The Unfortunate Nature of Road Accidents and Fatalities

“Accidents are not accidents but precise arrivals at the wrong time.” (Stojanovic,1991) With these resounding words being read aloud, it becomes easier for one to truly grasp the unappealing truth that road fatalities are as a result of precise arrivals at the most inopportune time regardless of age, religious beliefs or creed. Road accidents being the eighth leading cause of deaths worldwide is a feat that brokers minimal reactions from the populace unless it is that of astonishment. The number of fatalities resulting from them annually is even more horrendous and it is with this desire that the researchers deem the intervention of an econometric model accentuating the primary factors for these fatalities over a 35 year span in the United States of America, a territory with approximately 327.2 million people that equivocally loses about 40,000 people each year. What ramifications could this have on Guyana’s population over time? Is there even a correlation between these two separate random variables or is it just as fleeting as a gust of wind?

A multiple regression analysis will be explored with our dependent variable being the number of fatalities and it’s independent random variables or explanatory variables being the number of Driving Under the Influence cases, vehicle type in terms of their ability to protect their occupants, gender, age, successfully licensed drivers, speeding, lack of safety gears e.g. seat belt and helmet usage and the distance travelled. To be estimated also, is the sample covariance which illuminates the type of relationship between these variables and the dependence on the number of road fatalities. The time series data ranges from 1982 to 2017. Time series is best defined as “a set of regular time-ordered observations of a quantitative characteristic of an individual or collective phenomenon taken at successive, in most cases equidistant, periods/points in time” (OECD, 2005).

The error terms accounted for in this particular scenario are the times of day that are most likely to be accident prone, poor maintenance of vehicles, inattentiveness, aggressive driving and “poor motor coordination skills” (reaction time) which are almost impossible to decipher, hence their categorization as a continuous random variable as well as infrastructural properties that make road fatalities even more prominent.

The multivariate composition will comprise of the aforementioned variables with their respective measurements e.g. DUIs will be measured by every 100,000 deaths attributed to high alcohol consumption, vehicle type will be measured per 100,000 road users that fell prey to road fatalities, distance travelled will be measured in miles per hour and successfully licensed drivers will be measured by every 1000 driver that died.

The econometric technique that will assume dominance in the model created is the Ordinary Least Square (OLS) method whose chief initiative is to reduce the sum of residual squares derived from the observations. The assumptions of the OLS will be recurrent throughout the model to maintain all the relevant properties of a good estimator i.e unbiasedness, consistency and efficiency.