Is Thomas Piketty right about inequality in India?
Three years after writing a best-selling book on the growing problem of inequality in the Western world, the French economist Thomas Piketty has turned his attention to inequality in the developing world. In a recent research paper co-authored with Lucas Chancel of the Paris School of Economics, Piketty estimates that the share of the top 1% in India’s income pie is higher than ever before.
Based on their analysis of historical data—from tax sources, surveys and national accounts statistics—the duo shows that the share of the top percentile (or top 1%) in India’s national income pie is at its highest level (22%) since 1922. This share had declined from 21% in the late 1930s to less than 6% in the early 1980s. They also argue that the period between 1951 and 1980 witnessed a decline in the share of the super-rich in the national income pie, and a rise in the share of the bottom half of India’s population. Between 1980 and 2014, the situation has reversed, they argue.
The research paper thus presents an alarming picture of rising inequality in India during the precise period when India’s growth engine picked up pace. However, a Mint analysis suggests that Piketty’s conclusions regarding inequality may be exaggerated. While inequality may have risen over the past few decades, the alarming picture presented by Piketty may not be accurate.
The problem with Piketty’s estimates lies in his assumptions. Given that survey data often understates the extent of true incomes, Piketty relies on a combination of survey and tax data to estimate India’s income distribution. In doing so, he assumes that up to the 90th percentile, i.e. for the bottom 90% of the population, survey data reflect actual income levels. For the richest 5% (those above the 95th percentile), he uses income data imputed from tax filings. Between the 90th and the 95th percentile, he uses a combination of tax and survey sources. The upper percentile threshold changes slightly across years.
What this means is that the income estimates of the top end of the income distribution are fundamentally different from the estimates for the rest of the population. It also means that Piketty rules out—by assumption—any under-estimation (or under-reporting) of incomes by the bottom 90% of the population, an assumption that does not seem to draw support either from theory or common sense. While it may be fair to assume that the rich have a greater incentive to under-report incomes and may be under-sampled in household surveys, it is extraordinary to assume that only the rich have any incentive to under-report consumption and incomes in household surveys.
The second problem in the assumption is that it does not take into account changes in tax administration that may have led taxmen to measure and assess top-income groups better than before. This seems to have led to an exaggeration of the income estimates of top earners for the most recent years.
Piketty reports 54 variations in assumptions while reporting his results but all of those 54 variations in estimation strategy suffer from these fundamental problems. Thus, Piketty and Chancel under-estimate the incomes of the non-rich, and over-estimate the incomes of the rich. Both these effects have tended to exaggerate the estimates of inequality that they generate.
Does that mean we can dismiss the issue of inequality in India?
The evidence so far does not offer any room for such complacency. Even household surveys—which are likely to under-estimate inequality—present a worrying picture of inequality in India.
The extent of inequality reported by different household surveys differs. While official estimates of inequality present a benign picture of inequality in the country, income surveys such as the India Human Development Survey present a more disturbing picture suggesting that income inequality is far higher than what official estimates suggest. Another survey conducted more recently by the People Research on India’s Consumer Economy (PRICE) suggests that income inequality may be lower than what the IHDS suggests.
Unless there is an attempt to build an official and comprehensive income database for the country, based on officially-sponsored surveys or administrative data, or some combination of the two, it is unlikely that distributional issues will receive the attention they deserve.
However, there already exists one official database that provides some insights on trends in inequality in India: the all-India debt and investment surveys conducted by the National Sample Survey Office (NSSO). This does not provide data on incomes but provides data on wealth, which to some extent is easier to capture in a survey. In one of the few attempts to systematically analyze the level of wealth inequality in India based on that database, the economists Ishan Anand and Anjana Thampi showed in a 2016 research paper that the extent of concentration in wealth has increased sharply in India since the early 1990s.
Anand and Thampi’s estimates show that the top 1% in India accounted for nearly 28% of the country’s wealth in 2012, an increase of 11 percentage points since 1991. The share of the bottom 40% in the country’s wealth declined from 5 % to less than 4% over the same period. The wealth gap in the country is real, and has been increasing but it seems to be lower than what estimates from investment banks such as Credit Suisse shows—their estimates based on imputations from survey data suggests that the top 1% account for more than half of the country’s wealth.
Nonetheless the striking differences in wealth and income across social groups in India makes India’s inequality much more deep and layered compared to what aggregate measures suggest. Anand and Thampi show that the share of scheduled castes (SCs), scheduled tribes (STs) and other backward classes (OBCs) in national wealth are not only lower relative to their population shares, but have also deteriorated since 1991.
To sum up, the issue raised by Piketty and Chancel is indeed important but their estimates appear off the mark. We need both better data and better analysis to deal with the complex issue of inequality in India.