that made by the Federal Coordinator of Transportation covering the years 1924-33. That study covered the persons on the pay rolls of the first half of July 1924, the first half of July 1929, and the last half of December 1933. Because of the necessity for making a study that was within reason and that could be carried through, it was found impracticable to check the record of every individual who worked for those railroads during those periods of time. Mr. Lawler, I believe, in his testimony, named seven railroads that that study covered.

The rate of unemployment found from that study—that is, the rate of becoming unemployed for 1 or more days-expressed as a percentage of the total number of employees, ranged from a point of

13.01 in 1928 to 30.47 in 1932.

Now, the data on which these percentages were based did not include short-service employees, casual or otherwise, who did not appear on one of the three pay rolls.

Since the Coordinator's studies have been completed, the operations of the Railroad Retirement Act have yielded some additional data as to unemployment, and consequently it is desirable to use figures which probably, so far as the rate of becoming unemployed is concerned, are more accurate. We have not been able to tabulate all the data which we have from all the railroads, but we did take two railroads which we felt to be reasonably typical, one a large eastern railroad and a still larger one in the West. We found the number of employees who worked in all 12 months and those who worked in 11, 10, 9, 8 months, and so on, and tabulated the number of such employees by $100 wage groups.

I should like, if there is no objection, to insert those tables into the record, with some comment.

Now, there are several questions about those data. First, as to the degree to which those railroads are in fact typical of all railroads: The average annual employment on the 15th of the month for the eastern railroad was 9,573. The average of the number on the pay rolls was 10,420. The corresponding figures for the western railroad were 14,877 and 16,173. The total number of separate individuals on the eastern railroad was 14,014, and the total number on the western railroad was 22,444. In other words, the ratio between the average of the midmonth amount to the total number of separate individuals on the eastern railroad was 1.46, and 1.51 on the western railroad.

Now, in all class I railroads in the year 1937 the average number on the 15th of each month was 1,115,077, or, in round figures, 1,115,000. The total number of different individuals who worked in the whole industry during 1937 was approximately 1,950,000.

In addition to class I railroads there are covered by the Railroad Retirement Act, the class II and III railroads, switching and terminal companies, express agencies, sleeping-car companies, and a number of other companies owned or controlled by carriers by railroad (where they perform a service that is more than a casual service in the transportation of passengers and property by railroads) and certain railroad associations and labor organizations.

The pay roll on class I railroads was $1,985,000,000 and some odd hundreds of thousands. The total pay roll on which taxes were collected under the carriers' tax was approximately $2,300,000,000.

Using the ratio between the pay roll subject to the Railroad Retirement Act and the pay roll for class I railroads to expand the average number at the middle of the month on class I railroads indicates that the average number of employees at the middle of each month in 1937 subject to the Railroad Retirement Act, was 1,292,000. In other words, the ratio of the total number of separate individuals employed last year by all employers subject to the act to the average number at the middle of each month was 1.51 to 1. In other words, for every average employee that the railroads had, they employed half of another person, or for each 100 they employed 51 other persons for some part of the year.

That, roughly, is a measurement of the extent of seasonal and casual and cyclical and various other kinds of employment during the The CHAIRMAN. Have you examined this bill?

Mr. LATIMER. Yes.

The CHAIRMAN. What do you think of it generally?

year.

Mr. LATIMER. I would like to describe this first. It seems to me that the question as to whether or not the cost will support these benefits is perhaps more important than any other question. I want to describe in detail the process that we have gone through in trying to determine whether or not these benefits could be supported by the contributions which are levied.

The CHAIRMAN. All right.

Mr. LATIMER. Now, on the western railroad, which forms the main basis of these cost figures, the ratio between the unemployed or the groups who were employed sometime during the year, and the average employment was also 1.51. In other words, it looks as if we have picked out as nearly a typical railroad as we could possibly get, and in saying that I think it should be emphasized that that means that whatever seasonal and cyclical employment there is in the industry is allowed for, during the year, 1937, at least, in these figures which were used in trying to determine what the benefits would be and what would be required in the way of pay roll contributions to support them.

