The development of State and local government retirement systems in the last half of the nineteenth century represented one of the first efforts in this country to protect the worker against the major economic hazards of our modern society. The Social Security Administration, as part of its continuing concern with methods of providing protecting against economic insecurity, has long included studies of State and local systems among its research activities. This article summarizes the findings of the 1966 survey of the State and local government retirement systems whose members are not covered under the Federal old-age, survivors, disability, and health insurance (OASDHI) program. The full findings of the survey are to be published as a research report in the coming months.

SURVEYS of the State and local government retirement systems whose members are also covered under the Federal program of old-age, survivors, disability, and health insurance (OASDHI) have been made in recent years, and the findings of these surveys have been reported by the Social Security Administration.1

The present study of retirement systems not under the Federal program fills out the information on the protection afforded by these systems. The report, based on the study of 87 retirement systems with 1,000 or more members, includes discussion of the systems' provisions governing retirement for age and service, disability retirement, and death benefits. Special emphasis is given to the recent developments in providing survivor protection, and information is included on such matters of current interest as early retire

Division of Program and Long-Range Studies, Office of Research and Statistics.

'Joseph Krislov, State and Local Government Retirement Systems, 1965 (Research Report No. 15), Social Security Administration, Office of Research and Statistics, 1966, and Joseph Krislov, A Survey of State and Local Government Retirement Systems . . . 1961, Social Security Administration, Bureau of Old-Age and Survivors Insurance, 1962.

by SAUL WALDMAN*

ment, vesting of benefit rights, and provision for increases in benefits for retired persons.

SURVEY METHODOLOGY

The Social Security Amendments of 1954 made coverage under OASDHI available, at the option of the State, for most employees under State and local government retirement systems if the employees involved vote in favor of coverage. As a result, some OASDHI coverage of retirement system members has been implemented in nearly all States.

The survey included all retirement systems with 1,000 or more members who were not covered under OASDHI, according to the 1962 Census of Governments.2 Systems with incomplete OASDHI coverage resulting from a "divided retirement system" election were omitted from the survey, since eventually all the members will be covered. Systems with incomplete OASDHI coverage on a permanent basis, which is permitted under certain other provisions, were generally included.3 Excluded were (1) closed systems, which do not accept new employees, (2) supplementary systems, covering employees receiving substantial protection under another public retirement system, and (3) systems covering unsalaried employees such as volunteer firemen.

The systems selected were limited to those with 1,000 or more members. Though there is a relatively great concentration of small systems among those whose members are not covered under

2

Bureau of the Census, 1962 Census of Governments: Topical Studies, No. 1, Employee-Retirement Systems of State and Local Governments.

Under the "divided retirement system" provision, applicable in States specified in the Social Security Act, only those members who elect to be covered are brought under OASDHI, with all new members automatically covered on a compulsory basis. Under other provisions of the law, coverage may be arranged for only certain groups of members-the employees of various local governments who are members of a statewide system, for example-and the noncovered groups are permanently excluded, unless additional arrangements are made.

OASDHI, only an estimated 10 percent of employees were excluded for this reason.

The final listing of systems to be surveyed included exactly 100 retirement systems with an estimated 1.5 million members in January 1966 or about one-fourth of all State and local government employees under staff retirement systems. A questionnaire was mailed to the administrator of each system by the Bureau of the Census, acting as agent for the Social Security Administration. Sixty-three of the 75 general systems, with an estimated 94 percent of the members, responded to the survey. Only one of the 25 systems for policemen and firemen-a small one-failed to respond.

The systems for policemen and firemen are analyzed separately because their provisions tend to differ considerably from those of the general systems, though for the most part they are homogeneous among themselves. The term "general system" is used here to refer to all systems other than those for policemen and firemen (unlike the term as used by the Census Bureau). The terms "members" or "membership" refer to active members and are used synonymously with "employees."

with 1,000 or more members, have been brought under OASDHI. Measured by the number of employees, the survey is dominated by three States (Ohio, California, and Illinois) that together included 62 percent of all employees in the systems surveyed. Massachusetts contributed the largest number of systems-18 of the 63 systems followed by Illinois with nine systems and California with six systems. The fact that the retirement systems within a State tend to have similar provisions has a significant effect on the results of the survey. The Massachusetts systems, for example are virtually identical, and most of the Illinois and California systems are closely similar.

