New Delhi: A group of economists and statisticians recently came together to highlight the inconsistencies in the economic data reported for different sectors and raised serious questions about the reliability of current GDP measurement methods. Their discussion underlined the urgent need for reforms to restore confidence in India’s economic data.
“We must ensure that GDP estimates reflect economic reality. If they don’t, then the policies based on them won’t work either,” said Sharma.
At the event titled GDP Base Revision: Time to Regain Confidence, held at the India International Centre, leading Indian statisticians and economists gathered to discuss concerns about the country’s economic data.
The panel featured NK Sharma, former Director General (Statistics) at the National Statistical Office, Sanjay Kumar, former Additional Director General at NSO, and Amey Sapre, associate professor at the National Institute of Public Finance and Policy. The session was moderated by Siraj Hussain, former Union Secretary in the Ministry of Agriculture.
Among the audience were journalists, students, members of the UPSC, and a retired officer of the National Sample Survey.
The discussion comes at a time when the Centre is preparing to shift India’s GDP base year from 2011–12 to 2022–23. The revision, led by a 26-member advisory committee, is expected to be completed by 2026. The goal is to better reflect the structural changes in the economy and improve the quality of GDP estimates, aligning India’s statistical framework more closely with international best practices.
Where the numbers don’t add up
The GDP or GVA of public administration is often considered to be mainly the salaries and wages of government employees, covering public administration and defence. Sharma said that after adjusting for inflation, the data shows a surprising 77 per cent growth in the real GDP of public administration over time. This raises the question of where such growth is coming from.
“On average, with 6-6.5% or 7% growth per year, we expect steady growth, assuming employment levels remain constant. Employment itself also rises steadily, but this change (77% ) is too large,” said Sapre.
Kumar added that the number of government employees has not increased since 2011-12. “Unfortunately, official data on the number of government employees (or related details) is not fully available or reliable in economic surveys for the years after 2011-12, but it is a perception.”
Sharma raised several classification and reporting issues in the data. He referenced findings from the 74th round of NSSO data, where 12 per cent of businesses surveyed were not found at their listed addresses, and 20 per cent were misclassified. “This creates confusion and makes the estimates less dependable,” he said.
Kumar highlighted a key technical flaw in how GDP figures are adjusted for inflation. He explained that constant prices–used to measure real growth—are calculated using one type of price index, while current prices reflect another. Ideally, both should move in a similar direction once inflation is accounted for.
However, he pointed out that the two sets of numbers are often out of sync, raising doubts about the accuracy of the methods. “It’s very hard to imagine that constant and current price estimates behave so differently unless the methodology itself is flawed,” Kumar said. This discrepancy suggests that the way inflation is adjusted in GDP calculations may not be consistent or reliable, which could distort the real picture of the economy’s performance.
The panel also highlighted the anomalies in how some services were measured. For example, vehicle repair and maintenance were tied to new sales, rather than to the total number of vehicles in use. “Repairs are based on the whole stock, not on how many vehicles were sold in a year,” Sharma said, calling the assumption questionable.
A broader concern was that the statistical system may not be equipped to capture economic shocks or transitions effectively. “If we’re using indicators from the organised sector to estimate growth in the unorganised sector, we’re assuming they move the same way. That’s simply not true,” Sapre said.
Kumar said that for trade, the unorganised part is showing a growth rate of about 10–11 per cent per year at current prices. But if we look at other indicators–like household survey data–the growth is much lower, around 7 per cent, and in some years, as low as 3 per cent.
He added that when looking at own account enterprises (small businesses without hired workers), the gross value added per worker has declined over time. This doesn’t align with the high growth shown in the national accounts data.
“The unorganised communication sector is shown as growing faster than the organised sector, but that’s hard to believe. PCOs and internet cafés have nearly disappeared—from 2 million in 2011–12 to just 30,000 in 2023. Their share should be going down, not up.”
Kumar explained the main issue: national accounts use GVA per worker from larger, formal enterprises and apply it across all enterprise types, including small, informal ones.
“The GVA per worker in big firms is almost 2.4 to 3 times higher than in small units. Using the same productivity numbers for all leads to overestimation,” he said.
Also read: Virender Sangwan changed eye care in India. His methods took surgery from Rs 5 lakh to Rs 50k
Fixing what’s broken
There was broad agreement at the session that India’s GDP estimation system needs urgent reform, especially as more states set ambitious growth targets. Without reliable verification mechanisms, inflated numbers could easily pass as fact.
Citing Uttar Pradesh’s ambition of becoming a $1 trillion economy by 2027, Sharma said, “If states use consultants to present higher production data to meet targets, and if MoSPI has no mechanism to verify it, we could end up accepting unreliable figures.”
One major concern was the use of MCA-21—a corporate database introduced during the 2011-12 base year revision. While it aimed to improve coverage of the formal sector, speakers noted problems with classification, limited accessibility, and a lack of foundational research.
“We didn’t conduct enough research before adopting MCA-21. Now we’ve built the system around it and are stuck with its limitations,” Sharma said.
MCA-21, launched in 2006 by the Ministry of Corporate Affairs under the UPA government, is an online system that streamlines company compliance. It lets companies, professionals, and the public file documents and get corporate information easily and safely. This makes registering a company simpler and results in fewer visits to government offices.
Sharma also highlighted a lack of clarity in the methods used for estimation, pointing out that the government released only a summary of changes in the source and methods document, assuming users were already familiar with the old system—a move he considers a mistake.
Looking ahead, Sharma said, new annual surveys for unorganised sectors offer some hope. He emphasised that while updated methodologies or data sources can improve the accuracy of GDP and sectoral estimates, this accuracy depends critically on the correctness of the base year data. He warned that if the base year estimates or methodology are flawed or overestimated, those errors will get carried forward and persist throughout the subsequent period (usually five years), resulting in misleading or incorrect growth and sector share figures.
Sharma also called for revising key indices like the Wholesale Price Index (WPI) and the Index of Industrial Production (IIP) in sync with GDP updates.
“Otherwise, the system gets out of sync, and we start seeing unusual patterns in growth data.”
Rebuilding trust in the numbers, the speakers agreed, hinges on better transparency and accessibility. “We must make the system more transparent. Methods should be explained clearly. Data should be accessible,” said Sapre.
(Edited by Aamaan Alam Khan)