These illustrations, however, do not tell the whole story, as to the relative stability of railroad employment. About 1,700,000 individuals received pay from class I carriers in 1937, while only 1,067,000 received pay during the month of January 1938. Thus a total of 600,000 or more individuals who had some degree of railroad employment in 1937 were no longer on the pay rolls shortly after the turn of the year. Stated another way, the total number of employees who received pay from the carriers in 1937 exceeded the number on the pay rolls shortly after the turn of the year by nearly 60 percent. Insofar as 1937 is concerned, a year in which employment declined rapidly in the second half of the year, this 60 percent is the annual rate of employment separations.

Approaching the same matter from a different angle, Mr. Chairman, it is interesting to note the trend of railway employment by months during the year 1937. The committee will recall that from January to July there was a steady increase in railway employment, while from July to December there was a very marked monthly decline in railway employment. Taking July as the peak, therefore, and measuring the rate of increase to that peak and the rate of decline from that peak, it appears that from January to July there was an increase of 7.2 percent in railway employment, while from July to December there was a decrease of 14.4 percent. Therefore, during those relatively short periods within the confines of 1 year alone, the year 1937, there were these very marked fluctuations upward and downward in railway employment.

These facts and figures do not support the assumption that the railroad can afford to pay more liberal unemployment benefits than industry generally. Furthermore, as I have demonstrated, the pending bill is especially designed for the benefit of the casual, intermittent and irregular worker, and discriminates against the normal full-time railway employee. As the State laws were designed to afford the greater measure of unemployment protection to the normal fulltime worker, it is obvious that this bill departs radically from unemployment principles already adopted throughout the United States, that it would pay relatively higher benefits to a greater proportion of the employees, and that it would therefore lead to a relatively high basis of cost.

Let me summarize the general statements I have made regarding the ultimate cost of this bill.

First, we have the typical State law, which in the matter of duration of benefit payments is more liberal_than_contemplated by the Social Security Board. The even more liberal provisions of this bill are predicated on the unproved assumption that railroad employment is more stable than that in general industry.

Insofar as those workers who are most subject to unemployment are concerned, aggregate benefits payable under the bill would amount to at least twice as much as would be payable under the typical State law. My opinion is that the figure would be more than twice as great. The bars would be completely let down in the matter of paying benefits to seasonal, casual, or irregular workers. Of even greater importance, in the forecasting of potential cost, is the omission of any constant ratio between benefits and underlying earnings. Furthermore, we have the matter of so-called partial benefits. Under this bill, a railway worker who is employed on the same or

relatively the same schedule, year in and year out, would be as regularly entitled to partial benefits in any period of 15 days in which his employment did not reach 8 days. In fact, his employment in the benefit year may be substantially greater than that in the base year, yet he could still be entitled to regular receipt of benefits. This is in the face of the fact that, under the typical State law, his employment would have to decrease substantially from the base to the benefit year before he would be entitled to any benefit at all.

Thus two factors contribute to the liberality of benefit payments under the bill, and consequently to its cost. First, for individuals who could qualify for a limited amount of payments under the State laws, this bill would increase the amount payable up to 51⁄2 times. Second, many workers who, under the State laws, would not be eligible to receive any benefits, would be entitled to receive the full quota under this bill.

This sequence of facts leads to the inevitable conclusion that the enactment of this bill would impose a heavy cost, and that the contemplated system could not be supported by a 3-percent pay-roll tax. It appears certain that the cost would be far in excess of 3 percent.

In regard to the statistical and factual justification for these statements as to cost, I need make only a few observations. In the first place, definite cost figures that may later be presented to you will doubtless be founded on figures shown in the Federal Coordinator's report entitled "Unemployment Compensation for Transportation Employees," dated May 1936, and particulaly on tables 19 and 20, appearing at pages 98 and 99, showing estimated duration of unemployment among railroad workers. I believe it to be generally admitted that the Federal Coordinator failed to obtain representative records of the type of workers who would come in for the greater share of benefits under this bill-casual and irregular workers. It follows that his duration of unemployment estimates are not applicable with any degree of exactness to the bill before you.

