Two Trajectories
A fair question gets asked often: what would India look like in 2026 if the British had stayed? This page refuses to answer it — because nobody can, and anyone who says otherwise is guessing in an authoritative voice. What we can do instead is lay the two actual trajectories side by side: what the colonial decades measured, and what the decades since measured.
A speculative "British India in 2026" would be the most shareable thing on this site and the least defensible. Under our own rules every claim carries an evidence tag, and an invented timeline can only ever be tagged Unsupported. So we are doing the harder and duller thing, which happens also to be the more damning one: reporting the measured record on both sides of 1947 and letting the reader weigh it.
One deliberate choice about presentation. Everything below is a table, not a chart. A line graph of Indian literacy from 1901 to 2024 would look magnificent and would quietly lie, because the definition changes partway through — early censuses counted literates as a share of the whole population, later ones only those aged seven and above. A chart smooths that seam into a curve. A table can show it. Where the ground shifts under a number, we would rather you saw the shift.
| Period | Life expectancy at birth | Source |
|---|---|---|
| 1881–1920 (avg, males) | ~25 years | Official decennial life tables |
| 1941 (males / females) | ~32 / ~31 years | Bhagawati, generation life table |
| 1941–50 | 32.5 years | NSS Report No. 54, MoSPI |
| 1951–61 (males / females) | ~37.8 / ~37.0 | Milbank Memorial Fund Quarterly |
| 2018–22 (males / females) | 68.2 / 71.9 | ORGI, SRS Abridged Life Tables |
| 2019–23 (total) | 70.3 years | ORGI, SRS Abridged Life Tables |
Roughly 25 years in the 1880s–1910s; about 32 years at independence; 70.3 years for 2019–23. A gain of about seven years across the last six decades of colonial rule, and about thirty-eight years in the seven decades since. Documented
Honest limit: colonial vital registration was incomplete and uneven. Every pre-1947 figure is a reconstruction from census age-distributions, not a count of registered deaths — so treat the level and direction as documented and single-year precision as not. The 1911–21 trough (~20 years) sits in the decade of the influenza pandemic and rests on an interpolation, not a life table. Contested
| Census / survey | Overall | Male / Female | Definition |
|---|---|---|---|
| 1901 | 5.4% | 9.8% / 0.6% | crude (all ages) |
| 1911 | 5.9% | 10.6% / 1.0% | crude |
| 1921 | 7.2% | 12.2% / 1.8% | crude |
| 1931 | 9.5% | 15.6% / 2.9% | crude |
| 1941 | 16.1% | 24.9% / 7.3% | crude — wartime census, see note |
| 1951 | 18.33% | 25.0% / 7.9% | age 5+ |
| 1971 | 29.4% | — | age 7+ (definition changes here) |
| 1991 | 42.8% | — | age 7+ |
| 2011 | 74.04% | 82.1% / 65.5% | age 7+ |
| 2023–24 | 80.9% | 87.2% / 74.6% | age 7+ (PLFS survey, not a census) |
The honest headline, with its own caveat attached: about 5% (crude, 1901) to about 81% (age 7+, 2023–24) — a transformation, but one whose raw multiple is inflated by the change in definition. Anyone quoting "5% to 80%" without saying that is overstating a case that does not need overstating. Documented
Two further cautions. The 1941 census was taken under wartime conditions with heavily curtailed tabulation, and covers undivided India — its jump from 9.5% to 16.1% should be read as reported, not measured. Contested And there has been no decennial census since 2011; every figure after that is a sample survey. Documented
What is not in dispute: at independence a large majority of Indians could not read, and on the 1931 count more than 97 in 100 Indian women could not.
