The persistence of inequality across Indian states

The persistence of regional inequalities in developing countries is well recognised to be of great concern. In this paper I track stochastic convergence in relative incomes for Indian states between 1960-2011 with the intent to identify high persistence and mean reversal. Traditional unit root tests suggest that shocks to relative incomes across the Indian states are permanent thus contradicting the stochastic convergence hypothesis. Interval estimates of the largest autoregressive root for the relative incomes of 15 Indian states are very wide. However, confidence interval estimates of the half life of the relative income shocks, that are robust to high persistence and small samples, reveal that in most cases they die out within 10 years, suggesting mean reversion for a large number of states. Finally, I estimate a fractionally integrated model and obtain mixed evidence of mean reversion in the data, with six out of the fifteen states experiencing mean reversion. In sum, while the evidence obtained does not support the stochastic convergence hypothesis, our findings reveal that the relative incomes have a relatively short half life and that some states relative incomes are mean-everting. This result is encouraging and in contrast to earlier studies which indicate long term divergence and polarisation (Bandyopadhyay 2011).