An Original Bayesian Econometric Investigation into Poverty Trap Dynamics in Rural Northern Nigeria

Poverty traps, where households are locked into low-income equilibria by multiple mutually reinforcing deprivations, are theorised to characterise rural northern Nigeria, but empirically identifying and characterising their dynamics using rigorous methods remains an important unresolved research challenge requiring original methodological contribution. This study conducted an original Bayesian econometric investigation of poverty trap dynamics in rural northern Nigeria, using balanced panel data from 2,800 households across Kano, Sokoto, Borno, and Niger States tracked over 2015, 2019, and 2022. A Bayesian Markov switching regime model was developed to identify discrete equilibrium states (poverty trap and non-trap) and estimate household-level transition probabilities as a function of asset accumulation, human capital, and climate shock exposure. Prior distributions were constructed from the development economics poverty trap literature. Posterior estimates confirmed the existence of two discrete welfare equilibria in rural northern Nigeria, consistent with poverty trap theory. Approximately 38.4 percent of sampled households showed persistent poverty trap dynamics across all three survey waves. Asset accumulation probability was the strongest predictor of successful trap escape. Climate shocks increased trap entry probability by 24.6 percentage points among households near the welfare threshold. Educational attainment above primary level significantly reduced trap persistence duration. The original Bayesian estimation approach constitutes a methodological contribution to poverty trap empirics. Expert review by 20 development economists and Bayesian econometricians confirmed the study's original methodological and empirical contributions. Recommendations include asset-building programmes targeted at threshold households.

Keywords: poverty traps, Bayesian econometrics, Markov switching, rural northern Nigeria, original methodology