On the Maddison series, Indian income per head was the same in 1870 as in 1820, and was lower in 1950 than in 1940. Across the whole 130 years from 1820 to 1950 it rose by roughly 16% in total — a rate near zero, in the era of the fastest sustained growth the world had ever seen elsewhere. Contested (these are back-projected estimates, not measurements — but the direction is robust)
Since then, income per head has risen many times over: about US$274 in 1981, US$306 in 1991, and roughly US$2,700 in 2024. Documented
This is the most-repeated economic claim in Indian public argument, and it is wrong in a specific and instructive way. Here is the actual series it comes from:
| Year | India's share of world GDP | Note |
|---|---|---|
| 1 CE | ~32% | the figure most often misdated to the 1700s |
| 1500 | 24.4% | |
| 1600 | 22.4% | |
| 1700 | 24.4% | the actual eve-of-colonialism figure |
| 1820 | 16.0% | |
| 1870 | 12.1% | |
| 1913 | 7.5% | |
| 1950 | 4.2% |
There is no 27% figure for 1700. The real number is 24.4%. The circulating ~32% figure is the estimate for 1 CE — two thousand years earlier — routinely re-dated to the eve of colonial rule. Unsupported
And the deeper problem is worse for the claim than the arithmetic error. These shares are not measurements; they are back-projections that assume pre-modern incomes sat near a subsistence floor and then scale by population. Because India held roughly a quarter of the world's people in 1700, the method necessarily returns roughly a quarter of world output. The famous share is substantially a population statistic wearing an economics costume — and it does not show that Mughal India was rich per person. Maddison's own per-head series shows the opposite, and Broadberry and Gupta find Indian income per head was already falling relative to Britain from 1600, before the Raj. Contested
The same discipline applies to "India's share of world manufacturing fell from 25% to 2%" (Bairoch: ~24.5% in 1750 to ~1.7–2% in 1900). That fall is real — but part of it is arithmetic, because the denominator exploded as Europe and North America industrialised. Both mechanisms are true, and honest use names both. Contested
We publish the corrected figures because the sound version of this argument — a quarter of world output in 1700 down to about 4% by 1950, with income per head flat throughout — is devastating on its own. It has never needed the inflated number.
The Great Famine, 1876–78: 5.25–5.5 million deaths in British-administered territory; up to 6.1–10.3 million on the widest estimates including the princely states. Contested
The Bengal famine, 1943: 800,000 to 3.8 million across all published estimates, with 2.1–3 million the mainstream scholarly band — Dyson and Maharatna's 2.1 million measures excess deaths against a demographic baseline; Sen's widely-quoted 3 million counts famine-attributed deaths including later epidemic mortality. Contested The gap between those two numbers is not sloppiness. It is two different definitions, and we name which is which.
The correction: totals of "40 million" or "100 million" dead in colonial famines are produced by summing every famine from 1770 onward using each event's highest published estimate, adding non-famine excess mortality, and sometimes adding a counterfactual for suppressed population growth. Each step is arguable alone; combined they are not. Unsupported
Since 1947, no famine on the 1876–78 or 1943 scale has occurred in India. Serious food crises did — the mid-1960s droughts, Maharashtra in 1972 — without mass mortality. Documented Amartya Sen's explanation is that famines are failures of entitlement rather than of food availability, and that accountability prevents them: "no famine has ever taken place in the history of the world in a functioning democracy." Documented as Sen's argument — though the universal causal claim is itself Contested by researchers examining post-colonial Africa.
Foodgrain output ran from 50.8 million tonnes in 1950–51 to 357.7 million tonnes in 2024–25 — roughly sevenfold, against a roughly fourfold rise in population. Documented The Green Revolution's distributional record — regional concentration, groundwater depletion, the rainfed/irrigated gap — is a live debate and is Contested.
Printing only the improvements would be its own dishonesty. On the most recent verified figures: 35.5% of children under five are stunted and 19.3% wasted (NFHS-5, 2019–21). Documented The Global Hunger Index 2025 scores India 25.8 — "serious" — ranking 102nd of 123 countries, with a child wasting rate of 18.7%, among the worst in the index. Documented (the GHI's methodology is itself disputed, including formally by the Government of India — Contested)
The trend is genuinely improving: the GHI score has fallen from 38.1 in 2000 to 25.8 in 2025. The level remains in the "serious" band. Documented
India stopped having famines. India did not stop having hunger. Printing only one of those sentences is a form of lying — in either direction.
