Fiscal Consolidation: Narrowing the Deficit and Its Trade-Offs

Fiscal consolidation is narrowing the gap between what a government spends and what it takes in. Only three routes exist: spend less, take more, or grow the output the gap is measured against. The first two remove demand at the moment they are applied, so timing matters as much as size. Part of any narrowing also happens on its own when the cycle turns.

Two things already built are holding that sentence up. The first is that a government has exactly two instruments, what it spends and what it takes, and that a slice of the gap between them moves by itself with the state of the economy without anybody signing anything. The second is that the word deficit covers three different gaps, and which of the three a plan is narrowing changes what the plan actually is. Neither of those is repeated here. One further fact is narrower and, once it has been seen, hard to unsee: the number everyone quotes is a fraction, a fraction has two floors it can be moved from, and almost every argument about consolidation is really an argument about which floor somebody stood on.

What are the only routes to a narrower deficit?

Start with the arithmetic and keep the politics out of the room for a moment. The number that gets quoted in every discussion of a government account is not a rupee amount at all. The quoted number is a share: the borrowing for the year, sitting on top, divided by everything the economy produced in that year, sitting underneath. Sankhya, an invented republic used throughout these notes, spent Rs 4,00,000 crore, received Rs 3,00,000 crore, and therefore had to borrow Rs 1,00,000 crore. Against output of Rs 17,47,200 crore that is 5.72 per cent, and 5.72 is the number that ends up in the headline.

Now ask what could make 5.72 smaller. There are three answers and there is no fourth. Spending can fall, and the top of the fraction shrinks. Receipts can rise. The top is spending minus receipts, so the top shrinks again. Or output can grow and enlarge the denominatorThe number sitting underneath in a fraction, the thing a share is being measured against. Change it and the share changes even when the number on top has not moved at all. while leaving the top exactly where it was. The list ends there.

Spend less, take more, or grow the denominator are the only three routes to a narrower deficit share, and every consolidation plan that has ever existed is some mixture of the three. Which is why the mixture, and not the headline total, is the thing worth arguing about. Two governments can announce the identical share and have chosen wildly different mixtures to get there, and a reader who only ever looks at the total has no way to tell them apart.

The household version is quick. Suppose a house spends Rs 40,000/- a month against Rs 30,000/- coming in, so Rs 10,000/- has to be borrowed. Three things could shrink that hole. The house could spend less. Someone in the house could earn more. Or the house could stay exactly as it is while the street around it becomes richer. The same Rs 10,000/- hole is then a smaller thing relative to what the street earns. The third one sounds like a trick, and it is not: it is exactly what the growth route does to a national number, and it is the route people forget when they read a share.

Which deficit is the plan actually narrowing?

Here is a question that sounds pedantic and is not. When somebody says the deficit is being narrowed, the word deficit is carrying three different gaps, and the three were separated earlier in these notes for exactly this reason. Narrowing one of them is a different exercise from narrowing another, and the same rupee cut has a different meaning depending on which line it lands on.

Take the Sankhya spending of Rs 4,00,000 crore and open it up one level. Revenue spendingSpending that is used up inside the year it happens: salaries, pensions, subsidies, interest on past borrowing, running costs. What is in it and how it is classified is covered separately. is Rs 3,30,000 crore, and capital spendingSpending that leaves something behind at the end of the year, such as a road, a rail line or a building. How it is defined and where it appears in a government account is covered separately. is Rs 70,000 crore. Now suppose the plan is to reduce spending by Rs 12,640 crore. If it comes out of revenue spending, that line falls to Rs 3,17,360 crore and the assets still get built. If it comes out of capital spending, that line falls to Rs 57,360 crore and this year of consumption is left alone. Total spending is Rs 3,87,360 crore either way. The share of output reads 5.00 per cent either way.

Two completely different decisions produce the identical headline, so a plan quoted without naming which deficit it narrows and which line the adjustment lands on is not a plan a reader can check. The total misleads here for the first time, and not for the last.

Which way does consolidation push demand?

Every rupee a government spends is a rupee somebody receives. A salary is paid, a contractor is settled, a subsidy reaches a shop. So when spending falls by Rs 12,640 crore, that is Rs 12,640 crore of income which is not created, and the whole of it disappears from demand at the moment the cut takes effect. There is no leakage. The government was the one doing the spending.