The two largest systems in the survey (one California system and one Ohio system) include 28 percent of the total membership represented in the survey and the eight largest have almost 60 percent of the total membership. About half the systems had fewer than 5,000 members, and this group included 5 percent of all employees in the survey (table 2).

Major Findings

The survey findings indicate that, during the postwar years, the retirement systems have considerably strengthened the protection they provide for their members. The greatest improvements have been in provisions for monthly benefits for disability and death-types of protection offered by nearly all the systems in the survey.*

For a description of the retirement systems in the early 1940's, see Dorothy F. McCamman, The Scope of Protection Under State and Local Government Retirement Systems, Social Security Board, 1944. Although information is not available from that study on the specific systems in the survey, the improvements have been so widespread as to indicate revision of the great majority of systems.

The type of protection offered by OASDHI has been one of the factors influencing the direction of improvements of those systems whose members have not come under the Federal program (as well as the systems whose members were brought under coverage 5). The influence on noncovered systems is most apparent in new provisions for monthly benefits to widows and children that resemble the provisions for survivors in the OASDHI program.

The study also reveals an apparent modification of the principle of individual equity (under which benefits are based solely on the contributions or credits of the member) that has been an important element of the benefit provisions of many systems. Relatively few systems in the survey, for example, still use the money purchase method of computing benefits, under which benefits are actuarially based on the contributions credited to the member's account. Modification of this principle is revealed too in the growing tendency toward providing full or substantial benefits, without regard to contributions or length of service, to disabled members and, especially, to the survivors of deceased members.

employee may retire on his own volition and receive the full amount of retirement benefits to which he is entitled on the basis of his earnings and length of service. Similarly, where retirement based on length of service alone is possible, "normal" retirement is retirement at the earliest time full benefits become available. The earliest age at which retirement for age and service is possible, but with benefits reduced because of age, is defined as the "early retirement age."

To qualify for normal retirement, the member usually has to meet both an age and length-ofservice requirement, typically age 60 with 10-20 years' service, but many systems permit retirement on the basis of age alone (at age 60-65) and a few on the basis of service alone (with 30-35 years). Many systems offer alternative eligibility requirements for normal retirement. They may, for example, permit members with long service to retire at an earlier age or to retire on the basis of service alone, without meeting any age requirement (table 4).

The normal retirement age specified under the age or age-and-service provisions is usually 60, but age 65 is also frequently found. Many employees continue to work beyond the age that normal retirement first becomes available."

Involuntary retirement.-Although employees usually may continue to work beyond normal retirement age, under most systems they may be required, at the discretion of the employer, to retire at a specified age-commonly at age 70 and to a lesser extent at age 65. This type of provision is called a compulsory retirement provision. Some systems require the "automatic" retirement of their members, typically at age 70, with neither the employee or employer having any discretion (table 4).

Benefit amounts.-Except for one system with a money-purchase formula, the amount of the retirement benefit is always based on the member's average salary and length of service. The benefit amount is usually determined by com

The survey questionnaire requested information on the number of retirees who were under age 62 (an age when employees of most systems could take normal retirement). The 42 systems that reported on the subject indicated that only 20 percent of the men and 28 percent of the women were under age 62 at the time of retirement in the fiscal year 1965.

once.

296 880

54

183

84

1 Systems that provide alternative requirements are counted more than Earliest age required for normal retirement under the age or age-andservice provisions.

puting a percentage of average salary for each year of service, commonly 1% to 22 percent." The "final" average salary of the member is based usually on the 5 highest years of earnings, and many systems require that the years be consecutive or years during a specified period (usually the last 10 years) of employment (table 5). Some systems provide a minimum benefit amount, usually $30-$70 monthly or $6-$7 monthly for each year of service, payable for members who meet certain eligibility requirements, commonly 10-15 years' service.

