GST Compensation: Government Not In Breach Of Promise That Was Not Made, Says TV Somanathan - Exclusive

But the centre recognises the need for states to get adequate funding this year, says Somanathan.

The North Block of the Central Secretariat building, which houses the Ministries of Finance and Home Affairs, stands in New Delhi, India. (Photographer: Prashanth Vishwanathan/Bloomberg)
The North Block of the Central Secretariat building, which houses the Ministries of Finance and Home Affairs, stands in New Delhi, India. (Photographer: Prashanth Vishwanathan/Bloomberg)

The central government is not in breach of any promise made to states as part of the agreement on compensation for shortfall in revenue on account of the goods and services tax. That’s according to the government’s Expenditure Secretary TV Somanathan.

Speaking to BloombergQuint, Somanathan said discussions that took place in earlier GST Council meetings had debated a hypothetical shortfall in compensation cess collections. “Then chairperson Mr. Jaitley after listening to specific requests from many states that the central government should give an underwriting of payment of the compensation from the consolidated fund said (and it’s on record and you can check the minutes) that was not possible and that compensation would be paid only from the cess and if the cess became inadequate to pay the compensation, the council would meet and evolve a strategy for making up the shortfall.”

The solution could include an increase in cess, which is not an option in the current economic scenario, Somanthan said. The other option discussed was borrowing, but it was not specified in those discussions whether the borrowing would be by the centre or states. “I’m not verbatim, but I’m almost quoting his [Arun Jaitley’s] words, he said, additional borrowing through extension of the cess. I want to point out that he did not necessarily say additional central borrowing nor did he say additional state borrowing.”

So, we are not in breach of a promise that was not made, Somanathan said.

State governments have argued that the centre is reneging on its promise of paying full compensation to ensure the committed 14% annual increase in states’ GST revenue.

Somanathan said according to the GST Act, the centre is liable to pay this compensation only when the cess is available. However, the centre recognises the need for states to get adequate funding this year, he said.

The legal position, 50% of it is in favour of what the states would want—namely that the whole amount is payable. The other 50% is that it is only payable when the cess is available. So, strictly the legalistic interpretation is that the whole amount is payable but it’s only payable eventually when the money becomes available. That’s clearly not what we want to do this year.
TV Somanathan, Expenditure Secretary

The government has presented states with two options.

The first of the two options, which includes a borrowing of Rs 97,000 crore, will be done via special state government securities where the government provides a letter of comfort on repayment via the collections from compensation cess, routed through an escrow-like facility, Somanathan explained.

GST Compensation Financing May Debut A New Type Of Government Borrowing — Exclusive

Watch the conversation here:

India To Meet Spending Targets Despite Revenue Squeeze, Says Expenditure Secretary: Exclusive

Read edited excerpts below:

I will start with requesting for your analysis of the GDP numbers and what you took away from it in terms of the policy direction.

I must give you an answer that you would find surprising. I didn’t take much from the data because it wasn’t a surprise. So, the exact number whether it’s -23 or -20 or -8 or -17, we knew it would be a big minus. I mean it was obvious; if it wasn’t a big minus, then there’s something wrong with the statistics because we know what the economy went through in the months of April, May and June. It was only in June when the unlocking started. So it was a question of measurement and I don’t think the number is something that either invalidates anything or validates anything.

What is more important for the impact on the economy long run is what the recovery looks like in the remaining three quarters. So, in my view this is actually the non-surprise number. I mean I wouldn’t have known what the number was but I knew it was going to be bad. So, I must confess that that by itself is not a big input into our policy making. It’s the number for June. We are in September. With most of our decisions, unfortunately we don’t have the luxury of waiting for the statistics and we have to drive this into the fog. So, we are going through the fog as best we can.

Government spending is the only support to the economy right now if you look at data from the expenditure side. Consumer spending was down, you may recover a little bit in the next quarter. Investment predictably fell off a cliff. I guess the question then is -- is the government mindful of the fact that at least in the next two quarters it will have to be that support from government spending which keeps the economy going?

Certainly, I think we are extremely mindful of that. I think there’s no question that the government and the central government at that is the last resort of the economy in a crisis as big as this. We are acutely conscious of that.

