In the previous two parts of the series, Dr Pronab Sen presented a pathway to recovery from the Covid-19 shock in three distinct phases – survival, revival, and recovery. In the final part of the series, he discusses the financing of the recovery, addressing concerns around deterioration of the government’s fiscal position, and how resources can be raised for financing a larger stimulus package.

Financing the recovery

The results of the model presented in the previous parts of the series, no matter how approximate or contestable, face the government with a stark choice – either do nothing more than what has already been announced and prepare for a long period of sub-par per capita GDP; or act immediately with a large fiscal stimulus and hope to get back to a positive growth path in two years. It appears that the government is deeply concerned about, and therefore constrained by, the deterioration of its fiscal position. It seems to believe that this will affect its global standing as a responsible fiscal manager and thereby attract a down-rating by global rating agencies. It seems doubtful, however, that the government has actually estimated the longer-run effects of its current stand and that of any counterfactual. It is, therefore, necessary that the financial consequences of the alternative paths be examined so that the government can take an informed decision.

As far as financing of the recovery path is concerned, there are three distinct players that are involved: the financial sector, state governments, and the central government. Each of them has a key role to play at different phases of the recovery process. As has been described in the earlier sections, in the survival phase the main actors are the state governments and the financial sector; the former for lives and the latter for enterprises. All three players will be involved in the revival phase. In the recovery phase, it will be the state and central governments with the onus shifting towards the role of the Centre.

As far as the financial sector is concerned, it is practically impossible to predict the hit it will take on its profit-and-loss account and balance sheet at this stage. However, some qualitative assessment can be made. To begin with, this sector was already in trouble even before the pandemic and lockdown struck. NPAs were high at around 9%, and were expected to go up by around 2 percentage points on account of major slippages in MUDRA loans, for which forbearance had been granted by the government since September 2019 until 1 April 2020.¹ The asset quality recognition of these loans has now been further postponed by another six months. It is now almost a certainty that most of these MUDRA loans will become NPAs after 1 October 2020, when the forbearance is lifted.

An important component of the liquidity support provided by the government in May 2020 is a window of Rs. 3 trillion for non-collateralised loans to MSMEs with a tenor of four years. It is more than likely that this facility will also be used like the MUDRA loans earlier to kick the can further down the road. As a result, this will continue to mask the true damage contained in bank balance sheets for at least one year more when the recognition begins to happen. However, this is a small price to pay for keeping the standard loan accounts alive. Something similar is true for all the other liquidity measures as well.

Going forward, the degree of damage that will occur to currently standard accounts in the entire financial sector depends mainly on the pace at which the economy recovers its growth path. The quicker the recovery, the less will be the damage. If the recovery path is protracted as in the case of the baseline scenario, considerable damage will occur to the financial sector, which will seriously compromise its ability to support investments, thereby making the recovery even slower and more painful than shown in Table 1 in Part 2. The relatively quick recovery of the second scenario may actually be absorbed by the financial sector with some help from the government, and investment support may be forthcoming. In either event, setting the financial sector back on its feet will need large and sustained central government support.

It is fortunately possible to be much more precise with the financial position of the government. The first question that needs to be answered is: how much additional resources does the government (both Centre and states) need to raise to provide the fiscal stimulus in the two alternative scenarios outlined earlier? It should be remembered that a key assumption of the model is that the basic expenditure of government is held constant at the 2020-21 BE (budget estimate) level in real terms through the four years being considered (Assumption 3). Thus, the change in expenditure is only the stimulus provisions. Second, the change in fiscal deficit does not take into account revenue loss on account of revenue sources other than taxes. This is not of much consequence for states, but is significant for the Centre.² Third, it is assumed that no changes in tax rates are made by either the central or state governments.³ The results of this exercise are presented in the table below.

Table 1. Change in consolidated fiscal deficit from 2020-21 BE

(All figures in Rs. Trillion at 2019-20 prices)

As can be seen, in the current year the consolidated deficit will rise significantly with the proposed stimulus by about Rs. 5.8 trillion. This is despite the stimulus being Rs. 7.9 trillion higher. Thus, a part of the additional stimulus will be financed by additional tax collections arising from the higher growth generated. More importantly, in the subsequent years, the deficit will be much lower with the proposed stimulus. Thus, over the four-year time horizon being considered, the fiscal deficit will be nearly Rs. 6 trillion lower. What this means in effect is that the government’s view that an enhanced fiscal stimulus is unaffordable is completely misplaced if one takes a somewhat longer time-horizon than just the current year.

