Debating debt

Updated 10 Jan 2019


The writer is a member of staff.
The writer is a member of staff.

ONE headline from the current week has ricocheted around the national conversation in problematic ways. The headline was the one announcing that government borrowing increased by Rs2.24 trillion in the first five months of the fiscal year, according to data released by the State Bank on Monday.

On the face of it, the figure is alarming. Opponents of the government, as well as headline-hungry TV anchors, seized on it to make all sorts of analytical claims that the government has lost control of its fiscal effort and is borrowing in record amounts to make up for its failure to mobilise revenues.

A closer look reveals a different reality. Let’s put it in plain English first: almost half of the Rs2.24tr of borrowing in the July-November period is accounted for by the devaluation of the rupee, which caused the rupee value of external debt to shoot up. How do we figure this? Very easily, with a little bit of basic maths. Let me explain.

This is not the only time basic numerical literacy is required before launching an analytical ship into unknown waters.

The news gave us three data points. The total debt of the country rose by Rs2.24tr in five months, of which Rs907 billion was domestic debt and Rs1.334tr was external debt. At the end of the period under review, the total external debt was Rs9.13tr, the data release said.

Now do some elementary algebra. The total external debt at the start of the period under review (July) would be Rs7.796tr (that’s 9.13 minus 1.334). From one link on the State Bank website, you can see that the rupee fell by 15.5 per cent (from 121.5 to 140.3 to a dollar) in this time period. So what is 15.5pc of 7.796tr? The answer is 1.208, which is the amount by which our external debt, when denominated in rupees, shot up simply because of the devaluation of the rupee.

Next question: if Rs1.334tr was the total increase in external debt, how much of that is accounted for by the devaluation of the rupee? The answer is 90pc. And if the total increase in debt was Rs2.24tr, then how much of the total increase is accounted for by the devaluation of the rupee? The answer is 54pc.

This is not the only time such basic numerical literacy is required before launching an analytical ship into unknown waters. An accomplished historian of the interwar years, E.H. Carr, once wrote an interesting little booklet titled What is History? The simple title showed that he was writing about a subject that had been sucked into very intense partisan debates in a sharply polarised environment. When writing about the old school of history writing — some called it the Whig interpretation of history, which focused on a straight narrative telling of the facts as the historian’s primary job — he said it may be true that nothing speaks louder than facts, “but facts don’t speak for themselves, they only speak when they are called upon”.

It is very similar with numbers and data. Numbers are an integral part of the story. They describe the story, give it proportions and make its contours comparable to other places and times — but they are not the story itself. The story itself is the context in which the number has been produced. Sometimes that context is heavily political, and sometimes it requires other numbers to build.

For example, when you hear one political party saying “the GDP growth rate was higher in our time than it was in yours”, you should know that you are hearing a very political interpretation of a number. A high GDP growth rate may or may not be a good thing, first of all (it depends on the composition of the growth and how it has been achieved, among many other things); and even if one period sees higher growth, it could well lend itself in significant ways to work that was done long before the ruling party came into office.

For another example, take another headline that fluttered past recently, also related to debt. This one announced that the government borrowing from the State Bank between July and September surged past Rs5.4tr. This is alarming because borrowing from the State Bank is synonymous with the printing of money, which is highly inflationary and would denote that the government has run out of options to finance its operations and is basically printing money to pay its bills.

But this is one of those examples where the number by itself does not tell you the whole story. You will have to look at where this number has stood over a longer period of time to understand what is really going on. In other words, you will need other numbers to unlock the story.

Consider that total stock of government borrowing from the State Bank stood at Rs5.348tr in the first week of July 2018 — long before this government came in. A week earlier, it was Rs3.668tr, meaning it leapt by Rs1.68tr in one week. Sounds sensational, doesn’t it?

Not really, because from that spike of July 6, it declined steadily to reach Rs3.607tr by Aug 24, back to where it was end-June, if not slightly lower. Then it climbed again to hit a peak of Rs6.859tr by Nov 30, an increase of 90pc in three months. But then it fell again to Rs5.104tr by end-December, which is the latest figure available.

It should be obvious what is going on : government borrowing from the State Bank rises and ebbs in a quarterly cycle to accommodate short-term government liquidity requirements. Picking up one data point from the great cycle to build spectacular announcements upon is a highly misleading presentation of the facts.

A good rule of thumb to follow is: always be suspicious of a story built around a single number. Look for somebody who knows the right context and the right series into which to place that number before making up your mind. Politics is everywhere, but reality is sacred. Let’s work to preserve it.

The writer is a member of staff.
Twitter: @khurramhusain

Published in Dawn, January 10th, 2019