© The Indian Express Pvt Ltd
Tags:
Journalism of Courage
India’s GDP growth exceeding economists’ estimates by a wide margin for the second quarter in a row in July-September reignited the debate over the issues that plague the methods used by the Ministry of Statistics and Programme Implementation (MoSPI). These issues seem especially pertinent when the quarterly GDP numbers – particularly those of the manufacturing sector – don’t seem to tally with other, more regular data.
Take industrial production, for instance.
It’s worth adding that manufacturing’s share in India’s GDP has continuously fallen in the last few years, standing at 12.6 per cent in 2024-25 without adjusting for inflation, down from 16.1 per cent in 2011-12.
Several other high-frequency economic indicators also don’t seem to be in sync with the robust GDP data. Domestic air passenger traffic – an indicator of business and personal activity – was lower in each of the three months of July-September compared to last year. For the quarter as a whole, it was down 1.8 per cent after April-June had seen a 5.3 per cent increase. Fewer foreign tourists have entered India in each month of 2025 compared to 2024. Growth in Goods and Services Tax collections was negligible in November from an already weak 4.6 per cent in October.
“Consequently, there seems to be divergence between the GDP statistics and the evidence from high frequency activity indicators, making it difficult to gauge whether the GDP data are indeed reliably reflecting the state of the economy,” Nomura economists Sonal Varma and Aurodeep Nandi said in a note after the second quarter GDP data was released at the end of November.
The positives
This is not to say that all high-frequency data is suggestive of weaknesses. Lending by banks, including to industry, is picking up pace. As of the middle of November, all loans given by banks were up 11.6 per cent year-on-year, the highest growth rate in a year. Loans to industry were up 10 per cent at the end of October, compared to less than 5 per cent as at the end of May. More generally, Indian companies had a great July-September quarter – the best in two years, in fact – with earnings of more than two-thirds of them beating expectations. And though the global trade environment has been difficult, services exports have continued to do well.
On the government side, the Centre’s capital expenditure in the first seven months of 2025-26 is up 32 per cent. States, meanwhile, saw their capex rise by about 13 per cent in the first half of the year, according to CareEdge Ratings, with Central Public Sector Enterprises posting a growth of 14 per cent over the same period.
Early warning signs
While a mix of new destinations, transhipment, and a rapidly weakening rupee helped propel India’s goods exports 19 per cent higher in November, economists see the US’ 50 per cent tariffs pulling growth lower in the second half of 2025-26. Some signs of this are already visible, with industrial production up only 0.4 per cent in October – the lowest growth in 14 months – with manufacturing output up 1.8 per cent as against 5.6 per cent growth in September.
The fall in October industrial growth has been described as ‘industry wilts under tariff pressure’ (Nomura), ‘not looking good’ (Barclays), and ‘exhibiting weakness across sectors’ (Union Bank of India). And while fewer working days in October due to the Diwali holidays and frontloading of production in September ahead of the rollout of the reduced Goods and Services Tax (GST) rates on September 22 go some way in explaining the deceleration, the data was still worse than expected.
It is worth noting that IIP measures change in production and not the value of goods produced – and this is why there is a divergence between GDP data and monthly indicators.
Underestimation, overestimation
To find the value added for any sector, the value of inputs it uses is subtracted from the value of output it produces. But this is value added at current prices. To find the real GVA, the output and input values need to be adjusted by the respective inflation rates – inputs by the rate at which their prices have changed and outputs by the increase in costs for consumers. This is called double deflation.
However, MoSPI currently does this only for agriculture and mining and quarrying; for others, it uses the same rate of inflation for both input and output prices. Why is this a problem? As MoSPI Secretary Saurabh Garg and Chief Economic Advisor V Anantha Nageswaran themselves wrote in a newspaper column in early September, when input and output prices diverge, “estimates could be overstated or understated depending on the direction of divergence”.
This is exactly what has happened in recent months: input price inflation has been lower than that of output prices. As such, using the same inflation rate leads to dubious results. How?
Assume a sector uses Rs 100 of inputs to produce Rs 200 of output, and input inflation is 2 per cent and output inflation is 5 per cent. In this scenario, using double deflation gives a real GVA of Rs 92 (Rs 200 reduced by 5 per cent, minus Rs 100 adjusted by 2 per cent). But if the same inflation rate of, say 3 per cent, is used to adjust both input and output values, the real GVA is higher at Rs 97. Therefore, the growth rate will be higher.
Another issue is what prices are being used to deflate what activity. The difference in the services sector’s real and nominal growth rate was just 120 basis points (bps) in July-September, suggesting services inflation was 1.2 per cent, something which has not gone down well with economists.
Why was services inflation per GDP data so low? Because real GVA growth for services was arrived at by adjusting GVA in current prices with mostly wholesale inflation – which doesn’t include any services. So, when wholesale inflation is low due to soft commodity prices, the adjustment to nominal services GVA to get real GVA growth is not in sync with the actual services inflation in the economy. The result is “exaggerated real growth”, according to HSBC economists, who estimate real GDP growth in July-September to have been overstated by 120 bps due to the deflator issues.
Making changes
To be fair to MoSPI, its hands are tied to some extent. Ideally, inflation based on a Producer Price Index (PPPI), and not wholesale inflation, should be used to adjust nominal GVA. This has been a long-standing complaint of the International Monetary Fund too. However, India does not have a PPI. And the ministry responsible for the PPI is the commerce ministry’s Department for Promotion of Industry and Internal Trade, which hasn’t come around to finalising it.
The new GDP series, to be released in February 2026, proposes to make several changes that should make India’s national accounts statistics cleaner and more representative. Of course, what that may do to the growth rate is anybody’s guess and will likely raise new questions.
