This paper examines the geographical patterns of distribution in burglary crimes (residential and nonresidential) in Nigeria for the incidence year 2017. Our novel approach integrates spatial structures into the traditional regression framework and evaluates the spatial disparities in neighborhood effects and the socio-economic characteristics of burglary crimes at sub-national levels. The study proposed four spatially varying models to demonstrate the importance of incorporating spatial dependence components in the models. The determinant factors included in the model are; the unemployment rate, education index, and poverty indices, alongside demographic variables, to understand crime patterns. A Bayesian analysis was performed via Markov chain Monte Carlo simulations to estimate the model parameters. The analysis revealed that the proportional contribution due to neighborhood (clustering) effect was estimated as 24.7% for the house-breaking and the estimated neighborhood contribution as 29.0% for the store-breaking occurrence. This approach demonstrates superiority in model performance, as indicated by the lowest Deviance Information Criterion. Findings reveal negative associations between burglary and multi-dimensional poverty, while young male adults show a positive relationship with store-breaking incidents. Hot spot areas and spatial variations in crime patterns are identified, offering insights for criminologists and informing policing strategies for effective crime prevention.