The argument becomes even more compelling if one considers the consolidated fiscal deficit to GDP ratio (or the CFD ratio) – an indicator that the government sets much store by due to the prodding of the ratings agencies. Currently this ratio is budgeted to be 6.5%. This ratio will certainly rise whatever the government does, but it is the relative magnitudes that are of interest.

Although in 2020-21, the proposed stimulus will give a CFD ratio of 15.4% as compared to 13.2% for the current package; in 2021-22 it drops off to 12.2% compared to 15%; in 2022-23 these numbers become 10.1% vs. 13.1%; and in 2023-24 they are 7.5% vs. 9.0%. This should gladden the hearts of rating agencies since the fiscal consolidation process is much more pronounced for the proposed than for the current.

The other big question bothering the government is how it would raise resources to finance a larger stimulus package. This worry is being fanned by financial experts and economists who question the “fiscal space” available to government. What exactly this means no one seems to very clear about or for that matter agree on, but it is an evocative term. However, as has been discussed earlier, the notion of “fiscal space” does make sense for states since they are constrained in how much they can borrow by central diktat. But it certainly does not make sense for the central government. An entity which has the power to print money can never be “fiscally constrained” in an unqualified sense.

For this year (2020-21), the Centre has already announced an additional borrowing programme of Rs. 4.2 trillion and has permitted the states to borrow a similar amount in addition to their original borrowing limits. As things stand, with the current stimulus package, the additional Rs. 4 trillion that the states can borrow will not even cover their tax losses, which are estimated at Rs. 4.3 trillion. What this means is that the states will have to reduce their expenditure by Rs. 0.3 trillion, which in turn means the effective fiscal stimulus will not be Rs. 2.1 trillion, but Rs. 1.8 trillion. This will make the growth prospect even weaker.

In case of the central government, the projected tax loss is Rs. 3.1 trillion, which leaves Rs. 1.1 trillion for other purposes. However, as has been mentioned in footnote 1, there will also be losses in non-tax revenues; which may actually leave very little to finance the stimulus. It is almost certain that the Centre will have to borrow more to finance its current package.

Raising this quantum of additional borrowing (around Rs. 9 trillion for centre plus states) may not be very difficult or expensive in terms of the yields on government bonds. At present, the banks have parked nearly Rs. 8.5 trillion in the reverse repo account of the RBI. They should be only too happy to shift from the 3.45% interest they earn in reverse repo to at least 6% that will get from government securities. During the course of the year, as the liquidity components of the government’s package kick in, this amount will become larger.

It is, however, indubitably true that a major part of the additional Rs. 8 trillion required for the proposed package will put a strain on the market and will require yields to go up significantly.⁴ This will, in turn, push up all other interest rates – a disastrous prospect in a situation of already impaired investor confidence. This then is a text-book situation for monetisation of the deficit.⁵

The possibility of monetisation has been under active discussion in India during the last few weeks. There are three arguments that are put forward against monetisation. The first is that this will set a bad precedent and the government may get addicted to this form of financing its deficit into the indefinite future.⁶ The outcome, as the critics put it, would be to “jeopardise the hard-earned victory” against fiscal profligacy. Such a view can be described in another, less flattering, way – it is a willingness to sacrifice the economy and human lives at the altar of fiscal probity. It also does not reflect well on the government of the day.

The second argument is that monetisation is inter-generationally iniquitous in that it involves borrowing from future generations to pay for the present. While this is certainly true under normal circumstances, it does not apply in a crisis of this magnitude. If the absence of a money-financed stimulus leads to a sharp reduction in investments, as it is very likely to do, future generations will pay an even bigger price.