Taking more works differently and less completely. Suppose instead of cutting, Sankhya raises Rs 12,640 crore more in receipts. The extra receipts come out of somebody's disposable incomeWhat a household has left after tax has been paid, out of which it does both its spending and its saving. Take a rupee of it away and only the part that would have been spent leaves the demand stream.. Some of it would have been spent. Some of it would have been saved. The saved part was never going to show up as demand this year in the first place, so only the part that would have been spent leaves the demand stream at all. On an illustrative assumption that one rupee in five would have been saved, Rs 10,112 crore of demand is removed and Rs 2,528 crore is taken out of saving instead.

Consolidation acts on the cycle in the opposite direction from a spending package, and the spending route removes more demand than the receipts route for the same rupee of adjustment. Direction and size are all the comparison establishes. The comparison is not an argument for or against consolidating, and not a reason to prefer one route: taking more has its own consequences for whoever it is taken from, and those consequences are weighed elsewhere.

How can a share fall while the rupee gap stands still?

The growth route is the one readers forget, and forgetting it is what makes the failure described below so common. A share falls if the number on top gets smaller or if the number underneath gets bigger. Nothing in the arithmetic cares which. So there is a version of a narrowing deficit in which the government borrows exactly as much as it did last year, changes not one line of spending or taxation, and still reports a lower share.

Work out what that version needs, rather than asserting it. If the gap stays at Rs 1,00,000 crore and the share is to read 5.00 per cent, then output has to be Rs 1,00,000 crore divided by 0.05. The division gives Rs 20,00,000 crore. Sankhya produced Rs 17,47,200 crore, so the growth required is Rs 2,52,800 crore on Rs 17,47,200 crore, or 14.47 per cent in nominalMeasured in the rupees of the year in question, before any adjustment is made for prices having risen. The distinction between nominal and real is covered separately. terms. Fourteen and a half per cent is a large number, and the size of it is the point.

A deficit share can improve with the rupee gap completely unchanged, and the growth that does it is not something a government can simply order. Growth gets quietly assumed in plans, precisely because it is the one route that requires nobody to cut anything or tax anybody. The growth route is also the one nobody controls. Growth is an outcome of a great many things, most of which sit outside any budget document.

The denominator cuts both ways, and it is worth seeing that before moving on. If output shrinks by 4.00 per cent to Rs 16,77,312 crore while the gap stays at Rs 1,00,000 crore, the share rises to 5.96 per cent. Nobody spent more. Nobody taxed less. The number on top did not move by a single rupee, and the headline got worse anyway.

What do the three routes look like on one account?

Take the Sankhya account and narrow it three separate ways, each landing on exactly 5.00 per cent of output. All three land on the same destination, and the headline then carries no information at all: three economies that look nothing like each other report the identical number.

Route one is the spending route. To read 5.00 per cent against unchanged output of Rs 17,47,200 crore, the gap has to be Rs 87,360 crore, so spending falls from Rs 4,00,000 crore to Rs 3,87,360 crore, a cut of Rs 12,640 crore. Receipts stay at Rs 3,00,000 crore. Route two is the receipts route, and it needs the same gap of Rs 87,360 crore, reached from the other side: receipts rise from Rs 3,00,000 crore to Rs 3,12,640 crore while spending stays at Rs 4,00,000 crore. Route three is the growth route, and it changes neither. Spending stays at Rs 4,00,000 crore, receipts stay at Rs 3,00,000 crore, the gap stays at Rs 1,00,000 crore, and output is Rs 20,00,000 crore.

All figures in Rs crore, from the invented Sankhya account. In route one the gap falls by Rs 12,640 crore. In route two the gap falls by the same Rs 12,640 crore. In route three the gap does not move at all, and the growth of 14.47 per cent in the denominator does the entire job. One identical headline of 5.00 per cent sits on top of three economies that have nothing in common, and that is the strongest reason to distrust a share quoted on its own.

What is the trade-off between a faster path and a slower one?

