A Purple Passage to India

How Thomas Piketty\’s selective use of tax data exaggerates India\’s inequalities
My last Swaminomics column (`Fantastically Rising Inequality? Piketty Has Got It Wrong\’, TOI, goo.gl8X7GpB) addressed the claim of Lucas Chancel and Thomas Piketty , guru of inequality , that economic reform had led India to fantastic, unprecedented levels of inequality (`Indian Income Inequality, 1922-2014: From British Raj to Billionaire Raj?\’, goo.glgbPEde).Chancel and Piketty use incometax data for the richest 5%, but adjusted consumption survey data for others. He implies that the rich lie outrageously to surveyors about consumption, but are far more truthful to the taxman -something many will find hilarious.

Tax Data is a Bit Rich

Piketty views tax data as a novel way of measuring inequality , better than household data from surveys. He is wrong. Tax data have so many distortions that they mislead rather than enlighten. They both understate and overstate incomes, mostly the latter. Household surveys cover only humans. However, tax laws cover artificial entities like corporations and trusts. If a businessman owns 60% of a large company , his tax returns don\’t include a 60% share of company profits. They include only dividends declared by the company , which may be zero (some of the biggest global companies never declare dividends). Aga in, a businessman may control several trusts and societies, but their income is treated as separate by the taxman, not clubbed with the businessman\’s.

The businessman can keep all his profits in a corporation (avoiding dividends) and pay himself no salary , ending with virtually zero income as defined by the taxman. This was, indeed, the strategy of the rich when Indira Gandhi raised the top incometax rate to 97.75%. Piketty\’s calculations based on tax data show a huge improvement in equality in the `garibi hatao\’ era. But this contain illusions. The rich kept much money black or as undistributed corporate income. The bottom line: some tax data understate the income of the rich and, hence, inequality .

But in other ways, tax data grossly exaggerate inequality . Household surveys of the National Sample Survey Office (NSSO) or the India Human Development Survey (IHDS) estimate the consumption or income of households, and divide this by the number of family members to get a percapita figure. But income-tax data relate strictly to individuals and tell you nothing about family income.

If a rich man has a wife and collegegoing kids who earn nothing, the wife and kids will show up in tax data as paupers, even if they own Mercedes cars and go to London for weekend shopping. By not sharing family income among all members, tax data grossly exaggerate the living standard of the businessman and grossly underestimate that of his wife and children. This methodology produces a huge but artificial inequality .

Abetter procedure would be to club all family income and divide it by the number of family members, as is done in household surveys. By not using divided data for the rich but using divided data for the non-rich, Piketty greatly inflates the gap between them.

That\’s not all. Piketty estimates that the richest 1% have 22% of all in come. He doesn\’t say what the share of the richest 5% is. But let\’s conser vatively take it to be twice as much, or 56%. If taxed at the top rate of 35%, this should yield tax revenue of around 20% of GDP . In fact, total in come-tax revenue is less than 2% of GDP . Where has the rest disappeared?

Trust in Exemptions

In tax exemptions, of course. Most exemptions -e.g., for long-term sav ings or provident fund contributions -are trivial for the rich. A bigger ex emption is for donations to charita ble trusts, but is limited to 10% of gr oss income. The most important tax breaks by far are for capital gains, which have often been taxed at con cessional rates.

Long-term capital gains became tax-free in 2004 for sales through stock exchanges. This induced huge selling and buying. The Sensex has gone from 1,000 to 31,000 since 1991. Land values have risen as fast, so, all farmers ha ve gained too.

All capital gains (including exempt ones) are included in tax returns.

The taxman defines capital gains as income. But GDP or national income as defined in economics excludes capital gains, since these don\’t reflect value addition: they merely reflect the churning of assets.

Consider two rich men. One does not churn his portfolio at all. The second churns his portfolio every month. At the year of the year, the value of the two portfolios may be the same, but the first portfolio will have zero capital gains, while the second will show huge capital gains. This creates an illusion of a huge gap between the two, when their year-end asset values are identical.

Many investment gurus now say that an unchurned portfolio like an exchange-traded fund will intrinsically do as well or better as a churned one. So, there is no advantage in churning. Yet, the taxman treats the churning as massive income. GDP calculations do not, and they represent reality far better.

In sum, the use of tax data instead of household survey data produces enormous illusory inequalities. Using tax data for the rich and survey data for the non-rich compounds the illusion. Piketty\’s admirers think he has uncovered a treasure trove in tax data. Alas, all that glitters is not gold.

Leave a Comment

Your email address will not be published. Required fields are marked *