Therefore, on the basis of seasonal and other kinds of employment, it does appear that the railroads that we have picked out are typical. There are two further questions. The first one is whether in 1937 the rate of unemployment was reasonable, and what might be expected over a period of time, and whether or not the durations which we have used were also reasonably in line with what might be anticipated over a period of time.

We have tested both points by reference to the I. C. C. data, which 'cannot be construed as being measurements in themselves, but which relatively from year to year ought to be reasonably indicative. We find that in the year 1937 the rate of becoming unemployed was about 25 percent higher than it was for the 10 years 1924 to 1933 and substantially identical with the rate of becoming unemployed in the years 1929 to 1933, which latter were, on the whole, years of rapidly declining employment.

In making these cost calculations, therefore, we have taken the rate of becoming unemployed for the year 1937 and have combined them with the durations of unemployment for the years 1924 to 1933 as shown by the Coordinator's figures, and we feel that on the whole it is a fairly conservative statement, since it does indicate a rate of

unemployment which gives a considerably larger cost than would any actual figures which we would derive from experience over any 10-year period in the past.

Mr.

Now, I should like to add one further item which seemed to me to have considerable bearing on the reliability of these estimates. Lawler presented this morning, as he presented before the House committee, a statement as to the relationship between the benefits proposed under the Senate bill and the benefits which are now being paid under the State acts. He found that the benefits under this bill averaged 47.76 percent larger than under the State acts.

I asked our actuary to calculate, on the basis of these unemployment figures which I have just given, what would be paid under the State laws to railroad workers who became unemployed and what would be paid under this bill. He has found that, on the average, the railroad benefits as proposed would be 47.22 percent greater than under the State laws. He has arrived at that by making an estimate of the actual duration of unemployment.

The fact that that estimate which was made without any reference to date as to actual State payments, but purely on the figures which we assembled on unemployment and wages-was one third of 1 percent from the figure which has been derived from actual experience seems to be so striking that it lends much support to the validity of the data we have assembled.

These data—that is, the rate of unemployment based mainly on the western road for the year 1937 and the durations of unemployment shown by the Coordinator's study-indicate first that about 41 percent of the persons engaged in the industry during the course of the year will become unemployed. A part of that unemployment is to be eliminated for the reason that it arises because of retirement under the Railroad Retirement Act, deaths, and becoming totally disabled. On the other hand, among the group who served 12 months, there is some unemployment. Based on the studies of the wages received, our actuary estimates that 40 percent of the railroad employees will become unemployed.

The average daily benefit for all persons becoming unemployed, including both those eligible and those not eligible, is found to be $1.58.

The average duration of unemployment according to the Coordinators study would be 43 days. At the rate of 8 days' compensation per 15 days total unemployment. The actual duration of unemployment benefits would run, about 2% months.

The average annual compensation of all persons attached to the industry—perhaps that is not what is commonly known as the average annual wage, but is arrived at by dividing the total pay roll by the total number of persons in the industry is $1,150. The total amount of payments of unemployment compensation to each 100 individuals that is, 100 times 43 times $1.58-is $2,717.60. The total compensation for those 100 individuals is $115,000. So that the percentage of the pay roll required to support those benefits is 2.36.

The Coordinator's figures were worked out at a time when it was assumed that there would be some provision for some partial unemployment, something on the order of what is now provided for in the State laws. While it is not possible for us to lay our figure on any particular assumption which resulted from that particular provision, it may be wise to make some allowance for these unknown assumptions.

Consequently, for that and other reasons, we have added to that cost and allowance of one-eighth to cover these assumptions and other contingencies, so that the total cost of benefits is 2.66 percent of the taxable pay roll.

There is one factor in here for which no specific allowance has been made, and that is the fact that the base year is on a calendar year basis, so that the man who earns $250, $125 in December and $125 in January, would not qualify. Moreover, if he earned $150 in the early part of the calendar year and thereafter became unemployed and had no previous earnings, it would be approximately a year and a half before he could qualify for benefits. The chances of his getting a job in a year and a half are high, of course, so that both those factors would tend to reduce cost as here stated.

Our actuary says that it is virtually impossible to assign any specific figure to those items, although both would be in the direction of reducing cost.

