Fatality Analysis Reporting System (FARS)
Source
The Fatality Analysis Reporting System (FARS) data is provided by the National Highway Traffic Safety Administration (NHTSA)
This data is available for download here:
Documention about this data is provided in the "Fatality Analysis Reporting System Analytical User’s Manual, 1975-2021"
(including, in particular, the ability to lookup state codes)
Description of the Data
According to the NHTSA website (see) www.nhtsa.gov/resources-guide/glossary-highway-safety-terms-and-definitions
this is (quote) "The nationwide census providing public yearly data regarding fatal injuries suffered in motor vehicle traffic crashes, as published by NHTSA."
We can download the files as follows:
for i=1975; i⇐2021; i++; do wget static.nhtsa.gov/nhtsa/downloads/FARS/${i}/National/FARS${i}NationalCSV.zip; done
for i=1982; i⇐2021; i++; do wget static.nhtsa.gov/nhtsa/downloads/FARS/${i}/National/FARS${i}NationalAuxiliaryCSV.zip; done
The 1996 Auxiliary file is named incorrectly (CVS instead of CSV), so we need to download it separately:
(Dr Ward reported this to the NHTSA)
Now we can unzip the files and put them into directories:
for i=1975; i⇐2005; i++; do unzip FARS${i}NationalCSV.zip; mkdir ${i}; mv *.[Cc][Ss][Vv] ${i}; done
for i=2006; i⇐2015; i++; do unzip FARS${i}NationalCSV.zip; mkdir ${i}; mv FARS${i}NationalCSV/*.[Cc][Ss][Vv] ${i}; rmdir FARS${i}NationalCSV; done
for i=2016; i⇐2019; i++; do unzip FARS${i}NationalCSV.zip; mkdir ${i}; mv *.[Cc][Ss][Vv] ${i}; done
for i=2020; i⇐2021; i++; do unzip FARS${i}NationalCSV.zip; mkdir ${i}; mv FARS${i}NationalCSV/*.[Cc][Ss][Vv] ${i}; rmdir FARS${i}NationalCSV; done
for i=1982; i⇐2019; i++; do unzip FARS${i}NationalAuxiliary*.zip; mv *.[Cc][Ss][Vv] ${i}; done
for i=2020; i⇐2021; i++; do unzip FARS${i}NationalAuxiliaryCSV.zip; mv FARS${i}NationalAuxiliaryCSV/*.[Cc][Ss][Vv] ${i}; rmdir FARS${i}NationalAuxiliaryCSV; done
rm FARS*.zip
Transformations to the original data source
We build the 7581.csv file:
head -n1 1975/ACCIDENT.CSV >7581.csv for i=1975;i⇐1981;i++; do sed -e 1d ${i}/ACCIDENT.CSV; done >>7581.csv
and we build the states.csv file manually from the pdf file:
If the pdf file is not available, it is in the Wayback Machine: