Drawing on nearly 250 years of Indian colonial history, new research examines how direct British rule and landlord-based land tenure shaped access to schools, health centres, roads, and agricultural productivity across Indian districts. Updated analysis shows that targeted post-colonial policies have erased many of these gaps, but income and fertiliser-use differentials linked to the land system remain.
Editor’s note: You can listen to this podcast on Spotify, Apple Podcasts, or wherever else you get your podcasts. You can also watch this conversation on YouTube.
Can colonial rule still shape development outcomes seven decades after independence? This VoxDevTalks episode brings together history, economics, and policy to answer that question. Lakshmi Iyer discusses her research – co-authored with Coleson Weir – tracing how two key features of British rule in India, direct versus indirect colonial administration and contrasting land tenure systems, have shaped access to schools, health centres, roads, and agricultural productivity in the post-colonial period, and asking whether targeted policy can undo what empire left behind.
Why colonial legacies persist long after independence
British rule in India lasted nearly 200 years, from 1757 to 1947, reshaping legal systems, political structures, and economic organisation across what is now India, Pakistan, and Bangladesh. But why do historians and economists expect to find effects still visible today? Iyer points to both institutional and structural channels: formal systems like parliamentary democracy and English common law were carried directly into independence, whilst economic disruption – resource extraction, new technologies, altered land arrangements – changed the underlying structure of regional economies in ways that proved difficult to reverse.
The self-reinforcing nature of these legacies is central to understanding their durability. Drawing on Nathan Nunn's work on Africa, Iyer explains how colonial conditions can trap regions in sub-optimal equilibria:
"Once you get stuck in this sub-optimal equilibrium, it can be self-reinforcing. It is hard to break out. You get used to doing only these types of activities. Just because the original impetus goes away, you can't always snap back."
Direct versus indirect rule in British India
Not all of India was governed identically. Around a quarter of the population lived under Indian princes – rulers who retained internal autonomy in matters of law, currency, and education, whilst ceding foreign policy and defence to the British Crown. When independence came in 1947, approximately 680 of these ‘princely states’ were absorbed into the new republic, nominally on equal terms with formerly directly ruled areas.
To isolate the causal effect of direct rule, Iyer exploits a historical accident: the Doctrine of Lapse, introduced by Governor-General Lord Dalhousie between 1848 and 1856. Under this policy, any princely state whose ruler died without a natural son was annexed by the East India Company. Because this rule applied somewhat arbitrarily – based on the timing of a ruler's death rather than the strategic value of the territory – it provides a near-random mechanism for assignment to direct British rule.
What the original research found by 1991
Using this quasi-random variation, Iyer's earlier research compared the two types of areas in the post-colonial period up to 1991. The results were striking: directly ruled areas performed significantly worse on public goods provision than their indirectly ruled counterparts. They were less likely to have a middle school, road, or health centre – and these deficits translated into worse educational outcomes and higher infant mortality, nearly half a century after independence.
"When we compare them in the post-colonial period, in 1991, which is pretty late, we find that these areas, the directly ruled areas, actually look worse than the indirectly ruled areas... They have fewer health centres, they have fewer schools, they have less roads, and, consequently, also worse educational outcomes, higher infant mortality, and things like that."
The finding challenged any assumption that post-colonial governments would quickly equalise conditions across their inherited territories.
Closing the gap: What two more decades of data reveal
Updating the analysis with data through 2001 and 2011 tells a more encouraging story. The gaps that were large and statistically robust in 1991 had, by 2011, largely disappeared. Directly and indirectly ruled districts became similarly likely to have a middle school, road, and health centre. Neonatal mortality, which had been four percentage points higher in directly ruled areas in the 1970s, had converged to within one percentage point.
Iyer attributes much of this convergence to the design of post-colonial social policy. India's Minimum Needs Programme, the Sarva Shiksha Abhiyan education initiative, and the Accelerated Irrigation Benefits Programme all shared a targeting logic: prioritise places that lack provision. Because formerly directly ruled areas were precisely those lacking provision, they received proportionally more investment. By contrast, fertiliser subsidies – which persist as a gap in the data – were distributed universally rather than targeted, and the differential has not closed.
The land tenure legacy: A more mixed picture
The second dimension of the research examines the Zamindari system, under which the British collected land tax through large landlords rather than dealing directly with individual cultivators. In Zamindari areas, the landlord was responsible for delivering a fixed revenue sum; how it was extracted from farmers was left to him. In other regions, the Ryotwari system gave individual cultivators registered property deeds and direct tax assessments from the state.
Research co-authored with Abhijit Banerjee found that, as late as 1987, Zamindari areas had significantly lower agricultural productivity – 20% lower wheat yields – along with less irrigation, less fertiliser use, and lower uptake of Green Revolution seed varieties. Updating this analysis to 2011 produces a mixed picture. Irrigation gaps have closed entirely. High-yielding variety adoption for rice and wheat has converged, but fertiliser use gaps have widened rather than narrowed, and yield differentials, whilst smaller, persist – wheat yields remain around 17% higher in non-Zamindari areas.
"It's a very mixed picture in terms of agricultural investments. And I think because of this mixed picture, we find that the yield differentials are still there. They are smaller, but they are not completely gone."
Reference
Iyer, L, and C Weir (2025), "The colonial legacy in India: How persistent are the effects of historical institutions?" Journal of Development Economics, 177: 103576.