According to this Aviation Safety Network article and dataset there was a significant drop in the number of airline accidents in the early 80s comparing to the years before - from ~50 per year down to ~30 per year. However, in the late 80s and the 90s this number increased back to ~50 and stayed at this level until early 2000s. Is there any explanation for the sharp increase of the accidents in the late 80s and the 90s? I see the number of flights increased linearly throughout the years, so this isn't the factor that explains the accident statistics.

  • 1
    $\begingroup$ Just the pure number of accidents isn't a good indicator of safety. Accident rates have decreased steadily even through the 90's and 2000's. The better statistic would be to look at number of accidents per million flights or per million hours. That will correlate with the number of total flights. $\endgroup$ – Ron Beyer May 11 '19 at 14:37
  • $\begingroup$ Two factors come to mind. The 90s saw the rise of discount carriers, and Regional Jets, both involving sketchier maintenance and less experienced crews compared to mainline. $\endgroup$ – John K May 11 '19 at 18:05
  • $\begingroup$ The ASN dataset varies from the traditional definition of aviation accident by including intentional acts, namely hijackings, bombings and shootdowns. Therefore you have to consider the political situation at the time, i.e. Cold War, Soviet-Afghan war, wars after fall of the Soviet Union, etc. It also is worldwide, many statistics specifically exclude Soviet-made aircraft and only look at Western-built models. There was a mass retirement of Soviet types in the early 2000s. $\endgroup$ – user71659 May 12 '19 at 0:18
  • $\begingroup$ Also you're looking at the number of 14 passenger+ planes crashed and are dividing by number of air carrier flights. This misses trends in corporate aviation, many charter flights/cargo charters, air taxi operations, etc. $\endgroup$ – user71659 May 12 '19 at 0:39