to the question of safety, and it is referred to in Mr. Homer's memorandum (p. 3) only as a table that illustrates one of the "more important changes made in railroading over the last several decades." It shows nothing as to the average age of the Class I freight car fleet. Figures published by the American Railway Car Institute indicate that the average age of the Class I fleet declined by more than 10 per cent during the period 1946-1962 (the latter being the last year for which such information was published). This reduction in age was accomplished not only by new car acquisitions but also by retirements of substantial numbers of older cars. Retirements of such cars exceeded 45,000 in each of the years 19461962. In each of the years 1963-1967 Class I freight car retirements have exceeded 80,000 (1967) estimated).

Table 2 is called "Motive Power Utilization" for Class I railroads during the years 1936 to 1963. It shows the percentage of total road locomotive miles and yard locomotive hours operated by various types of motive power. It reflects the disappearance of steam and the ascendancy of non-steam power on American railroads. This constitutes no justification for the bill because non-steam power is demonstrably safer than steam power. Table 2 contains no parallel figures of casualties attributed to locomotive defects; but if it had done so, the marked superiority of the diesel and electric in this respect would have been evident. Some indication of this superior safety can be found in the following tabulation of the frequency rates of casualties attributed to defects in steam and non-steam locomotives per 100 million unit miles:

1 The years 1945-1959 were chosen because they are the only period in which U.S. railroads simultaneously operated significant numbers of both steam and non-steam locomotives. Locomotive unit miles were chosen as the measure of frequency because only locomotive casualties are covered and because a defect on any locomotive unit may cause a casualty. Locomotive unit miles therefore measure the exposure to the risk of a casualty. The markedly better safety record of non-steam power was owing to the characteristics of the machines themselves-as developed by and for the railroads--and not to any governmental regulation. This is evident because during this period both the steam and the non-steam locomotives were inspected and tested by the some governmental agency, operating under the same broad regulatory statute, the Locomotive Inspection Act.

Table 3 shows new rails and cross ties laid by Class I railroads during the period 1930 to 1966. It shows that fewer replacement ties have been installed in the later than in the earlier years and that the same is true of new rail laid. The purpose of the table is apparently to suggest a decline in the condition of track but it fails in this purpose for five reasons: (1) The table takes no account of the improved treatment methods for tie preservation that have been developed since 1930 which assure a longer life for ties. Ties so treated require less frequent replacement. (2) The table takes no account of the 15 per cent decline in miles of track owned by Class I roads, which were 383,076 in 1930 and 322,582 in 1966. (3) The table takes no account of the laying of used but serviceable rail to replace worn-out rail. (4) The table takes no account of improved methods of rail manufacture which virtually eliminate certain rail flaws and thus permit longer and safer rail life. (5) The table takes no account of the trend to welded rail in recent years. Such rail is safer and longer-lived than conventional rail and also reduces stress on ties.

Table 4 shows mileage "protected by" centralized traffic control during selected years from 1929 to 1966. This shows a steady increase in miles of track and miles of road so protected over the years. The Interstate Commerce Commission has found that centralized traffic control increases safety. This table therefore demonstrates the extent of a safety activity which the railroads have voluntarily engaged in, not any reason why a safety bill is needed.

The same applies to Table 5, which shows mobile radio installations on certain railroads during the period 1948 to 1962. The Federal Communications Commission has found that mobile radio, both end to end and point to train, is an aid to safety.

Table 6 shows the percentages of freight cars and locomotives found defective by the Interstate Commerce Commission during selected years from 1931 to 1966. Defects on cars as referred to in this table relate solely to so-called safety appliances (such as handholds, sill steps, running boards, couplers and, especially since 1958, when the Power Brake Act was adopted, brake parts and appurtenances). Defects on locomotives relate to all parts of each unit. The suggestion is that all such defects have a direct and necessary relation to safety, but they do not. A handhold bent 1/4" is a "defect," for example, and almost every one of the almost 2 million freight cars and locomotives that operate in the United States has many handholds. A single brake piston on a car that travels farther than 10'' is a "defect" but long piston travel on several cars has no measurable effect on the braking capacity of a train in which they are included. A locomotive that does not carry tags or cards showing that certain tests have been made within specified periods is "defective" even though the tests have in fact been made and passed. Absence of a cab card is a ground for ordering the locomotive out of service. One of the most common defects found on locomotives is oil on floors or passageways. This defect does have a relationship to safety, since men can slip on oil spots; but it is not a defect that is peculiar to locomotives or even railroads. It occurs wherever machinery is used.