Illustrative median benefits payable to members aged 65 with 10, 20, and 30 years of service would replace about 21, 40, and 54 percent of salary, respectively. Based on the actual experience of the system, the average benefit amount awarded to men in the fiscal year 1965 was about $235 monthly, for the 44 systems that supplied benefit data.

Early retirement.-Most systems permit employees to choose early retirement at a reduced benefit amount. The early retirement age is almost always age 55, and usually 20-30 years of service is needed to qualify (but several large systems require 5 years). Retirement for service alone (20-35 years) is offered, usually as an alternative requirement, by many systems (table 4). Several methods are used to reduce the benefit amount, but usually the reduction is, in effect, roughly actuarial.

Adjustment of benefits.-In recent years, there has been increasing interest in the problem of protecting the purchasing power of pension benefits after retirement. The laws of a few of the systems surveyed provide for an automatic increase in benefits to persons on the retirement rolls (usually annually) without requiring additional legislation. The automatic increase is usually specified as 1 percent or 12 percent annually. For almost one-half the systems in the survey, additional legislation was enacted that provided an across-the-board increase in benefits during the 3 years 1963-65.

2.00-2.17.

2.50.

Two percentage factors.. Other

Provision for minimum benefit.. Monthly benefit amount:

Less than $30..

$30-70.

$70-100.

Monthly benefit amount per year of service: Less than $4.00. $6.33-6.67.

$7.50-7.80. Other.

Number of systems

Number of members (in thousands)

62

1,295

883

299

51

900

96

5

479

157

149

68

716

3

28

Vesting.-Unless a retirement plan has a provision for vesting of credits, a worker who leaves employment before retirement age loses any rights to a pension at retirement age. Vesting refers to the right of members to "all or part of their accrued pension benefits at retirement age, regardless of their employment status at that time." Vesting provisions usually require that the employee leave his contributions in the fund. when he leaves his job. Systems with nine-tenths of the membership had provisions for vesting. Eligibility for a vested right is dependent usually on completion of 10-20 years of service, but several of the large systems require 5 years (table 6).

Retirement for Disability

Eligibility. Disability retirement was among the first types of protection provided by State and local government retirement systems. The 1944 report of the Social Security Board 10 indicated, however, that many systems did not provide for non-service-connected disability, and benefits were often limited to members with long

President's Committee on Corporate Pension Funds, Public Policy and Private Pension Programs, January 1965.

10 Dorothy McCamman, op. cit.

periods of service. Sometimes the provisions were designed mainly to facilitate the premature retirement of older employees.

The present survey indicates a broadening of the scope of protection. Non-service-connected disability retirement, provided by all but one system, depends usually on meeting a length-ofservice requirement of 5-10 years with the larger systems usually requiring 5 years; many small systems, however, require 15 years (table 7). Most systems use an occupational definition of disability under which the member must be unable to perform his usual job. The others generally require a determination that the member be unable to perform any type of work.

Benefit amounts.-Most systems compute the benefit in the same manner as for age-and-service retirement, often using, however, a smaller percent of salary for each year of service. There is somewhat greater diversity in the computation methods (table 7). Some systems use a modified formula or other method of computation under which the benefit amount is figured without regard to length of service or they provide a minimum benefit to all eligible members. This type of computation is used by 15 systems that include two-thirds of the employees." An illustrative median benefit for a member aged 50 with 20 years of service would replace 39 percent of the member's salary. The actual benefit amount awarded, in the fiscal year 1965 (for non-serviceconnected and service-connected disability combined) averaged about $185 monthly, based on data for 37 systems reporting on disability benefits.

Survivor Benefits

Monthly payments for survivors of active members who died from non-service-connected causes, now provided by nearly all systems in the survey, have been introduced comparatively recently into

"This analysis was made by selecting those systems in which the length of service does not substantially affect the benefit amount, including systems (1) providing a percentage of the member's salary for each year of service but basing this calculation on assumed service to or near retirement age; (2) providing a benefit amount based on that payable at normal retirement; (3) providing a specified percent of salary-and by selecting those systems with a minimum benefit amount that is not based on length of service.