By Nana Dadzie Ghansah
In spite of the existence of federal databases to record homicides, it has been known for decades that while the reporting of civilian-on-civilian ones is close to 99% accurate, the police killings of civilians are grossly underreported.
A study by Sherman in 1979 found killings of civilians by law-enforcement officers were underreported by about 50% when they looked at data from 1970–76. Loftin et al looked at data from 1976–88 and found significant underreporting too.
In both studies, two of the main reasons for this problem were failure to report law enforcement killings by states and municipalities and the false designation of causes of death on death certificates by coroners and medical examiners. In the latter instance, such cases were often reported as homicides, giving the impression that they were civilian-to-civilian homicides.
The two main databases are CDC’s National Vital Statistics System (NVSS), that was established in 1890, and the FBI’s Uniform Crime Reporting Program (UCR) that was established in 1908. In all, the NVSS database gets more reports from all states and municipalities versus the UCR program.
In 1949, the ICD coding system (a system used internationally to code for diseases) introduced a code for law-enforcement-related deaths. It was called “injury by intervention of police”.
It was defined as “injuries inflicted by the police or other law-enforcing agents, including military on duty, in the course of arresting or attempting to arrest lawbreakers, suppressing disturbances, maintaining order, and other legal action.”
It was later renamed “legal intervention”.
Even with this designation, there was significant underreporting. As far back as the 1950s, the role coroners and medical examiners play in this underreporting was known. Though they should be independent, most coroners and medical examiners are not. Thus, they often would state a cause of death that suited the official line.