Did the NDA Government Inherit “Empty Coffers”?

Prime Minister Modi’s comment (see here) that the UPA government left the government coffers empty for the NDA government predictably drew a strong response from the former Finance Minister, P.C. Chidambaram (here). The context for the controversy was that the fiscal deficit for the first two months of the fiscal year 2014-15 had reached 45% of the budget estimates. Some commentators – example: Madan Sabnavis of CARE Ratings – also felt that the UPA government had used up an excessive amount of the budgeted fiscal deficit in the first two months of the fiscal year 2014-15 (here).

A question that arises here is whether a government, certain of defeat in an election, strategically leaves its successor government with a high deficit in order to constrain its manoeuvrability (See Hodler). If such strategic behaviour has, in fact, taken place, how can we find evidence for it? What we need to examine is whether the outgoing UPA government spent recklessly during April and May of 2014. More importantly, did it do this knowing fully well that it would be defeated and wished to deliberately leave government finances in a poor state for the incoming NDA government? I seek to answer these questions by comparing government finances during April and May of 2014 with the finances during April and May in previous years. If the finances during April and May 2014 are significantly worse than in earlier years, we could entertain the suspicion of reckless behaviour by the UPA government.

However, we should not rush into any judgement of reckless behaviour. It could well be the case that the time profile of revenue generation (via taxes, for example) and expenditure requirements are inherently mismatched. If this is, indeed, the case, there could well be significant gaps between expenditures and revenues (leading to deficits) in the early months of the year while revenues catch up with expenditures as the fiscal year rolls on.

I investigate these issues by looking at detailed fiscal data on a monthly/quarterly basis from the year 2003-04 (such data are not available for earlier years) till the first two months of  fiscal year 2014-15. All data used in this note are sourced from the Ministry of Finance, Government of India.

Figure 1 gives information on the government’s Revenue Expenditures (RE) (note below Figure 1 explains this term) incurred during the first two months of each fiscal year of the time-period under consideration as well as Revenue Receipts (RR) generated over the same two months of each fiscal year. This information is reported, not in terms of actual expenditures or receipts, but as percent of announced Budget Estimates (BE) for the fiscal year. For instance, RE is defined as:

Revenue Receipts are defined in an analogous manner.

It is clear from the figure that, in April and May of each fiscal year, expenditures are much greater than receipts. It is also seen that this gap is increasing over the years indicating that governments are increasingly front-loading expenditures to the beginning of the year. The main reason for the gap between expenditures and revenues is that the government’s expenditure commitments begin as soon as the fiscal year commences but revenues (especially tax revenues) begin to flow into government coffers much later.

Does the mismatch between revenue expenditures and revenue receipts disappear as the fiscal year progresses? Figure 2 shows the position for the second quarter (July – September) for the fiscal years 2003-04 to 2013-14. RE is once again measured in a manner similar to what was done for Figure 1 above.

Figure 2 is dramatically different from Figure 1: the line for revenue receipts lies almost completely above the line for revenue expenditures. It may be recalled that the situation was the exact opposite in Figure 1. This shows that, as the year progresses, revenues start flowing into government coffers and the mismatch between expenditures and revenues observed in the earlier part of the year gets rectified.

My analysis so far seems to suggest that the behaviour of expenditures and receipts during April and May 2014 (under the outgoing UPA government) has not been any different from the behaviour of these two items during April and May of earlier years. Having looked at RE and RR, I now turn my attention to Revenue Deficits.

Revenue Deficits (RD) are defined as revenue expenditure (RE) minus revenue receipts (RR).  As per Indian definitions, when RE is greater than RR there is a deficit and RD is a positive value; when RE is less than RR there is a surplus and RD is a negative value. If RE is greater than RR, then the government must borrow funds to meet its expenditure requirements. The data that we use is quarterly (three month period) and extends only up to 2013-14 since RD data for April and May of 2014 were not available. Figure 3 shows RD for Quarter 1 and Quarter 2 for each year under consideration. RD is measured as:

As would be expected from what we have seen in Figures 1 and 2, RD is much higher in Quarter 1 (Q1), namely April, May and June, as compared to Q2. In Q1, due to the paucity of revenues, RD is much higher leading to high borrowing requirements. However, in Q2 as revenues start flowing in, RD falls, leading to much lower borrowing requirements. Once again, the analysis does not seem to suggest any reckless behaviour on the part of the outgoing government: borrowing requirements (due to RE being greater than RR) have always been higher in Q1 as compared to Q2. Having examined RD, I finally turn to Gross Fiscal Deficits (GFD) which began the whole brouhaha about empty government coffers.

