Many
Indias.
India is not one country, statistically speaking. Pick the state you live in below. In a few seconds you’ll see what your India actually looks like, on ten indicators that define everyday life, compared against the national average or any other state you want to stack it up against.
Then drag the decade slider to see how the country has moved over the past decade. The chapters at the top go deeper on each theme.
…
Select a state above.
India, over a decade.
Six national indicators, 2014 to 2025. Many of these numbers have shifted more in the past decade than in the four decades before. Drag the slider to see any year.
The story in your data.
Quiet shifts.
Several indicators that define everyday life have moved substantially over the past decade. Not every number has improved, and progress varies sharply by state, but the direction of travel on most is real. The figures below are rounded, draw from the sources cited on each card, and are meant to give a one-glance picture of the decade rather than a point estimate.
Figures are rounded; sources are cited on the corresponding metric cards above and in each chapter. Not every indicator has moved in the right direction — groundwater stress in the North-West has worsened, graduate youth unemployment remains stubborn, urban air pollution is improving but from very high bases. The overall picture is not triumphant. It is also, for most people, better than it was a decade ago.
A quarter of a billion fewer Indians live in multidimensional poverty than a decade ago.
By the most policy-relevant measure of deprivation across health, education, and living standards, the change of the past decade is genuinely without parallel anywhere in the world. The drivers are not mysterious: tap water (Chapter 02), sanitation (Swachh Bharat), insurance (PM-JAY, Chapter 04), housing (PMAY), gas connections (Ujjwala), bank accounts (Jan Dhan), and direct transfers (DBT) operating in concert. The state did not solve every problem; it solved enough of them in stacked combination to lift a number of people out of measurable destitution that exceeds the population of most countries.
Sources: NITI Aayog Discussion Paper on MPI 2024; World Bank Spring Meetings briefing 2025; Government of Kerala (2025). The MPI is the Alkire-Foster method developed by UNDP & OPHI, Oxford. India’s National MPI uses 12 indicators across health, education, and living standards. Different methodologies (income vs deprivation, World Bank vs national lines) report different magnitudes; the direction is consistent across all of them.
Eight states, half the economy.
Most national-average figures mask a steep gradient. A small group of states — Maharashtra, Tamil Nadu, Karnataka, Gujarat, Uttar Pradesh, Delhi, Telangana, and Haryana — produces disproportionately more of India's output than their population share suggests. They are also, on several of the indicators on this dashboard, significantly ahead of the national average. This is not new, but the gap has widened over the past decade, not narrowed.
One way to read the dashboard above: the "India average" for most indicators is pulled sharply in one direction by the performance of a handful of states. Your state’s position in that gradient matters more than the national figure, which is why the compare-with dropdown lets you pick any state as your benchmark. The point is not that eight states are "carrying" the country — every state has its own story — but that the national number is an average of many Indias, not a picture of one.
Read the chapters.
Each chapter goes into the history, the global peers, the best practices, and the policy levers for one theme. Pick one.
A note on the data.
Indian policy data is a mixed bag, and this dashboard reflects that. The last comprehensive Census was conducted in 2011. For a country whose population is estimated to have added over 20 crore people since then, that is a serious limitation. Most headline population-linked figures in public policy today — from state-wise demographic shares to per-capita estimates — are extrapolations. At best, educated guesstimates. We work with them because they are what the policy community works with.
Other datasets are genuinely current and high-frequency. The PLFS is now quarterly. ASER releases annually. CPCB publishes real-time air-quality data. NHA National Health Accounts comes out every two years. PM-JAY and Jal Jeevan Mission maintain live dashboards. Where the underlying data is timely and well-documented, we have used it directly.
A note on the Census: The 2021 Census has been indefinitely deferred and the 2026 enumeration is yet to begin. Many social and demographic indicators in Indian policy debates are still anchored to a 15-year-old baseline; treat all such figures as directional. Data sources used here: CPCB and IQAir for air quality; AQLI, University of Chicago for life-expectancy impact; CGWB for groundwater; Jal Jeevan Mission Dashboard for rural tap coverage; MoSPI PLFS 2023-24 for female labour force participation; ASER 2024 for Class 5 reading fluency; National Health Accounts 2021-22 for out-of-pocket expenditure; CEEW Heat Risk Index 2025 for heat risk; Census 2011 (extrapolated) for literacy and urbanisation — with the important caveat that the 2021 Census has been indefinitely delayed and many social, demographic, and urban indicators here ride on a baseline more than 15 years old.
State-level values are drawn from the most recent available rounds of each source. Where different authoritative sources report different numbers for the same indicator, we have rounded conservatively. Treat every figure here as a directional indicator of scale and trend, not a point estimate. The first decimal matters less than the order of magnitude.
The air
we breathe.
If you live in one of India’s northern cities, air pollution is not abstract. It is the grey haze on your terrace. It is the child who wakes up coughing in October. It is the schools that shut for days in November. And yet — for most of Indian history — it was an invisible problem, because there was no infrastructure to measure it.
That has changed, and quickly. India today runs one of the most extensive ambient air-quality monitoring networks in the developing world — around 1,500 continuous and manual stations, up from under 500 a decade ago. The CPCB’s SAMEER app publishes real-time data. The National Clean Air Programme (2019) set city-level PM reduction targets for 131 cities. The Commission for Air Quality Management (2021) coordinates the North Indian airshed across five states. The measurement and policy infrastructure is in place. The enforcement muscle is still building.
Approximate annual average PM2.5 concentration by state/city, 2024. Sources: CPCB; IQAir 2024 World Air Quality Report. Station coverage varies by state; figures are directional. Byrnihat’s prominence is partly an outlier driven by local industrial sources.
India’s population-weighted average PM2.5 sits far above what most of the developed world experiences day to day. For rough comparison: Pakistan ~67 µg/m³, Bangladesh ~66, India ~49, China ~29, United States ~8, European Union ~12, WHO guideline 5.
The peer example everyone watches is China. In 2013, Beijing’s annual average PM2.5 was around 90 µg/m³. A decade of sustained enforcement — coal controls, emission caps, real-time plant monitoring, ultra-low-emission retrofits, and air-quality-indexed bureaucratic accountability — brought Beijing below 40 µg/m³ by 2023. The trajectory is possible. It took China about a decade.
The scaffolding, assembled.
Over the past decade, India has built a serious policy apparatus for air. Not a perfect one; not one matched by uniform enforcement. But the scaffolding that didn’t exist in 2012 is now real.
The burden, in four numbers.
Annual India Fact Sheet, 2024
Four case studies.
No country has solved air pollution without sustained enforcement. These are four cases that Indian policymakers, journalists, and citizens should know in detail.
Ahmedabad’s Heat Action Plan
Though primarily designed for heatwaves, Ahmedabad’s early-warning-and-response model has been formally evaluated as averting over 1,000 premature deaths per year. The template is now in 140+ cities across India and has been studied by China, Pakistan, and the UAE. A demonstration that Indian cities can design and export public-health infrastructure.
Beijing’s air transformation
Annual PM2.5 in Beijing fell from ~90 in 2013 to ~35 µg/m³ by 2023. Mechanisms: coal out of the city core, ultra-low-emission retrofits at power plants, bureaucratic accountability tied to air targets, and an enforcement regime that imposed real costs on violators. Sustained political will over a decade is the active ingredient.
Graded Response Action Plan
The GRAP framework in Delhi-NCR — staged measures (from dust control to construction halts to school closures) triggered by AQI thresholds — is itself an Indian innovation now studied elsewhere. Its limitation is reactive design; its achievement is institutionalised response.
The United States’ Clean Air Act
The 1970 Clean Air Act and its 1990 amendments cut US SO&sub2; emissions by over 90% and PM2.5 by ~40%, while the economy tripled. The lesson most cited: strong standards plus credible enforcement work; voluntary programmes alone do not.
Six things that would move the curve.
On India’s air-quality framework, 2024
Ambient-air data in India relies on the CPCB monitoring network, state pollution-control boards, and independent validators like IQAir and AQLI. Figures cited here are population-weighted annual averages; individual stations vary widely.
A 2024 Nature Sustainability study (Mukherjee et al.) finds that part of the recent multi-year improvement in PM2.5 over the Indo-Gangetic Plain is attributable to favourable meteorology rather than to policy alone — wind patterns, monsoon timing, and atmospheric mixing have been atypically helpful. The policy architecture is real; the improvement story should be read with that caveat.
Attribution of deaths to air pollution is always based on statistical models; different research groups (ICMR, IHME, Lancet, Global Burden of Disease) produce estimates in a broadly similar range. Treat such figures as indicative of scale, not point estimates.
The most recent Census of India is 2011; some population-linked metrics extrapolate from that baseline. The series notes caveats where particularly material.
Primary data: Central Pollution Control Board; IQAir World Air Quality Report 2024; Air Quality Life Index (AQLI), University of Chicago; Commission for Air Quality Management.
Research: State of Global Air 2024; ICMR & Lancet Planetary Health India studies; CEEW Air Quality work; Shweta Narayan (HEAL); Down To Earth.
Policy: National Clean Air Programme (MoEFCC); Graded Response Action Plan documentation; Ex-Situ Paddy Straw Management Scheme; PM e-Bus Sewa guidelines.
The water
that isn’t there.
The imagery of the Indian water problem is usually Cape Town — that is, a city running dry. The Indian reality is structurally different. There is, in aggregate, enough water. The problem is that it is being extracted at rates the aquifers cannot sustain; that it is concentrated in the wrong places; that it is polluted in many of the places where it exists; and that the delivery infrastructure, until recently, covered only a fraction of rural households.
The Central Ground Water Board classifies assessment units by how much of annual recharge is being extracted. Below 70% is Safe. Between 70% and 90% is Semi-Critical. Between 90% and 100% is Critical. Above 100% is Over-Exploited — meaning the aquifer is depleting year-on-year.
Roughly one in four of India’s ~7,000 assessment units falls into one of the three concerning categories. The concentration is geographically stark: the Green Revolution heartland of Punjab, Haryana, Western Uttar Pradesh, and Rajasthan are the hotspots. Groundwater levels in parts of Punjab have declined by over 10 metres in the past two decades.
Groundwater extraction as a percentage of annual recharge, approximate 2024. Source: Central Ground Water Board, Dynamic Groundwater Resources of India 2024. Above 100% indicates the aquifer is losing water year-on-year. Figures are directional; the CGWB updates methodology periodically.
Per-capita annual freshwater availability: India ~1,400 m³, China ~2,000, United States ~8,800, Brazil ~28,000, Israel ~90, Bangladesh ~6,700. The international thresholds are 1,700 (water-scarce) and 1,000 (water-stressed). India sits just above the scarcity line and falling.
Two peer examples everyone watches: Israel — which built the world’s most advanced drip-irrigation and desalination system on half India’s per-capita water; and Spain, which has operated statutory aquifer-recharge law (Real Decreto 907/2007) with meaningful effect on its over-drawn basins.
15 crore homes, connected.
The Jal Jeevan Mission is not a small thing. Providing functional tap-water connection to an additional 12 crore rural households in six years is among the largest drinking-water-infrastructure programmes in global history. State-level execution has been uneven — Goa, Telangana, Puducherry, Himachal achieved near-100% coverage early; Jharkhand and West Bengal lagged. But the direction and pace are unambiguous.
Former Planning Commission member; independent JJM evaluations, 2023-25
Water, in four numbers.
The playbook, in four cases.
India’s water challenge has both domestic and international precedent for what works. These four cases each illustrate a different piece of the solution.
Drip irrigation and desalination
Israel’s water economy is the global benchmark. Drip irrigation (invented by Netafim, 1965) uses 30-50% less water than flood irrigation for similar yields. Israel now sources 85% of municipal water from desalination. Agricultural water productivity is roughly 3× India’s. Indian drip adoption has grown but remains under 10% of irrigated area.
Statutory aquifer management
Spain’s Real Decreto 907/2007 made river-basin plans legally binding, including groundwater abstraction caps. Over-drawn basins (La Mancha, Segura) stabilised over a decade of enforcement. The political economy was difficult; the outcome is the model India’s Atal Bhujal Yojana partly draws from.
Chennai’s rainwater harvesting rule
Tamil Nadu’s 2001 mandate requiring rainwater harvesting in all new buildings was one of the first such laws globally. Chennai’s groundwater levels measurably improved in the years following. Enforcement is uneven; the principle is sound. Replication in other water-stressed Indian cities has been limited.
Jal Jeevan Mission
From ~3 crore rural tap connections in 2014 to ~15.7 crore by March 2026 (~82% coverage). Execution varies by state but the aggregate pace has been among the fastest of any comparable programme globally. JJM 2.0 (launched 2026) explicitly pivots to quality, functionality, and O&M financing — the harder, less visible second half of the work.
Six things that would work.
Dynamic Groundwater Resources of India 2024
India’s groundwater data comes from the Central Ground Water Board’s Dynamic Groundwater Resources Assessment, updated periodically. The 2024 assessment is the most recent; methodology has been refined over successive rounds, so trend comparisons across decades require care.
JJM coverage data is collected at the state level and aggregated by the Ministry of Jal Shakti. Independent evaluations (NITI Aayog, World Bank) have broadly corroborated the pace, with notes on quality and functionality.
The most recent Census of India is 2011. Rural household totals extrapolate from that baseline. Treat all household-count figures as directional.
Primary data: Central Ground Water Board; Jal Jeevan Mission Dashboard; Ministry of Jal Shakti; Census of India 2011 village-level water sources.
Research: CEEW water security work; Mihir Shah Committee on Water Reforms; ATL Foundation; CPR water programme; Himanshu Thakkar, SANDRP.
Global reference: Netafim / Israel Ministry of Agriculture; Spain Ministerio para la Transición Ecológica; OECD water resource reports.
Where the
jobs aren’t.
Picture a 28-year-old engineering graduate from a tier-2 college in Madhya Pradesh. She has a BTech. She has a driving licence. She speaks English, Hindi, and passable Marathi. She has applied, over two years, for 140 jobs. She has received three offers — a BPO in Noida paying ₹18,000 a month, a bank-branch sales role in Indore paying a variable salary topping out at ₹25,000, and a data-entry role with no stated upper bound on working hours. She took none of them. She is studying for the SSC Combined Graduate Level examination for the third year running.
Multiply her by several million. That, statistically, is the Indian youth-employment problem.
It is easy to miss if you look only at the headline PLFS numbers. National unemployment: 3.2%. Labour Force Participation Rate: ~60%. Female LFPR: ~42% (up sharply from ~25% in 2019). Informal employment above 80%. These numbers, taken on their own, are not bad for an economy at India’s income level. The distress lies in the age- and education-disaggregated slice.
Unemployment rate by educational attainment, ages 15-29. Source: ILO-IHD India Employment Report 2024, based on PLFS 2022. The pattern looks counterintuitive: the more education, the higher the recorded unemployment. The explanation is that less-educated youth work more, but in subsistence or unpaid family labour — their distress is invisible in the statistic. Graduates are visibly "unemployed" because they are searching for the kind of formal job that should match their qualification.
India’s overall labour force participation rate of ~60% compares with Bangladesh ~60%, Vietnam ~74%, Indonesia ~68%, China ~68%, OECD average ~72%. The catch-up country for India is not the US or Germany; it is Vietnam and Bangladesh, both of which have absorbed rural-to-urban labour into formal manufacturing at a pace India has not yet matched.
Vietnam’s manufacturing share of GDP is ~24%, Bangladesh ~22%, Thailand ~27%. India’s has been stuck near 16-17% for a decade. This is the link between Chapter 15 (Growth) and Chapter 03 (Jobs) — the two are versions of the same question.
Scaffolding, assembling.
The policy machinery for India’s labour market has matured substantially over the past decade. The Skill India programme, the Labour Codes, PLI schemes, the Code on Social Security, e-Shram, ELI, and the 2024 internship scheme together represent a more comprehensive response to India’s employment challenge than any period since independence. Outcomes are uneven; the intent and architecture are stronger than before.
On India’s employment data, 2024
The arithmetic of Indian work.
Four case studies.
The Indian jobs question has both international and domestic precedent for what works. These four cases illustrate different pieces of the answer.
The manufacturing-as-employment model
Vietnam moved roughly 30 million workers from agriculture to formal manufacturing in 30 years. The ingredients: bilateral and multilateral trade openness, disciplined FDI attraction, high female labour participation, and TVET (technical and vocational education) integrated with industry demand. India’s NEP 2020 vocationalisation draws explicitly from this model.
Dual vocational training
Germany’s dual system integrates classroom learning with paid apprenticeships in 300+ recognised trades. Around 50% of German youth enter skilled work through this route rather than academic university. Youth unemployment has stayed below 10% even in downturns. India’s NAPS is the adapted framework; scale and quality are the gaps.
PLI and the mobile phone story
From ~2 mobile-manufacturing units in 2014 to ~300 in 2024; from 75% imported phones to near-zero; India is now the world’s second-largest mobile manufacturer. PLI created ~12 lakh direct and indirect jobs across 14 sectors. The employment multiplier has been highest in labour-intensive sub-sectors — textiles and assembly — where the next-phase PLI extensions are targeted.
Karnataka’s Shakti Scheme (indirect)
Free bus travel for women across Karnataka from June 2023 has correlated with a ~12 percentage-point rise in the state’s female LFPR. Transport access is labour-market infrastructure. The lesson: mobility and safety are often the binding constraints on female workforce participation.
Six things that would move the curve.
On India’s employment challenge, 2024
Labour-market data in India primarily comes from the Periodic Labour Force Survey (PLFS, ongoing since 2017) replacing the older NSSO Employment-Unemployment rounds. Independent estimates from CMIE, ILO, and academic researchers often vary; each methodology has its own strengths.
PLFS uses a seven-day reference period for activity status, which undercounts underemployment and captures subsistence work as "employed." CMIE uses a different reference window. Treat all employment and unemployment figures as directional; the relative patterns (age, education, gender) are more robust than the absolute levels.
The most recent Census of India is 2011; labour-force totals extrapolate from that baseline. The delayed 2021 Census will eventually revise many of these estimates.
Primary data: PLFS 2023-24 (MoSPI); ILO India Employment Report 2024; e-Shram portal; CMIE Consumer Pyramids.
Research: Santosh Mehrotra (JNU); Ashwini Deshpande (Ashoka); Jean Drèze on rural labour; Radhicka Kapoor (ICRIER); Amit Basole (Azim Premji University).
Policy: Ministry of Labour & Employment; Labour Codes 2019-20; Code on Social Security 2020; PLI scheme documentation; Budget 2024-25 and 2025-26 employment provisions.
Enrolled,
but learning?
Picture a nine-year-old in Class 5 in rural Bihar. She sits in a government school, attends most days, has a textbook, has a teacher. By almost every statistical measure, Indian education has reached her: she is enrolled, she is not dropping out, she receives a midday meal, she is counted in the system.
Hand her a Class 2 Hindi storybook and ask her to read aloud. According to the 2024 Annual Status of Education Report, she has about a one-in-two chance of doing so fluently. Give her a two-digit subtraction problem. A similar probability applies.
This is the central empirical fact of Indian education today. The access problem — how do we get children into schools — has been substantially solved. The learning problem — what happens once they are there — has not.
It is important to register that the access achievement is not small. Dropout rates at the secondary level have fallen from ~17% in 2012 to under 10% today. Girls’ enrolment has caught up and exceeded boys at every level below higher secondary. The mid-day-meal scheme reaches ~12 crore children. Samagra Shiksha has consolidated three earlier schemes. PM Shri has begun upgrading ~14,500 schools as model institutions.
Percentage of Class 5 children in government schools who can read a Class 2-level text fluently. Source: ASER 2024 (released January 2025). Figures are from a volunteer-administered, nationally representative household survey; ASER methodology is well-documented but inherently carries a survey range. Kerala and HP consistently outperform; Bihar and UP consistently lag.
India spends around 2.9% of GDP on public education. For comparison: Brazil ~6%, South Korea ~5.1%, Vietnam ~4.2%, OECD average ~5.2%, US ~5.4%. The 1966 Kothari Commission recommended 6%. India has missed that target for six decades.
The Vietnam example is particularly instructive: Vietnam spends ~4.2% of GDP on education at roughly a tenth of India’s per-capita income, yet consistently outperforms middle-income peers on the PISA international assessments. The lesson: spending isn’t destiny, but under-spending makes the structural problem harder.
The architecture, rebuilt.
The policy machinery for Indian education has undergone more structural change in the past 6 years than in the preceding 60. NEP 2020 reoriented the framework. NIPUN Bharat made foundational learning the stated first priority. Samagra Shiksha consolidated scheme delivery. PM Shri created a model-school network. The vocationalisation of the curriculum has begun. Enrolment has stayed near-universal. The challenge now is converting architecture to outcomes.
ASER 2024 field report
Enrolment vs learning.
Four models.
The PISA over-performer
Vietnam’s PISA performance is the most-cited developing-country education story. Despite per-capita income roughly a tenth of developed economies, Vietnamese 15-year-olds score comparably to OECD averages in mathematics and reading. The ingredients: disciplined teacher training, intensive focus on foundational skills, and centralised curriculum quality with local implementation. India’s NEP 2020 draws from similar principles.
The Teaching at the Right Level model
Pratham Education Foundation’s Teaching at the Right Level (TaRL) methodology — grouping children by learning level rather than age/grade — has been rigorously evaluated by J-PAL and others. TaRL interventions produce learning gains equivalent to roughly 60-100% of a grade level in 3-6 months. Scaled across multiple Indian states through government partnerships; adapted in Zambia, Cote d’Ivoire, and elsewhere.
Teacher professionalisation
Finland’s education outcomes are built on a structural commitment: every teacher holds a master’s degree, teaching is a prestigious profession with strong pay and autonomy, and curricula are locally adapted within national frameworks. The lesson: no education system exceeds the quality of its teachers. India’s challenge is more scale than Finland faces, but the principle holds.
Universal foundational achievement
Kerala’s Class 5 reading fluency at 66%, well above the national 50%, reflects decades of sustained investment in primary education, female literacy, and accountability in government schools. The model is neither fast to replicate nor a matter of a single policy — it is a multi-generational commitment to basic education being the state’s first priority.
Six things that would close the gap.
Report of the Education Commission, 1966 — recommended 6% of GDP on education
India’s education data comes from multiple sources with slightly different methodologies: UDISE+ (administrative data, universal coverage), ASER (household survey, volunteer-administered), National Achievement Survey (sample assessment). Each has strengths and gaps.
ASER 2024, the most recent systematic household-based learning-outcomes dataset. Enrolment figures from UDISE+ are near-complete; learning-outcome figures are survey-based and carry inherent sampling uncertainty.
The delayed Census (last comprehensive in 2011) affects some extrapolations of school-age population totals. Treat specific figures as directional indicators of the pattern rather than point estimates.
Primary data: ASER Centre (Pratham); AISHE — All India Survey on Higher Education; UDISE+ data; NCERT National Achievement Survey; Economic Survey 2024-25 education chapter.
Research: Rukmini Banerji (Pratham); Karthik Muralidharan (UC San Diego); Wilima Wadhwa (ASER); Abhijit Banerjee and Esther Duflo on remedial learning (J-PAL).
Policy: Ministry of Education; NEP 2020; NIPUN Bharat guidelines; Samagra Shiksha framework; PM Shri scheme documentation.
Sick, and
paying for it.
For most of Indian history, catastrophic illness in a poor household meant economic ruin. Estimates from the 2000s suggested 5-6 crore Indians were pushed below the poverty line each year by medical expenses alone. This was not a rare outcome; it was the baseline condition of the Indian health system.
That baseline has been shifting, and at pace. Out-of-pocket expenditure (OOPE), the share of health spending that households bear directly, has fallen from ~64% in 2013-14 to around 39% in 2021-22. Fewer Indians today fall into poverty from a single hospital admission than at any point since independence.
Several programmes have contributed simultaneously. Pradhan Mantri Jan Arogya Yojana (PM-JAY), launched 2018, provides cashless in-patient cover up to ₹5 lakh per family per year to around 55 crore Indians — the world’s largest government-funded health insurance scheme. Ayushman Arogya Mandirs (formerly Health & Wellness Centres), of which over 1.7 lakh are now operational, have expanded primary care well beyond the historical reach of the public system. Jan Aushadhi stores — around 14,000 of them now — offer generic medicines at a fraction of branded prices. None of these eliminates the Indian health challenge; all of them meaningfully narrow it.
Out-of-pocket expenditure as % of total health expenditure, India. Source: National Health Accounts (NHA) estimates, Ministry of Health & Family Welfare. The 2021-22 round was released September 2024. The drop from 64.2% to 39.4% is one of the steepest documented improvements in Indian health economics in the independence era.
Out-of-pocket health spending as % of total health expenditure: Thailand ~12%, United Kingdom ~13%, Indonesia ~30%, China ~35%, India ~39%, Bangladesh ~69%, Pakistan ~55%. India has closed most of the gap with middle-income peers and is now broadly in China’s range.
The peer example Indian policymakers watch is Thailand’s 2001 Universal Coverage Scheme, which drove OOPE from ~34% to ~11% in a decade. The Thai model combines tax-financed universal coverage with a fixed capitation payment to providers, no co-payments, and tight cost controls. PM-JAY shares some features and differs in others (eligibility, financing mix). The trajectory is possible.
Architecture, assembled.
Over the past decade, India has built the scaffolding of a universal-coverage health system. PM-JAY for catastrophic care. Ayushman Arogya Mandirs for primary care. Jan Aushadhi for affordable medicine. Ayushman Bharat Digital Mission for records. The mix is imperfect, varies by state, and leaves real gaps — but it is meaningfully different from what existed in 2012.
On the NHA 2021-22 data
The improvement, quantified.
Four models.
Universal coverage done well
Thailand’s 2001 Universal Coverage Scheme drove OOPE from ~34% to ~11% in a decade. Tax-financed, zero co-payments, fixed capitation to providers, tight cost controls. Thailand achieved universal coverage at ~3.8% of GDP on health — an efficiency India should be studying closely.
Community-based health insurance at scale
Rwanda’s Mutuelles de Santé community-based insurance now covers ~90% of the population. Under-five mortality has fallen from ~220 per 1,000 in 1995 to ~40 in 2022. A small country, but an instructive case in coverage-expansion under resource constraints.
CoWIN and the vaccine build-out
India administered over 220 crore Covid vaccine doses by 2022 — the fastest and largest vaccination campaign in human history. CoWIN has since been offered as DPI to other countries. The case demonstrates that the Indian state can execute at extraordinary scale when the political and financial commitment align.
Public health infrastructure as state priority
Tamil Nadu has for decades invested more per capita in public health than most Indian states. Result: infant mortality rate ~14 per 1,000 vs national average ~32; institutional delivery rate ~99%; OOPE below national average. Kerala’s story is similar. The conclusion: sustained state-level commitment compounds.
Six things that would complete the transition.
On India’s universal coverage architecture, 2024
India’s health-expenditure data comes from the National Health Accounts (NHA), produced by the National Health Systems Resource Centre. The 2021-22 round was released September 2024 and is the most recent comprehensive estimate. Previous rounds use comparable but not identical methodology.
PM-JAY beneficiary counts include state scheme convergence and vary across sources; the figure of ~55 crore covers the core AB-PMJAY programme with an additional coverage layer from state-level schemes.
The most recent Census is 2011; household-level estimates of health-expenditure burden rely on NSSO rounds and NFHS, each with their own sampling frame. Treat all such figures as directional.
Primary data: National Health Accounts 2021-22; NFHS-5 (2019-21); National Health Authority (PM-JAY dashboard); Ayushman Bharat Digital Mission portal.
Research: Public Health Foundation of India; K. Srinath Reddy’s health systems analyses; Lancet Countdown India; Observer Research Foundation health team.
Global reference: WHO Global Health Expenditure Database; Thailand Universal Coverage Scheme evaluations; Rwanda Ministry of Health reports.
The heat
is on.
Phalodi, Rajasthan, 26 May 2024. The mercury hits 50.5°C. Across May and June that year, much of North India runs above 45°C for days on end. Construction workers collapse. Electoral workers die on polling duty. Delhi registers its highest-ever night-time temperature: 35.2°C at 3 am.
2024 was, officially, India’s hottest year since the India Meteorological Department began keeping records in 1901. Nine of India’s ten warmest years on record have fallen in the past fifteen.
What is different about heat, as a climate story, is that it is immediate. Air pollution’s toll is modelled. Groundwater depletion is slow. Heat is the child who cannot concentrate in an unventilated classroom; the vegetable vendor whose stock perishes by noon; the delivery worker whose phone screen shuts off.
India’s response has been unusually pioneering for a developing economy. Ahmedabad’s 2013 Heat Action Plan was the first in South Asia; a 2018 Lancet Planetary Health study found it averted over 1,100 deaths annually. The template has since been adopted by 140+ Indian cities. In 2024, the Ministry of Home Affairs formally notified heatwaves as a disaster under the State Disaster Mitigation Fund, unlocking structural funding. Kerala in 2025 became the first state to mandate binding work-hour restrictions during heatwave days.
Heatwave station-days per year, IMD data. Source: IMD Annual Climate Summaries 2015-2024. A single station recording 10 heatwave days counts as 10 station-days. The 2024 figure is more than 2× any previous year on record.
India sits among the world’s most heat-exposed large economies. Per-capita heat exposure is comparable to the Gulf and parts of the Sahel, but at a population scale none of those regions approach. Roughly a quarter of global heatwave-related deaths since 1990 are estimated to have occurred in India (figures vary across studies).
