A consolidated database of police-reported motor vehicle traffic accidents in the United States for actuarial applications

# GENERAL INFORMATION

A consolidated database of police-reported motor vehicle traffic accidents in the United States for actuarial applications.

Author Information

    A. Author
        Name: Carlos Andrés Araiza Iturria
        Email: caraizai@uwaterloo.ca
    
      B. Co-author
        Name: Mary Hardy
        Email: mary.hardy@uwaterloo.ca

      C. Co-author
        Name: Paul Marriott
        Email: pmarriott@uwaterloo.ca
    
    Institution: University of Waterloo
        Address: 200 University Ave W, Waterloo, ON N2L 3G1

Funding granted by the Natural Sciences and Engineering Research Council of Canada. Hardy: RGPIN-2018-03754, Marriott: RGPIN-2020-04015.

# DATA & FILE OVERVIEW

The script to create the database was written with Python and to simplify its readability it was split into 4 different files available. The files do the following:

  1 FARS_GESCRSS_Construction.py : Download the annual NHSTA files and merge them into two databases, one for GES/CRSS and another one for FARS.

  2 FARS_GESCRSS_Filtering.py: Filter relevant MVTA for personal automobile insurance.

  3 FARS_GESCRSS_Standardizing.py : Standardize the NHTSA historical coding practice to that of 2019.

  4 FARS_GESCRSS_Sample.py : Pooling procedure to create the final database.

NOTE: The ‘Auxiliary files’ folder needs to be downloaded and the ‘.zip’ files inside need to be uncompressed for the code to work.

Estimated run time of the 4 codes:
~9 minutes
using a MacBook Pro (2020), M1, 8gb RAM.