How good is this, actually?

what works

Growth is real and steep — a 34-fold rise in a decade, off an almost non-existent base, with private firms now the majority contributor and roughly 80 destination countries.

what does not

The absolute number is small, much of it is components and sub-assemblies feeding foreign primes rather than complete Indian platforms, and exports are still only about 15% of what India produces. India remains the world's second-largest arms *importer* while celebrating export records.

strongest objection to this grade

Base effects cut both ways: a country that exported ₹686 crore in 2013-14 was starting from nothing, and component exports are how every arms industry begins. Judged as a trajectory rather than a level, this is closer to a B.

what works

Production has more than tripled since FY2014-15, the private share is rising, and Russian dependence in new procurement fell from 72% to 36% in a decade. Supplier diversification is a genuine structural achievement.

what does not

Production value counts output, not indigenous content — a platform assembled from imported kits counts in full. R&D spending at roughly 0.65% of GDP against China's ~2.4% is the binding constraint, and no amount of assembly capacity substitutes for it.

strongest objection to this grade

Tejas, BrahMos, the SSBN programme and the Agni series are real design-and-integration capabilities that most middle powers lack. The critique of 'assembly not design' understates what has actually been engineered.

what works

Capital outlay was fully utilised in FY2025-26 — historically, unspent modernisation money was a chronic embarrassment, and that has been fixed.

what does not

Pay and pensions crowd out equipment. Pensions alone are roughly a fifth of the budget and buy no capability. Headline budget growth of 175% since FY2014-15 substantially overstates capability growth once inflation and the revenue share are stripped out.

strongest objection to this grade

A large standing army on two contested land borders is a strategic choice, not an accounting failure. Personnel-heavy spending is what that posture costs, and comparisons with maritime powers are not like-for-like.

what works

Access-controlled corridor length grew 55-fold from 93 km in 2014. Four-lane-plus highway length more than doubled. On freight-relevant capacity rather than raw kilometres, this is the strongest infrastructure story in the dataset.

what does not

Annual construction peaked at 13,327 km in FY2020-21 and has since plateaued near 10,000-12,000 km. Headline network length keeps rising partly because existing state highways are re-designated as national highways — growth on paper, not on the ground.

strongest objection to this grade

Land acquisition in a democracy with strong property litigation is genuinely slower than in China. Comparing build rates without comparing legal constraints flatters the comparator.

what works

Operational airports more than doubled from 74 to over 160, with 93 aerodromes brought into service under UDAN and 663 routes started. Genuine geographic broadening of air access.

what does not

A material share of UDAN routes lapsed once the three-year viability-gap subsidy expired. Cumulative routes inaugurated is not the same as routes flying, and the headline is usually quoted as though it were. Several regional airports see minimal scheduled traffic.

strongest objection to this grade

Route churn is normal in regional aviation everywhere, and even discontinued routes leave behind a usable airstrip and terminal. The infrastructure outlives the subsidy.

what works

Turnaround time roughly halved to 48.84 hours, which now beats several developed-economy averages outright. Throughput hit a record 915 MT at major ports. This is a real operational win, not an announcement.

what does not

Turnaround improvement has stalled since FY2021-22 — four years to shave four hours, against a halving in the decade before. Capacity utilisation near 59% means further capacity spending has weak near-term returns. India's largest port still handles a fraction of Shanghai's volume.

strongest objection to this grade

Utilisation headroom is deliberate: ports built to just-in-time demand become bottlenecks the moment trade grows. Spare capacity is a feature when you expect to need it.

what works

From 248 km in 5 cities to over 1,000 km in 23 cities. Crossing 1,000 km in October 2025 made India the third-largest metro network in the world.

what does not

Ridership grew 4x while network grew 4.1x — passengers per kilometre have been flat. Several systems run far below the projections used to justify their capital cost. Building track is not the same as moving people, and the financing of most systems assumes ridership that has not materialised.

strongest objection to this grade

Metro ridership typically matures over 10-15 years as feeder networks and land use catch up. Judging lines opened in the last five years on current ridership is premature.