The comparison frame is not the US or Europe. It is Pakistan, Bangladesh, Nigeria, Egypt. Among these, India’s Heat Action Plan framework has become the reference adapted by others. Miami, Athens, Phoenix, Santiago have appointed Chief Heat Officers; no Indian city has yet done so at that formal level.
Heat Action, institutionalised.
India’s response to heat has been one of the clearest examples of adaptive policy in a developing country. HAPs expanded from one city (2013) to 140+ by 2024. The formal notification of heatwaves as a disaster in 2024 was a structural shift that unlocks longer-term financing. Kerala’s 2025 occupational-heat rule is the first of its kind. None of this makes India heat-safe; all of it makes India measurably more heat-prepared than it was a decade ago.
On the 2025 Heat Risk Index
The climate frontline, arrived.
Four case studies.
The first HAP in South Asia
Designed by the Indian Institute of Public Health with NRDC and the city administration. Colour-coded alerts, SMS warnings, training of ASHAs, cooling-centre maps, worker-hour protections. A 2018 Lancet Planetary Health evaluation estimated ~1,100 deaths averted annually. Adapted by 140+ Indian cities and studied by China, Pakistan, UAE.
The accountability model
Miami (2021) appointed the world’s first Chief Heat Officer. Athens, Phoenix, Santiago, Dhaka followed. A single accountable official with budget and KPIs for city-wide heat resilience. No Indian city has yet appointed one at the same formal level; institutionalising this role would replace goodwill with structure.
Cool Roofs at scale
First Indian state-level subsidised cool-roof commitment. Cool roofs cost ₹50-100/m² and reduce indoor temperatures by 2-5°C. The policy targets 300 million square feet by 2028. Replication across other states would be among the cheapest public-health interventions available.
Heat warning systems
After the 2003 European heatwave killed ~70,000 people, France, Spain, Germany, Italy built integrated heat-warning systems. European heat mortality has fallen markedly despite rising temperatures — an adaptation success. India’s HAP framework draws on the same architecture.
Six things that would save lives.
Updated Heat Wave guidelines, 2024
India’s heat mortality is systematically undercounted. The government toll is based on state-level reporting; independent monitors use hospital admission data and media reports and arrive at higher figures. The real number is almost certainly higher than the official one.
Temperature and heatwave data from IMD use specific definitions that differ slightly from peer services abroad; comparisons across countries require care.
HAP counts include formal plans notified by city administrations; quality and implementation vary. The ~140+ figure is a count of plans, not necessarily of functioning ones.
Primary data: IMD Annual Climate Summary 2024; CEEW Heat Risk Index 2025; HeatWatch Struck by Heat Report 2024.
Research: Azhar et al, Lancet Planetary Health 2018 (Ahmedabad HAP); Center for Policy Research HAP audit (2023); Yale Program on Climate Change Communication.
Half the
workforce?
India has 73 crore women. In 2019, about one in four of the working-age ones was in the labour force. In 2023-24, it was about two in five. Roughly 10 crore Indian women have entered or re-entered the labour force in a five-year window. That is the single largest shift in the composition of the Indian workforce since the 1970s.
The caveats are real. Much of the increase is concentrated in rural self-employment — own-account agricultural work, animal husbandry, home-based enterprises. Urban FLFPR has risen more modestly. The Indian woman who is visibly in the labour market — the female-headed household, the returning migrant, the gig worker, the shopfloor operator — remains underrepresented relative to Vietnam or Indonesia.
But the direction is clear. The 106th Constitutional Amendment was passed in 2023 and gazetted on 16 April 2026 (now in force) reserving one-third of Lok Sabha and state assembly seats for women, the first major structural reform after 27 years of attempts. Implementation is still tied to post-Census delimitation (targeted 2029 at earliest); the 131st Amendment to delink failed in Parliament on 17 April 2026. The reform is law; the implementation is the unresolved question.
Female Labour Force Participation Rate (15+), PLFS 2023-24. Note: Himachal Pradesh’s top ranking reflects a large share of rural women in family-owned agriculture. The pattern in Delhi and Bihar is low participation, not low work — Indian women work extensively; much of it is unpaid and unrecorded.
Female Labour Force Participation Rate: Rwanda ~85%, Vietnam ~70%, China ~61%, United States ~57%, Indonesia ~53%, Brazil ~53%, India ~42%, Bangladesh ~38%, Pakistan ~24%. India has closed the gap with Bangladesh; Vietnam and China remain the aspirational benchmarks.
The Vietnamese and Bangladeshi cases are the most structurally comparable. Both absorbed large rural female workforces into formal manufacturing (garments in Bangladesh; electronics, garments, agro-processing in Vietnam). India’s next FLFPR leg is likely to require the same pattern — and the PLI, National Manufacturing Mission, and textile-focus-product schemes are the policy instruments being prepared for it.
Ten crore women, returning.
The Indian women’s-workforce story is the clearest case of rapid structural change in the past decade of labour-market data. From ~25% FLFPR in 2019 to ~42% in 2023-24. From ~6% of Lok Sabha MPs being women in 2004 to ~14% in 2024, with 33% reservation now legally required. From mandatory maternity leave of 12 weeks to 26. None of this makes India gender-equal; all of it makes India measurably more gender-inclusive than it was a decade ago.
On Indian panchayat reservations evaluations
The missing half, returning.
Four models.
The gender-quota maximalist
Rwanda is the only country where women hold the majority of parliamentary seats (~61%). Constitutional minimum quota is 30%; actual is much higher. Rwanda’s FLFPR is ~85%, the highest in the world.
Manufacturing-driven FLFPR
Bangladesh’s garment industry employs ~4 million women directly. FLFPR rose from ~10% in 1980 to ~38% today, primarily through formal manufacturing absorption. The lesson for India: labour-intensive manufacturing is the historic on-ramp for women’s workforce participation at scale.
Transport as labour infrastructure
Karnataka’s free state-bus-travel scheme for women (June 2023) correlates with a ~12 percentage-point rise in state FLFPR over the following year. Mobility, safety, and transit access are often the binding constraints on women’s workforce participation. Tamil Nadu, Delhi have parallel schemes.
Women’s workdays at scale
Women have historically accounted for ~55% of MGNREGA workdays nationally. For a significant share of rural women, this rural employment guarantee was their only direct monetary income. The largest formal-work channel for rural women in independent India’s history. Whether the late-2025 transition to VB-G RAM G (125 days, but Centre-State 60-40 funding) preserves this share is the open empirical question.
Six things that would close the gap.
Speaking in the Constituent Assembly, 1947
FLFPR is from PLFS, replacing older NSSO rounds. Methodology shifts mean precise comparisons with pre-2017 figures require care — but the directional trajectory (~25% in 2019 to ~42% in 2023-24) is robust. The PLFS 2023-24 round also reclassified women collecting firewood, fetching water, and working on family farms as participating in the labour force, which mechanically lifts the rural number; the urban figure has barely moved. Both are signal: women are doing more economic work, and urban formal-employment opportunities for women remain stubborn.
India’s sex ratio at birth, per NFHS-5 (2019-21), is 929 females per 1,000 males — an improvement from 919 in NFHS-4, but still well below the natural figure of around 952. Son preference is weaker than it was a generation ago; it has not been eliminated. Northern states (Haryana, Punjab) continue to lag the southern (Kerala, Tamil Nadu) by margins that have proved durable across surveys.
Much of the recent FLFPR rise is in rural self-employment. Urban FLFPR has risen more modestly. Informal work is harder to measure; treat specific state-level figures as directional.
The most recent Census is 2011; population totals extrapolate from that baseline.
Primary data: PLFS 2023-24 (MoSPI); WEF Global Gender Gap Report 2025; MoSPI Time Use Survey 2019; NCRB Crime in India 2022.
Research: Ashwini Deshpande (Ashoka); IWWAGE; Esther Duflo on panchayat reservations; Rohini Pande on representation.
The farm
question.
The Indian farmer exists in the political imagination as a single figure: a small landholder, at the mercy of the monsoon. The reality is more varied. The Indian farmer is also a tenant cultivator invisible to most government schemes; a landless labourer whose household runs on MGNREGA wages; a member of a Farmer Producer Organisation negotiating contract supply; a drip adopter who exits paddy for horticulture.
The Agricultural Census of 2015-16, the most recent complete count, recorded 146 million operational holdings at an average size of 1.08 hectares. The 2021-22 Census has been delayed; 2015-16 remains the baseline most current analysis extrapolates from. About 86% of operational holdings are under 2 hectares.
Share of operational holdings by size class. Source: Agricultural Census 2015-16 (latest available). Fragmentation through inheritance has compressed holdings to a point where modern investment — tubewells, tractors, cold chains — is economically unviable for most single operators.
India’s agricultural sector employs ~45% of the workforce but produces ~18% of GDP. In peer economies the gap is narrower: China (~22% workforce, ~8% GDP), Brazil (~8%, ~7%), Vietnam (~30%, ~12%), US (~1%, ~1%). The transition China, Vietnam, and earlier Japan managed — moving rural labour into formal manufacturing — is the structural shift India is still completing.
Serious scaffolding, assembled.
The policy machinery for Indian agriculture has expanded dramatically over the past decade. Not all of it has worked well. Some of it has worked far better than anticipated. All of it combined represents the largest farmer-welfare architecture independent India has assembled.
On Indian agricultural policy, 2024
The rural arithmetic.
Four models.
High-intensity, high-value agriculture
The Netherlands produces roughly $100 billion of agricultural exports annually from a country smaller than Tamil Nadu, through glasshouse horticulture, intensive dairy, and value-added food processing. The lesson: agricultural prosperity is about value-add, not land area. Indian horticulture (already 350+ million tonnes annually) is the closest structural parallel.
Drip and water productivity
Israel’s Netafim-pioneered drip irrigation uses 30-50% less water than flood irrigation. Israeli agricultural water productivity is roughly 3× India’s. India’s drip adoption has grown but remains under 10% of irrigated area. The biggest-lever intervention available, given the water chapter.