Suppose Sankhya decides on the Rs 12,640 crore adjustment and has to choose how quickly to make it. Take two paths, both of which end at the same place. The faster path makes the whole adjustment in year one: the gap is Rs 87,360 crore in both years. The slower path makes half of it in year one and the rest in year two: the gap is Rs 93,680 crore, then Rs 87,360 crore.

Now count both consequences. Over the two years the faster path borrows Rs 1,74,720 crore and the slower path borrows Rs 1,81,040 crore, so the faster path borrows Rs 6,320 crore less. Over the same two years the faster path removes Rs 12,640 crore of demand in year one against Rs 6,320 crore on the slower path, so the faster path removes Rs 6,320 crore more demand in the year the economy has to absorb it. The same Rs 6,320 crore turns up as the gain and as the cost. Arithmetic rarely gives a cleaner picture of a trade-off.

Consolidating faster reduces borrowing sooner and removes more demand sooner; consolidating slower does the reverse; and both of those sentences are true at the same time. Which of the two costs matters more depends on the state of the economy, on what the borrowing funds, on who the demand belongs to and on what a government is willing to accept. Each of those is a judgement rather than a calculation, and the trade-off is where the arithmetic stops. Anyone claiming the arithmetic settles it has stopped doing arithmetic somewhere without saying where.

Why can a deficit narrow with nobody deciding anything?

Now the part that undoes a great deal of confident commentary. When output recovers, a government does not have to lift a finger for its gap to shrink. Non-tax receiptsMoney a government takes in without levying a tax: dividends from what it holds, fees, charges, and the like. Where receipts come from and how they are classified is covered separately. and tax receipts both climb. There is more income and more spending to take a slice of. At the same time, fewer people are in difficulty, so the payments that exist because people are in difficulty fall away. Nobody signed anything. The gap narrowed anyway.

Put Sankhya through it. Output grows 4.00 per cent to Rs 18,17,088 crore. Assume receipts move one for one with output, so they rise from Rs 3,00,000 crore to Rs 3,12,000 crore. Assume need-based payments fall by Rs 3,000 crore, so spending drops from Rs 4,00,000 crore to Rs 3,97,000 crore. The gap is now Rs 85,000 crore against output of Rs 18,17,088 crore, and the share reads 4.68 per cent. The fall of more than a point from 5.72 is larger than the entire decided consolidation worked earlier, and that consolidation took the share only to 5.00.

Receipts rise and need-based payments fall on their own once output recovers, so a narrowing deficit is not evidence that anyone consolidated. The automatic part of a government account was set out at the start of these notes, and how to tell which phase an economy is in belongs with the business cycleThe repeating swing of an economy between expansion and contraction, and the phases in between. How to tell which phase an economy is in is covered separately., covered separately. Both of those are worth having in hand before any single year of a deficit is read as a decision.

What does an analyst separate before reading a narrowing?

Here is what somebody doing this for a living actually does with a falling deficit share, whether they are covering the economy for a lender, sizing a construction order book for an equity view, or assessing what a government is likely to do next. The analyst does not read the change. The analyst splits it.

Suppose the Sankhya account for the following year prints receipts of Rs 3,20,000 crore and spending of Rs 3,95,000 crore against output of Rs 18,17,088 crore. The gap is Rs 75,000 crore and the share reads 4.13 per cent, down from 5.72. The 4.13 per cent is the headline. Now split it. The recovery alone, worked in the previous block, would have produced a gap of Rs 85,000 crore. So of the Rs 25,000 crore by which the gap narrowed, Rs 15,000 crore is what the cycle produced and Rs 10,000 crore is what somebody decided. Look inside the decided part and it separates further: receipts came in Rs 8,000 crore above what the recovery alone would have given, and spending came in Rs 2,000 crore below it.

One technical note that matters. The share has a denominator that moved between the readings, and differences in points would not add back to the whole, so the split is done in rupees rather than in points of the share. The rupee gap has only one moving part, so a rupee decomposition holds together. The same fact the whole account rests on arrives here one more time from a different direction.

And the reason for doing any of this: only the decided part carries anything about intent. The cyclical part says the economy recovered, and the growth figure already said that. The Rs 10,000 crore that somebody chose carries information about how far a government is willing to go, and therefore about what it might do next year.

The Republic of Sankhya and its government account are invented.
Educational material. Not advice on any investment, tax, budget or market position.