levels and include workers' compensation records and claims, death certificates, newspaper articles, aircraft and highway transportation data, and other administrative records. Demographic information, the circumstances of each incident, the industry in which it occurred, the involvement of any equipment or machines, and the victim's occupation are recorded in CFOI. Data are available at the national, state, and metropolitan statistical area levels. BLS and the states share the costs and data-collection responsibilities.
From 1992 to 1996, CFOI recorded a total of 339 deaths of children and adolescents. There were annual averages of 27 work-related deaths of youngsters under the age of 16, and 41 work-related deaths of 16-and 17-year-olds (Bureau of Labor Statistics, 1997b; Derstine, 1996). The most common causes of death during the period from 1992 to 1995 (latest year available) were motor vehicle or other transportation-related incidents and homicides. Agriculture was the most deadly industry for younger teens: Some 80 percent of the fatalities of youngsters under the age of 14 were in agriculture, as were 46 percent of the fatalities of 14-and 15-year-olds (Derstine, 1996). For 16-and 17-year-olds, about a quarter of the fatalities occurred in agriculture and a quarter in retail. More than 25 percent of all fatalities of youngsters under the age of 18 occurred in family businesses (Derstine, 1996). As is discussed at greater length in Chapter 6, agriculture is exempt from many child labor and health and safety laws.
In 1993, homicides accounted for 28 percent of the work-related deaths of 16-and 17-year-olds recorded in CFOI (Toscano and Windau, 1994), compared with 10 percent of the work-related deaths from 1980 to 1989, as recorded in the National Traumatic Occupational Fatality surveillance system (NTOF) (Castillo et al., 1994). The differences between NTOF and CFOI, however, make it impossible to determine whether homicides at work have been increasing for this age group or were underreported in NTOF during the 1980s.
A strength of CFOI is its use of multiple data sources to develop a relatively complete count of fatalities. However, the final count still depends on the recognition, in individual cases, that the deceased was working at the time of the fatal incident. Young people are not typically thought of as workers. It is, therefore, reasonable to assume that deaths among children and adolescents who are fatally injured on the job may not be identified as work-related deaths