employees who would be covered by Social Security.  According to
the National Conference of State Legislators, the legislators of
seven states, including Texas, meet biennially.  Therefore, the
initial legislation could require 2 years in those states. 


   CONCLUSIONS
------------------------------------------------------------ Letter :6

In deciding whether to extend mandatory Social Security coverage to
state and local employees, policymakers will need to weigh numerous
factors.  On one hand, the Social Security program would benefit from
the decision.  The solvency of the trust fund would be extended for 2
years, and the long-term actuarial deficit would be reduced by about
10 percent.  Mandatory coverage would also address the fairness issue
raised by the advisory council and simplify program administration. 

However, the implications of mandatory coverage for public employers,
employees, and pension plans are mixed.  To the extent that employers
provide total retirement income benefits to newly hired employees
that are similar to current employees, retirement costs would
increase.  While the increased retirement costs would be phased in
over several years, employers and employees would also incur
additional near-term costs to develop, legislate, and implement
changes to current pension plans.  At the same time, Social Security
would provide future employees with benefits that are not available,
or are available to a lesser extent, under current state and local
pension plans. 


   AGENCY COMMENTS
------------------------------------------------------------ Letter :7

SSA stated that the report generally provides a balanced presentation
of the issues to be weighed when considering mandating coverage.  SSA
provided additional technical comments, which we have incorporated as
appropriate.  SSA's comment letter is reprinted in appendix II. 


---------------------------------------------------------- Letter :7.1

We are sending copies of this report to the Commissioners of the
Social Security Administration and the Internal Revenue Service and
to other interested parties.  Copies will also be made available to
others on request.  If you or your staff have any questions
concerning this report, please call me on (202) 512-7215.  Other GAO
contacts and staff acknowledgments are listed in appendix III. 

Sincerely yours,

Cynthia M.  Fagnoni
Director, Income Security Issues


SCOPE AND METHODOLOGY
=========================================================== Appendix I

To examine the implications of a decision to extend mandatory
coverage to newly hired state and local employees for the Social
Security program, we reviewed documents provided by SSA and IRS and
held discussions with their staff.  We examined SSA estimates
concerning the increase in taxable payroll and Social Security
revenues and expenditures attributed to extending mandatory coverage
to newly hired state and local employees and discussed data sources
with SSA officials.  We did not assess the validity of SSA's
assumptions.  SSA estimates used the intermediate assumptions
reported by Social Security's Board of Trustees in 1998. 

To examine the implications of mandatory coverage for state and local
government employers, employees, and their pension plans, we reviewed
the 1980 study by the Universal Social Security Coverage Study Group,
which was prepared for the Secretary of Health, Education, and
Welfare at that time and transmitted to the Congress in March 1980. 
We discussed study results with the study's Deputy Director for
Research and examined supporting documents for the study. 

We also held discussions and reviewed documentation of state and
local government employer, employee, or pension plan representatives
in the seven states that account for over 75 percent of the
noncovered payroll.  We examined financial reports for nine state and
local retirement systems:  the California State Teachers' Retirement
System, the Public Employees' Retirement Association of Colorado, the
Teachers' Retirement System of the State of Illinois, the Louisiana
State Employees' Retirement System, the Massachusetts State
Retirement System, the Massachusetts Teachers' Contributory
Retirement System, the State Teachers Retirement System of Ohio, the
Public Employees' Retirement System of Ohio, and the Teacher
Retirement System of Texas. 

We also identified a number of states that have changed, or have
considered changing, plan benefits in ways that are similar to those
that might be made by states and localities with noncovered employees
in response to mandatory Social Security coverage.  We discussed the
potential impact on plan finances of changing plan benefits with
pension plan representatives in those states and examined study
reports provided by them.  For example, we contacted representatives
of pension plans in Arizona, Kansas, Montana, South Dakota, Vermont,
Washington, and West Virginia that have implemented or considered
implementing defined contribution plans to replace some or all of the
benefits provided by their defined benefit pension plans. 

Additionally, we reviewed survey reports addressing pension benefits,
costs, investment practices, or actuarial valuation methods and
assumptions prepared by BLS, PPCC, and the Society of Actuaries.  We
discussed the implications of mandatory coverage for public pension
plans with actuaries at the Office of Personnel Management, the
Pension Benefit Guarantee Corporation, the American Academy of
Actuaries, and in private practice. 

To analyze differences between public pension costs and benefits for
covered and noncovered state and local employees, we used PPCC survey
data.  We used the 1995 survey, which covered 1994, because the 1997
survey, which covered 1996, did not include some of the required
data.  Despite some limitations, the PPCC data are the best
available.  The data cover 310 pension systems, representing 457
plans and covering 80 percent of the 13.6 million active members in
fiscal year 1994.  The survey questionnaire was mailed to 800
systems, which were selected from member associations.  Due to the
nonrandom nature of the sample, no analysis can offer
generalizations, nor can confidence intervals be calculated. 
Nevertheless, the survey describes the costs and benefits of a
substantial majority of public pension plan members. 

For our analysis of PPCC data, we classified pension plans as (1)
Social Security covered if 99 percent or more of the members
participated in the Social Security program or (2) Social Security
noncovered if 1 percent or less of the members participated in the
program.  We did not adjust cost and contribution rate data to
standardize actuarial cost methods and assumptions.  State and local
governments may have legitimate reasons for choosing various cost
methods, and we did not evaluate their choice. 

To identify potential legal or other problems with implementing
mandatory coverage, we reviewed relevant articles and current case
law.  We conducted our work between September 1997 and May 1998 in
accordance with generally accepted government auditing standards. 




(See figure in printed edition.)Appendix II
COMMENTS FROM THE SOCIAL SECURITY
ADMINISTRATION
=========================================================== Appendix I


MAJOR CONTRIBUTORS TO THIS REPORT
========================================================= Appendix III

Francis P.  Mulvey, Assistant Director, (202) 512-3592
John M.  Schaefer, Evaluator-in-Charge
Hans Bredfeldt, Evaluator


*** End of document. ***