Dairy cooperatives at scale
Operation Flood (1970-96) took India from milk-scarce to the world’s largest milk producer through Amul-style cooperative architecture. The FPO model for agriculture is the direct descendent of this blueprint. The institutional technology exists; replicating the Amul trajectory in other commodities is the challenge.
Township-village enterprises
China moved roughly 200 million people out of agriculture in 30 years, largely through rural non-farm enterprises co-located with agriculture. The lesson: the farm problem is often solved not on the farm but at its edge — in food processing, logistics, agri-services. India’s Agriculture Infrastructure Fund targets the same layer.
Six things that would work.
People’s Archive of Rural India
Agricultural data in India comes from multiple sources: the Agricultural Census (2015-16 is the most recent complete round; 2021-22 is delayed), NSSO Situation Assessment Surveys (2019 is most recent), Ministry of Agriculture administrative data, and NCRB for accidental deaths/suicides. Different sources produce slightly different figures.
The farmer-household-income figure of ~₹10,200/month (NSSO 2019) is the most recent authoritative household-level estimate. Cost of cultivation and net income vary dramatically by state, crop, and tenure status.
NCRB farmer-suicide data has been contested on methodology; the figure has been broadly stable at ~11,000 farmer and farm-labourer suicides per year for two decades. Treat all such figures as directional.
Primary data: Ministry of Agriculture & Farmers’ Welfare; NSSO 77th Round (2019); Agricultural Census 2015-16; NCRB Accidental Deaths & Suicides 2022.
Research: Ashok Gulati (ICRIER); Himanshu (JNU); P. Sainath; People’s Archive of Rural India; PRS Legislative Research; Mekhala Krishnamurthy on agricultural markets.
Global reference: Netherlands Ministry of Agriculture; Netafim / Israel Ministry of Agriculture; World Bank agricultural productivity reports.
The city is the
future.
The twenty-first-century Indian story is, increasingly, an urban one. The growth story, the jobs story, the climate story, the middle-class story — all of it concentrates in cities. And yet India’s cities, in both physical and political terms, have long been the orphans of the Constitution. The 74th Constitutional Amendment Act of 1992 was meant to empower urban local bodies with 18 specific functions; three decades on, implementation remains uneven.
What has changed in the past decade is the structural seriousness of the Centre’s urban engagement. Smart Cities Mission. AMRUT I and II. Jal Jeevan Mission Urban. Swachh Bharat Urban. PM-AWAS Urban. PM e-Bus Sewa. National Infrastructure Pipeline’s urban component. The Ministry of Housing and Urban Affairs’ allocation in the 2025-26 Budget rose ~17% year-on-year to around ₹96,800 crore. Urban, in policy terms, is no longer an afterthought of state-level politics.
India’s urban population share over time, by official Census definition. Source: Census 1951-2011; World Bank and NSO estimates for 2024 and 2036 projections. Two important caveats: the 2021 Census has been indefinitely delayed and 2024 estimates are from sample surveys and projections; and India’s administrative definition of urban is narrower than spatial-density definitions used elsewhere, suggesting the actual urban share already approaches 50%.
India’s Total Fertility Rate fell to 2.0 in 2024 — below the replacement level of 2.1. Urban TFR is around 1.6; the South and West are well below replacement; only Bihar (~3.0), Meghalaya (~2.9), Uttar Pradesh (~2.4), and a few others remain above. India is projected to peak around 2060 and then begin a slow demographic decline. The implications for school capacity, urban planning, the labour market, the care economy, and old-age support are all built into the next 15 years of policy choices, even if they are barely visible yet in public debate. The country urbanising fastest is also the one ageing fastest.
Municipal own-revenue per capita (approximate USD equivalent): Mumbai $540, Delhi MCD $380, average Indian ULB $98, vs Singapore $7,800, Shanghai $8,900, New York $14,200, Jakarta $1,600, Bangkok $1,400. The gap is not population or wealth alone — it is legal authority, property tax bases, and fiscal devolution.
Property tax as % of GDP: India ~0.15%, OECD average ~1.1%, low-income countries average ~0.3%, middle-income ~0.5-0.6%. India collects property tax at a fraction of peer rates. The 16th Finance Commission has an opportunity to change this.
Urban, reclaimed.
The past decade’s urban missions represent the most concerted Centre-led urban programme in independent India’s history. Smart Cities. AMRUT I and II. Jal Jeevan Mission Urban. Swachh Bharat Urban. PM-AWAS Urban. Each has had uneven execution. Taken together, they have fundamentally altered the physical fabric of many Indian cities and the policy expectation of what cities can deliver.
Chair, High-Powered Expert Committee on Urban Infrastructure, 2011
The urban arithmetic.
Four urban models.
Land-finance and urban planning
Singapore finances much of its urban infrastructure through land-value capture — the government leases state-held land for 99 years and recycles revenue into infrastructure. The Delhi Mumbai Industrial Corridor and GIFT City draw from this template. Legal authority to do this at scale remains the Indian constraint.
Municipal execution at its best
Indore has ranked #1 in Swachh Survekshan for seven consecutive years. Door-to-door segregated waste collection covers nearly 100% of households. The model was built through empowered mayors, stable municipal commissioners, and sustained political commitment. Surat’s post-1994 plague transformation is the closest parallel.
Doubling revenue through GIS
Bengaluru’s 2008 Self-Assessment-Scheme + GIS-based property rolls doubled property tax collections in four years. Patna and Lucknow have adapted the model; several other cities are in progress. The technology is not novel; the administrative will is.
Municipal bonds at scale
Tamil Nadu’s TNUIFSL pools smaller cities’ borrowing capacity to issue investment-grade municipal bonds. Roughly 10 Indian cities have issued bonds since 2017, raising ~₹7,600 crore cumulatively. Small relative to need; proof of concept at scale.
Six things that would unlock the urban future.
Urban data in India is famously fragmented. The Census (2011, with 2021 delayed) is the foundation. MoHUA maintains city-level scheme dashboards. World Bank and OECD produce comparative data with their own methodologies. State-level urban statistics vary in quality.
Urban population projections (including the 2036 figures) are from World Bank and NSO estimates, using demographic transition assumptions that may need revision after the next Census.
Municipal fiscal data is patchy; the NIPFP and Janaagraha City Finance Reports are the most authoritative independent sources. Figures are directional.
Primary data: World Bank, Gearing Up for India’s Urban Transformation (2024); Ministry of Housing & Urban Affairs; 15th & 16th Finance Commission documents.
Research: Isher Judge Ahluwalia HPEC 2011; CPR (Partha Mukhopadhyay, Mukta Naik, Shubhagato Dasgupta); IIHS; Janaagraha City Finance Report; Citizen Matters.
Justice
delayed.
The Indian Supreme Court sits on around 88,000 pending cases. The 25 High Courts, on approximately 62 lakh cases. The subordinate courts that most Indians actually encounter — district, magistrate, family, special courts — sit on over 4.7 crore. Altogether, roughly 5.5 crore pending cases, or one for every 26 Indians.
The problem is not that courts have stopped working. In 2023 alone, Indian courts disposed of around 1.8 crore cases and received around 2 crore new filings. Disposals are running at historically high rates. The gap of ~20 lakh new cases per year adds to the backlog; it does not subtract from it.
What has changed in the past decade is the seriousness of the reform. eCourts has computerised over 21,000 subordinate courts. NJDG provides real-time pendency dashboards. The Mediation Act 2023 institutionalised pre-institution mediation. The 2024 criminal codes replaced the IPC, CrPC, and Indian Evidence Act with timeline-bound procedures. In 2025, the Supreme Court unlocked ad-hoc judges under Article 224A for up to 10% of High Court sanctioned strength. Whether any of this compounds into actual backlog reduction is the empirical question of the next five years.
Pending cases across all court tiers, in lakhs (hundred-thousands). Source: NJDG 2024 estimates, state-level judicial statistics. Uttar Pradesh accounts for roughly a fifth of all Indian pendency. The Allahabad High Court alone is among the most burdened courts in the country.
Judges per million population: United States ~150, United Kingdom ~51, Australia ~41, Germany ~38, South Africa ~24, India ~21. The 1987 Law Commission recommendation of 50 per million has not been met in four decades.
On case clearance rate, peer benchmarks diverge widely. Countries with functional mediation and specialisation systems (Singapore, Germany) dispose of civil cases in months; India’s average extends to years. Singapore’s 1990s judicial reform package — case management, mandatory mediation, specialised courts, performance metrics for judges — is the most-cited peer case.
Reform, compounding.
Indian judicial reform has historically been incremental. The past decade broke that pattern. The IBC was a structural reset for commercial law. The Mediation Act 2023 institutionalised a pathway that had long existed informally. The 2024 criminal codes were the largest overhaul since the colonial-era originals. The 2025 ad-hoc judges ruling unlocked a constitutional provision that had lain dormant. Whether this reform momentum outpaces the underlying filing growth is the measurable test.
On undertrial detention, 2024
The scale problem.
Four reform models.
The judicial reform package
Singapore went from a backlog-ridden system in the late 1980s to one of the world’s most efficient judiciaries in a decade. Case management, mandatory mediation, specialised courts, performance metrics for judges, aggressive caseflow technology. The most-cited peer reform example globally.
The one reform that worked at scale
The Insolvency and Bankruptcy Code 2016 + NCLT infrastructure resolved in five years what High Courts had failed to for a decade. Over 8,000 cases resolved; ~₹3.5 lakh crore in value recovered; average resolution time compressed from 4+ years to under 18 months for completed cases. The lesson: specialised benches with focused mandates work.
Alternative dispute resolution at scale
Lok Adalats, run by state legal services authorities, disposed of ~1.5 crore cases in 2023-24. Pre-litigation settlement of compoundable disputes, motor vehicle claims, cheque-bounce cases, family matters. A uniquely Indian institutional innovation now being studied abroad.
Digital administration done right
The Delhi High Court has the most advanced e-filing and case-management infrastructure of any Indian High Court. Average case disposal time has fallen ~18% since 2019; mandatory e-filing for several categories; near-universal digital cause lists. The gap is uneven adoption across the other 24 HCs and 17,000 subordinate courts.
