STM Article Repository

This paper models and estimates disaster occurrence in Nigeria using time-series data covering the period of 1970 to 2012.An autoregressive model was employed and the effects of stochastic shocks of each of the endogenous variables were explored using error correction model (ECM). The static long run estimate shows that D(LOG(UBR)) and TER are statistically significant. The multivariate regression results of the over-parameterized error correction model indicate that DIS(-1), D(LOG(UBR)), D(TER) D(LOG(UBR-1)) and D(TER(-1)) are statistically significant while other variables were not. The short run result reveals thatDIS(-1), D(LOG(UBR)) D(TER), D(LOG(UBR-1)) and D(TER(-1)) have significant effect on the occurrence of disaster in Nigeria.Among other recommendations, Nigerian government should develop a comprehensive disaster risk management plan that goes beyond a predominantly response-focused approach to disaster planning by incorporating it in the national planning process.