A large number of studies in the empirical growth literature in India have studied the growth dynamics of the states by dividing the period into the pre- and post-reform periods (e.g. Bhattacharya and Sakthivel, 2004; Kar and Sakthivel, 2007). The authors argued that it is difficult to think of the dynamics in the pre- and post-reform periods as resulting from the same transition process. While the most of these studies clearly identify different dynamics of the per capita income distribution in the pre- and postreform period, the method used to divide the entire sample into parts was based on an a priori judgment regarding the historical timing of policy reforms (i.e. 1991) and no statistical method was applied to identify the sub-periods corresponding to different transition processes. Moreover, the historical record of what we know regarding policy reforms in India suggests that the process of gradual liberalization of the economy may have started almost a decade before the comprehensive reforms in 1991 (Panagariya, 2003). Thus, the effect of liberalization in terms of increase in regional polarization (or inequality) may have begun much before 1991. There is also the possibility that the observed trend in the post-reform period might actually date to before 1991 and therefore be related to factors other than the comprehensive economic reforms.