In the second place, I do not believe that information exists as to the extent of the railroad earnings of casual workers, except on one or two railroads. As the casual nature of their employment indicates that during a year they may have been employed for short periods by several different railroads, their earnings from one railroad cannot be taken as an indication of their total earnings from two or more railroads.

I repeat that unemployment records are yet too meager and unreliable to supply the basis for an absolute cost estimate, even as to time past, let alone the future.

To support that general conclusion, may I read to the committee another quotation from the report of the Committee on Economic Security, to which I have already referred, at page 12? The committee said:

We

Unemployment compensation does not lend itself to actuarial determination of benefits of the same precision as is possible in other forms of insurance. have now in this country only very limited statistics of unemployment. One of the values of a Nation-wide system of unemployment compensation will be the collection of accurate and comprehensive unemployment statistics which it will make possible.

The very system to which the committee's report referred, Mr. Chairman, is the system now set up under the Social Security Act in cooperation with all of the States, and it is the working of that system

over the next few years which will produce what is now not existentin other words, a body of information and statistics which may produce reliable estimates of unemployment-compensation costs. Today no such estimates are possible on any basis of information available

to us.

At the beginning of my testimony, I briefly reviewed the fundamental principles inherent in the system of unemployment compensation existing in the United States today, among which are the accumulation of reserves, and incentive for stabilization of employment through merit rating. I now direct attention to the probable results of making no provision for building up reserves, and the lack of provision for merit rating.

Senator TRUMAN. Is it possible to amend this bill so as to include this merit rating?

Dr. PARMELEE. Mr. Chairman, that question was asked me yesterday, and my answer was that I do not believe the principle of merit rating can be superimposed on the general plan of unemployment compensation set up in the bill. However, it would be possible, of course, to so completely revamp and revise the bill that it would include that principle.

It is possible, but very improbable, that during times of normal business activity the revenue from a 3-percent tax would support the measure. But just as inexorably as the economic cycle exacts its toll, so would the fund from a 3-percent tax be rendered insolvent soon after the downward turn of the cycle.

Insolvency of the fund would present two alternatives, namely: Increase in the tax rate to meet the cost, and the increase would have to be very great in times of decreasing employment and consequent increase in the claims for benefits; or denial of the payment of benefits to long-service employees who become depression-unemployed.

There is serious question whether this bill would protect the individual who in normal times finds substantial employment in the railroad industry. The bill is not even drawn for his benefit. By its very terms, the emphasis is placed on the regular supplementing of the income of the worker who has not established a permanent attachment to the industry-the casual, irregular, and intermittent employee. Senator TRUMAN. How do you account for the endorsement of this bill by the various brotherhoods?

Dr. PARMELEE. Mr. Chairman, I find it difficult to answer that question.

Senator TRUMAN. They have endorsed it.

Dr. PARMELEE. They have said, very frankly, that their effort has been to provide larger benefits to the casual, short-time worker. Even if it were conceded that there should be some scale of benefits in their favor, we believe that the scale set up in this bill is out of all reason when related to the merits of the situation and particularly to the question of economic cost. That is, we believe that this very heavy weighting in favor of the short-service man and against the long-service employees will undoubtedly lead to insolvency of the fund.

But what of justice to the carriers, as employers? All but two of the State laws lay down the principle of merit rating, under which an employer who keeps unemployment in his own establishment within certain limits may obtain a reduction in his tax rate. Thirty-nine of

the laws provide for the automatic operation of the merit provisions, with consequent reduction of taxes to those employers who merit it. Employers in the State of Wisconsin already have been granted tax reductions. In other States reduction in taxes will start at various periods from 1939 to 1942.

Of the 39 States which have automatic merit rating, the tax rates in 10 States may be reduced to nothing; in one, to one-half of 1 percent; in 13, to nine-tenths of 1 percent; in 11, to 1 percent; in 2, to 11⁄2 percent; while in 2 States there is no set minimum.

The possibility of reduced cost to participating employers in those 39 States is certainly great. To pass this bill, however, will cut the railways off from any possibility of saving in taxes in those States, a possibility they now enjoy. To contend that the possibility does. not exist is to renounce the system of unemployment compensation which has been erected as a result of an act of Congress. To say, before these merit-rating provisions have even been put to the test, that they will not work, and to condemn principles so fundamental as the accumulation of reserves and operation of merit rating, is certainly premature.