Table 7 relates solely to locomotive defects, which have been discussed briefly above. As stated in the testimony of the Department of Transportation, H.R. 16980 would provide little or no additional jurisdiction over locomotives, which are now regulated under the broad authority of the Locomotive Inspection Act.

Table 8 shows the decline in employment of maintenance personnel from 1933 to 1967. The conclusion is sought to be drawn that this reduction proves undermaintenance of way and equipment, but it does not. It is common knowledge that during these 30-odd years, mechanization of maintenance has enabled the same job to be done better by fewer employees. The table, then, is no more than a showing of how technology can improve productivity.

Table 9 shows the average speed of freight and passenger trains on Class I railroads for selected years during 1921 to 1967. There is no showing here or elsewhere that this modest increase in average speeds has any relation to safety.

Table 10 provides similar information on the length and weight of freight trains during selected years since 1921. It would appear to be equally without relevance, particularly in view of the numerous revolutionary improvements in track and equipment made during those 47 years.

Table 11 provides the same information for passenger trains.

Table 12 shows the number of train accidents, by class of accident, between 1930 and 1956, both in totals and in frequency per million motive power miles. Because all but a few of the subsequent tables in Mr. Homer's memorandum also relate to train accidents and their frequency, it is well to summarize here some of the distortions inherent in his approach: (1) "Train accidents" are defined solely by reference to monetary damage to railroad property. During the years 1930 to 1956 there were 7 different measures employed in various years; but since January 1, 1957 the amount has remained unchanged at $750. One "cause" of increased train accidents in recent years, therefore, is inflation. (2) Because train accidents are based on property damage, they have no necessary relation to casualties. In each of the year 1961-1967, for example, less than 9 per cent of all train accidents resulted in casualties. Only 6.6 per cent of the train accidents experienced in 1967 resulted in casualties. Moreover, between 1961 and 1967, train accidents never accounted for more than 5.9 per cent of all railroad casualties, and in the last three of those years they were less than 4 per cent. (3) A substantial portion of the casualties that do occur in train accidents result from grade crossing accidents. The Interstate Commerce Commission, as the result of a lengthy proceeding, recently said:

"Summarizing the evidence in this proceeding it is apparent that while mechanical difficulties encountered by the motor vehicle are the cause of some of the grade-crossing accidents, by far the main cause is the failure of the motor vehicle operator to exercise due care and caution or to comply with the existing laws or regulations in the operation of his motor vehicle." (32 I.C.C. 63.) (4) Casualties resulting from train accidents have decreased by more than 24 per cent between 1961 and 1967. For these and other reasons Mr. Homer's many tables relating to train accidents and their frequency do not provide justification for H.R. 16980.

When Mr. Homer shows the frequency of train accidents in this and later tables he uses the measure of motive power miles. Such a measure is valid

when he is discussing motive power failures and a few other types of accidents; but it is clearly invalid when it is used-as it is-to measure the frequency of car failures, or roadway failures, or employee error. When it is so used it creates distortions because the measure has no logical relation to the risk exposure purported to be shown.

Table 12 has all the foregoing deficiencies. It has one other in addition, for statistics covering only 1930 to 1956 have no relevance to a bill being considered in 1968.

The same comments apply to Tables 13 and 14, which again survey train accidents, measured by motive power miles (Table 13), and covers only the period 1930 to 1956.

Table 15 shows merely the number of train accidents by general cause that occurred 8 to 10 years ago. The foregoing comments therefore apply to this table as well.

