8

railroad retirement system while retaining all its earlier staff pension elements.

Railroad workers who spend a full career in the industry receive benefits that are about twice as high as social security benefits. They also pay taxes which are about twice as high as social security taxes. In 1972, the railroad worker and his employer each pay 9.35% of the first $750 of covered monthly wages, or a maximum of $841.50 a year. This tax is divided 4.6% ($414.00) for OASDI and 4.75% ($427.50) for the Railroad Retirement Account. In addition, the carriers finance the entire cost of a supplementary benefits program for career employees enacted temporarily in 1966 and made permanent in 1970.

Second, the railroad retirement system has become administratively and financially interwoven with social security in a number of extremely complex ways. It is almost impossible for the Railroad Retirement Board to make a final award today without first checking with the Social Security Administration. Many of the benefits awarded, particularly for survivors, are based on combined railroad and social security service and wages; also, many railroad benefit awards must be computed by a different formula if the beneficiary receives an additional benefit directly from OASDI.

A close linkage also results because of the financial interchange, enacted in 1951. The statutory purpose of the interchange is to place the social security system in the position in which it would have been if there had been no separate railroad retirement system and railroad workers had always been covered by social security. On the revenue side, the Railroad Retirement Account transfers to the social security trust funds the taxes which social security would have collected directly from railroad workers and their employers. On the benefit side, the social security system transfers to the Railroad Retirement Account that portion of the railroad retirement benefits that social security would have paid for combined railroad and social-security-covered service, minus the sum of any OASDI benefits paid directly by the Social Security Administration to railroad retirement beneficiaries. The net effect of this offset is that the Railroad Retirement Account bears the excess cost of dual benefits. This excess cost is a large and increasing amount because of the large and growing number of dual beneficiaries. It is a cost that could have been avoided by full coordination of railroad retirement with social security. The financial interchange settlement is now carried out annually on a current pay-as-yougo basis, except for an average 18-month lag between accrual and transfer, which is covered by an adjustment for interest.

Had it not been for the financial interchange, the Railroad Retirement Account would be entirely exhausted today. Retroactive application of the 1951 financial interchange from 1937 on has resulted in a net flow to the Railroad Retirement Account from 1937 through June 30, 1971 of $5.9 billion. Even so, the Railroad Retirement Account is seriously underfunded.

The formal exclusion of railroad workers from regular social security coverage has led to costly structural problems. Because railroad employment is not covered by social security, and because social security permits qualification for substantial benefits after relatively short service, many railroad workers are able to get both social security and railroad retirement benefits. In 1971 38% of the railroad bene

9

ficiaries were receiving benefits from both systems. Typically, these dual benefits involve an excess or windfall element. They are weighted in favor of low earnings and are disproportionate to contributions and to length of social security service.

Under the financial interchange, the excess cost of these dual benefits is charged to the railroad retirement system-estimated at an average annual level cost of $480 million a year for the years 19702000. Although short-service railroad workers qualify for dual benefits more frequently than full career workers, all railroad workers must pay taxes that help finance these excess costs.

BASIC TRENDS AFFECTING RAILROAD RETIREMENT

The financial condition of the Railroad Retirement Account depends directly on railroad employment trends which, in turn, result from the forces affecting the economic situation of the industry. Because pension commitments are intergenerational in nature and extend from a worker's job entry to his or his wife's death, it is necessary to look many years ahead to assess fully the system's actuarial solvency, even though long range projections involve much uncertainty. The Commission made many projections for periods up to 75 years ahead. Projections of annual receipts and expenditures year-by-year are also necessary to diagnose cash flow problems that are not revealed by conventional actuarial present-value calculations. Inasmuch as the critical problems of this system lie in the next 30 years, the Commission has analyzed projections of receipts and outlays to the year 2000 with particular care.

The Commission carried out two major in-depth studies. One analyzed and projected the future of railroad employment against the backdrop of total transportation and GNP prospects through the year 2000, using alternative assumptions. The second study developed and programmed a dynamic computerized actuarial model of the railroad retirement system which projected its receipt and benefit flows in considerable detail for periods up to 75 years and also made the conventional actuarial present-value calculations under alternative sets of assumptions regarding economic and benefit trends. All the projections confirm the finding that the present system is in financial trouble and will soon go broke, unless immediate corrective action is taken to increase the tax rate and/or the wage base or reduce benefits.

These studies were supplemented by eight additional related projects. The Commission's reports provide a comprehensive array of data and other information for assessing the status of the railroad retirement system.

Declining Railroad Employment

Levels of railroad employment running as far back as 50 years determine the present and future number of retirees and other beneficiaries. Present and future railroad employment and payrolls also determine current and future tax collections to support the system.

