and reserve financing, may possess the disadvantages of both of those courses without the advantages of either. In particular, there is no actuarial balance whose maintenance could serve as a safeguard against unsound liberalizations of the program while at the same time the large reserve to be accumulated will serve as ample temptation for such liberalizations. In fact, it is very clear that the present liberalization proposals are implemented by the common illusion that social-security benefits are cheap, an illusion that is inevitable when a program of this nature has the great bulk of its costs long deferred and at the same time has no level premium reserve financing. This is not to argue for reserve financing, however, since the accumulation of a large reserve (regardless of any actuarial virtue) may have serious effects on the economy and may become largely useless in the face of marked or continuing inflation.

The proposed legislation would depart even further from any relationship between the value of the individual worker's benefit and the combined employee and employer taxes paid with respect to his earnings. Under H. R. 2893 average old-age and survivor benefit amounts would be roughly doubled immediately, yet immediate pay-roll tax rates (exclusive of those attributable to costs of permanent and temporary disability benefits) would hardly rise (though they would be applicable to a larger earnings base), and such increase as there would be would of course, not be retroactive. The result is that, for individuals becoming entitled to retirement benefits in 1955, for example (see the attached chart), the aggregate employee contributions paid, or even the aggregate combined employee and employer contributions, would bear a very negligible ratio to the value of benefits. Even in the case of workers now entering the program, the great political improbability of a substantial early rise in pay-roll tax rates almost certainly means that their benefits, too, will fail to be paid for by pay-roll taxes.

The proposed legislation would greatly increase the differential between benefit amounts of high-paid and low-paid individuals. The excess of the high-paid individual's benefits over the low-paid individual's cannot be justified on the basis of individual equity, assuming such equity is measured by employee taxes, except perhaps in the remote future and even then improbably. It is true that the higher-paid employee pays a higher proportion of the cost of his larger benefits than the lower-paid pays of the cost of his smaller benefits, but since the proportion will generally be negligible anyhow, the higher-paid employee will derive a much greater profit from public funds than the lower paid. (See the attached chart.) Those persons who support the general structure of the program but oppose the increase of the annual wage base from $3,000 to a higher figure (say $3,600 or $4,800) point out that the differential between the $3,000 employee's benefit and the higher-paid employee's benefit could not be justified on equity grounds; but they apparently overlook the fact that the same type of anomaly exists as between the $1,200 and the $3,000 employee, under both proposed and existing legislation. The proposed increase in wage base and change in benefits formula under H. R. 2893 are designed primarily to take account of recent increases in wage levels. Since 1939 wage levels have almost doubled, an annual wage of $1,800 today may correspond to $2,400 in 1939. The benefit formula under H. R. 2893 is such that the ratio of the $4,800 individual's benefit to his earnings will be only moderately higher than the ratio of the $2,400 individual's benefit to his earnings under the 1939 formula.

There is no satisfactory machinery, however, to take satisfactory account of future changes in wage and price levels, either upward or downward. While any significant downward wage and price trend is admittedly improbable, it can be seen that were such trend to occur the ratio of benefits to earnings might increase to an unfavorable extent. The drafters of H. R. 2893 hoped to take care of the upward-trend situation by basing benefits on wages of the five highest consecutive earnings years, though presumably in case of a very marked trend they would seek further increase in the annual earnings base and revision of the benefit formula. This difficulty with the highest 5-year approach is that, while it is of encouragement to the individual who has not yet retired or died, it is of no help to him or his survivors in the very likely event of a continued upward trend after his retirement or death.

A remedy for this difficulty might be to have some special adjustment made periodically in the benefits of individuals already on, the benefit rolls, an adjustment possibly reflecting the change between the purchasing power of the then current dollar and that of the dollars on which their benefits are based. But to do this, it would seem, would emphasize increasingly the absence of relationship between contributions and benefits and would seriously question the basing of benefits in any way on past earnings.

