Long term dynamics of poverty transitions in India

Abstract: We model the dynamics of poverty using a stochastic model of Geometric Brownian Motion with reallocation (RGBM) and explore both transient and persistent poverty over 1952-2006. We find that annual transitions in and out of poverty are common and show a rising trend, with the rise largely being driven by transitions out of poverty. Despite this promising trend, even toward the end of the time frame, there is a non-trivial proportion of individuals still transitioning annually into poverty, indicative of the economic fragility of those near the poverty line. We also find that there is still a marked persistence of poverty over time, though the probability of poverty persistence is slowly declining. Particularly concerning in this context are the poverty trajectories of those at the very bottom of the income distribution. The choice of poverty line appears to impact the dynamics, with higher poverty lines corresponding to lower transitions and higher persistence probabilities. The distinct nature of emergent transient and persistence dynamics suggests that the approaches to counter these phenomena need to be different, possibly incorporating both missing financial markets and state action.