|M.Sc Thesis||Department of Industrial Engineering and Management|
|Supervisor:||Prof. Avner Bar Ilan|
|Full Thesis text - in Hebrew|
Road accidents are inevitable and force a heavy burden on individuals and society in general. The records of the transportation ministry - the department of economics and planning, show that the accumulated cost of road accidents between the years 2000-2002 to the Israeli economy amounted to approximately 12.6 milliard dollars a year, which are 2.5% of the GDP. For individuals, the burden comes in terms of physical injury, financial damage and time loss. Individuals not involved in road accidents also suffer a certain cost, due to increased fees of the insurance premium.
The goals of this work are as follows.
This essay is based upon the data of 29,198 random micro observations of licensed drivers who committed/didn’t commit red light/stop sign violation during the years 1992-1999.In order to identify the characteristics of the involved drivers, we use Probit type model in which the explained parameter is the probability to commit red light/stop sign violation. The explaining parameters are: age of the driver, years of driving experience, driver’s criminal record, sex, religion and whether he is a new immigrant.
The regression results show that:
Following the identification of the characteristics, we calculate the insurance premium for the various drivers while the basic principle is that each insured driver will pay insurance premium according to the characteristics that influence his insurance risk. The more characteristics that increase the probability of the driver committing traffic violations, the more dangerous the driver, and therefore the higher insurance premium he will be required to pay.
Low R2 values indicate relatively poor fit. Other variables that influence the probability of committing traffic violations, such as income level, residential area, driving distances, vehicle manufacture and age should also be taken into account. Future research should attempt on getting and analyzing this additional information.