The social budgeting approach

Years ago the well-known actuary Miles M. Dawson pointed out that under a compulsory social-insurance system the many rigidities and limitations of private insurance need not be present. Thus, in a system where the existence of future generations of participants was assured through the State's taxing power, there was not only no need for a reserve, but more important it would be both possible and preferable at the commencement of the compulsory program to bring into benefit status all those who would have been receiving benefits had the program started long before. Such a program would be truly contributory, for though the individual would not be paying for the benefits he himself was to receive in the future, he would be paying an equivalent cost by paying for the benefits of those now aged or disabled or orphaned or widowed. In return for his paying for the benefits of these others, he is assured that when in the future he becomes aged or disabled or deceased, he or his survivors' benefit will be paid for by those then in their productive years. As Dawson says, such an approach would say to every man: "In consideration of a contribution of your full share of the cost of providing this indemnity for all other contributing members of society, society will cover the entire hazards of your having the re sponsibilities resting upon you which make the insurance requisite or desirable." (Proceedings, Eighth International Congress of Social Insurance (1908), p. 192) The adaptation of this approach, as many students of the subject visualize it. would be (1) to determine each year the amount which could be raised through some predetermined form of direct taxation specifically earmarked for social security purposes, (2) to determine the number of persons eligible for social security benefits and their respective proportionate shares, and (3) to divide (2) into (1) to obtain the amount of each share. As visualized, each of the aged would receive the same size share, not only because of the difficulty of devising variable shares on any pertinent basis but primarily because a uniform individual benefit would be considered socially preferable, for reasons partially indicated below. On the other hand, it might be desirable to pay benefits to some other category, such as orphans, at some uniform level different from the uniform level for the aged.

To illustrate, assume that the social-security tax method decided upon was to yield a total of $8,700,000,000 in 1951. This amount, incidentally, is just one half of the mean of the high and low estimates of the cost (assuming present earnings levels (in the year 2000 of H. R. 2893 plus public assistance, a cost to be paid by a population much less than twice the present population. Assume further for 1951 a total of 11,000,000 aged, to have one benefit unit apiece, 2,000,000 under 65 but chronically disabled also to have one unit apiece, and 2,000,000 dependent children to have three-quarters units apiece. The number of benefit units would total 14,500,000, so that each benefit unit would be $6 annually, or $50 monthly.

Social budgeting is distinct from current Townsend proposals in that soci budgeting avoids any form of hidden tax and keeps costs fully in the open. Từ Townsend form of tax, misnamed "gross income" tax, is really a gross-receipts tax, which could be included several times, and to an immeasurable extent, in the price of every commodity. An appropriate tax for social budgeting would be a flat percentage added to the individual normal income tax rate.

A system of uniform benefits under the social budgeting approach may be compared with the present or proposed old-age and survivors insurance legisla tion, with respect to the various problems under consideration, in the following way:

Present program (or proposed liberalizations thereof, retaining present basic structure)

Suggested system of uniform denefiti independent of previous wage histor

ies

1. COVERAGE OF PRESENTLY EXCLUDED GROUPS OF WORKERS

Administrative problems relating to wage reporting make coverage of some new groups very difficult. Any new coverage that is effected necessitates elaborate and confusing "new start" provisions.

Wage-reporting and new-start pro lems are obviously absent, as no wage records would be maintained for a0fone under this system.

Under suggested system there would

which inflexibility could result, and benefits could be made to change automatically with economic and fiscal developments.

Benefit amounts largely dependent on wages received in long-past years lack be no wage-benefit relationship from flexibility to meet changing price levels and other conditions. Adjustments of benefit formula designed to remedy this situation do not meet problems of all beneficiaries, provide only temporary solution, and result in increased complications.

8. ADMINISTRATIVE EFFICIENCY AND ECONOMY

Despite the efficiency with which wage reports are processed and wage records are maintained, the cost of these operations is appreciable, and relatively higher costs and less efficient operations may be expected with respect to some of the excluded groups whose coverage is now contemplated.

Which is the cooperative method?

These operations would be eliminated, and there would also be savings in other operations. The existing income-tax machinery would suffice for collecting contributions.

If we are to accept as our aim cooperative welfare instead of paternalistic welfare, and if we seek to preserve the ingredients Dean Brown speaks of, namely individual incentive and mutual responsibility, we should ask ourselves which of these two social-security methods is more nearly the cooperative method and which is more nearly the paternalistic. The following are some of the questions whose answers may guide us to a solution:

1. Which is the more paternalistic in its effect, a program under which for some decades to come a substantial proportion of those in need are left to the mercies of needs test public assistance, or a system under which each individual in the categories where need is presumed received a benefit which is his own to use as he sees fit?

