Senator BARTLETT. And I wanted to clear up my position on that because of course I realized that an agreement as such was not within the province of the Board.

Mr. HEISS. I was sure the chairman didn't mean that. Thank you, sir.

Senator BARTLETT. Now, Mr. Homer, if you please.

Mr. HOMER. My name is Winfield M. Homer. I must apologize, Senator Bartlett, I don't have very much of a voice today, but I hope I will get through with this electronic help.

Senator BARTLETT. You are coming over loud and clear.

Mr. HOMER. I am glad to know that. I am going to deal today with the question of railway safety. I am going to respond specifically to the presentation made by the carriers through their witness Mr. H. E. Greer.

I have on my left Mr. Jack Fry, who assisted me in my studies. It may be necessary at some point to refer a question or two to him. Senator BARTLETT. Before you go on, Mr. Homer, your rebuttal statement says it is made by you by name, representing, as I infer, the Labor Bureau of the Middle West.

What is that bureau?

Mr. HOMER. The Labor Bureau of the Middle West is an association of professional men who provide economic counsel to labor unions, primarily to unions in the public utility and transportation fields. We have been in existence something over 40 years. We have offices in Chicago and here in Washington, D.C. We are the economic advisers for the Railway Labor Executive Association, and have worked over the years with almost all of the standard railway organizations.

Senator BARTLETT. Thank you.

Mr. HOMER. Perhaps I should say I appeared here earlier in this proceeding, on August 10, 1965. I have asked to have presented to the members of the committee three documents. The first of them is a prepared statement, the second one a series of statistical exhibits, and the third is a technical memorandum.

I do not intend to read the prepared statement, rather I shall try to talk from the exhibit.

The technical memorandum deals with a number of the more complex sophisticated elements in computing accident rates and the effect of such items as weather, and traffic on accidents, and I felt that would be best dealt with in a written statement and I shall not discuss it except to the extent it may relate to some of the other things I shall discuss in my general statement.

At the outset I would like to say there is a degree of unanimity on railway safety trends. The railroads, the Interstate Commerce Commission, and the brotherhood all agree that railways accidents have increased. They differ only as to the cause.

The Brotherhood contends that the removal of firemen in freight and yard service was a major cause of the tremendous increase we have seen in accidents and casualties; the railways contend the increase was due to other causes; and the Interstate Commerce Commission in effect has been noncommittal with respect to any such relationship.

61-927-66-pt. 2-5

Today I shall deal only with this central issue, that is, with the facts that bear most directly on the effect of removal of firemen on the accident and casualty trends.

In this connection I would like to turn to my exhibit, to page 1. On page 1, I have reproduced an excerpt from the testimony of Commissioner Bush, who appeared here on behalf of the Interstate ! Commerce Commission early in the proceedings.

Mr. Bush commented on the fact that crew consist; that is, the presence or absence of firemen or brakemen, and so forth, may have; some bearing in collisions and in other train accidents. He commented he saw less relationship to the type of accidents we call. derailments.

Without agreeing completely that the presence or absence of firemen is unrelated to the incidence of derailments, I would like to present such information as we have as to what has happened to railway collisions.

I think it is quite clear that the presence or absence of firemen has its greatest effect in that area. About 90 percent of the collisions, train accidents, are caused by employee failures, not by broken tracks or defective equipment, but by errors in judgment, inability to see, inability to pass a signal in actual operations.

Thus, taking the collisions, we have the most direct evidence of! the extent to which the removal of firemen may have increased accidents in the railroads.

On page 2 of my exhibit I have shown the number of collisions of all U.S. railways for each month for the years 1961 through 1963. I have shown each month opposite the 3 years showing the 3-year average and showing the year 1964, and then the first 2 months of

1965.

I regret that we have no information beyond February of 1965, with this degree of detail.

If you will observe these figures, it will be noted-and I can do this in a summary fashion-that there was very little increase in the number of collisions on American railroads in the first 4 months of 1964. As a matter of fact, the number of collisions was very close to what it had been in the average of the preceding 3 years.

