Q1 GDP pow-wow: Why we need to "double-deflate" both the ra-ra comments and the Cassandras

It is a pity that national accounts statistics (NAS) have become a political football and subject to biased commentary. If the government side would like to put the best interpretation on the latest numbers, the opposition would like to do the opposite. This cannot but be demotivating for those who actually work very hard to compile these statistics painstakingly year after year.

The latest data print on first quarter GDP (April-June 2026), which showed that GDP grew by a healthy 7.8 percent in real terms (10.3 percent in current terms), has been celebrated by the government as a huge achievement in the face of global headwinds. The opposition and ideologically inclined economists would like to pretend that this is a fake story, since the new numbers are based on a deflated 2025-26 first quarter base in the new GDP series, which uses new methods to estimate GDP, including double deflation.

Double deflation means that output/revenue numbers are deflated twice to remove the effects of inflation: once for inputs and again for final products. The GDP number in Q1 last year was lower by Rs 6 lakh crore compared to the previously used index. The new index revised the base year to 2022-23, and redid all the back numbers.

The problem is not in the statistics, but in what you choose to see based on your predilections. While government ministers are over the moon at the 7.8 percent GDP growth rate, critics choose to note the lowered base in Q1 of 2025-26 which helped bump up the Q1 figures for this fiscal. But they make a fundamental error, not least among them Subhash Garg, a former finance ministry secretary. Garg said that on the old base of Rs 86 lakh crore in Q1 last year (in current price GDP), this year’s numbers represent only a 2.5 percent rise. In current prices, Q1 GDP this year grew 10.3 percent against 7.8 percent in real terms.

Both supporters and critics should chill. Those who are merely watching the fracas from the sidelines, should consider doing a double-deflation on the credibility of the arguments put forth by both supporters and critics. The truth will, as always, lie somewhere in-between.

Government, critics and neutrals should remember the following realities.

First, GDP numbers are always estimates, not actual numbers. They are good at indicating trajectory, but may not indicate the whole truth. In a dynamic economy, where many parts and prices are moving all the time, to try and capture a snapshot in any one quarter is a statistical nightmare. It gets better with time, and old numbers are revised with more and better data, but there is no need to see any kind of deliberate attempt to manipulate data to suit the government in power.

Second, any new method of calculating GDP (or inflation or industrial production) will need time to settle, and they may be impacted by new and better sources of data that are made available. Earlier, we needed to calculate GDP from data reported by various ministries and companies; today, we get real time data from goods and services tax, sample surveys, and other databases, improving our ability to predict. Also, estimating growth in the informal sector is another nightmare, but this is being improved by more regular surveys of informal sector output. In short, how good or how bad the current quarter GDP numbers are will be known only when the data improves, and later data are incorporated to fill the gaps in the current data.

Third, two years ago, the International Monetary Fund (IMF) rated the quality of Indian national accounts statistics at C – the second lowest grade. Every critic was jumping with joy, as it seemed to suggest that our data was fudged. However, this too is wrong. Not-so-good data caused by the use of outdated methods is not the same as claiming that the data was mostly cooked up. They are not. It is a disservice to our statistical services to claim that the data they produce, whether good, tolerable or not-so-good, is the result of deliberate distortions. The new GDP series is, if anything, an attempt to improve on the old methods, and needs to be lauded, even while we wait for time to correct any distortions created by inadequate availability of timely data from various sources.

There is, however, a smell test that can be useful to have beyond the numbers. If you think the GDP numbers are so good, why is there a sense of despondency about jobs and incomes? Conversely, if growth is seriously overstated, why does the Japan Credit Rating Agency raise our sovereign rating to level A- after nearly four decades? Why do Indians abroad want to invest $127 billion with Indian banks, when the Reserve Bank was hoping for only half that amount to stabilise the rupee? Why does high-frequency data (bank credit growth, exports, tax revenues) show such large increases in double digits?

The only way to make sense of opinions is to “double deflate” them for known biases. If you are a government minister or even a journalist inclined to believe the government, your comments on the GDP numbers need to be taken with a pinch of salt. If you are always on the side of the critics, your opinion too is worth less than you think it is.

Maybe, just as stock market experts are asked to disclose their own investments while making recommendations, commentators ought to declare their basic ideologies and political preferences before we take their opinions at face value.

Personal disclosure: I have been largely supportive of the Modi government and its policies, though I also have serious differences on some of them, which I have voiced.