So, we saw government expenditure in the first quarter grow by 11%. As a matter of perspective, one can look at it and say oh that’s only 11%, but it’s also 11% against a backdrop of a precipitous fall in revenue, against the backdrop of an economy where spending money on projects was difficult, against an economy where even spending money on welfare schemes was difficult because people were not venturing into rural-development employment schemes in April and May. So, in that sense I think 11% is not, in aggregate, a bad figure. As a ratio of GDP it’s much higher, because remember the GDP fell by 24%. So, if you look at numerator versus denominator, there is a big increase in government spending as a share of quarterly GDP.

In addition, I think if you break that 11% down by sector, you will find that we have done a considerable amount of reprioritisation. There is a huge increase in food, rural development, in health and even in transfers to states. There are decreases in several elements where we deliberately sought to reprioritise by squeezing a number of ministries. That was a conscious decision. We classified all of the government into three categories, A, B and C, and we put the C category under a very tight leash so that they basically were kept on a survival ration and we diverted money. So, 11% is the aggregate increase but the increase in specific sectors that have greatest impact on employment, on welfare and on the poor, has been more. Those have actually risen much more than 11%.

Within the budget, at least the outside world is working with a number which is based on the Rs 12 lakh crore borrowing number, is there enough to keep government spending support high at least in Q2 and Q3? Unless there is a replan of that borrowing number which shall be revealed at an appropriate time I’m guessing.

There may be. As I said, this is not a situation where you can do five-year plans. We are working on five-week plans from time to time. But the point is that government spending will be sustained and we will certainly hit our budget targets regardless of the revenue environment. So,we will not be reducing government expenditure, absolutely not. We will keep the budget space of government spending up. But how much up?

I know lots of people want us to press the accelerator very hard and hit the floor with it. I understand that but we have many constraints that we have to manage under. So, expenditure will be sustained. How much will it be increased is something that we will take as it comes.

India To Meet Spending Targets Despite Revenue Squeeze, Says Expenditure Secretary: Exclusive

We are talking about the central government so far. One can’t deny that state governments have a large role to play. I think in that context what has been surprising is the GST tussle. There are reasons behind it but if you zoom out and take a broader economy view, is this perhaps the worst time for states to get squeezed? The last thing you need is underfunded states even if that helps to Centre to an extent?

Since you’ve asked the question, you must allow me to answer it. As you said, let me zoom out a little bit from GST. I mentioned the transfers to states this year in the first quarter were 35% higher than last year. You need to check that number, that’s an actual number you can see it in the government accounts.

Now, what we have done is you know that under the Finance Commission, we are supposed to pay 41% of central revenues to the states. What we’ve done in the first three months of the year is that we have paid 41% of the budget estimate of revenue and not 41% of what we collected in the first quarter. That was a deliberate and a conscious strategy because we understand what you understand, namely that states have to keep performing in this economy to keep things going. So we deliberately released money on the basis of the budget estimates even though we knew on the first of April, that the budget estimates were completely unrealistic because of the lockdown on March 22. So, we’ve done our bit.

When it comes to the GST, again, I think one should look at what is the definition of fair treatment. I think just in terms of context, is this something that the Centre has suddenly invented? This reasoning that if there is no cess, we don’t pay you and we will not pay from the consolidated fund -- is this some afterthought that we in the Ministry of Finance have dreamt of after March 22 or April 1? So I draw your attention to the GST Council discussions in December 2016 and January 2017 when this, at that time, hypothetical question came up of what happens if the cess falls short of what needs to be paid as compensation? At that time it came up -- not in the context of a big macro-economic shock -- it came up in the context of what if the 14% was much more than the actual growth of GST revenue. Then chairperson Mr. Jaitley after listening to specific requests from many states that the central government should give an underwriting of payment of the compensation from the consolidated fund said (and it’s on record and you can check the minutes) he said that was not possible and that compensation would be paid only from the cess and if the cess became inadequate to pay the compensation, the Council would meet and evolve a strategy for making up the shortfall. It might include raising cess, which of course is not an option right now, let’s be very clear on that.