The difference between RD and GFD is that the former considers only revenue expenditures while GFD considers revenue and capital expenditures. Hence, GFD is the broadest measure of the borrowing requirements of the government and encompasses RD. The announcement that the outgoing UPA government had used up more than 45% of the budgeted GFD during April and May of 2014 led to much criticism that the UPA had behaved irresponsibly and had deliberately left government finances in bad shape. The analysis so far has suggested that there was probably nothing sinister at work on the part of the UPA government and that such mismatch between expenditure requirements and revenue generation at the beginning of the fiscal year has been common in India. Figure 4 gives the profile of GFD for April & May and for Quarter 2 for each year in our dataset. GDF in the figure is measured as follows:

The figure reveals that, generally, GFD as percentage of budget estimates in the first couple of months is higher than what it is in Q2. The need for borrowing, as shown by GFD, falls during Q2 as tax revenues begin to flow into government coffers. The UPA using up 45.6% of the budgeted GFD during April and May of 2014 is, in fact, not exceptional. This percentage was higher in 2006-07 (48.5%) and in 2008-09 (54.9%).

The analysis in this note suggests that passing judgment on the finances of the government on the basis of data of the first two months of a fiscal year is hasty and premature. This is not to suggest that government finances as inherited by the NDA government are in fine order. Clearly, the finances are in a critical state and this fact has been recognised by the current and previous governments. The limited point I am making is that dependence on borrowings in the first quarter (or first couple of months) of the fiscal year is inevitable since taxes and other revenues begin to flow to the government later in the fiscal year but expenditures have to be incurred right from the beginning of the year. Prime Minister Modi’s allegation regarding empty government coffers was probably an exaggeration, designed to convey the message that government finances are not in a healthy state and that the fiscal room available to his Finance Minister is quite limited.

Does an outgoing government in India behave strategically and leave its successor with less fiscal room by running up high deficits? Do the fiscal data point to any such election effect? In the time period under consideration, there have been three elections and two changes of government. Both changes of government were similar in that the outgoing government continued to rule for the first two months of the fiscal year before the new government took over. Focusing on these two changes in government, the analysis in this note has shown that the data during April and May of these two years have not behaved particularly differently as compared to April and May of non-election years. However, reaching an empirically sound verdict on strategic behaviour of an outgoing government will probably require more data which include more numerous changes of government. Given the limited data that are available at this point, it is not possible to discern any such strategic behaviour on the part of an outgoing government.

Monsoon Failure: A major challenge to the Government of India

The Indian Meteorological Department has warned that rainfall during the 2014 south-west monsoon season (June to September) is likely to be below normal to the extent of 90-96% of the long period average. It is well-known that the SW monsoon accounts for a significant percentage of the total rainfall in India. The long period (1901 to 2012) average contribution of the SW monsoon to the total rainfall in a year has been 75%, with the highest being 84% (in 1908) and the lowest being 69% (in 1987 and 1990). It is generally the case that whenever monsoon fails, Indian agriculture is immediately affected and this impacts the entire economy. Even though the share of agriculture (excluding allied activities) in total GDP of India has come down from 41% in 1950-51 to under 12% in 2012-13, the number of persons dependent on agriculture is still very high. Almost 50% percent of Indian workers are still employed in the agriculture sector (See here). Consequently, whenever agricultural production fails, income levels in the sector drop, leading to a fall in consumption expenditure. This, in turn, reduces the demand for products and services produced in the other sectors of the economy causing these sectors to also suffer.

To some extent there has been a decoupling of the agriculture sector from the rest of the economy, so that it is to some extent insulated from the adverse performance of agriculture. The chart below plots 10-year moving correlations between rate of growth of GDP and rate of growth of GDP from agriculture. (See note below chart for explanation of moving correlation.)

There is a clear break around 1994-95 (see vertical line at this point) when the correlations begin to dip sharply. The lowest point is reached in 2001-02 after which the correlations again begin to rise. It is possible to put forward the conjecture that the economic reforms, which began in 1990-91, represented a significant change in the inter-relationship between agriculture and the rest of the economy. Even though it is not as sharp in 1994-95, it is possible to discern a fall in the correlations beginning around 1990-91.