2. Which is the more cooperative and which the more paternalistic, a virtually noncontributory program promising expensive benefits for token considerations, hiding true costs and therefore stimulating the demand for even more expensive benefits, or a system under which the present worker in paying for the benefits of those who can no longer contribute is contributing an amount equivalent to the real value of his own benefits?

3. Which is the more paternalistic, a program whose benefits may be tied in unrealistically to economic conditions of years already past and can be revised only through congressional wrangling over complicated and discriminatory formulas, or a system under which both income sources and benefit levels can vary automatically with economic changes?

4. Which is the more paternalistic, a program which uses public funds to subsidize the continuance of the differential of the high-paid worker's income over that of the low-paid worker, or a system which recognizes that the high-paid worker is better able than the low-paid to supplement his flat social-security benefit through nonsubsidized thrift and insurance channels, thus maintaining his differential through his own efforts and the medium of private enterprise?

The above are some of the many questions which can and should be asked in choosing our pattern of social security. As Dean Brown points out, the decisions made in the year 1950 may have far-reaching effect.

The CHAIRMAN. The next witness is Mr. John D. Battle, vice president of the National Coal Association.

Mr. Battle?

STATEMENT OF JOHN D. BATTLE, EXECUTIVE VICE PRESIDENT, NATIONAL COAL ASSOCIATION

Mr. BATTLE. My name is John D. Battle. I appear here on behalf of the National Coal Association. The National Coal Association is the trade association of bituminous coal producers, with membership comprising about 75 percent of the commercial production of bituminous coal in the United States and with members in each of the major coal-producing States in the Nation.

As a representative of the bituminous coal-mining industry, I do not appear here today as a specially qualified social-security expert, I do appear for the purpose of explaining to the committee some of

VALUES OF INDIVIDUAL EMPLOYEE'S BENEFITS AND PAYROLL TAXES UNDER H. R. 2893

NOTE. Values of benefits are actuarial values, as of retirement age, of employee's own (primary) benefits and of wife's and widow's benefits which may become payable after his retirement. Computation of wife's and widow's benefits makes allowance for probability of there being no wife or widow qualified to receive such benefits and also makes allowance for effect of benefit maximum.

Pay-roll tax values are interest accumulations of net employee taxes to retirement age, the net tax after 1949 being one-half percent of pay roll less than the assumed total employee tax. Benefit of survivorship is ignored, and it is assumed that this saving plus one-half percent of pay roll is applied to the cost of survivor and disability claims incurred before retirement. Total employee tax rates assumed are 2 percent for 1950-59, 21⁄2 percent for 1960-69, 3 percent for 1970-79, and 3% percent for 1980-84, though H. R. 2893 provided no tax rate higher than 2 percent. Steady earnings of $1,200 a year for one employee and $4,800 (or more) for the other are assumed from 1950 on, while before 1950 the assumed earnings are $750 and $3,000 (or more), respectively. Fluctuating earnings could produce the same benefit values with lower tax values.

United States 1939-41 white population mortality and 22 percent interest are assumed.

our industry's special problems under old-age and survivors' insurance, and also to express our position on the proposals of broadening the wage base, adding disability benefits, and sharply increasing benefits. The recommendations which we shall make are of vital importance to the coal industry.

You have before you a bill which in 1951 will impose old-age and survivors' insurance taxes at double the rate which has been in effect up to the beginning of this year. In addition, it would apply to $3.600 of an employee's annual wage instead of $3,000. The Social Security Commissioner has asked that the base be made $4,800, and some union representatives have demanded an even broader base.

You gentlemen will appreciate the difference between the $30 annual maximum per employee we have been paying, and the $72-over twice The old maximum-which we shall be paying next year if H. R. 6000 s enacted as presently written. You will understand the impact of he $96 maximum for next year proposed by Mr. Altmeyer-more han three times our maximum before this year.

Under Mr. Altmeyer's proposal, 15 years from now we would pay 144 per year tax on our high-wage employees-nearly five times the nitial $30 maximum per employee. The withholding tax for social