I’m not verbatim, but I’m almost quoting his words, he said, additional borrowing through extension of the cess. I want to point out that he did not say necessarily say additional “central borrowing” nor did he say additional “state borrowing”. So what I’m trying to say is, this is not something that has just been imagined now. This question came up in 2016, states wanted a promise and the promise was not given by Mr. Jaitley. So, we are not in breach of a promise that was not made.

Let’s come to what the Act says. So, the Act has two elements according to the Attorney General. One element is that the condition is payable regardless of the source of the problem. So, it is not a claim that we are only liable to pay compensation for the portion that is GST related and that we are not responsible to pay for what is an act of force majeure, that’s not what we are saying. What we are saying is, we are due to pay the whole amount, and the attorney general has also confirmed, that we are only due to pay it when the cess is available. The legal position, 50% of it is in favour of what the states would want--namely that the whole amount is payable. The other 50% is that it is only payable when the cess is available. So, strictly the legalistic interpretation is that the whole amount is payable but it’s only payable eventually when the money becomes available. That’s clearly not what we want to do this year.

Therefore, we’ve gone by the spirit of the Act, which is that we are liable to pay compensation for the shortfalls arising from the implementation of the GST.

Let’s remember, if you do a thought experiment and if this year was a year with VAT and not GST, would the states have suffered no revenue or no shortfall at all? I mean it’s not true that they wouldn’t have, I would assume that with a 23% GDP shortfall, if you had an alternative tax system which was value added tax run by the states;that would have suffered a revenue shock. So the question is, what proportion of this is the centre obligated to pay this year, in the middle of our serious financial situation? That’s where the two options have been given. One is the Rs Rs97,000 crore, which in our calculation is attributable to the GST system, and would have existed with or without the Covid-19 scenario and the rest is attributable to this year’s special situation.

So, we’re discussing the options with the states, we are open to finding a solution, but if there is some meaning in that distinction, it’s not an invented or contrived thing that we’ve done. I think we need to be clear about the facts.

Can you explain how you would have arrived at that Rs 97,000 crore number?

It is very easy to calculate and you can, of course, come up with an alternative calculation but within a range because there’s not too much uncertainty in that calculation.

That calculation is: What is the sum that was due to the states at a compound growth rate of 1.14 to the power of n? Starting with 2015-16 as the base year, 1.14 to the power of n (I think it’s 5) times the 2015-16 revenue, is the revenue of each state. You then subtract from that. If we didn’t have this situation you simply subtract the actual revenue and that gives you the shortfall. This year, what we’re trying to estimate is what would the actual revenue this year in GST have been if this was not a post-Covid situation? So, how have we estimated it in our estimate of Rs 97,000? We took the actual collections in 2019-20, which is available because that’s a known figure, and we inflated it by 10%. Now why 10%? That is because the budgeted nominal growth was 10%. So, even if you take 5% real and 5% inflation that would have given you 10%. So, we said, let’s take last year’s actual GST collection; you collected it so that was possible in the GST system; so to that extent it is feasible GST revenue, we’re not imagining that figure. Inflate that by 10%, subtract the 1.14 to the power of n and then you get this shortfall. That figure is Rs 97,000 crore.

Now,this year’s actual collection, should it be 1.1 of last year or should it be some different number? That’s a fair argument and some of the states in our meeting with the finance secretaries have said that we shouldn’t take 10% and we should take the last few years’ average growth which is closer to 7% so on. Now, that will produce a slightly different number but it won’t produce a hugely different number, whatever growth estimate you plug into that. Even if you theoretically plug a 0% number, it would still be much less than the full shortfall.

Under the first proposal offered to states, this Rs 97,000 crore gets borrowed from some sort of special window from the RBI which we don’t have details of yet.

There was some confusion on this. We are not necessarily indicating that the RBI would directly buy the securities of the states, that’s not the implication. This window is, and we’ve said it in the written proposal that we circulated, we would coordinate this operation to get as attractive rates as possible for the states on this Rs 97,000 crore rather than leave the states to form their own borrowing calendars and go to the RBI as they normally do for their market borrowing.