Does this mean that the economy as a whole is now insulated from an adverse monsoon? Such a conclusion would be unwarranted. Growth rates are still affected by an adverse monsoon and the consequent deceleration in agriculture. Adverse agricultural performance is still likely to shave off 1-2 percentage points from the rate the growth. The chart below shows moving 10-year moving correlations between (a) rate of growth of GDP and deviations from normal July-September rains and (b) rate of growth of agricultural GDP and deviations from normal July-September rains.

For a considerable number of years, the two correlations have been high – in excess of 0.6. However, during the period 1997-98 and 2002-03 (the segment within the two vertical lines) the two correlations dip to levels as low as 0.1. During this period, monsoon was below normal in 1998-99 (-6%), 1999-2000 (-9%), 2000-01 (-7%) and 2001-02 (-20%) and yet GDP growth rate remained persistently high except for 2001-02 when it fell to 3.8%. Growth rate of agriculture was negative in two of the 4 years: -0.6% in 1999-2000 and -8.5% in 2001-02. It is not clear why there was a delinking between monsoon rains and growth rates during this period and additional research is required to come up with answers. Various reasons are possible: reduction in importance of monsoon dependent kharif crop, better irrigation facilities, reduced dependence of rural households on agricultural incomes. None of this is conclusive and it is not clear why these effects did not become permanent. After 2002-03, the strong association between monsoons and growth rates has been re-established.

The likelihood of poor July-September rains for this year (2014-15) is being blamed on the El Niño weather system which affects the world every few years. Of course, there is little that can be done about the weather system since the fault, literally, lies in our stars, or more precisely, in our star – the sun. Numerous studies have shown close connection between sunspot activity, El Niño weather systems and rainfall (See J. Curt Stager and others or David Rind). Increasing sunspot activity is associated with increasing rain while decreasing sunspot activity is associated with decreasing rains. This is true for the Indian monsoon rains as well (See Hiremath and Mandi). It is expected that sunspot activity is likely to remain weak for a few years in the future, which means that the south-west monsoon might also remain weak in the years to some. This certainly does not augur well for India.

It is unfortunate that the El Niño weather system has hit India within a few months of the BJP government taking office. The problems that are likely to emerge are going to be quite challenging:

1. Poor agricultural performance depresses incomes in the rural areas. It is likely that many will be pushed into poverty.
2. Lower agricultural income impacts the buying power of the rural sector which will hurt other sectors of the economy that produce products and services bought in the rural sector
3. Domestic food prices will be expected to rise and, worse, since the El Niño is a global phenomenon, world food prices are also likely to rise. It has been estimated that El Niño has accounted for 20% of commodity price inflations over the years (See Allan D. Brunner).
4. Rural poverty is especially susceptible to inflation in general and food inflation in particular. Hence, in addition to point 1 above, this will also likely raise poverty levels in India.
5. If food imports become necessary, this will impact India’s import bill especially in a year when uncertainty in the Middle-East could drive up oil prices

It is difficult to prepare a comprehensive list of what needs to be done to combat the drought-like situation that seems likely. Some of the following seem important:

1. Water is likely to be very scarce and prioritising its use for drinking and irrigation purposes will be essential
2. With the kharif crop likely to be adversely affected, wherever possible, compensatory rabi cultivation needs to be undertaken.
3. Allied activities in the rural sector – animal husbandry – will also be affected. Provision of fodder to animals will be required to avoid distress sale of such animals.
4. Rural incomes are likely to be adversely affected. In such a situation, employment guarantee schemes, such as MNREGA, will be able to provide a lifeline to the rural sector. Despite its grave reservations about this scheme, the government would be well advised not to wind it down, especially this year. Providing income to rural areas will not only help sustain livelihoods but will also sustain rural demand. This will help limit the adverse impact that the drought might have on the rest of the economy.
5. Farmer suicides have been the tragic bane of agriculture in numerous states in India. There is a danger that this might see a strong resurgence due to the drought. The government must stand ready to assist those who might be unable to repay the loans they have taken.
6. Food shortages are likely to be grave. The government has already stated that it will use buffer stocks of food to keep inflation in check. This must be followed up by adequate distribution of foodgrains wherever these are likely to be most required.