But if you will look at the number of collisions which occurred after April of 1964-and these are the figures which on the exhibit are bordered in black-you will notice a very substantial increase.

For the last 8 months of 1964, there were 109 collisions on an average each month. You will note that figure, averaging for the 8-month period, is as high as was known for any single individual month prior to May of 1964. There was only one month where the rate or where the number of collisions reached that level. And in this 8-month period immediately after the removal of firemen, they; averaged 109 collisions.

On the next page we show the collision rate. This is a more precise method of making the judgment as to the increase in collisions.

Again, the format is the same. I have shown the information for the first 4 months of 1964, which may be readily compared with the same months of the earlier years and of the last 8 months of that year. I would like at this point to ask that the first of my charts be put on the board.

I have shown here the number of collisions for each year on the average, or rather the collision rate for each year on the average.

In 1961, it was 1.07; 1962, 1.09; 1963, 1.18; and in the first quarter of 1964, a slight drop, to 1.17.

Here is where they removed the firemen [indicating]. The rate immediately shot up to a record level, 1.39 was the average rate of 8 months, and not in any 1 month prior to that time did the level ever reach that particular figure, 1.39.

And yet immediately after the removal of the firemen, the average for the whole period of 8 months in that year went to that level.

In January 1965 it stayed at 1.39, and in February 1965 it rose to an almost unprecedented level of 1.59.

Senator BARTLETT. You say, Mr. Homer, you can't get figures after February 1965. Is that the usual lag in time?

Mr. HOMER. No, it is not the usual lag, Senator Bartlett. We normally have figures available much earlier than, say, a year or a little more than a year. My understanding is that the Interstate Commerce Commission now has received the figures through May. But they have not received the March report for some reason, and consequently they didn't distribute any of the figures.

I think I should point out this chart is reproduced in my exhibit at page 5.

I now ask that the next of the charts be put on the board.

I spoke of the fact that at no time prior to the removal of firemen did the collision rate ever reach 1.39. There has been, as you see, very substantial variation month by month through the years. Collisions, unlike some other accidents, don't have any real regular method of falling. They depend so much on human error that it is a variable concept.

I have taken, however, each month for this period 1961 through 1963 and plotted them on this chart.

The 1961, 1962, and 1963 Januaries are all in this same box, so they can be seen together. This is true of each of the months. And the yellow background is the average of each month. Thus, this is the average January of those 3 years, the average February, and so forth.

During this period in 27 out of the 36 months the rate was under 1.2. In 5 of the months it was under 1, 1.0. There were five that were between 1.2 and 1.3, and there were only 4 months in which the rate went above 1.3.

Now I have taken this same yellow background and reproduced it on exactly the same scale on the next chart, in order to show what happened when the firemen were removed from service.

This yellow background is plotted exactly the same as it was on the preceding chart. The red bars are an indication of what happened after firemen were removed.

The black bars here represent the first 4 months of 1964, before the arbitration award took effect. In this period you will note every one of these months exceeded the average of the earlier period.

I am speaking, that is, of the red bars, those which occurred after firemen were removed.

In the first 4 months it will be noted that one of the bars, one of the rates, was below 1. The rest were slightly above the average for the preceding years, although April did rise rather substantially. But

when we get to May and June and July, they are all substantially above the average of the earlier years.

And since April of 1964, first of all, they have all been above 1.20 There were very few above 1.20 before that. There are 7 out of 10 months that were over 1.30, and there were 4 out of 10 that were over! 1.50, a level that they never even came close to prior to the time of the removal of firemen.

I want to recall to you this is the type of accident most closely associated with the duties of the firemen and engineers, the accidents that arise because of failure to see, failure to understand or abide by hand signals, or other signals. I would like to point out the last two charts I have been discussing appear on pages 6 and 7 of my exhibit.

Thus far I have been referring to the accident experience of all carriers. Now these summarize the total collision record of the industry, and they include the experience of carriers who had laid off substantial proportions of their firemen, but they also include the experience of carriers who have laid off a much smaller proportion of their firemen, and perhaps some that laid off no firemen whatever Now if the removal of firemen has had the effect that we say it, has, that these figures indicate it has, we would expect that on those carriers where they removed a great number of their firemen, they! would have a worse experience than on those removing very few firemen.

