pothesis since the social security benefit formula is "weighted" so that individuals with lower income receive a higher return on their social security contributions than individuals who earn higher salaries (as explained in chapter 8).

The results of our survey do support this assumption, but only to a degree. The data shows that, while 64 percent of the groups that have decided to remain under social security have average wage levels below $15,000, only 50 percent of the groups responding that have recently terminated social security coverage have average wage levels below this amount. This does not appear to be a convincing pattern, however. For instance, among groups in categories I and II, the only group with average salary above $20,000 has opted to remain in the program. This group stated that a majority of its employees see social security as better protection than any alternate plan. It is worth noting that this was one of only three groups surveyed that employed a private actuary to study the issue of termination.

Adding to the uncertainty of using salary levels as an indicator of the likelihood of termination for any particular group, is the fact that three groups in category III have average wage levels below $10,000. While it is possible that these groups may not actually leave the social security program, the employees in two of these groups have already voted to terminate coverage, primarily because they believe they can obtain better protection for the same cost elsewhere. This may reflect the fact that, although social security may be a superior value for low-wage earners, these workers probably have the greatest need to maximize take-home pay. It should be remembered that, although the social security benefit formula is weighted to favor workers with low earnings, the social security tax is regressive-people with low earnings pay the same percentage of their income as employees with higher salaries (up to the maximum amount of wages subject to the tax).

The third group in category III with average wages below $10,000 presents an interesting and disturbing case that illustrates the kind of problems that can occur under the voluntary nature of social security coverage for State and local workers. This group of 83 employees is being forced to leave the social security system because the State has refused to continue paying the employer portion of the social security tax, and the local system is not able to absorb this cost. This is a particularly unfortunate event for this group in light of the low average salary and the fact that the average age of the employee population is over 51. Worse yet, the only insurance plan these employees will be covered by once social security is dropped is a teachers retirement system to which the employer makes no contribution, and under which cost-of-living adjustments in benefits are made only when the State legislature approves them.

It is apparent that, in this case, social security protection is being terminated as a money-saving device for the employer at the expense, of the employees. Although this is the only group that reported such a situation, the committee is quite concerned that State and local governments, being increasingly pressured to cut spending, will be tempted to eliminate vital protection for employees. Under the present law, of course, there is little recourse for the employees who are adversely

AVERAGE AGE AND LENGTH OF SERVICE

Regarding average age and length of service, the committee found that the groups that have opted to remain in the social security program have a somewhat older population of employees, with a somewhat longer average length of service. It is not surprising that employees who have worked under the social security system for a longer period of time and are closer to retirement would be more likely to want to remain in the program. However, there were no striking differences in these parameters between the three categories of groups, and the overwhelming majority of groups in both categories report average employee age between 31 and 40, and average length of service as 6 to 10 years.

Length of service should not be discounted entirely as a parameter of what a group might do, however, since after 10 years of covered employment, workers have enough quarters of coverage to be eligible for social security retirement benefits, even if social security coverage is then terminated. Thus, after 10 years of covered work, employees might be quite eager to terminate coverage.

NUMBER OF EMPLOYEES IN THE GROUP

One unexpected conclusion supported by the data is that large groups are apparently more likely to choose to remain in the program after the notice of termination has been given than are small groups.

The groups selected for this study range in size from 2 to 11,725 employees. Of the groups in category I-those that have already left the program-the average number of employees is 82.5. Of the groups in category II-those that withdrew their notice of termination-the average number of employees is 1,987-24 times the average size of groups in the first category. This surprising difference seems to be a real phenomenon; the averages do not appear to be skewed by a few very small groups in category I, or a few very large groups in category II. In fact, the largest employee population in the first category is smaller than all but four of the groups in the second category.

To verify that this size difference is not a result of sample selection, the committee determined the average size of all groups that have terminated social security coverage. This number is 185. Unfortunately, records on the number of employees in all groups that have rescinded notice of termination were not available. However, the average size of all groups with termination notices pending as of January 1980 is 335, almost twice the size of the groups already terminated. It seems reasonable to assume that this difference derives from the same phenomenonnamely that, of the groups with termination notices pending, the groups with relatively large numbers of employees will choose to remain in the program.

The committee believes there are at least two logical reasons that larger groups are more likely to withdraw a termination notice. Perhaps the simplest explanation is that, in smaller groups, there is prob ably a greater chance that a consensus about termination is reached before the notice is sent. Thus, small groups would probably not submit a request to terminate coverage unless the majority seemed to favor the

• See question No. 8 in chapter 1 for an explanation of quarters of coverage.

idea. In larger groups, often with a more diverse employee population, it would be more difficult to gage the attitude toward termination. As an example, the committee found one group within category II, with 3,000 employees, that withdrew its notice of termination because a significant faction of employees objected to withdrawal. The group is made up of both certified teachers, eligible for benefits under the State teacher's retirement system, and classified employees who are ineligible for membership in that system. When a vote was held, a majority of the teachers, who had been the most vocal about the idea, voted to withdraw from social security, but 86 percent of the classified employees opposed termination, since social security is their only form of insurance protection. In this case the classified employees have retained social security coverage. However, it is obvious that there must be cases where the outcome is not as fortunate. This is another example of a serious problem with present law protection of State and local government employees.

A second possible explanation for the fact that large groups seem more likely to opt to remain covered is that large coverage groups probably have more resources with which to study the issue. It is simply too large a burden on a group of 10 or 20 employees to assign even one employee to work full time on such a project. Moreover, it would also be too heavy a burden for a small group to hire a private firm to do the analysis. The committee found that an actuarial analysis of benefits under social security and an alternate plan costs upwards of $7,000.

WHY IS OPTING OUT CONSIDERED?