Past funding methods did not provide for the accumulation of sufficient reserves to tide the system over its coming peak outlay period. The need for such funding was not brought out in the published actuarial reports. The valuation methods used relied very heavily on future

10

tax collections to pay the benefits, and calculated the costs on a level basis "in perpetuity". These past actuarial studies have usually assumed that railroad employment would continue indefinitely at a level close to the actual employment at the time. Their results have therefore been consistently over-optimistic and the system was not funded adequately for its period of peak costs in relation to covered payroll. Through its studies of past trends and its projections of possible future trends in railroad employment the Commission has reached two principal findings:

(1) Railroad employment has trended almost steadily downward for the past 25 years; and,

(2) Employment will probably continue to decline.

The general trend in railroad employment for the last 25 years, despite minor fluctuations, has been almost steadily downward. Indeed, the all-time peak of railroad employment occurred more than 50 years ago, during World War I. When the railroad retirement system was established in 1937 there were 1.28 million railroad workers (average of mid-month monthly counts). The number rose again to 1.68 million during World War II, and thereafter declined almost steadily. Average railroad employment in calendar year 1971 was only 611,000, only 43% of the 1950 level. In the first quarter of 1972 it was 582,000, compared with 616,000 in the first quarter of 1971.

This trend has occurred despite a doubling in the real GNP and an increase of nearly 30% in ton-miles of intercity freight carried by the railroads during the last twenty years. Two significant forces account for this reduction of railroad employment in a growing economy. Technological improvements in the railroad industry and changes in the composition of the freight carried have increased output per worker and enabled growing amounts of freight to be moved by sharply decreasing numbers of employees. In addition, increased competition since the 1940's from highway, pipeline, waterway and air transportation have cut into the share of intercity freight and passenger traffic moved by the railroads. In 1940, for instance, railroads accounted for 61% of all freight ton-miles, but in 1970 their share was 40%.

Passenger service has also declined sharply. At their all-time peak in 1944, railroads carried 75% of the intercity public transport load, airlines about 2%. In 1969 railroads carried about 8% of these passengers, airlines about 73%. (These figures do not include the large proportion carried by private automobiles.)

The likelihood of continuing future decline in railroad employment is suggested by projections made for the Commission by the National Planning Association (NPA). Of several alternative projections, the one deemed most probable indicates that between 1970 and 2000 in a growing economy the revenue traffic units (largely freight ton-miles) carried by railroads will rise from 784 billion to 1,396 billion. This is a 78% increase in 30 years, or 1.9% a year. In this span of years it was assumed that output per employee will increase by 4.8% yearly at first but will gradually fall off-to a rate of 3.6% by 2000. This compares with a 4% increase in the 1950's and a 5.8% increase in the 1960's, calculated without adjustments for the changing mix of traffic.

82-813 72 10

400

CHART 1.-Indexes of Actual and Projected GNP, Railroad Revenue Traffic Units, and Railroad Employment, 1935-2000

400

1935

1940

12

Based on the foregoing assumptions, employment as defined by the Railroad Retirement Act is projected to decline from the 640,000 in 1970 to 503,000 in 1980, 397,000 in 1990, 327,000 in 2000, and to 300,000 by 2015 where it would remain through 2044. Chart 1 shows the actual and projected trends for GNP, for railroad traffic units, and for railroad employment for the period 1935 to 2000.

A Heavy, Rising Beneficiary Load

High levels of employment in the past account for the large number of railroad retirement beneficiaries now. Declining employment increases the ratio of beneficiaries to active workers and raises benefit costs as a percentage of payroll.

At its inception the railroad retirement system assumed responsibility for pensions to the retirees then on the company rolls and also granted up to 30 years of credit for railroad service prior to the Federal law. These generous provisions resulted in a rapid growth in the number of beneficiaries and in higher monthly benefit awards. During 1940 there were 173,000 individuals on the beneficiary rolls. After benefits were initiated for spouses and widows, the number of beneficiaries rose to 461,000 in 1950, then to 883,000 in 1960 and 985,000 at the end of June 1971. Of the latter, 444,000 were aged or disabled retirees and 541,000 were principally spouses and aged widows, plus other survivors. In 1971 those 985,000 individuals represented 740,000 households.

The rapid decline of employment coupled with the rapid increase in beneficiaries has made it increasingly difficult to support the system from current taxes. In 1940 there were approximately seven active workers to each beneficiary family unit. By 1950 the ratio had dropped to about 4 to 1; and by 1971 there was only 0.8 of a worker for each family drawing benefits. In contrast, in 1971 the social security system had 73 million active workers paying taxes for benefits to 21 million families, a ratio of 312 to 1. Thus, the ratio of families on the benefit rolls in railroad retirement in relation to active workers was more than 4 times that in the social security system.

The number of railroad retirement beneficiaries is now nearing its peak. Despite a substantial projected decline it will continue to exceed employment until well beyond 2000. For the year 2000 the Commission projects 442,000 beneficiaries as against 327,000 active workers in the industry, as chart 2 indicates.