PIB Tracker21% of firms put most of their capex into robotics; 79% put almost none
MoSPI is running its third forward-looking capital expenditure survey. The last one found no enterprise anywhere between those two positions - and a panel growing 1.9 per cent.
What happened
- The NSO is running the third Forward-Looking CAPEX Survey, October to December 2026, under the Collection of Statistics Act, 2008.
- CAPEX 2025 found manufacturing at 50.17% (Rs 5,73,900 crore), info and communication 16.38%, electricity and gas 8.96%, transport and storage 5.55%.
- On robotics: 21.16% of enterprises planned more than 80% of capex there; 78.84% planned less than 20%.
- Realisation ratio 96.3%; actual capex Rs 173.5 crore per enterprise in 2024-25.
- A fixed panel of 3,819 enterprises shows growth of 1.9% over actual 2024-25 capex.
For Prelims
- Forward-looking survey: it asks firms what they intend to spend, not only what they did. That is why it carries a realisation ratio - intentions measured later against outcomes.
- Collection of Statistics Act, 2008: the statute under which the survey is conducted. It makes response a legal obligation and guarantees confidentiality of unit-level data - which is what lets NSO ask firms about future plans at all.
- The frame: built from active enterprises registered with the Ministry of Corporate Affairs, subject to turnover criteria - the same registry problem the MSME-MoSPI Statistical Business Register MoU of 29 September was signed to fix.
- Fixed panel: 3,819 enterprises tracked across 2024-25, 2025-26 and 2026-27. A panel removes composition effects, so its growth rate is the cleaner one.
- Realisation ratio: actual spending against stated intention. At 96.3%, Indian corporate capex plans are turning out to be close to reliable - which is what makes the forward-looking question worth asking.
- What it is not: this is the private corporate sector only. Central and State government capex, public sector enterprises and unincorporated business are all outside it.
- Why it matters now: private capex is the one component of demand that policy has struggled to revive, and until 2024 India had no forward-looking survey of it at all - this is the third round of a three-year-old instrument.
- The sister releases: the IIP (monthly, industry), the Index of Services Production (monthly, trial, from 29 September) and the Annual Survey of Industries (annual, registered manufacturing, released 30 September).
For UPSC: Private capital formation is the central question in Indian growth policy and this is the only instrument that asks firms directly about it. Use the survey design on statistical capacity and evidence-based policy, the manufacturing share on where investment is actually going, and the robotics split wherever automation and employment are the question - it is the first official Indian data on how firms are allocating to it.
What it is NOT: The release gives no aggregate capex figure, only sectoral shares and four absolute values, so the total has to be inferred. It gives no sample size for CAPEX 2024 or 2025 beyond the fixed panel, and no response rate. The robotics question is reported in two buckets with nothing in between, and no bucket definitions are published, so a genuinely bimodal distribution cannot be distinguished from a two-option questionnaire. No state-wise or firm-size breakdown appears. And no figure is given for how much of the intended capex is new capacity rather than replacement, which is the distinction that decides what it means for growth.
For Mains
Syllabus: GS3.1 · GS3.8 · Linkage L1
Anchor
Asked how much of their capital expenditure would go into robotics, 21.16 per cent of enterprises said more than 80 per cent and 78.84 per cent said less than 20. Those two figures add to exactly 100. Nothing was reported in between. Either Indian firms are genuinely divided into automators and non-automators with no middle, or the question had only two boxes - and the release does not publish the bucket definitions that would tell you which.
Substantiation (data)
The rest of the survey is solid. Manufacturing took 50.17 per cent of provisional 2025-26 capex at Rs 5,73,900 crore, up from 43.80 per cent in the first round; information and communication 16.38 per cent, electricity and gas 8.96, transport and storage 5.55. All four shares imply the same total of about Rs 11.44 lakh crore. The realisation ratio is 96.3 per cent, with actual spending of Rs 173.5 crore per enterprise in 2024-25.
Position
The instrument is the achievement. Until 2024 India had no forward-looking survey of private corporate investment intentions at all - the single most-discussed variable in growth policy was tracked only backwards, through completed accounts. Three rounds in three years, with a fixed panel and a published realisation ratio, now allow intentions to be checked against outcomes. A 96.3 per cent realisation ratio is the finding that makes the whole exercise credible.
Counterpoint
The headline from the fixed panel is modest: 3,819 enterprises report capex growth of 1.9 per cent over their actual 2024-25 spending. That sits oddly beside the August IIP, where capital goods production grew 16.9 per cent. The two measure different universes - IIP output includes exports and public-sector and government demand, while this panel is large private corporates - but the gap is the question anyone writing about an investment cycle has to answer.
Problematisation
A survey built on an MCA-derived frame inherits the quality of that register, which is the precise weakness MoSPI and the MSME Ministry signed an agreement to address two days earlier. Add that no response rate is published for any round, and the honest status of these numbers is: a well-designed instrument drawing on a frame the government is currently rebuilding, reporting a growth rate with no confidence interval attached.
Conclusion
Watch two things in CAPEX 2026: whether the panel growth rate stays near two per cent while capital goods output grows at double digits, and whether the robotics question reports a middle. The first decides whether the investment cycle is real or public-sector-led; the second decides whether that striking split was a finding or an artefact.
Deploys into: Private capital formation and the investment cycle · Statistical capacity and forward-looking surveys · Automation and employment · Reading survey design before survey results
Ministry of Statistics and Programme Implementation · 2026-10-01 · PRID 2317606 · PIB source ↗