Six things that would unclog the system.
On Indian judicial reform, 2023
Pendency figures vary across sources (NJDG, Supreme Court annual reports, India Justice Report, PRS, state-level data). NJDG is updated daily; aggregate totals differ slightly across reporting dates.
The ~50% figure for government as litigant comes from LIMBS and is widely cited but not updated in real-time; treat as directional.
Undertrial figures come from NCRB Prison Statistics and India Justice Report, using different methodologies. The ~76% undertrial share is consistent across sources within a 3-4 percentage-point band.
Primary data: National Judicial Data Grid; India Justice Report 2025; Department of Justice vacancy data; NCRB Prison Statistics.
Research: DAKSH; Vidhi Centre for Legal Policy; PRS Legislative Research; Law Commission of India Reports 120, 230, 245.
Policy: Bharatiya Nyaya Sanhita 2023; Bharatiya Nagarik Suraksha Sanhita 2023; Mediation Act 2023; eCourts Mission Mode Project documentation.
Where the
rupee goes.
Every February, the Union Budget is laid before Parliament. What people remember from the speech is typically two or three numbers — the tax-slab change, the fiscal deficit, one or two new schemes. What the Budget actually is, however, is a 1,500-page document of allocations that determines how the central government spends roughly ₹53 lakh crore over the next year. Twenty-six paise of every rupee the Union spends goes to interest on past borrowings. Twenty-two to states as their share of taxes. Eight to subsidies. Eight to defence. Six to pensions. Everything else is allocated within the residual.
The past decade’s fiscal story has been one of structural reform layered on cyclical shock. GST (2017) is the largest indirect tax reform in independent India’s history — 17 central and state taxes collapsed into one regime. The Insolvency and Bankruptcy Code (2016) gave creditors a legal framework that cleared a decade of bad-loan accumulation. Direct Benefit Transfer via Jan Dhan-Aadhaar-Mobile is estimated to have saved over $27 billion in welfare leakage by 2022. Capital expenditure has roughly tripled as a share of GDP. The fiscal deficit has travelled from pre-Covid 3.3% to post-Covid 9% to FY26 target of 4.3%. The shape of the Indian fiscal state has changed more in the past decade than in the two before it.
Composition of Union expenditure by major category, BE 2026-27. Source: PRS Legislative Research analysis of Union Budget 2026-27.
Tax-to-GDP ratio (approximate): India ~18%, China ~28%, Brazil ~33%, United States ~27%, OECD average ~34%, UK ~33%, South Africa ~27%. India’s tax base is narrow by peer standards. GST has formalised large parts of the informal economy; direct-tax filers remain only about 2-3% of the population.
Interest payments as % of revenue receipts: India ~40%, US ~15%, Germany ~2%, China ~10%. India’s high ratio reflects the accumulated cost of past fiscal deficits. Every year real GDP growth exceeds interest rates, this ratio improves automatically. The FRBM glide path toward 50% debt-to-GDP by 2031 is the policy vehicle.
Reform, compounding.
The past decade’s fiscal reforms — GST, IBC, DBT, capex push, fiscal consolidation — together represent the most substantial structural reform in Indian public finance since liberalisation 1991. The outcomes are uneven but the trajectory is unambiguous: a state that collects taxes more efficiently, spends more on productive capital, and is on a credible debt-to-GDP glide path.
On the fiscal constraint of legacy borrowing
The shape of the state.
Four fiscal reforms.
Direct transfers at scale
Brazil’s Bolsa Familia covered ~14 million households at peak — the template for conditional cash transfers that India’s DBT architecture draws from. Indian DBT now covers far more beneficiaries (~90 crore across 300+ schemes) with digital plumbing more sophisticated than the Brazilian original.
Structural creditor-debtor reset
Insolvency and Bankruptcy Code 2016 + NCLT cleared ~₹3.5 lakh crore of stressed-asset recovery in five years. India’s World Bank Resolving Insolvency rank improved from 136 to 52 in four years. The single most consequential commercial-law reform of the decade.
Digital tax infrastructure
GSTN is among the most sophisticated tax platforms in any G20 country. Real-time invoice matching, e-way bills, input tax credit reconciliation at scale. The formalisation of SMEs through mandatory GST filing is an under-appreciated structural shift.
Return-less tax filing
Estonia’s e-Tax Board pre-fills tax returns based on employer and financial-institution reporting. Filing takes 3-5 minutes; compliance is near-universal. India has the data infrastructure (TDS, AIS, Form 26AS) to do this; the policy design to complete the pre-filing shift is the remaining step.
Six things that would reshape the budget.
Annual Union Budget Analysis
Budget figures come from Ministry of Finance documents. PRS Legislative Research produces a concise analysis within a week of each budget — the most accessible third-party source. CAG audit reports assess actual vs budgeted spending.
BE (Budget Estimate) figures are revised to RE (Revised Estimate) mid-year and ultimately to Actuals 18-24 months later. All percentages of GDP are based on projected nominal GDP; actual GDP for FY26 will be known later.
Tax-to-GDP comparisons across countries vary slightly by methodology (whether social security contributions are included, etc.). Treat as directional.
Primary data: Union Budget 2026-27; PRS Union Budget Analysis; Economic Survey 2024-25; 15th and 16th Finance Commission reports.
Research: NIPFP (Pinaki Chakraborty, M. Govinda Rao); Rathin Roy; CAG audit reports; Accountability Initiative Budget Briefs; CBGA (Centre for Budget and Governance Accountability).
From coal
to sun.
In 2014, India had about 2.8 GW of solar. By early 2026, about 150 GW. That is a 53-fold rise in eleven years, a pace no other large economy has matched. Total installed electricity capacity crossed 530 GW at end-FY26, of which 53% is now non-fossil. India hit the Paris Agreement NDC target of 50% non-fossil capacity in June 2025, five years ahead of schedule, and now ranks third globally in installed renewable-energy capacity (IRENA 2026), having moved past Brazil.
The policy machinery has matured in parallel. The Production Linked Incentive for solar modules has driven domestic manufacturing capacity from 2.3 GW in 2014 to about 172 GW in 2026, a 75-fold expansion. PM Surya Ghar Muft Bijli Yojana targets 1 crore rooftop-solar households by 2026-27, with about 45 lakh installed by early 2026. Battery Energy Storage Systems have seen about 43 GWh sanctioned under two viability-gap-funding schemes. The Green Hydrogen Mission targets 5 MMT annual production by 2030. The September 2025 GST cut on renewable equipment from 12% to 5% measurably lowered project costs.
What hasn’t fully closed: coal still produces around 70% of actual electricity (capacity is not generation). Transmission bottlenecks are causing peak-hour curtailment in some states. Discom finances remain fragile, with cumulative losses above ₹6 lakh crore. A formal just-transition framework for coal-dependent states is still being built. The transition is real and measurable. Its completion depends on storage, transmission, and discom reform.
Cumulative solar installed capacity in India (GW). Source: MNRE. The 2030 target of 500 GW non-fossil includes ~293 GW of solar. Costs have fallen dramatically — Indian solar PPAs now clear below ₹2.50/kWh, lower than new coal in most cases.
Per-capita CO&sub2; emissions (tonnes/year): US ~14, Russia ~12, Japan ~8, China ~8, Germany ~8, World average ~4.7, India ~2.0, Bangladesh ~0.5. India is the 3rd-largest absolute emitter and among the lowest per person in the G20. India’s cumulative share of global emissions since 1850 is ~3%.
Non-fossil capacity share: India ~53%, China ~52%, Germany ~65%, Brazil ~83%, US ~43%. Among large developing economies, India’s pace of transition is near the front. Brazil has more non-fossil, but Brazil’s grid is hydro-dominated; India’s transition is the harder one because it combines solar scale-up with a coal-base country.
Transition, accelerating.
The past decade has been the most consequential in Indian energy policy since independence. Solar capacity up 50-fold. Wind crossing 50 GW. Domestic solar-module manufacturing from near-zero to ~144 GW/year. PM Surya Ghar targeting 1 crore rooftop households. Battery storage operationalising. Green Hydrogen Mission funded. Collectively, the most ambitious energy-transition architecture of any large developing economy.
Press release, July 2025
The acceleration.
Four transition models.
Manufacturing and deployment at scale
China built solar manufacturing and deployment concurrently. Annual additions now exceed 200 GW. The Chinese lesson for India is industrial-policy coordination: module manufacturing, cell manufacturing, polysilicon, and deployment subsidies moving together. India’s PLI for solar modules is the adapted template.
The solar-park template
Charanka Solar Park in Gujarat was India’s first large-scale solar park (2012). The template — state-aggregated land, transmission pre-built, competitive auctions — drove costs below ₹2.50/kWh by 2017. Now replicated across 14 Indian states; Bhadla (Rajasthan) is the world’s largest at over 2 GW.
Distributed solar at household scale
Germany’s Energiewende brought solar to a third of all households. Feed-in tariffs, net metering, community-owned generation. The Indian PM Surya Ghar scheme is the adapted architecture for distributed rooftop deployment at household scale.
The rapid-deployment cautionary tale
Vietnam added ~19 GW of solar in 3 years, then faced grid-integration and curtailment problems. The lesson: deployment without matching transmission and storage capacity produces stranded assets. India’s Green Energy Corridor and Battery Storage schemes are the deliberate answer to this risk.
Six things that would complete the transition.
On India’s solar trajectory, 2024
Installed capacity figures from CEA and MNRE are updated monthly and are authoritative. Generation figures (how much electricity is actually produced) are reported separately and lag capacity additions. All figures are directional.
Per-capita emission comparisons use Global Carbon Atlas / Our World in Data. Methodologies and base years vary slightly across sources.
PLI investment figures include both realised and committed investment across 14 sectors; cumulative figures are reported by Ministry of Commerce & Industry.
Primary data: Ministry of New & Renewable Energy; Central Electricity Authority; Ember Energy India Report; Our World in Data emissions; IRENA Renewable Capacity Statistics 2025.