Our aim would be to get as good a price on this borrowing as we can manage, we coordinate, we will assist, we may even group it and we make sure that they get as low a cost as possible. And, if necessary, we were willing to subsidise it to make it equal to the G-Sec rate.

How would that happen? There are only two possibilities; there is ‘Ways and Means Advances’ or there is the RBI perhaps directly buying. If not, then does the Centre backstop these bonds? I don’t understand what coordinated would mean and how the rates would be managed here?

It’s an administrative coordination.

So, you could make a special issue of state bonds where we undertake and give a letter of comfort that the interest and principal will be serviced from the compensation cess account held by the Government of India. The receipts on that cess account are more than adequate to service Rs 97,000 crore. The annual collection in that cess, in a normal year, is in excess of Rs 95,000 crore. So, the servicing cost per annum would be somewhere in the range of Rs 6,000 crore per annum and if it’s agreed that the cess would be extended beyond year six, there are adequate revenues, there’s a very specific revenue stream that would be earmarked for the repayment of this debt. That would be underwritten by the central government because the money is actually credited in the GST Compensation Fund, which is controlled by the central government.

So, the central government underwrites that these monies would be a charge on the GST Compensation Fund and this will be a bond issue of the states but with this coordinating arrangement and with a dedicated repayment mechanism.

So, while you’re not calling it guarantee but effectively the market sees this as the central government-guaranteed state bonds.

It would be state bonds, if you like. I’m not using this in a formal sense (and you shouldn’t say that I said that this is what we will do) but I’m giving an analogy. It is as if we are securitising against future cess receipts. So, there is an escrow account into which a predictable statutorily levied set of revenues will be credited every month and that will be available to service these bonds.

What is the advantage of going via this circuitous route rather than the Centre just going in and borrowing a little bit more directly from the RBI and move into direct monetisation?

That’s an alternative but an alternative that we don’t prefer because to emphasize on your earlier question we need to spend a lot too. We need to spend a lot, we haven’t seen the back of this pandemic. We are still in month 6 of a 12-month year. So, there are lots of uncertainties. We are not taking a preferred option of doing that. There is also some legal ambiguity about whether its tax which is essentially fully assigned to states. I mean, the central government doesn’t get one paisa from the compensation cess. So, for us to securitise a receipt that doesn’t belong to us...

What you’re suggesting is simply borrow it as a regular security. I mean that’s a feasible option but it is not an acceptable option right now.

I’ve also heard this argument that if one borrows via state facilities rather than the central government, there won’t be an increase in yields. Analytically that doesn’t sit right. The bond market will always look at all public sector borrowing requirements and then if that is is excess of what they were expecting or well in excess of households financial savings, yields will rise.

Let’s look at this very clearly. There is a certain supply of government securities in the G-Sec market, there’s a supply of State Development Loans also but they’re slightly differentiated. They’re not identical. There is a differentiation between the G-Sec and the SDL. There’s also an interstate differential between Kerala SDL and the Punjab SDL and the Manipur SDL. There is no differentiation between one Government of India G-Sec and the other if they have the same tenor in the same terms etc. So, it’s a segmented market. It’s not one uniform product, it is two very similar products with slight differentials.

If you add supply into this broad market and if you consider this as one broad market there is a particular impact on yields which will happen and I take what you say. So, whether we add SDLs or we add G-Secs, you are increasing the demand for funds in this market and there will be a delta on the yields, other things being equal. So, let’s assume there’s no neutralising action by anyone, there will be a delta. That point is well taken.

However, if you act on that differentiated SDL market, there will be delta on the whole market but it will be a little bit more skewed. You will have a higher delta on the SDL and somewhat less on the G-Sec market. Now, the G-Sec is the only part of it that is a benchmark for everyone else. The SDL is not a benchmark for anyone other than state corporations and PSUs of individual states. The Government of India security is a benchmark not only for states but for the entire private sector bond market too because no one can have a rating better than the sovereign. So, if the sovereign yields go up, it has knock on effects for people down the chain.

We just think that it would be more differentiated if it were done to SDLs, which is why we have said that we will actually pay an interest subsidy for the difference between the two and we’ve offered to pay upto 0.5%. To our knowledge, the average gap is 0.43% last week.

