Former Chief Statistician Pronab Sen on Friday commended the methodological improvements made by the Ministry of Statistics and Programme Implementation in computing the gross domestic product (GDP) under the new 2022-23 base year series, but he also raised some concerns. In an interview with Kuldeep Singh, he highlighted the dichotomy between the sluggish investment rate and (high) GDP growth, and finds it surprising that consumption growth aligns with the GDP rate. He noted that a 32% investment rate (as in FY26) supports a maximum GDP growth of about 6.5% in the medium term. Conducting the population census and economic census promptly is crucial for better frames and reducing inaccuracies, Sen stressed. Excerpts
Q: How can the significant difference—around Rs 100 lakh crore each—in real GDP estimates between the old and new series for FY26, and the previous three years, be explained?
The main reason is the change in deflators between the old and new series. Real GDP is affected by these deflators, while nominal GDP doesn’t change very much.
Q: Based on the new series, the investment rate is stuck around 32% in FY26. How do you see this?
Yes, the investment rate (gross fixed capital formation as a share of GDP) is around 32%. Generally, a 32% investment share supports a maximum of about 6.5% GDP growth in the medium term. Yet we’re seeing 7.8% growth. So where is this extra 1 percentage point plus growth coming from? That’s a key puzzle. We have a situation where investment is stuck, even as consumption has gone up.
Q. The new series shows a substantial jump in the manufacturing sector’s growth estimates. What do you attribute this to?
It’s hard to say anything definitively without deeper analysis, but one likely factor is lower global commodity prices (raw materials). Manufacturing is heavily impacted by these prices. With double deflation in the new methodology, lower input prices can significantly boost the real value-added estimates for manufacturing.
Q: Consumption shows 7.7% growth in FY26, which is almost in line with overall GDP growth.
Yes, that’s quite surprising. Normally, consumption grows slightly slower than GDP. Typically, with 7% GDP growth, you’d expect consumption growth of around 5.5–6%. Here, consumption is growing roughly at the same rate as GDP. The share of household consumption in GDP is also at 57%, which is higher than the usual 53–55%.
Q: Is this 57% share unusual, or has it been consistent with the revised series?
The share itself isn’t entirely unusual, it can vary, but what’s striking is that consumption growth is matching or nearly matching GDP growth, unlike previous patterns. For example, in FY23 and FY24, GDP grew around 7.2% and 7.1% respectively while consumption was at 5.8%.
Q. Nominal GDP is lower in the new series compared to the old one. Why is it so?
It’s difficult to pinpoint exactly. Nominal GDP is calculated from the same sources in both series, and there isn’t much difference overall. The one change is that multi-product firms can now have their output split more accurately between manufacturing and services (which wasn’t possible earlier). But that should only affect the sectoral allocation, not the overall GDP.
Q. We still don’t have recent census data—the last one is from 2011, and the next is delayed. How reliable are these GDP numbers given outdated population figures?
The new series assumes population growth of about 0.9%. We don’t know the exact source, but it’s likely an extrapolation (probably from the Home Ministry or Registrar General’s projections handed to MoSPI). It raises the question: Has India’s population growth really fallen below 1%? That’s worth examining.
Q. What do you think of the methodological improvements MoSPI has made in the new series?
The efforts are commendable. They’ve tried to address long-standing issues. The base year was overdue (it should have been revised around 2017–18, but was delayed until now). Using more recent data helps reduce questions about outdated benchmarks.
Q. Are there still areas where India’s statistical system needs improvement?
Improvements are always ongoing. The real challenge is having reliable, timely mechanisms to collect data. Relying too heavily on surveys can lead to issues if the sampling frame is outdated (e.g., without a recent census). Conducting the population census and economic census promptly is crucial for better frames and reducing inaccuracies.