摘要:The paper attempts to investigate into the prevalent notion that higher speed limits lead to increase in fatalities using Vector Auto Regressive (VAR) model for bivariate time series. The paper used Fatality Analysis Reporting System (FARS) time series data set from 1975 to 2010. The paper is the first one to make use of Vector Autoregressive Model (VAR) on this time series dataset to study the potential dual causality between speed and fatalities. The results of the paper partially support the causation between speed limit and fatalities, thus, complementing the results of the existing literature relating speed limits and fatalities such as “Influence of speed limits on roadway safety in Indiana” [1] , “Driver Speed Behavior on US Streets” [2] , and “Speed Limit Laws in America: Economics, Politics and Geography” [3] . However, instead of determining causation through various regression models, the paper establishes the causality by inducing a speed shock and analyses its impact upon fatalities. The paper concludes that in states with higher per capita automobiles and limited public transportation facilities such as Georgia, a sudden increase in speed limit has an immediate and higher impact upon accident related deaths, whereas, in states with lower per capita automobiles and higher public transportation services such as New York, the sudden increase of speed limit has minimal impact on accident related fatalities.