what works

Two corridors are operating with land allotted and units producing. Eleven nodes across UP and Tamil Nadu give the policy real geographic anchoring.

what does not

Roughly ₹10,855 crore grounded against ~₹70,000 crore committed — about 15%. 170 MoUs converted into 57 land allotments. 13,000 jobs against a stated potential of 50,000. Every headline number quoted for these corridors is a commitment, and commitments are being reported as if they were construction.

strongest objection to this grade

Industrial corridors run on 15-20 year horizons and these are seven years old. Early-stage conversion rates of one-in-three on MoUs are not obviously bad by international standards.

what works

Absolute manufacturing output has grown substantially, and specific sectors — pharmaceuticals, automotive components, and more recently electronics assembly — are globally competitive.

what does not

Manufacturing's share of GDP has been essentially flat since 2001 and has never approached the 25% target, through two governments and a decade of Make in India. Merchandise exports grew 0.93% in FY2025-26. Over 40% of the workforce is still in agriculture. India industrialised out of order — services grew before factories did — and the factories have not caught up.

strongest objection to this grade

Services-led growth is not automatically inferior; it delivered India's fastest sustained growth and its largest export surplus. The premise that every economy must pass through a manufacturing phase is itself contestable.

what works

Record $863 bn total exports. Services exports at $421 bn are a genuine global strength, with India a dominant supplier of IT and business services. Remittances are the world's largest.

what does not

Growth is almost entirely services-led — 8.71% for services against 0.93% for goods. Roughly half of services exports sit in one sector now directly exposed to AI automation of exactly that work. Concentration risk in the strongest part of the external account is the most under-discussed vulnerability here.

strongest objection to this grade

Services concentration is a strength being described as a weakness. India's IT sector has repeatedly absorbed technology shocks — offshoring, cloud, automation — and grown through each.

what works

Launch tempo roughly quadrupled. Cost-efficiency is genuinely world-leading — Chandrayaan-3 and Mangalyaan delivered results at a fraction of comparable mission budgets. A hundred missions and a working interplanetary capability.

what does not

Absolute cadence is an order of magnitude below China. Heavy dependence on PSLV is a single-point risk, underlined by third-stage failures on PSLV-C61 and PSLV-C62. Human spaceflight has slipped repeatedly.

strongest objection to this grade

Cost-per-outcome, not launch count, is the right metric for a programme with India's budget. On that measure ISRO outperforms nearly everyone.

what works

Electricity access moved from roughly 60% to effectively universal, and basic sanitation access rose faster than almost any comparable country has managed. On any honest reckoning these are the largest achievements in this dataset — they touched more lives than every defence and infrastructure line combined.

what does not

Access is not reliability: connection counts say nothing about hours of supply or voltage stability. Per-capita electricity consumption remains roughly a third of China's, which is the number that actually tracks industrialisation.

strongest objection to this grade

None worth making. This is the strongest part of the record and is routinely under-claimed relative to the infrastructure projects that photograph better.

what works

Infant and maternal mortality fell steeply. Life expectancy rose. Extreme poverty fell by hundreds of millions of people — the single most consequential number on this site.

what does not

Out-of-pocket health spending is among the highest in the world and medical costs remain a leading cause of household impoverishment. Female labour force participation is weak against every peer and is a direct drag on growth. Public health and education spending as a share of GDP remain low for the income level.

strongest objection to this grade

Poverty reduction at this scale is historically rare and arguably outweighs the composition critiques. A billion people got meaningfully richer.

what works

A functioning triad — land, air and sea-based delivery — which few states possess. Canisterised systems and MIRV development indicate a maturing second-strike capability.

what does not

Estimated at ~180 warheads against China's ~600 and growing. Every figure here is an external estimate with a wide error bar — India publishes nothing — so confidence in this entire assessment is low by construction.

strongest objection to this grade

Minimum credible deterrence is a deliberate doctrine, not a shortfall. Warhead parity was never the goal, and grading against China's count imports a yardstick India explicitly rejected.