Before You Cheer the GDP Number, Ask What They Divided By

Every year around this time, India picks a number and fights over it. This year it is 7.8%, the growth the National Statistical Office reported for the April-to-June quarter. Three of the country’s most serious economists, Subhash Garg, Kaushik Basu and Raghuram Rajan, have all raised a hand. They are not saying the same thing, and the differences are the whole story.

Start with what “real growth” even is, because it is less solid than it sounds. I teach this in the first month of macroeconomics. The statistical office can see spending, output and income in rupees. Add them up and you get nominal GDP, which grew 10.3% this quarter. But some of that is just prices rising, not more goods made. To strip prices out, you divide nominal GDP by a price index, the GDP deflator. What is left is real growth, 7.8%. So real growth is a leftover. It is whatever survives after someone decides how much of the rupee rise was inflation.

This quarter that judgment was about 2.3%. And here is the oddity that set the economists off. That 2.3% is lower than the roughly 4% inflation households met at the shop, and far below the 9% rise in wholesale prices. The number they divided by came in under almost every other price gauge in the economy. Divide by a small number and the real figure it leaves behind looks big.

Why so low? Mostly one technical choice. India now deflates manufacturing by adjusting output prices and input prices separately, a method called double deflation. This year input costs rose faster than the prices factories could charge, so on paper manufacturing’s price index actually fell and its “real” output jumped. The method is textbook-correct. But it books a squeeze on factory margins as a burst of real growth, which is not what most people picture when they hear the sector boomed.

Now the critics, in order of how far they go.

Subhash Garg goes furthest. The former finance secretary points out that last year’s April-June GDP, first announced at ₹86 lakh crore, was revised down to ₹80 lakh crore under the new base year. This year’s figure is ₹88 lakh crore. Against the revised ₹80, nominal growth is 10.3%. Against the ₹86 first announced, it is only 2.6%. “If you had not revised last year’s GDP,” he says, “the growth in current prices would have been only 2.6 per cent.” He is careful to add that he is not disputing the deflator. What bothers him is that last year’s number moved.

On this the government has the better argument. The ₹86 lakh crore and the ₹80 lakh crore sit on two different base years, 2011-12 and 2022-23, and you can no more subtract one from the other than you can subtract feet from metres. Hold the base fixed and growth is 7.8%, so Garg’s 2.6% does not survive contact with how the series is actually built.

Kaushik Basu is more cautious, and more interesting for it. He says he has not studied the numbers closely, but that Garg’s is the sharpest critique he has heard, and his own worry sits further back. India’s investment rate has fallen from above 38% to below 34% and stayed there, which is not the profile of an economy about to speed up.

Rajan comes at the same doubt from the demand side. If output really grew 7.8%, he argues, the jobs and the private investment and the foreign money should be showing up, and they are not. He also keeps returning to the price index itself, which he thinks is set too low to trust the real figure that falls out of it.

Put the three together and the honest verdict is narrow. Garg’s 2.6% is wrong. But Basu and Rajan are pointing at something true. In the accounting sense, 7.8% is real. In the lived sense, it is generous. Had the deflator matched the roughly 4% households actually paid, real growth would land closer to 6% than to 8%. On an economy this size, that gap is close to ₹5 lakh crore of output a year. It is the difference between wages that outrun prices and wages that barely keep up.

There is a fix, and it is not fewer revisions. It is sunlight. Publish the full deflator, sector by sector, so anyone can see where 2.3% came from. Publish a clean bridge between the old base and the new, so no one can line up two incompatible numbers and no one can cry manipulation. Do that and most of this argument dissolves, because it was never really about honesty.

It was about a price index almost no one outside the statistics office can see. So before the next number lands, ask the one question that settles it. What did they divide by?