Research: CEEW; IEEFA; Prayas (Energy Group); Brookings India energy work.
Built on
trust.
In 2009, India did not have a system of digital identity covering its population. In 2016, it did not have a real-time payments system. By early 2026, Aadhaar had issued over 142 crore numbers (the world’s largest digital-identity programme), UPI was processing ~22 billion monthly transactions worth ~₹29 lakh crore (about half of all global real-time payments), DigiLocker held over 800 crore digital documents, and ONDC had crossed several hundred million cumulative orders. The AI Impact Summit at Bharat Mandapam, hosted in early 2026, was India’s coming-out as a multilateral convener of the AI policy conversation. Account Aggregators had begun mediating financial-data flows at scale. None of these existed fifteen years ago. Their combined effect on Indian life is the largest single technological transformation of the past two decades.
The India Stack is a set of interoperable digital public goods, each addressing a specific foundational gap and working with the others. Aadhaar provides identity. UPI provides payments. DigiLocker provides document storage. DEPA / Account Aggregator provides consented data sharing. ONDC provides an open protocol for e-commerce. Bhashini provides language translation. GSTN provides tax infrastructure. Together they constitute the most ambitious attempt by any state to build population-scale digital utilities.
Monthly UPI transactions (millions). Source: NPCI. The 2016-20 ramp was partly driven by demonetisation and pandemic cash avoidance; 2020-onward growth is structural adoption across merchants, P2P transfers, and increasingly cross-border use cases.
Share of global real-time payment transactions: India ~50%, Brazil (Pix) ~15%, Thailand (PromptPay) ~8%, China ~6%, South Korea ~3.5%, UK ~2.3%, US ~1.5%. Card-dominant legacy economies have been slower adopters. Brazil’s Pix launched 2020 and grew rapidly; India’s UPI started in 2016 and is now both the largest and the fastest-growing.
Number of countries adapting the India Stack template: 10+ MoUs signed, including Armenia, Sierra Leone, Suriname, Papua New Guinea, Trinidad & Tobago, Tanzania, Kenya, Cuba, Colombia, Antigua & Barbuda. Brazil, Nigeria, and Ghana have built adapted variants. The India Stack is India’s largest digital export.
The India Stack, assembled.
The past decade-and-a-half has produced the most consequential set of public digital infrastructure anywhere in the world. Not the most advanced technically — Estonia’s is arguably ahead — but the most consequential by scale. The India Stack has become a template studied, adapted, and in some cases directly imported by countries across the Global South.
Co-founder, UIDAI / Aadhaar
The India Stack, quantified.
Four DPI cases.
The Pix parallel
Brazil launched Pix (central-bank-operated instant payments) in November 2020. By 2024, Pix accounted for ~15% of global real-time payments. Pix and UPI together now handle about two-thirds of global real-time payment volume. The structural choice — public-utility payment rails over card networks — is now visibly global.
Data interoperability done right
Estonia’s X-Road infrastructure is the most sophisticated government-data interoperability system globally. ~99% of government services are digital; citizens interact with the state through a single digital identity. India’s ABDM, ULI, and DEPA frameworks draw from X-Road principles.
National digital identity at scale
Singpass covers all Singapore residents with biometric + password + app-based multi-factor identity. Used for tax filing, driving licence, medical records, banking, voting. The Singapore model is more integrated with private-sector services than Aadhaar is.
DPI at crisis scale
CoWIN administered over 220 crore Covid vaccine doses — the largest vaccination campaign in human history. Built in six weeks on the India Stack backbone; later offered as DPI to other countries. The case study most cited globally for what Indian DPI can do at emergency scale.
Six things that would protect the stack.
Global Risks Report 2024
UPI transaction data is reported monthly by NPCI and is considered authoritative. Volume and value figures are updated in real time and widely cross-referenced by RBI.
Aadhaar figures are reported by UIDAI. 138 crore is cumulative Aadhaar numbers generated; active/linked Aadhaar is a different (slightly lower) figure.
DBT leakage savings ($27 billion by 2022) is a government estimate; independent academic estimates fall within a broader range. Treat as directional.
The ballot
box.
Between 19 April and 1 June 2024, India conducted its 18th general election. 642 million voters cast ballots. 969 million were eligible. The exercise unfolded over 44 days, across 7 phases, at 1.05 million polling stations, administered by 15 million election workers and security personnel. For scale comparison: more Indians voted in 2024 than the combined populations of the United States, Russia, Germany, and Japan.
Every one of those numbers is a statistic. Together, they are a claim: that an electoral system of this scale can function, in a country with 22 scheduled languages, 28 states, 8 union territories, a geography that includes the Himalayas and the Andamans, and income levels that span from the poorest villages in Bihar to the richest neighbourhoods in Mumbai. It worked. Again.
The story of the past decade’s electoral reforms is a mix: some genuine improvements (home voting for elderly and disabled voters, record female turnout, accessibility infrastructure, the 2023 Women’s Reservation Act); some open questions (the 2023 CEC & ECs Appointment Act’s executive-majority selection panel; the post-electoral-bonds political-finance vacuum). The process is strong. The reforms that compound it further are the ongoing work.
Voter turnout in each Lok Sabha election, 1952-2024. Source: Election Commission of India historical data. Notable: 2014 marked the start of sustained 65%+ turnout; 2019’s 67.4% was India’s highest-ever.
Voter turnout in recent national elections: India 2024 (~66%), Australia 2022 (~91% — compulsory), Brazil 2022 (~79%), United Kingdom 2024 (~60%), United States 2024 (~62%), France 2022 (~72%), Germany 2021 (~77%).
India’s turnout is comparable to established Western democracies — at a population scale none of them approach. The closest comparison is not another country but the rest of the democratic world combined: India conducts, in a single election, more voting than every other country south of the Tropic of Capricorn put together.
The process, expanded.
The mechanics of Indian elections have been steadily refined over the past decade. VVPAT universal since 2019. Two EVMs per booth to reduce queues. Home voting for elderly and disabled from 2024. Accessibility infrastructure at every polling station. Women’s Reservation Act on the books. Record female participation. None of these solves the deeper political-finance and representation questions; all of them make the mechanics more inclusive and functional than at any point before.
Press conference, June 2024
The largest democratic exercise.
Four electoral models.
Domestic innovation at scale
Indian EVMs are stand-alone, not networked, and paired with VVPAT audit trails since 2019. The Supreme Court rejected 2024 petitions seeking a return to paper ballots, citing the EVM-VVPAT system’s integrity. The model has been studied by several developing-democracy electoral bodies.
The digital-court model
Brazil’s Superior Electoral Court runs electronic voting for 150 million voters and is notable for quick result-declaration and public-facing transparency dashboards. The Brazilian model is more vertically integrated than India’s; each has strengths.
Civic-tech candidate transparency
Association for Democratic Reforms (ADR) successfully petitioned the Supreme Court in 2003 to require candidates to disclose assets, criminal records, and education. MyNeta.info aggregates these disclosures for every candidate in every election since 2004 — one of the most consequential civic-tech interventions in Indian democracy.
Electoral assistance at scale
IFES (International Foundation for Electoral Systems) has advised on electoral reform in 140+ countries since 1987. The India election is their most frequently cited reference point for population-scale logistics.
Six things that would deepen the democracy.
On Indian electoral reform, 2024
Election data comes from the Election Commission of India’s official Statistical Reports, released after each general election. The 2024 data was released as 42 statistical reports in late 2024.
Voter turnout is reported as polling-station turnout; gross turnout (including postal ballots) is slightly higher. Independent organisations (Lokniti-CSDS, ADR) produce parallel analyses.
Campaign expenditure estimates from Centre for Media Studies combine candidate, party, and state machinery spending; different methodologies produce figures within a broad range.
Primary data: Election Commission of India 2024 Statistical Reports; OGD Platform India; IndiaVotes.
Candidate / finance: Association for Democratic Reforms; MyNeta; Centre for Media Studies.
Research: Lokniti-CSDS post-election surveys; Yogendra Yadav, Suhas Palshikar; Pradeep Chhibber and Rahul Verma.
The growth
story.
In 2014, India was the world’s ninth-largest economy. By 2026, it is the world’s 4th-largest by nominal GDP and 3rd-largest by purchasing power parity, having overtaken the UK, France, and Japan in nominal terms (the IMF April 2026 ranking briefly placed India 6th nominal due to rupee depreciation against the dollar; underlying growth in rupee terms remains strong). The economy has added more nominal GDP in eleven years than in the sixty that preceded. This is not a gentle trajectory; it is a structural acceleration.
Infrastructure has followed the macro story. 149 airports today vs 74 a decade ago. Over 45,000 km of 4-lane and above highways, including the Delhi-Mumbai Expressway, Ganga Expressway, Dwarka Expressway. Expanded port capacity. Metro networks in over 20 cities. Foreign exchange reserves above $687 billion. FDI inflows have grown from ~$36 billion in FY14 to ~$81 billion in FY25 — and manufacturing FDI alone has risen ~69% cumulatively to cross $165 billion over a decade.
The manufacturing push is the central economic bet of this period. India aims to raise manufacturing’s share of GDP toward 25% (from around 16%) by 2035 under the new National Manufacturing Mission. PLI schemes in 14 sectors have attracted ~₹2 lakh crore in cumulative investment, with incremental production/sales exceeding ₹18 lakh crore and over 12 lakh direct and indirect jobs created. Mobile-phone manufacturing has gone from ~2 units a decade ago to ~300 today. Import dependence has fallen from ~75% of domestic demand to essentially zero. India is now the world’s second-largest mobile manufacturer.
Manufacturing value-added as % of GDP (approximate, most recent available). Source: World Bank national accounts; Economic Survey 2024-25 (new base-year GVA series shows India rising from 14.5% to 16.3% between 2022-23 and 2024-25). Different sources use different definitions of manufacturing; figures directional.
Real GDP growth rates, FY 2024-25: India ~7.4%, China ~5.0%, United States ~2.5%, EU ~0.8%, Japan ~1.0%, Brazil ~2.5%, South Africa ~1.0%. India is the only large economy consistently growing above 6%.