On the assumption of a 10 percent administrative cost, it does seem to me that such data as are available do lend validity to the view that the contributions as set forth in this legislation would support the benefits which are to be provided—that is, I say, taking into account the seasonal employment in the industry.

Second, there is the question of whether or not the allowance of 10 percent for cost of administration is sufficient or is excessive. It is bound to be said at the beginning that before actual experience it is difficult to set any precise figure for cost, for this reason: Obviously the Railroad Retirement Board could not possibly expect to duplicate the agencies which the States have set up for the administration of unemployment compensation throughout the 48 States and the District of Columbia.

I think it is also obvious that some functions which are now being performed by existing agencies could be utilized in the work of administering this act in such a way as to reduce administrative cost, and I would assume also that such extra expenses as these existing agencies would be called upon to expend in their performance would be paid for out of the administration funds provided by this bill. I do not assume that the Congress would pass or the railroad labor organizations are in here to advocate a bill in which they do not pay for the total cost by whatever agency that cost happens to be borne.

Furthermore, I would assume that a number of functions could be performed by the railroads themselves, and I would assume in that also that the additional cost involved to the railroads, beyond what they would normally be required to have in the way of making reports to a governmental agency where the Government itself administered these functions and the only requirement was that the railroads furnish a reasonable amount of reports, would also be defrayed by the administrative agencies.

The bill, as I read it, does give authority for such expenses to be defrayed. I refer particularly to the functions of registration and to the functions of keeping the so-called negative records if they should be required.

The CHAIRMAN. You are assuming that if these negative records are required the Board would pay for that expense?

Mr. LATIMER. Would pay for that, because that seems to be what you might call a governmental function being performed.

The CHAIRMAN. How could you say that the Government would pay back to the railroads the extra cost? I think that would be a very difficult thing to do.

Mr. LATIMER. I do not think you could figure the exact cost. Nevertheless, it does seem to me that there are certain functions for which a reasonable estimate of cost could be made and which would be so arranged that no substantial extra cost would be involved on those who performed them. I say that on the assumption that the labor organizations who are sponsoring this legislation are not doing so with the intention of adding to anyone's cost. It would seem to me that the easiest way of handling the administration of a piece of legislation of this sort is to integrate it as far as possible into the existing machinery in the industry.

I would not say that for any other industry, for the fact that there exists in most local departments and divisions of the railroad industry a committee of adjustment of labor organizations who stand ready to protect the rights of their members makes it safe to adopt the procedure in the railroad industry, whereas it might be unsafe to adopt it in a great many others. I believe that the possibilities of action in that manner are very great.

In that connection, I would like to mention the statement made the other day by Dr. Burns of Columbia University, who was asked last summer to make a study of European systems of unemployment compensation, with a view to making available to administrators in this country the experience in Europe. Dr. Burns reports finding great astonishment by most of the administrators in Europe on learning of the failure of American administrators to take advantage of existing industrial relations organization in this country. They feel that there is a very definite possibility of integration such as has been practiced for a good many years in Europe, particularly in Denmark and Switzerland. There is less of this integration in England, although there is a good amount of it there. The kind of integrated organization possible in this legislation is in line with what has been done in other countries.

The point I want to make is that until there is a specific piece of legislation and somebody to administer it, with authority to act, there will not be the opportunity to go to the railroads and to these other agencies and say: "We want so and so done. Can you do it? And if you will do it, how much will it cost you compared to what you are now spending?" And look at these estimates and study them and make suggestions here and there as to where the procedure may be shortened, and sit down around the table and confer. I do not think it is possible now to say how much it will cost.

I do think, however, that there is some precedent to indicate that 10 percent is not altogether a hypothetical figure; that is, it is not altogether picked out of thin air. In the first place, there is the precedent of the Social Security Act itself. Title IX of that act taxes certain pay rolls, including those of the railroads, 3 percent. The contributions of employers to the State unemployment-compensation agencies may be offset by employers against that tax up to 90 percent of its amount.

The Social Security Act does not specify what the purpose of the remaining 10 percent collected under title IX is, but in the hearings before the Ways and Means Committee which was considering the

72656-38--10