The equity of omitting merit rating from this bill is not apparent. The proponents say that railroad employment is more stable than employment generally. If this be so, why should not the railroads, as employers in 39 States, be permitted to benefit from such a situation? If not true, why should not the carriers be given the incentive to make it true?

The incentive that merit rating provides for stabilization of employment was recognized by the President's committee on economic security on page 18 of its report. The opportunity for the States to include the principle within their acts was afforded by title IX of the Social Security Act, which stipulated the conditions under which additional credits against the Federal tax would be granted to employers who qualify for reduction in their tax rates under the State

Included in the 39 laws which provide for the automatic operation of merit rating is the District of Columbia law, enacted by the Federal Congress itself. It is working in Wisconsin today. Yet, with all of this precedent, with the Federal Congress having twice gone on record in its favor through actual legislation, and with no evidence that it will not work in the States, this bill recognizes the principle in neither fact nor theory.

I turn now, Mr. Chairman, briefly, to the very important question and problem of finding work for the unemployed, as to which the bill makes certain provisions, among others, that the Railroad Retirement Board, which is to be charged with the administration of the bill, shall either set up its own reemployment offices or shall work cooperatively with the State employment offices now existing in the several localities. The question in our mind is as to whether a cooperative arrangement of the kind suggested, when you consider that railroad employees have been set apart from the employees of other industries, will actually be effective.

The day has not arrived, Mr. Chairman, when workers in America are confined in their employment to one industry; particularly is this so in regard to younger men, not yet settled in their work, and older men, cut loose from their employment in industries which have passed the peak of employment. These are the men whom unemployment

compensation is intended to benefit, and these are the men for whom State employment offices have been established, so as to bring them in touch with new sources of employment. On the other hand, men who have become firmly entrenched in their jobs are little concerned with unemployment compensation. Even as to these men, if they do become unemployed, their chief interest also is finding new sources of work. What is a hundred dollars of unemployment compensation, compared with the opportunity to work? It must not be forgotten that these men, who have had steady railroad employment, would not receive any more benefits under this bill than they would receive under the typical State law.

Unemployment compensation, no matter how generous, is only a palliative; it eases but does not cure the impact and consequences of unemployment. To be effective in its working, it must be part and parcel of a system which directs the unemployed to new sources of employment. This the State employment offices, coordinated in their functions under the Wagner-Peyser and Social Security Acts by the cooperative undertaking of the United States Department of Labor and Social Security Board, attempt to do. What effect would a similar effort, to achieve the same purpose, be under this bill?

It is only logical to believe that for the extremely liberal benefits which they would receive from the railroad system they would pay an excessive price in loss of working opportunity elsewhere; that they would be discriminated against when they apply for work through a State employment office.

So far as placement of unemployed railway men in contact with new jobs in the railway industry is concerned, it seems doubtful whether the employment agencies that the bill authorizes the Board to set up could accomplish a great deal, desirable as their objectives would be. Men with longer periods of service, who have built up a substantial degree of seniority, are now automatically called back to work when the opportunity offers; no additional agencies are needed for them. Men with shorter periods of service might receive some measure of assistance, although even these men have built up a seniority status that makes it easier for them to secure reemployment with their former employer, and at the same time makes it more difficult for them to transfer to another railway company; where they come into competion with the former employees of that company, whose seniroity would of course take precedence. This fact is evidently recognized in the bill, section 4 (c) (v), which authorizes an employee to refuse to accept work if such acceptance of work would subject him to loss of substantial seniority rights elsewhere. Finally, we have the group of casual or irregular workers, for whom the new employment agencies might find new jobs, provided they are willing as in many instances they are not to move from one locality to another. Furthermore, the bill specifically directs the Board, in section 12 (i), to operate any of its employment facilities so as to be "directed primarily toward the reemployment of employees who have theretofore been substantially employed by employers.'

In other words, the effort would be to place the regular, rather than the irregular, workers in new jobs.

Summarizing the reemployment problem as sympathetically as possible, it seems questionable whether an elaborate set-up of agencies such as the bill contemplates-and these agencies would be entirely