Table 16 is concerned with train accidents only, those attributed to "roadway defects." Even if such accidents were a reliable index to railroad casualties (which they are not), the frequency of such accidents is measured on the table and accompanying chart by motive power miles. This may be valid as to the few accidents caused by signal defects; but it is clearly invalid as to all the other accidents caused by "roadway defects." The exposure to risk from defective rails, ties, and the like must be measured by total traffic over them—ton miles— not merely locomotive traffic.

Table 17 is a mere restatement of Table 16 broken down into numbers of accidents resulting from particular causes.

Table 18 shows train accidents caused by defects in equipment from 1961 to 1966 and also the frequency of such accidents measured by motive power miles. Here again, motive power miles is a most unrealistic measure when it is used to show the frequency of accidents resulting from defects in equipment. All equipment, not merely locomotives, is exposed to risk.

Table 19 merely restates Table 18 by absolute numbers subdivided into attributed causes.

Table 20 shows the number and frequency of train accidents caused by employee error "and other factors." An appended chart of frequency (again measured by motive power miles) shows a very steep rate of climb in both the frequency of accidents caused by employee error and in the total number of such accidents. Nothing in the bill or the statements of its proponents suggests that rules to regulate employee error would or could be adopted under it.

Table 21 shows train accident collisions and their frequency measured by motive power miles.

Table 22 is a restatement by causes of the employee error train accidents set forth in Table 20.

Table 23 shows wage rates paid, measured by train miles and car miles, to two classes of maintenance employees during recent years. It would seem designed to suggest that inflation has not been a great "cause" of increased train accidents despite the establishment of the $750 definition in 1957. The effect of inflation on this "cause" in the face of the inflexible limit, however, can best be gauged by standard indices of cost increases. It would be hard to deny that considerable inflation has taken place in the last 10 years.

The same comments apply to the price of certain railroad materials listed on Table 24.

Table 25 summarizes grade crossing accidents and casualties during 1961 to 1966. It serves to demonstrate the substantial share of all railroad fatalities and injuries resulting from this cause, which admittedly has little relevance to any fault on the part of the railroads. Moreover, representatives of the Department of Transportation have testified that they do not intend to regulate grade crossing safety even if the bill is enacted.

Table 26 breaks down the types of grade crossing accidents for a single year. Table 27 shows all train accidents in 1965 and 1966, the frequency thereof measured by motive power miles.

Table 28 takes the same data for these years, plus another, based on preliminary data. This is evidently with a view to showing that in two years, at least, the preliminary data differed from the final.

Table 29 shows total casualties in railroad accidents for 1961 to 1966 broken down by types of accidents. It shows that by far the greatest number of both

fatalities and injuries arise in train service accidents, not the train accidents to which Mr. Homer elsewhere refers almost exclusively.

Table 30 breaks down these casualties among classes of persons killed and injured. It is notable that although employees on duty form the largest class of persons injured, the largest class of persons killed is non-trespassers, that is, neither employees nor passengers. Most of these are killed in grade crossing accidents, a type of accident which the Department of Transportation does not expect to be affected by passage of the bill.

Table 31 shows employee casualties in a single year by class of employee. It is chiefly of interest as showing the wide distribution of casualties which the railroads must report. Among such classes are executives, chief clerks, ticket agents, stenographers, traveling auditors, switchboard operators, office boys, motor vehicle operators, and janitors, to name a few. There is no parallel in the reporting of accidents by other modes of transportation.

Table 32 shows damage to various classes of railroad property from various types of train accidents from 1961 to 1966. Table 33 is an estimate of the cost of railroad accidents for one year. Both tables show only that accidents are a heavy financial burden on the railroads. This is conceded. It is one reason why the railroads try so hard to promote safety.

The CHAIRMAN. Mr. Kornegay.

Mr. KORNEGAY. Thank you, Mr. Chairman.

Mr. Menk, I appreciate very much your coming back today. I did not hear all of your testimony the other day. I was busy in the office but I did hear the part of your testimony with reference to the ways and means and methods of testing trackage. As I recall, you have two track inspection cars that roll up and down the line and pickup the defect in the track, is that correct?

Mr. MENK. Well, first of all, all of your mainline is gone over by track foreman every day to inspect the general condition of the mainline.

