The letter, authored by analysts Venugopal Garre and Nikhil Arela, the letter warns that if these constraints are not addressed, India risks under‑delivering on its potential.
Bernstein has laid out eight critical themes – from agriculture and energy to artificial intelligence, manufacturing, transport, subsidies, innovation and taxation – as a roadmap for investors to think about India’s next decade of risk and opportunity.
1. AI: From Data‑Centre Hub to “Permanent Consumer”?
On artificial intelligence, Bernstein argues that the real risk for India is still “not visible today, but… likely to be profound over time.”
India is positioning itself as a data centre hub, but the report warns this may overstate domestic value capture because a “significant portion of data center investment is import‑intensive” and global AI workloads are unlikely to shift materially to India given latency and regulatory constraints.
Crucially, “India does not own frontier AI models, unlike the US and China,” a gap Bernstein says is “not accidental-it reflects the absence of first‑generation technology platforms historically.” If Indian data continues to train global models without a domestic capability build‑out, “India risks becoming a permanent consumer in the AI economy.”
Bernstein has called for a China‑style protectionist approach in “all newer tech areas”, noting that “China’s development of its tech industry was through protection. India should follow this route.”
It also recommended that any global tech company offering AI models should be pushed to list in India, with half the value shared with the public.
2. Agriculture: Phase Down Rs 3-4 Trillion Subsidies, Restart Reforms
Calling agriculture “perhaps the clearest example of… structural inertia,” Bernstein notes that 42-45% of India’s workforce is still tied to a sector that contributes only ~15-16% of GDP. Average landholdings are below one hectare, nearly half the cultivated area remains monsoon‑dependent, and policy responses continue to revolve around loan waivers and input subsidies that “are not solutions; they are recurring responses to a system that has not been reformed.”
The rollback of the farm laws has made future reform “more difficult, but it has not made it any less necessary,” the note stresses, calling for a restart of the process “with a more calibrated approach.”
Key recommendations include scaling up irrigation to reduce weather dependence and tighter monitoring of capex‑to‑capacity conversion to prevent leakages, phasing down input subsidies on power and fertilisers, which now cost Rs 3-4 trillion annually, while protecting farmer incomes via higher and more predictable procurement prices and post‑procurement transfers that avoid distorting input use and retail prices.
3. Energy and EVs: “Auto OEMs Do Not Need PLI”
In energy, Bernstein spotlights a growing contradiction: India is wooing data centres and advanced manufacturing, yet power supply remains unreliable while distribution companies (discoms) continue to post Rs 5-6 trillion in accumulated losses. Industrial users, it notes, “bear a disproportionate burden through cross‑subsidisation, effectively funding inefficiencies elsewhere in the system.”
On the external front, India’s dependence is stark as the country imports 88% of its crude oil, with transportation accounting for more than half of demand, a vulnerability the brokerage calls “strategic”.
The report calls for a clear phase‑out timeline for internal combustion engine (ICE) vehicles, re‑design of incentives away from PLI benefits for capital‑rich incumbents towards demand‑side support for new entrants, subject to localisation thresholds.
“Auto OEMs do not need PLI. They are cash‑rich and should be held responsible for driving the transition,” it says, pointing out that even leading OEMs still import battery packs/cells from China instead of investing in R&D and capex.
“The government should consider taking stakes in cash‑rich companies who solicit PLI in the future,” it said.
4. Manufacturing: Late Entrant, Thin Supply Chains
Despite flagship schemes like PLI, Bernstein argues that manufacturing still reflects “intent without adequate depth.” Manufacturing’s share of GDP is stuck at ~16-17%, while employment remains concentrated in low‑productivity informal services, underscoring limited gains from the China+1 narrative.
Bernstein warns that “India cannot continue to enter industries decades after global leaders and expect different outcomes,” and argues for a decisive shift towards early identification of emerging sectors, including automation, robotics, advanced materials and AI‑integrated manufacturing, with capital and policy support committed before global supply chains are fully formed.
5. Rail, Roads and Aviation: “Design for Our Own Constraints”
On transport and urban infrastructure, Bernstein argues that India has over-indexed on aviation while underinvesting in railways, despite a clear comparative advantage in the latter.
The pace of high-speed rail (above 300 kmph) is described as “slow,” with just one major bullet train project underway, even though India has the financial capacity and engineering capability to build multiple corridors connecting key economic centres.
“If capital allocation continues to favour cars and planes over metros, buses and rail, India risks hard-wiring a transport system that is congested, expensive to run, and exclusionary for a large share of its citizens,” the brokerage warns.
6. Cash Transfers vs Capex: The “Very Expensive Way to Buy Growth”
In a sharp critique of ballooning welfare schemes, Bernstein says the renewed focus on cash transfers and subsidies—where a “small productive group of taxpayers is used to provide cash to the broader population to win votes”—reflects either “the ills of democracy or the economic failures of the nation.”
The note traces the rapid proliferation of unconditional cash transfers to women across “roughly a dozen-plus states,” estimating annual outlays of about Rs 1.7–2.5 trillion, or ~0.5% of GDP.
Bernstein acknowledges that these programmes are “not pure ‘freebies’” and that evidence shows well-designed cash transfers can lower vulnerability, smooth consumption, and boost local demand in the short term. “The issue is not that they do nothing; it is that, for an investment-starved emerging economy, it is a very expensive way to buy growth,” it says.
7. R&D, Innovation and Merit: Ambition Outpacing Spend
On R&D and innovation, Bernstein argues that India’s ambition significantly outstrips its investment effort. Total R&D spending at ~0.6–0.7% of GDP remains well below global benchmarks.
“Innovation ecosystems require sustained capital, high-quality talent, and institutional rigor,” the note stresses, calling for greater openness to global expertise and a sharper focus on merit-based hiring of students and professionals at top educational and research institutions.
8. Taxation, Cash and State Capacity: High Burden, Weak Services
The final theme centres on taxation and public service delivery, where Bernstein sees a “widening disconnect.”
“India’s tax burden is not low, yet the quality of public goods, healthcare, education, and urban infrastructure, remains weak in most cities,” the report notes, pointing to public health spending at ~2% of GDP and education at ~3% as being below what a country of India’s scale needs.
Broadening the tax base by removing exemptions, including for “economically significant but currently untaxed institutions such as certain political, religious, and sporting bodies,” to improve equity and create space for lower rates.
“Tax compliance ultimately depends on trust, and that trust is built when taxpayers see consistent improvements in public services,” the brokerage writes, underlining that improving service delivery is as important as widening the net.
Threading these eight themes together is an overarching concern about employment quality and the future of India’s services‑led growth model.
For two decades, India’s IT services, global capability centres (GCCs) and BPOs have employed 10-15 million people, powering the aspirational middle class that buys homes, flies more and drives consumption. But “Gen AI now challenges that template,” Bernstein warns.
“A meaningful share of the roles that lifted this cohort are directly exposed to automation,” while most of the surplus AI value in terms of models, platforms, and IP remains concentrated in the US and China, it notes. The risk: “India becomes a user of these technologies without capturing a commensurate share of the upside.”
The key question, Bernstein says, is whether the next leg of the growth story will “create more engineers, product builders and innovators, or does it mostly create more drivers, delivery staff and domestic help?” The answer, it argues, “will determine whether India’s rise remains a story of potential, or becomes one of realised capability.”