Such a comparison can be made and the basis of it was furnished to the committee in the material and testimony submitted by the carrier witness, H. E. Greer.

Mr. Greer submitted in his appendix 12 and appendix 21 information on 2 groups of carriers, 1 group of 33 carriers, which he said had conducted more than 35 percent of their operations without firemen, and they averaged about 50 percent of their operations without firemen, and the other group operated with less than 35percent without firemen and averaged about 25 percent without firemen.

Now we have 2 groups of carriers then, the 33 over 35 percent, and the 43 under 35 percent.

With reference to these particular groups of carriers, Mr. Greer commented as follows, and this appears on page 641 of the printed record of this committee:

Following the effective date of the award, the 33 carriers who conducted nearly 50 percent of their freight and yard operations, without firemen, had lesser increase in their train accident rates than the 43 carriers who conducted only 25 percent of their train operations without firemen.

Now I don't know what influenced Witness Greer to select that 35 percent dividing point. I do know when you take all types of ¦ accidents, it might be possible to select a breaking point between them that would cause a degree of balancing out, when you take all kinds of accidents. But I think it would be of interest to this committee to see how the collision experience of the two groups of carriers. turned out. And I have done that on pages 9 and 10 of my exhibit. On page 9 I have shown the aggregate numbers of collisions. think that is less useful to us than the collision rates, which are shown on page 10, so I will just refer to page 10 now.

I have shown here the 33 carriers which had laid off over 35 percent or had operated over 35 percent without firemen and averaged around 55 percent. And the 43 carriers with the so-called firemen-on philosophy, who did not go as far as the other group in laying off firemen. If you look at these two rates you will notice in 1961 the 33 carriers had a rate of 1.15-these figures are on page 10 of the exhibit-the 43 carriers a rate of 1.16.

In 1962 the 33 carriers had 1.15; the 43, 1.13.

The next year, the 33 carriers had an increase, but not a substantial increase in light of the variations that you will see for all carriers in their collision records, in the earlier years, 33 carriers had a rate of 1.24, and the 43, 1.15.

The average for the 3-year period you will notice is not very far apart. One is 1.18, and the other is 1.15.

The first quarter of 1964, the average for the 33 stayed about the same level. But for some reason the 43 carriers in the first quarter dropped considerably to 1.02.

Now, the second quarter in 1964, that was the first period for which we had an indication of the effect of the removal of firemen. Two months in that second quarter were affected by that fact. You will note that the 33 carriers who had removed the larger proportion of their firemen, their rate jumped up immediately to 1.41, whereas the other group did rise, because they did take some firemen out of service, but much less, only to 1.27.

In the last half of 1964, the last 6 months, you will notice the 33 carriers that had operated so much service without firemen, their rate jumped to 1.56, for a 6-month period, compared to 1.30 in the 43

carriers.

Now, as Mr. Greer said with reference to his figures, there is no difference, but when you look at the area where the presence or absence of firemen is most likely to be felt, you find out there is a very substantial difference.

I have had these figures also put on charts. I need not, I think, discuss them in any detail.

Now, the significant comparison is between the top chart and the bottom chart. The experience of the two carriers you will see was very much the same for the period prior to the removal of firemen.

There is a tremendous difference in the red bars after firemen were removed. The big increase took place in the 33 carriers where approximately 50 percent of the operations were without firemen. I should say that this chart is reproduced on page 11 of my exhibit. I have shown on page 12 this same information on a seasonal basis to show that even when it is compared on that basis the result is exactly the same.

I think I need not discuss the figures themselves.

Clearly there was enough divergence between these two groups of railroads to indicate when firemen were taken off, it had a very substantial effect on railway safety.

On pages 13 to 22 of my exhibit I have included some brief reports of collisions which occurred in operations where firemen helpers were not employed. The original source of this information is a series of reports made to the brotherhood by its officials and members throughout the country.