Spatial Analysis of Distribution of Inmates in Nigeria Prisons

Abstract

The number of prison facilities available in Nigeria differs from state to state as well as across the regions. Hence the distribution of the inmates across these correctional facilities differs. This study investigates the distribution of the inmates across Nigeria correctional facilities (prisons) and the likely factors that may influence the distribution considering the spatial effect of the  distribution. The considered factors are; year, month, state, region, gender and state population. Ordinary Least Square (OLS) was used to obtain the multiple regression model and Moran’s I test of spatial autocorrelation showed that there is presence of spatial autocorrelation in the distribution of inmates across Nigeria prisons. The Spatial Autocorrelation Regression Model (SARM) was therefore adopted to fit the regression model. While OLS approach indicated that all factors are significant except state, the SARM approach found both state and region not to be significant among the factors considered in the model. The SARM did not only give good compact spatial distribution of the inmates across Nigeria prisons it also gave a better predicted (forecast) values of inmates’ distribution than the OLS approach.