Monday, January 30, 2017

Income Tax Reform II: A feasible Negative Income tax/Net Income Transfer (NIT)

with Surjit S. Bhalla

    In our Jan 14 article, Towards an Income Tax Revolution, we had outlined what we believe is a major opportunity for the Modi government to bring in a structural reform to our income tax structure. What we suggested, and expand today, is an integrated approach to both taxation and redistribution. The latter can take, and has taken, many forms over the years. Technology (Aadhar), and political will (demonetization) allows India to finally begin to think BIG, and efficiently, in terms of redistributive policies.

But the two are necessary, not sufficient, for successful implementation. There is money involved, and the last thing the Modi government should do is to revert back to the bad old days of “in the name of the poor” corrupt policies like PDS, NREGA, loan waivers, fertilizer subsidies, etc. There is buzz around that the Budget might contain a Basic Income policy. One of us (Surjit) had offered a discussion of how Rs. 1000 per person per month could be transferred to every one in the bottom half of the population, and that the cost would be Rs. 5 lac crore (trillion). As Subhashis Banerjee has correctly pointed out (A Contentious Proposal, Indian Express, Jan 14, 2017), the calculations are incorrect if transfers have to be made to the bottom 50 %; the calculations are correct if transfers are to be made to the bottom 20 %, the percentage poor according to the upwardly adjusted poverty line of Rs. 1525 per person per month. The Tendulkar poverty line for 2016-17 is a lower Rs. 1250 per person per month. The error is unfortunate, and regretted.

Negative Income Tax: We are proposing, in lieu of basic income for all or even for a targeted population, that each non-farmer worker (hereafter just worker) receive a transfer upto a maximum of Rs. 15000 a year. The negative income tax is obtained according to the formula 15000 – 0.05*income of worker (upto an income of Rs. 3 lacs). In the aggregate, for about 240 million workers (72.7 % of the worker population) the total outgo for NIT is Rs. 1.70 lakh crores or an average of Rs. 7100 per earner recipient of NIT.
The Tendulkar poverty line, for a family of five, in 2016-17 prices, is Rs. 75000 a year. According to NSS data, the average number of earners in a poor family are close to 3.1; even for a two earner household, the transfer will be equal to an average of Rs. 16000  (Rs. 8000 per earner). The average earnings for the bottom 25 % of earners is Rs. 90,000. Thus for these 25 %, the take home post NIT income is 2*(90000+8000) or Rs. 196,000 a year – i.e. well above the five person poverty level income of Rs. 75000.

Proposed Income Tax system: In several comments received over our previous proposal [Towards an Income Tax Revolution, January 14], the one overwhelming response was that while the flat tax rate of 12 % was appreciated, most felt that it was politically unrealistic in the Indian context. Hence, we now propose a new revenue neutral system. This will be a two rate structure, 10 and 20 %. For the income range Rs. 3 to Rs. 6 lacs, the tax rate is 10 %; for those above Rs. 6 lacs, the tax rate is 20 %. Details are presented in the Table. Note that there is a loss of Rs. 1181 billion in the new system with compliance unchanged. What is noteworthy about tax compliance in India in 2013-14 is that while tax compliance for those earning less than Rs. 10 lacs was close to 30 %, the rate for those earning more than Rs. 10 lacs is as low as 20 %. As the PM had mentioned, in 2013-14, there were only 2.4 million earners with income above Rs. 10 lacs – our NSS distribution for 2016-17 suggests that the number of workers in this category is around 12 million.
Why should our numbers/estimates be believed? For two reasons: the mean income of our constructed distribution matches the mean income as obtained from national accounts (this is by construction!). Second, and more importantly, official tax receipts for each of the years 2011/12 to 2015/16 broadly matches (within 10 percent)  the receipts obtained via our synthetic distribution. Note that compliance rates are estimated via the number of official taxpayers (MOF data, 2013/14) and the taxpayers as per the constructed distribution.
There are significant gains for all workers in the new system. A person earning Rs. 4 lacs has her tax liability reduced by 35 %; for those earning Rs. 8 lacs, the tax liability is reduced by 20 %, and for those earning Rs. 16 lacs, the reduction is 27 %. A minimum set of compliance changes assumed by us are as follows: all compliance in the new tax regime is estimated to be 40 percent.

With this change in compliance, the new system will still entail a loss of Rs. 640 billion. This will still mean that only 4 out of 10 workers actually pay taxes in India. In the US, 85 of every 100 people pay taxes. We will still have a very long way to go – but we can begin to get there!
This loss can be made up either by another few percentage points increase in compliance (even with the reduced tax rates, we are assuming only a 40 % compliance rate) or by removing all budgeted tax incentives. In 2015-16, total incentives for tax payers was Rs. 550 billion, with Section 80C (mutual fund investments for rich taxpayers!) accounting for a fat Rs. 450 billion. There is no need for this payment to the rich in the new system – hence, removal of this exemption, is able to reduce the tax loss in the new system to practically zero; actually, a small loss of 90 billion.

What remains is the financing of the NIT. As outlined above, the cost of NIT is Rs. 1.7 lakh crores. If NIT is part of official tax policy, then there seems to be precious little need for anti-poverty programs like PDS or NREGA. These programs are not only costly, but also involve a lot of corruption -as most analysts (and us), and politicians, have pointed out over the years. Starting with Rajiv Gandhi, who in 1985 stated that only 15 % of the money meant for the poor actually reached the poor! Currently, total expenditure on these two wasteful programs is Rs. 1.75 lakh crores. If the government decides to eliminate other in the name of the poor (but not benefitting the poor) programs, then the net income transfer (NIT), per worker, can be increased.

With the adoption of this new thinking, and new anti-poverty policies, absolute poverty will be zero in India, and according to a much higher poverty line. And with no extra cost. The only assumption we are making in the tax reform is that tax compliance rates have to increase to 40 percent. The de-monetization policy has provided the stick for increased tax compliance – our proposed policy provides a much needed carrot.

Revenue Neutral - a Two (10 & 20 %) Income Tax Policy
Existing Tax
New Tax
Reduction in Tax Paid
Number of Workers
Revenue Loss
With Greater Compliance
(in lacs)
(At upper end of range)
(in per cent)
(in million)
(per worker)
(in billions)
(in billions)

3 - 4
4 - 8
8 - 16
> 16


Note: Proposed Tax Rates : 0-3 lacs : No Tax; 3-6 lacs: 10%; and above 6 lacs: 20%; average income >16 lacs estimated to be Rs. 28 lacs
1. Reduction = 100 x ("Existing Tax" - "New Tax")/"Existing Tax"
2. Total Loss = Number of Workers x Average Loss per worker
3. Tax revenue loss with greater compliance has been calculated with compliance increase of 10 % points for ranges 3-4 lacs and 4-8 lacs, while a compliance increase of 20 % points for the range 8 -16 lacs and 16 lacs & above.
A version of this article appeared in the  Indian Express, of  January 19, 2017 under the banner, " Taxing your way to popularity."