Statistical analysis of landmine fatalities in Kurdistan

This study analyzes mines and unexploded ordnance (UXO) victim data in the Kurdistan Region of Iraq during the period of the 1960s to 2005. In addition to descriptive analysis of the data, we use regression analysis to identify the determinants of the probability of getting killed by a mine or UXO and to estimate the determinants impacts. The rates of killed and injured victims are explained using a set of socioeconomic variables. Since the data is repeated cross sections in which individuals are observed when they are subjected to accident and in order to account for the dynamic aspect of the process and heterogeneity by location, we create pseudo panel data where districts are observed over the entire study period forming almost a balanced panel data. This allows for analysis of both heterogeneity and dynamics across locations and over time, as well as to control for unobserved location and time effects. The result can be useful in planning, monitoring, and resource allocation for mine action and labor market programmes and rehabilitation activities.