Now, the two cars that possibly you are referring to are rail detector cars. They are electronic devices that are constantly on our railroad, constantly moving over the rail at slow speed diagnosing the condition of the rail, its strength or weaknesses and, if it finds a weakness, it marks the rail and the rail is taken out.

This is the practice on all major railroads of the United States. We happen to own our own equipment. Other railroads contract it with the Sperry Rail Service. This was what I was referring to. This is to continually assure ourselves that our rail is in safe condition.

Mr. KORNEGAY. In other words, it is a continuous operation and goes on all the time?

Mr. MENK. Twelve months a year; yes, sir.

Mr. KORNEGAY. What means do you use to inspect the roadbed itself, the crossties and the foundations?

Mr. MENK. These are the track foremen, track supervisors we call them, who go over the railroad every day, 6 days a week. Mr. KORNEGAY. How long have these railcars been in use?

Mr. MENK. How long?

Mr. KORNEGAY. How long?

Mr. MENK. Twenty years I suppose.

Mr. KORNEGAY. The proposition that disturbs me is that with all of this inspection going on, with the repairs and maintenance that would naturally come as a result of weak rails, weak or wornout crossties, why the incidence of accidents caused by the tracks has increased from 449 in 1961 to 1,038 in 1966. Is there any explanation for that rather substantial and graphic increase which is well over 100 percent?

Mr. MENK. Well, I don't have the explanation for it myself. This hasn't been our experience. I have no explanation as to why or what the criteria were or what the criteria are or what type of accidents these are. These are not casualties I might point out.

Mr. KORNEGAY. Well, they are labeled improper maintenance of way construction and that is broken down into track, other structures and/ or signal systems.

Mr. MENK. I don't have an explanation of the increase. There has been some increase in volume which could contribute to it and also the length of trains has increased.

Mr. KORNEGAY. Has there been any appreciable increase in the speed of trains from 1961 to 1966 ?

Mr. MENK. No.

Mr. KORNEGAY. And under "other way and structure," and I guess that includes trestles, bridges, and grade separations, maybe grade crossings and things of that sort, there is an increase from 137 in 1961 to 390 in 1966. There again you have a great jump.

Mr. MENK. This includes grade crossing accidents you say?

Mr. KORNEGAY. I don't know. The only designation on this chart is "other way and structure." I don't know whether that includes grade crossings or not.

Mr. MENK. This is Mr. Lang's testimony?

Mr. KORNEGAY. Yes, sir. There has been no dispute as to his figures as I understand it.

Mr. MENK. I am not qualified to dispute them, no. Trucks, 449 to 1,038. I have no opinion as to the reason for that increase. I don't know what he classifies as other way and structure. I don't know what that is.

Mr. KORNEGAY. On your line, the Northern Pacific, you say there has been no appreciable increase in accidents caused by rail failure.

Mr. MENK. As a matter of fact, last year we didn't have a major derailment.

Mr. KORNEGAY. Now, what consideration is being given by the railroads in moving from wooden ties to concrete ties?

Mr. MENK. Well, if you are going to build a new railroad out of face it is practical and economical to put in concrete ties, but concrete ties are put in on 31-inch centers as a result of the capacity for the bearing weight.

Wooden are I believe on 19-inch centers so that, if you are going to disperse them, you have the problem of tie centers and it isn't economical. It costs about $12 a tie to put in a concrete tie versus about $6 for a wooden tie. The concrete tie is no more safe than is the wooden tie. The economics are that it lasts about 11 years longer so far as we can calculate. Hopefully we will be building, well, we will in the very near future be building a 55-mile line in Mr. Adams' State, as a matter of fact, and we are studying the matter of putting in concrete ties because it is a new railroad and it will be probably economical to do so. Mr. PICKLE. Will the gentleman yield?

Mr. KORNEGAY. I see my time is up. Otherwise I would be glad to yield.

The CHAIRMAN. The gentleman from North Carolina, Mr. Broyhill. Mr. BROYHILL. I have no questions, Mr. Chairman.

I will yield to my friend from North Carolina, Mr. Kornegay.