Manufacturing share of GDP is the structural gap. China, Vietnam, Bangladesh, Thailand, and Malaysia are all above 22%; India sits near 16%. The gap is the link between Chapter 03 (Jobs) and Chapter 15 (Growth). Without a larger manufacturing share, the youth-employment problem cannot be solved durably.
The climb, compounding.
The past decade is the first time in independent India’s history that growth has been accompanied by a comparable structural-reform package. IBC, GST, corporate tax cut, PLI, DPI, logistics policy, DBT, fiscal consolidation, manufacturing focus. None of these individually explains the growth; together they constitute the most coherent growth-policy architecture India has ever had.
On India’s development trajectory, 2024
The ascent, quantified.
Four growth models.
The manufacturing-led miracle
China moved from ~10% of GDP from manufacturing to ~32% in two decades. Absorbed 200 million workers out of agriculture into formal manufacturing. The original template for what India’s National Manufacturing Mission is attempting, at a different starting point and with different political conditions.
The FDI-enabled convergence
Vietnam went from manufacturing share of ~16% to ~24% in two decades, largely through disciplined FDI attraction and trade-agreement-driven market access. Vietnamese exports have grown at a pace India’s trade policy is trying to replicate through bilateral FTAs with Australia, UAE, UK (under negotiation), EU (under negotiation).
Services-led growth at a billion people
India’s IT services export story (~$200 billion annually) is the clearest example in developing-country economics of a services sector generating formal middle-class jobs at population scale. The question for the next decade is whether manufacturing can do the same at an even larger employment scale.
From importer to second-largest producer
India went from 2 mobile-manufacturing units in 2014 to ~300 in 2024. From 75% imports to ~0%. From insignificant exports to ~₹2 lakh crore in FY25. The clearest case of PLI delivering structural transformation in a single sector at scale. The template other sectors (textiles, electronics components, semiconductors) are being designed to follow.
Six things that would close the last mile.
On India’s growth outlook, 2024
GDP figures come from the Ministry of Statistics & Programme Implementation. A new base-year (2022-23) GVA series was introduced in 2026 with slightly different methodology from the earlier 2011-12 series; some figures reported here blend both. Manufacturing-share figures vary by source: old series, new series, and World Bank all report slightly different numbers.
FDI and export figures come from RBI and Ministry of Commerce; PLI investment figures from Ministry of Commerce & Industry. All are directional.
Growth-rate projections vary across IMF, World Bank, ADB, and domestic sources; the ~7.4% figure for FY26 is from the Economic Survey and broadly consistent across sources.
Primary data: Economic Survey 2024-25; Ministry of Commerce & Industry; MoSPI National Accounts; RBI Annual Report 2024-25.
Research: Ideas for India; India Brand Equity Foundation; World Bank India Development Update; IMF World Economic Outlook; Arvind Subramanian, Raghuram Rajan, Pinaki Chakraborty.
Commentary: Mint; Business Standard; Hindu BusinessLine.
The build-out.
If a single Indian of ordinary means were to describe what changed most visibly in their everyday environment over the past decade, it would not be a policy or a slogan. It would be the road. The fact that a bus from their village reaches the district headquarters in two hours, not five. That the phone they bought for under eight thousand rupees streams video at speeds that would have been premium-tier broadband in 2015. That their train has stopped running on diesel. That the neighbouring town now has an airport.
This is the quiet infrastructure story of the past decade, and it is an unusually consequential one. Infrastructure rarely makes headlines except when something goes wrong. But under the noise, a set of structural build-outs have happened simultaneously: national highways, rail electrification, regional airports, metro rail, installed power capacity (see the Energy chapter), the payments stack, the broadband stack, ports and waterways, and the supply-chain logistics network.
None of this is complete. Highway quality varies sharply by state. Railway modernisation is uneven. Airport expansion has concentrated in tier-1 and tier-2 cities; tier-3 connectivity lags. Urban metro networks are partial in most cities. Broadband last-mile quality varies by district. The direction of travel is strong; the finish line is not close.
Road network lengths in India, 2024, approximate. Sources: MoRTH Annual Report 2023-24; Economic Survey 2024-25. Road category definitions vary; figures are directional. State highway data varies across reporting frames.
India’s national highway length grew from around 91,000 km in 2014 to approximately 146,000 km by 2024. In absolute terms, India is building highway kilometres at the second-highest annual pace globally after China, and has done so for most of the past decade. Daily construction rates have ranged between 20 and 37 km per day over the last five years, depending on the measurement window.
For comparative scale: China’s national expressway network is roughly 180,000 km; the US Interstate Highway System is about 80,000 km. India’s operational access-controlled expressway network, a subset of total NH, has expanded from around 90 km in 2014 to about 3,000 km at end-2025, with another ~10,000 km under construction or awarded. The government targets 18,000 km of operational expressways by 2028-29. The scale is comparable to major build-out decades in China and mid-century US.
The scaffolding, extended.
The pace of infrastructure build-out over the past decade is difficult to overstate, even for those who track it closely. What sits below is a compressed timeline of the structural shifts — roads, rail, air, urban transit, power, digital — that together reshape what is physically possible for people and firms operating in India.
On PM Gati Shakti and the infrastructure decade, 2024
The decade, in four figures.
Four case studies.
PM Gati Shakti: the planning layer
The Gati Shakti National Master Plan integrates infrastructure project data from 16+ ministries and all states into a single GIS platform. Roads, rail, telecom, pipelines, power transmission, and logistics clusters can be planned with visibility to each other rather than in silos. The old practice of building a road only to dig it up for a water pipeline six months later — and both again for a telecom cable — has measurably declined in the projects that use the platform. A governance innovation as much as an infrastructure one.
UPI and the digital payments rails
The Unified Payments Interface — a public digital infrastructure layer operated by NPCI — enabled real-time interoperable payments across banks. From zero in 2016, UPI now processes roughly half of the world’s real-time payment transactions by volume — about 22 billion a month, formally recognised by the IMF as the world’s largest real-time payment system. The marginal cost of sending money from one Indian to another has fallen to effectively zero. The second-order effects on small-merchant commerce, gig-work payouts, and MSME working-capital cycles are still being worked out.
The high-speed rail decade
China built the world’s largest high-speed rail network, from roughly 700 km in 2008 to over 40,000 km by 2020. The lesson most cited: sustained political commitment over a decade, plus predictable financing, plus technology transfer that became indigenous capability. India’s Vande Bharat, RRTS, and upcoming Mumbai-Ahmedabad bullet train are the adapted framework; the scale is meaningfully smaller, the institutional architecture comparable.
Metro rail: from 9 to 24 cities
Operational metro networks expanded from 9 cities in 2014 to 24 by 2024, with Delhi-NCR, Mumbai, Bengaluru, Chennai, Hyderabad, Kolkata, Kochi, Jaipur, Lucknow, Nagpur, Ahmedabad, Kanpur, Pune, Navi Mumbai among the active networks. Over 1,000 route km now operational nationally; another ~1,000 under construction. Ridership patterns vary widely; Delhi Metro alone carries over 7 million passengers per day on most working days.
Six things that would extend the decade.
Chapter on Infrastructure and Capital Formation
Infrastructure metrics in India come from multiple ministries with distinct reporting frames — MoRTH for roads, Ministry of Railways for rail, DGCA and AAI for airports, MoP for power, CEA for generation capacity, DoT and TRAI for telecom, MoHUA for metros. Where state-level data exists, it varies in comparability across reporting periods.
Headline figures here are rounded and drawn from the most recent government releases. “Kilometres built” numbers can vary across sources depending on whether they count additions, widening, upgrades, or the total network. Treat all specific figures as directional indicators of pace and scale.
The infrastructure build-out is broad-based but not uniform; state-level gaps are substantial. The data here captures the national picture.
Primary data: MoRTH Annual Report 2023-24; Indian Railways Annual Statistical Statements; Airports Authority of India; Central Electricity Authority; TRAI quarterly performance indicators; NPCI UPI statistics.
Research: EAC-PM working papers on infrastructure; Economic Survey 2024-25 infrastructure chapter; PRS Legislative Research Union Budget analyses; Observer Research Foundation infrastructure series.
Policy: PM Gati Shakti National Master Plan; National Logistics Policy 2022; National Infrastructure Pipeline; Bharatmala, Sagarmala, UDAN; Dedicated Freight Corridor Project documentation.
The people
behind this.
A brief note on who built this, why, and what else they do.
Yash Agarwal
Yash is a technology and internet policy practitioner based in New Delhi. He works in Global Stakeholder Engagement at ICANN, focusing on internet governance in South Asia. He previously served as Twitter India’s first Public Policy Fellow from South Asia, held government affairs advisory roles at Chase India, and was a LAMP Fellow in the Indian Parliament.
He teaches as guest faculty across 30+ institutions in India, serves on Boards of Studies at multiple universities, and advises organisations working at the intersection of public policy, technology, and education.
He is also the founder of Public Policy India (PPI).
Find him at yashagarwal.info or on LinkedIn.
Public Policy India
Public Policy India (PPI) is India’s largest community for public policy and social impact. It runs across 30+ city chapters, hosts a newsletter, a large WhatsApp community, and in-person capacity-building engagements across India and South Asia.
The community exists to make public policy less abstract and more accessible for students, early-career professionals, and anyone curious about how India actually works. Many Indias is a PPI data project — part of a broader effort to put credible, sourced data about the country into the hands of people who want to understand it without wading through three PDFs and two dashboards to do so.
Find PPI at publicpolicyindia.com.
About this project
Many Indias is a single-page data dashboard and reference series covering fifteen themes that shape everyday life in India. The underlying data is drawn from credible, publicly available government and independent sources (listed on the dashboard itself and at the bottom of each chapter). The project is opinionated about how to read the numbers but tries to be transparent about the sources and their limits.
Corrections, suggestions, and arguments are welcome. The best way to send them is through the PPI website or LinkedIn.
— Delhi, 2026 —