The intuitive move is to compare India with somewhere that stayed colonised longer. It does not work, and saying so is more useful than forcing it. The territories held substantially past 1947 were small, differently administered, and mostly plunged into war on independence — Angola and Mozambique (1975), Zimbabwe (1980), Namibia (1990). None is a subcontinent of 350 million with a century-old railway network and an existing industrial base. Setting them beside India is not a controlled experiment; it is two different objects on the same table. Contested
Nor do the territories that decolonised alongside India help: Indonesia, Korea, Taiwan diverged for reasons — Cold War alignment, foreign capital, land reform, war — unrelated to how long they were ruled. The one genuinely close comparison is India, Pakistan and Bangladesh, which inherited the same administration, law and rupture. That tells us something real about post-independence choices — and nothing whatever about whether independence itself helped. We use it for what it can answer and refuse it for what it cannot.
Attribution is genuinely hard, and pretending otherwise would undermine everything above. The decades after 1947 were also the decades of antibiotics, mass vaccination, oral rehydration, malaria control and high-yielding seed — global advances that raised life expectancy in nearly every country on earth, including those whose political status never changed. Not every Indian gain since 1947 is attributable to independence, and saying so plainly is what makes the rest trustworthy.
What the data does support is narrower and firmer: life expectancy rose about seven years across the six decades before 1947 and about thirty-eight in the seven decades after; income per head was flat from 1820 to 1950 and has multiplied since; famine mortality in the millions was recurrent under colonial rule and has not recurred. What it does not support is any claim about the rate at which those gains would otherwise have arrived, or a clean division of credit between medicine, policy and self-government.
The reverse caution matters equally: the stagnation of Indian income from 1820 to 1947 cannot be waved away as merely the condition of the age, because the same era produced the fastest growth the world had ever seen elsewhere — a world in which India's share of output fell from roughly 16% to roughly 4%.
And finally: this is a comparison of two actual trajectories, not a counterfactual. We do not know what a British India of 2026 would look like. Nobody does. We would rather show you the trend lines and let you weigh them than sell you a guess in a confident voice.
Sources
- Life expectancy & mortality: Office of the Registrar General of India, SRS Based Abridged Life Tables (2018–22, 2019–23) and SRS Special Bulletin on Maternal Mortality 2020–22; NSS Report No. 54, MoSPI; IIPS Working Paper No. 23; Bhagawati, "Generation Life Table for India, 1901–1951."
- Literacy: Census of India, 1901–2011; Periodic Labour Force Survey 2023–24 and Comprehensive Annual Modular Survey 2022–23, MoSPI.
- Income & world share: Maddison Project Database; Bairoch, "International Industrialization Levels from 1750 to 1980," Journal of European Economic History (1982); Broadberry & Gupta, "An Anglo-Indian comparison of GDP per capita, 1600–1871" (2012); Sivasubramonian, The National Income of India in the Twentieth Century (2000); Alagh (1992); Bank for International Settlements (2026).
- Famine: Dyson & Maharatna (1991); Maharatna, The Demography of Famines (1996); Sen, Poverty and Famines (1981) and Development as Freedom (1999); Davis, Late Victorian Holocausts (2001).
- Nutrition & food: NFHS-5 (2019–21), MoHFW; Global Hunger Index 2025; Ministry of Agriculture & Farmers' Welfare, Final Estimates of Crop Production 2024–25.
- Every claim is tagged in the Claims Ledger. Figures we could not verify to an original table — including a precise Sivasubramonian growth rate, the 1951 infant-mortality figure, and famine ranges for 1896–97 and 1899–1900 — are deliberately absent rather than approximated.