In lieu of an abstract, here is a brief excerpt of the content:

Appendix2 CasualtiesandAccidentsfromInterstate CommerceCommissionStatistics,1888–1965 The basic data to measure railroad safety from 1888 on derive largely from the statistics collected by the ICC. This appendix presents the series from which most of the tables and graphs in the text derive. It also contains the rail failure data derived from the AREA and employed in chapters 7–10. From 1888 to 1900 the ICC gathered no statistics on numbers of train accidents but it presented railroad casualties (fatalities and injuries) in its Statistics of Railways. From 1889 on these data are divided into injuries and fatalities to employees , passengers, and others. The data are presented for the United States as a whole and for ten geographic divisions. Employees are sometimes divided into broad groups of occupations (trainmen, switchmen, flagmen, watchmen) while casualties to “others” include trespassers and nontrespassers. Casualties are also classified by broad class and cause of accident (casualties from collisions, from train and nontrain accidents; casualties from car coupling).1 Beginning in 1901, under authority granted by the Accident Reports Act of that year, the commission also began to gather data on train accidents and a separate set of casualty statistics, both of which it presented in its quarterly (later yearly) Accident Bulletin. Thus numbers of collisions, derailments, and other train accidents date from 1901 (fiscal 1902), as does the definition of a reportable train accident, which was one doing damage of $150 and/or resulting in a casualty. Until 1910, the casualty statistics presented in the bulletin include only those to passengers and employees that result from movement of trains. Employee injuries were defined as those causing incapacitation for more than three days in the succeeding ten.2 There are many deficiencies in these early data. Those in Statistics of Railways do not separate casualties to employees on duty from those to employees not on duty. Employee injuries are grossly underreported and casualties by type of accident do not always agree with those in the Bulletin. Beginning in 1910 the commission ended the dual system of reporting and thereafter all statistics on accidents and casualties derive from its Accident Bulletin. In that year it divided reportable casualties into three groups. There were those arising from train accidents (and the $150 damage figure was now revised to include the cost of clearing wrecks). Train service accidents were those arising from movement of trains but doing less than $150 of damage, while there were casualties 321 from nontrain accidents, such as passengers who fell in stations or workers injured in shops. In the text I have grouped accidents differently. Because contemporaries placed so much emphasis on collisions and derailments, I treat them separately. All other accidents and injuries, including some train accidents such as boiler explosions , I call “little accidents.” The ICC revised the definition of what constituted a reportable accident or injury many times from 1910 on, and its enforcement of reporting accuracy also varied. In the period under consideration the most substantial revisions occurred in 1957 and data from that period on are not entirely comparable with those of earlier years.3 PassengerCasualties Table A2.1 contains passenger casualty rates per billion passenger miles, from 1888 to 1965. Over the years the ICC presented passenger casualties using different definitions of who constituted a passenger and what constituted an accident. For the sake of continuity I have chosen a series from the appendix to the commission’s Accident Bulletin, for these are the only data that span the entire period under consideration. Passengers exclude individuals carried under contract (e.g., postal employees) while the data include only those passengers killed in train and train service accidents. They exclude a small number of passengers killed or injured in nontrain accidents, for example, by falling down the stairs in a station. For the early years I have been unable to discover the precise definition for either fatalities or injuries. By 1901 a passenger injury included anyone hurt suf- ficiently to miss a day of work while anyone who died within 24 hours of the accident counted as a fatality. Since the commission presented the data in Table A2.1 as one series, and since it usually noted any inconsistencies, I assume that these definitions obtained from the beginning. A more serious difficulty is that the definition the commission employed counted as an injury anyone who died more than 24 hours after the accident. In 1922 the commission began to require...

Share