INCOME AND WEALTH INEQUALITY IN INDIA, 1922-2023: - THE RISE OF THE BILLIONAIRE RAJ

First, while there is consistency in the use of the NSSO AIDIS across all papers, there is irregularity in terms of the use of rich lists, and when used, in terms of the methodology used to combine the two sources. [...] By comparing the total filers in the tax tabulations with the total effective taxpayers in the ITSD, we get an estimate of the non-filers which are missing from the tax tabulations which we use to 11 the adult (20+) population to define the fractiles (ranks) corresponding to the respective income thresholds in the tax data based on the number of returns in each bracket. [...] After more-or- less uninterrupted execution and release of the CES between 1951-2011, the Government of India suppressed the report and micro-data of the 2017-18 CES round on rather unqualified grounds of poor data quality.17 An analysis of the numbers in a leaked summary of the report suggests that the consumption-levels in real-terms may have fallen across the distribution (Subramanian, 2019b). [...] In the log-linear method, we regress the log-wealth on the log of the normalized rank and estimate 𝛼𝑙 using ordinary least squares as the slope of the line of best-fit.27 On the other hand, the constant Pareto method assumes a constant inverted Pareto coefficient, leading to 𝛼𝑐 = 𝑤/(𝑤 − 𝑤0), where 𝑤 is the average wealth of the rich list individuals. [...] Since the issue of non-coverage of the rich worsened in the 2018 AIDIS round as described above, we use 𝑝0 = 0.995, implying that we assume the survey is non-representative for the top 0.5% of the population.