So it’s not as if we don’t understand that but we still think there is some benefit to this. I think you or someone had made the point that the general government debt is going to be the same regardless of these two consequences. That is true but it’s also our experience with both rating and other agencies that they also look at the differentiated levels of central debt and state debt. They look at all three. They look at the Centre, they look at the state and they look at the sum of the two. You may disagree but we think there is some benefit to some optimisation between these two.

So it looks like option one is what is being seriously discussed, which is Rs 97,000 crore borrowing via perhaps this mechanism. The second one, we were not very clear on what was being suggested but that seemed like states were being told to go to market and fend for themselves.

In the second one too, we have said that we will provide a letter of comfort that the principal will be deducted from the cess account. So, that security will still be a little different from the normal SDL where you are dependent on the normal budget of the state. In this case we are committing that the principal will eventually be paid from proceeds of the cess. The central government will issue a comfort letter to that effect but the interest will be serviced from the state budget.

In this case, the state would be eligible to borrow the entire shortfall without making this distinction of whether the revenue loss was GST related or not. But yes, then there are some countervailing adjustments which are not favourable to the state. So, states would actually make this choice based on whether they are in a very difficult cash position and what the whole shortfall is needed this year is or they’re willing to take the Rs 97,000 crore this year and get the rest year six onwards.

It’s possible that one or two states might actually prefer option two and we are not necessarily averse to that. There are certain countervailing adjustments in the overall debt ceilings which we want to ensure so that we don’t have an unsustainable total level of debt on the states.

So this goes to debt to GDP of the states in either situation?

In option two, it does. But in option one, it doesn’t.

In option one, where would it get counted?

An academic would count it into the general government deficit and so would I, but it wouldn’t be in the state deficit for finance commission purposes. So, the Finance Commission has prudential norms. They sometimes have, we don’t know what this finance commission will do, but some Finance Commissions have some things that are linked to state debt to GDP ratios. So, in option one this would not be counted as state debt because it’s paid from a dedicated revenue pool. I don’t like to use the word, but I’m using it because it may be more familiar, it is like a securitised kind of tax source.

So since this will be paid from a different source and it’s not on the balance sheet of the state, we would not consider this in the state debt-to-GDP ratio.

It would show up somewhere in the central government budget, I’m not sure if it’ll show up in the headline?

It’s neither centre nor state, but it is India. So, it will be somewhere.

There is a general sense that this is a time for the government to be spending on infrastructure. There is a national infrastructure pipeline, there are shovel-ready projects where the government can speed up implementation to be able to provide job and income support to the economy. Does that analytically makes sense to you?

That makes a lot of sense and we are doing our best. The Finance Minister has had three meetings with central public sector enterprises in terms of stepping up public sector enterprise capital expenditure. They are really going as fast as they can. In some cases, they’ve had difficulty with restarting because of health-related constraints in terms of getting workers onto the site.

So, we have pressed the accelerator to the floor as much as we can, but the engine transmission, even if you press the accelerator, the fuel isn’t getting transmitted fully, but certainly that’s something that we are doing. As far as the government is concerned, the National Highways Authority, Railways—capital expenditure is one of our priorities. It has the highest multiplier.

So, when I was speaking of re-prioritisation, we have ensured that budget availability is not a constraint on the central government capital expenditure. Now you can look at the numbers. There’s no reduction in capital expenditure but in Q1 you wouldn’t see a massive increase either because there are constraints which are not cash. I mean, it’s non-cash constraints but in the department of expenditure, we are making sure that capital expenditure doesn’t get delayed for want of cash or want of budgetary support. So, that is a commitment.

We agree that capital expenditure should be a priority and we will try to ensure that budgetary constraints don’t stand in the way. Central public sector enterprises have also been pushed to maximize expenditure on shovel-ready or already shoveled projects. I agree with you, conceptually that that is something we should and we will do our best.

Even a few former central bank governors, like Dr. Rangarajan, have said the current situation may merit direct monetisation of a part of the fiscal deficit to ensure availability of funds...

Purely on the economic merits, I agree with him but these are matters for the central bank and the Department of Economic Affairs to calibrate. But I think just as an academic point, I don’t disagree.