Decoding 15th FC's estimate of mismatch in two sets of salary income data

The Commission asserts salary incomes shown by taxpayers are understated by 50%, finds scope to extract more direct tax by chasing evasion without changing tax rates; experts aren't convinced

Through a tucked away piece of calculation in the fat (1,427-page) report of the 15th Finance Commission, its chairman N K Singh has again fanned the topic of tax evasion. 

The Commission has matched the data on income tax returns for salaries with comparable numbers in National Accounts Statistics. It reaches the conclusion that salary incomes shown by taxpayers are understated by almost 50 per cent. It adds a caveat that the difference is not as stark as the data shows, yet argues that there is scope for unexplained large differences between the two data sets. “..considering that we have accounted for only organised sector incomes in this ratio, it is desirable that the income tax base is widened and deepened through simplification of deductions and exemptions”. Arguments on tax evasion and how to deal with them have pervaded the previous term of the Narendra Modi-led government, leading to the demonetisation exercise. It has welled up in previous governments too.  

Significantly, in Budget FY22, finance minister Nirmala Sitharaman has not narrowed the base of the direct tax. She has also refrained from expanding it.

In Annex 5.2 of Volume II of the Finance Commission report, the Commission has run three exercises on the possible dissonance between what tax data tells and what GDP data shows. Other than the one for salary income, it conducts one for non-salaried income. It has compared non-salary incomes in the household sector as reported under income tax returns, divided by operating surplus or mixed income of the self-employed in the household sector. Here too, it pastes a caveat saying full consistency between the national accounts and tax returns cannot be established in this segment owing to data limitations. “However, this ratio should improve considerably, given the bunching of incomes noticed in the tax avoidance bracket”. The level of difference it finds is even higher at 64.4 per cent.  The reference to tax evasion or tax avoidance by the Commission is pronounced in the exercise. The third of the exercise matches corporation tax returns against the operating surplus of the private corporate sector and public sector as reported in the National Accounts Statistics.

The study of tax data has become a contested territory worldwide, of late. Governments have become leery of offering tax cuts when faced with data of already persisting tax leakages. The Finance Commission data could add to it for India. It makes no bones about why it has studied the data. It presumes there is scope to extract more direct tax by chasing tax evasion without any changes in rates. “With improvements in tax information system, administration and policy, the direct tax to GDP ratio can be raised significantly,” it says. 

However, national accounts experts are not convinced the difference is as substantial as the Commission makes out. Deep Narayan Mukherjee, visiting faculty at IIM Calcutta, who has studied GDP data for several years, is one of them. “These are standard problems when one looks at sum totals and averages without explicitly adjusting for multiplier effect, tax bracket and distribution of earnings by tax liability,” he notes. As a thought experiment, he says, let’s assume everyone in India earns the same annual salary of Rs 2.5 lakh. If all the salary is paid through cheques, making the trail transparent, the total tax receipts shall be zero. Yet, it would show up in the GDP calculation and would be also great news for ensuring equality.

A somewhat similar argument was offered by Devendra Pant, chief economist and head of public finance at India Ratings. He also said if the salary distribution shows that an overwhelming majority have nil tax liability—a case could be of a company with just 10 workers, none of whom earn more than Rs 1.5 lakh per annum—the discrepancy would show up. 

Singh, however, has strong theoretical support from the work of OECD, the organisation of mostly developed countries, that has studied tax issues deeply. An OECD paper notes that tax returns data should broadly coincide with that of national accounts. If they do not, “the differences between the figures can be assumed to be a consequence of tax evasion”. The only mismatch the paper considers valid are those for statistical and legal causes such as bankruptcies, discharges and individual arrangements. In fact, it argues that comparing the two is “especially useful if the national accounts do not use the income tax statistics as a data source”. In principle, the difference between the national accounts and income tax statistics, according to the OECD research, should be an indication of the size of economic activity that is concealed for income tax reasons and that has been included in the national accounts. Budget FY22 for India estimates gross tax revenue receipts would climb to just 9.9 per cent of GDP, against 9.8 per cent in FY21.

For India, the ministry of statistics defines personal income, as shown in the GDP data, as a measure of the actual current income receipt of persons from all sources. “It is derived from private income by subtracting the savings of the private corporate sector and the corporation tax,” the ministy notes. It differs from private income in that it excludes the undistributed profits which accrue to the private sector, but are not received by persons. It also excludes the expenditure tax paid to government by the private corporate sector.

Right since Independence, successive governments have examined whether they were getting the tax due to them. An RBI analysis of each of those beginning, with the Mathai committee report of 1953, finds a common thread. Each noted that there are too many exemptions which diluted the tax levied. The Parthasarathy Shome committee in 1988 suggested the tax system was “lacking the design that would automatically yield higher tax revenue with growth in GDP”. If the rates do not track changes in income of those taxed, the government does not get a higher income share even when GDP climbs. According to Shome, to get the improvement in tax elasticity, there has to be a regular adjustment in rates on inflation and reasonable progressiveness in the system as a whole. Else, the level of tax evasion expands. A recent report on tax evasion was written by Sacchidananda Mukherjee and Kavitha Rao of NIPFP. The 2015 study concludes that the estimated share of unaccounted GDP in the economy is about 25.4 per cent, using FY12 for their estimates. 

The 15th Finance Commission under N K Singh has not dived into estimates of unaccounted income. But it has suggested the need to pore over the tax data deeply and widen the net from the centre to the municipalities. It echoes the view of OECD where it says the data from tax records “should approximate each other with appropriate adjustments”. Mukherjee, however, offers his caveats. There is a need to look beyond sum totals and averages in making those assumptions.