Once it has been decided that age alone is not to qualify one for a pension, we are brought to a sharp dividing of the ways. Should the added condition of eligibility be destitution, on the one hand, or should it, on the other, be past earnings which insofar as they indicate financial condition suggest affluence more than they suggest poverty? The former answer to this question—that governmental aid should be conditioned on need—is hallowed by tradition and justified on the simple ground that people ought not to be allowed to starve. The latter answer is more modern in its origin and justified on more sophisticated grounds.
The "means test" method of paying pensions establishes a standard of need, and its objective is to bring up to that standard all applicants whose income falls below it. In practice, of course, the process is complicated by many factors, such as obligations of support, nonproducing assets, and the special needs of the sick, as well as by lack of enough funds to meet the standard in full. To the extent that the objective is achieved, however, this method actually attains a greater measure of equality than do flat pensions which disregard the private income of recipients.
To complement contributory old-age insurance, which could become effective only gradually, the Social Security Act provided for grants to the states for old-age assistance based on a means test. These federal grants have resulted in much enlargement and many improvements in state assistance programs, but they have generated pressures which have led to repeated increases in the federal payments, now some $900 million a year. Though case loads seem finally to have leveled off, many persons believe that public assistance has got politically out of hand, with the federal government footing too much of the bill for payments controlled by the states. This belief largely motivates proposals for drastic revision of old-age and survivors insurance, so that it could fill at least a part of the gap that would be left if federal old-age assistance grants were ended.
By contrast with public assistance, old-age and survivors insurance as now constituted selects its beneficiaries and determines the amount of their benefits on the basis of their past earnings. It does so, and as a practical matter it could only do so, in conjunction with the tax levied on those earnings for the specific purpose of providing funds for the payment of benefits. Without the tangible justification of contributions paid in, this method of paying benefits would run into insuperable objection from the people who found themselves at the lower end of the benefit schedule or excluded from it altogether—frequently the people most seriously in need of help. It is safe to say that this method of distributing benefits would not have been adopted by Congress apart from the enactment of payroll taxes to pay for them and would not be continued in force if those taxes should be repealed.
It is equally safe to say that the pay-roll taxes would not have been enacted but for the simultaneous adoption of a system of benefits geared to past earnings and would not survive if that system of benefits were abolished. A tax on employers as such, a tax measured by the wages they pay, would be difficult to justify for any purpose other than social insurance. Still more difficult would be the pay-roll tax on employees which, considered merely as a tax, is a regressive form of income tax limited to a particular kind of income. Except as a part of a social insurance system, why tax only the first $4,200 of wages and exempt the additional wages of the more well to do? Why tax wages and not unearned income? Why impose a tax with no personal exemption when, even under the stress of war, the exemption from the general income tax was not reduced below $500? On all counts, the pay-roll taxes are special taxes, imposed for a special purpose and for that purpose only.
It is perhaps an odd circumstance but it is a crucial one for the purposes of this discussion that two pieces of legislation, neither of which would make economic or political sense if it stood alone, make in cornbination very good sense indeed. The relationship between the two, commonly known as the "contributory principle" in social insurance, will be found to create most of the limitations on the direction and scope of changes that can safely be made in the system, and we shall therefore have occasion to examine it in detail.
With the addition of the contributory feature, complaint from those who are excluded from the system becomes much less persuasive, and we are free to consider on their merits the arguments for gearing benefits to past earnings. Such benefits, in contrast to flat pensions, adjust the payments to the loss insured against and thus recognize both individual and geographical variation in standards and costs of living. In contrast to public assistance, they seek to ward off destitution rather than wait for it to occur; they encourage rather than discourage thrift; they minimize governmental scrutiny of intimate personal affairs and the risk that with scrutiny will go control; they lessen rather than increase pressures upon adult sons and daughters to support their aged parents. Earned by contributions, payable as a matter of right without regard to the recipient's means, they promote the worker's self-reliance and self-respect. And finally, because they give the worker a greater assurance that the promised benefits will actually be paid when the time comes, contributory benefits serve better than can any noncontributory system to allay the haunting fear of penniless old age.