The rising price tag of social security coverage and the concern over a perceived lack of long-term financial stability are by far the most often quoted reasons for considering opting out of social security. These concerns were the predominant ones in every category, with 28 groups citing one or both of these issues as primary reasons for sending a notice of termination. It is interesting to note, however, that groups with termination notices pending offered somewhat more precise and sophisticated reasons for terminating coverage than did the groups that terminated coverage only a few years ago. For example, groups with notices currently pending were more likely to explain that the employees are seeking a better investment for their money, rather than simply stating that social security costs too much.

Of course, there are other considerations that are somewhat more involved than cost or solvency of the program. One entity in category I explained the reasons for the decision to opt out as follows: Currently, the group is covered under social security and the State public employee retirement system (PERS). Benefits under the two systems are coordinated. Thus, employees pay a combined tax rate of nearly 14 percent of wages, yet upon retirement, the PERS benefit is reduced by a portion of the social security benefit. By leaving the social security program, employees will reduce their insurance tax burden by nearly 50 percent without losing eligibility for social security benefits, since almost all of the current employees have enough work credits to remain eligible for social security benefits upon retirement. Moreover, these employees will collect social security and PERS benefits under more favorable conditions, since, in this case, the PERS benefit

will no longer be offset to take account of receipt of social security benefits. (This will almost certainly lead to an eventual increase in the cost of the PERS.)

It is impossible to argue with the logic of terminating social security coverage for current employees in this group. The only negative effect the choice might have is that it may adversely affect new employees, since they will not obtain social security coverage and may be left with inadequate alternate protection. This can be particularly dangerous since benefits under the PERS may have been designed to be supplemented by social security.

Groups in category I listed a variety of other reasons for leaving the program. One group terminated social security coverage because many employees would be eligible for social security on their spouse's record and did not want to pay social security taxes on their own behalf. Two groups stated that they were dropping social security in favor of the State PERS because of the earlier retirement age and more generous retirement benefits under the PERS. Another group in category I was merged with a larger political subdivision that is not covered by social security and had no option to continue social security coverage. This is another example of a situation that can arise under present law which can adversely affect State and local government workers.

Although a majority of groups in both categories cite cost and solvency of social security as their major concerns, groups in category II were even less divided about their reasons for considering termination than the groups in category I. Twelve groups in category II gave cost as the primary reason for considering termination of social security coverage. One group cited concern about the long-term solvency of the program as its only reason for submitting a notice of termination.

Another entity in category II considered termination because one faction of employees-police and firefighters desired to leave social security, since they would be eligible for good employment-related insurance benefits even without social security protection, although the rest of the municipal workers would not. The group did not withdraw since the city felt that "while withdrawal may benefit some employees, employees for the most part would have been adversely affected," and the respondent stated that there was strong opposition to withdrawal from the majority of employees. Although it did not actually occur, this is another case in which many employees might have been disadvantaged because of the desires of one faction of workers within the group.

As in the other two categories, cost and solvency of the program are the primary concerns of the majority of groups in category III—those with notices of termination currently pending. Of these groups, seven stated that they have sent in a notice of termination because they are trying to develop a better investment.

Of course, as with the other two categories, there are groups in category III that expressed different reasons for terminating coverage. For instance, one group, discussed previously, wants to remain covered by social security, but cannot afford to because the State has refused to continue paying the employer's matching share of the social security tax. Another group cited, as primary reason for leaving social

Legislation recently enacted prevents this situation (refer to chapter 5).

security, the fact that employees in that group have relatively high salaries ($15,000 to $20,000) and "feel that they are paying benefits for those in lower income brackets." These employees believe they can obtain adequate protection at a lower cost from a different retirement plan. In a similar vein, one group in category III, citing cost and solvency as its main concerns about social security, went on to add that employees in the group view many social security benefits as welfare payments, and wish to end participation in the program because they oppose this "watering down of the social security program with welfare add ons."

WHERE DID THE IDEA TO OPT OUT ORIGINATE?

In order for the committee to gain more insight into the decision to terminate social security coverage, groups were asked whether the idea to end social security coverage originated with the employer or the employees. The responses showed an interesting pattern. Among groups in category I-groups that have withdrawn from the program-the idea to terminate coverage originated with the employees by a ratio of 2 to 1. However, in category II-the groups that chose not to leave the program-the idea to terminate coverage originated with the employer by a ratio of 3 to 1.

There are two explanations for this difference that seem most plausible to the committee. First, if the idea to terminate enjoys popular support with the employees, it is unlikely that the employer would prevent termination of social security coverage-particularly if it results in a savings to the employer. Thus, by the time notice of termination is sent, the decision may be a "fait-accompli." On the other hand, when the suggestion originates with the employer, it may be subject to more scrutiny and perhaps even suspicion.

Another explanation for the fact that, in category II, the idea to terminate coverage most often originated with the employer may be related to the finding that groups in this category are larger, on average, than the groups in category I. It should be kept in mind that a higher percentage of groups in category II cited cost concerns as the primary reason for considering withdrawal than in either of the other two categories. And with a large employee population, even a small per capita savings that the employer might realize under an alternate plan may be quite significant, even if the change would not be costeffective at all with a small employee population. However, if savings to the employer is the motivation for considering termination, employees would likely be reluctant to approve such action.

WHAT INFORMATION WAS USED IN MAKING A DECISION?

In order to determine whether decisions about termination are made in a thoughtful manner, the questionnaire used in our survey contained a series of questions regarding the sources of information that were used in the decisionmaking process. Specifically, the survey asked whether or not the following three informational sources were used: (1) The Social Security Administration (SSA); (2) the State Social Security Administrator; and (3) other State and local government employee groups. The groups surveyed were also asked what