The third, and most popular, argument is that monetisation of the deficit will lead to excess money supply in the economy and thereby to inflation. The 2009-10 experience is often offered in support of this view.⁷ While this sounds eminently reasonable, it has a fatal flaw in that it does not consider the counterfactual. The notion of excess money supply is not about the absolute supply of money, but the money supply relative to the GDP. In the present context, the choice is between (a) not providing the extra stimulus and keeping the money supply constant, on the one hand; and (b) providing the stimulus and allowing the money supply to rise, on the other. The relevant figures for the two options 2020-21 are as follow:

Option (a): Change in money supply = 0; Change in GDP = - Rs.25.2 trillion; Excess money supply = Rs. 25.2 trillion⁸

Option (b): Change in money supply = Rs. 8 trillion; Change in GDP = - Rs.11.6 trillion⁹; Excess money supply = Rs. 19.6 trillion

In other words, the excess money supply will be far higher (by Rs. 5.6 trillion) if the additional fiscal stimulus is not given than if it is. By the theory and logic of the Quantity Theory of Money espoused by the critics of monetisation, inflation should then be higher in the former case than in the latter.¹⁰

In short, there is no compelling reason why a substantial fiscal stimulus should not be given, but many why it should. The government needs to take heed of the consequences of not providing sufficient stimulus and shed its diffidence. Too many lives and livelihoods are at stake.

APPENDIX: The model structure

The model used in this series uses the simplest possible specification that can address the issues under consideration. It is based on one key assumption: the investment rate (Investment/GDP ratio) is constant. This assumption essentially means that the growth rate of potential output is also constant. The model is calibrated to yield an annual growth rate of 6%, which we refer to as the core momentum.

Thus, the potential supply side becomes a 6% annual growth rate applied to the actual GDP of the previous year. The annual estimate of potential supply is then allocated across the four quarters by using seasonality factors computed from historical data. The actual output of the economy is assumed to deviate from the potential supply only if there are exogenous shocks to demand. Thus, in the absence of any demand shocks, the economy would be in a steady-state growth path of 6%.

Since the investment rate is held fixed, demand shocks can originate in consumption, exports, or government expenditures.¹¹ These shocks have been computed separately and allocated across the quarters depending upon their point of origination and duration of persistence. Thus, the negative lockdown and export shocks are limited to the first quarter 2020-21, while the positive fiscal stimulus shocks are spread out over multiple quarters.

Each shock is then projected to propagate through the next six quarters through the multiplier process. The quarterly break-up of the propagation process is pre-specified so that the total across the quarters adds up to 2. This quarterly break-up is not the same for negative and positive shocks. The negative multiplier peaks in the third quarter following the shock and reduces thereafter, whereas the positive tapers off steadily.

The net effect of all the shocks in each quarter is deducted from the potential supply to yield the estimate of actual GDP in that quarter. This procedure is then iterated over the entire time horizon to yield the actual GDP series. All other variables are computed from these GDP estimates using historical ratios.

Full paper available here.

Notes:

1. It may be recollected that a large proportion of these loans were taken by MSMEs to roll over their past loans during 2017, in the aftermath of the demonetisation.
2. The central government budget has large non-tax revenue projections from dividends and disinvestments. These too are likely to be adversely affected, but are not included in these calculations.
3. This is of course not strictly accurate since the Centre has raised taxes on petroleum and some states have raised taxes on potable alcohol after the budget. While the former has not materially affected the retail price of petroleum (due to softening of global oil prices), the latter has raised the retail price significantly. This could have a further depressive effect on aggregate demand depending upon the price elasticity of demand.
4. It is conceptually possible to create some more borrowing space by reducing the Cash Reserve Ratio (CRR) of banks, but it is very unlikely that the banks will actually allow their cash liquidity position to go down significantly since cash withdrawals from banks have gone up sharply in the last two months as people draw down their savings to finance their consumption needs.
5. Monetisation means that the government sells bonds directly to RBI for cash and not go to the market.
6. The great advantage of monetisation over market borrowings is that it is virtually costless to the government. Although the government does pay interest to the RBI, practically all of it comes back in the form of higher dividends. Thus, the only cost that is borne is the repayment of the principal, but that would happen many years later depending upon the tenor of the bonds. To address even this cost, Dr. Rathin Roy of the National Institute of Public Finance and Policy (NIPFP) has been advocating the use of ‘consoles’, which are bonds into perpetuity, and therefore the principal is never paid back. The problem with this is that consoles cannot be traded in the market and, therefore, cannot be used to reduce the money supply if needed.
7. Inflation did spike in 2009-10 when the government increased its fiscal deficit by 3 percentage points. But, there was no monetisation of the deficit.
8. See Table 1, Part 2.
9. See Table 1, Part 4.
10. None of this is actually true, since the actual outcome will probably be deflationary in both cases, but it does highlight the dangers of fuzzy theorising.
11. In National Income Accounts, the external sector has imports as well. However, in this context, imports are not a source of shock and are determined endogenously.