Is income concentration in India growing? Two views have recently been put forward on this issue.
On the one hand, the research arm of the Congress
states that income inequality is rising rapidly (let’s call this the prosecution’s view). It uses data from the Central Board of Direct Taxes (CBDT) to make the following claim: between 2012-2013 and 2015-2016, the conventional Gini coefficient for India went up by 15 percentage points. It further states, “India’s top 5% today earn as much as the remaining 95% combined.” Read another way, this makes it seem as though India has grown swiftly more unequal under the Bharatiya Janata Party (BJP) government.
On the other hand, Surjit Bhalla, an economist associated with the Economic Advisory Council to the Prime Minister,
believes that there is no rapid rise in inequality (let’s call this the defence’s view). Such commentary is not new. But where lies the reality? How do we separate imagination from evidence? The basic problem is that in the Indian debate on inequality, each view picks a separate slice of data and makes its stance accordingly.
MYTH | Indian inequality has gone up by 15 percentage points between 2012-13 and 2015-16.
From the outset, those propagating the prosecution’s view picked income tax data which applies to a fraction of individuals and generalised those findings to the whole population. In 2012-13, 28.7 million individuals (2.28% of the population) filed tax paperwork. In 2015-16, the number was 40.6 million (3.10% of the population). The Congress research division conflated companies and other associations with individuals in their population estimates and thus produced a story of rising inequality. Just for the sake of comparison, the prosecution’s view reports 273 ‘people’ earning more than ₹500 crore in 2016. In fact, there was only one person who made the cut; the rest were companies.
Now, consider the evolution of the individual tax base between 2012-13 and 2015-16. The tax data speak one main fact — there has been a disproportionate growth in the total number of taxpayers. This growth has been predominantly driven by people with salaries between ₹1 lakh and ₹10 lakh a year. If the tax base were a population, then it is as if there has been an influx of relatively low-income persons into this category, thereby giving an image of increasing inequality within the population. This is shown in the figure. Such entry from the lower end could be because people have recently become employed or because inflation pushed up incomes into higher tax brackets. The data clearly do not reflect increasing income concentration within the rich.
REALITY | The Indian tax base is, in fact, a deeply unequal society split into the elites and the working class.
The Gini coefficient is a tool for population-level estimation of inequality. It is not suitable for fractions such as the tax base. So, what can be done? One straightforward approach is to calculate for each year the average income of the tax-filer and then see what percentage of people earn more or less than this measure of central tendency. I did this and unsurprisingly found a striking statistical regularity in the CBDT data.
The mathematical ‘laws’ to estimate the level of inequality change about two-three times the average income of the Indian tax base. The incomes of 94-96% of individuals in 2012-13 and 2015-16 can be explained accurately by a simple formula (an exponential distribution). Although this model implies some inequality, one is not likely to see extremes. In 2012-13, people who earned less than ₹10 lakh a year were part of this social class. The cut-off in 2015-16 was around ₹15 lakh. Scaled for each year’s income, tax-filers fall on almost the same fitted line. If one was to calculate the Gini coefficient for the working class, it would be unchanged for this three-year period.
Above this limit sat an elite class of taxpayers (top 3-6% of individual taxpayers), earning tens, hundreds and thousands of times the average income. A ‘power law’ formula explains this class with 99.9% precision and little change in inequality amongst elites for each year. In this group lie India’s well-paid professionals, managers, successful businesspersons, and speculators.
The disparity predicted by the appropriate mathematical formula depends on stock market returns and profit rates. The more modest working-class distribution, on the other hand, is closely determined by the level of incomes. In this three-year period, the income share of the elite class in the tax base fell from 32% to around 25%. Thus, the tax data alone do not present evidence of growing disparities within elites and the working class. They simply confirm a dual-class reality which is the more pressing long-term issue in terms of social mobility and income inequality.
The same data could predict two stories depending on who wields it because the reality of India is that inequality is omnipresent, but incomes are also rising.
As political history has taught us, had the surge in inequality been of the speed and scale incorrectly predicted, then social turbulence would have followed much more swiftly.
The main lesson from this data is that one has to be very careful when trying to compute inequality measures. This is exactly the kind of training social statisticians are equipped to handle. The same data could predict two stories depending on who wields it because the reality of India is that inequality is omnipresent, but incomes are also rising. A politically defensive view could, for example, state that inequality is falling because elites are lower in both number and income share in the transition from Congress to BJP rule. However, the Opposition could just as easily state that inflation and tax policy are penalising the lower-income classes.
While the rich reap the benefits of super tax rate ceilings, the burden of taxation will instead fall on the taxman’s less-fortunate new subjects. It is hoped that future debates are more prudent of these issues.
