Tamoghna Halder criticized one of my writings on nineteenth-century Indian famines. Halder distorts my views and wrongly implies that I suppressed data. He misreads the very nature of the Indian famine debate, thinking it is about facts. It is not. It is about method, about how economic historians and development scholars should read the history of climatic shocks. The piece demands a response and a clarification of the issues involved.

Which famines?

The Indian famine scholarship consists of two clusters. One of these discusses three late-nineteenth-century famines in peninsular India. Extreme drought conditions triggered these. Droughts of such intensity were not uncommon in the region. What was unusual is that droughts ceased to cause mass mortality after 1900, bringing about India’s demographic transition.

The second cluster forms of the Bengal famine in 1943. Although climatic shock had a role in causing it, the famine was not mainly a result of climate. And although a traumatic event for the Bengalis, the shock did not seriously impact India’s demographic transition because no other region in India in 1943 saw a sharp rise in death rates. These two clusters, therefore, should not be read together. The events were too dissimilar. I will concentrate on the former cluster and return to the Bengal famine.

Of the three “Deccan” famines, the first one in 1876 affected the region locally known as Rayalaseema and, at the time, Madras-Deccan (see map). The following two, in 1896 and 1898, occurred in the Deccan Traps to the northwest of the peninsular. The region is semi-arid. A long dry and hot season is broken by a monsoon much weaker in comparison with northern and eastern India. Agriculture and water-supply in the peninsular depended on the strength of the summer monsoon. Its geology of hard rocks made it expensive to harvest water from the underground. The great rivers of the region did not carry much water in the dry seasons. And therefore, unusual rainfall deficits, like those in 1876-1898, could cause a widespread water shortage, killing people directly via food and water deficiency and indirectly via cholera.

But droughts were not unknown in this climate and geology. These famines were not unusual events. They still have a special place in Indian economic history because they generated enormous official documentation. Some of that corpus asked the question why some droughts caused mass deaths. Based on the answer (that these followed climatic shocks), actions like relief codes, railways to carry food faster and canals to irrigate dry lands were recommended. Before these measures had an effect, two more famines followed.

These later famines showed that state intervention was not only slow and weak but would never be enough. In India’s caste society, upper castes had access to safe water, and the depressed castes were discriminated against to avoid the risk of ritual pollution. During droughts, discrimination took extreme forms, and many poor people died due to a shortage of water and dependence on common sources infested with the cholera bacterium. The state could not change caste sentiments. Other types of action were needed for that.

India was ruled by the British when these famines happened. Not all of it. Princely states ruled a big part (see map), but their record of dealing with famines remains largely unknown. Around the time the Deccan famines happened, pro-empire intellectuals had written books claiming that the British helped economic growth in India. Indian nationalists challenged that claim and said that British colonial rule caused famines via its indifferent and ideological attitude to Indian suffering.

Map: The area affected by the Deccan famines. Rainfall zones shown in background. Borders are those of major princely states.

Colonial-mentality theory

I will call this the “colonial-mentality theory” of famine causation. It is plausible that mentality did matter. The influence of Thomas Malthus, who taught political economy to students who would join the East India Company’s administration in India, was considerable. His theory told them that famines were nature’s response to overpopulation and did not call for deep intervention.

Later, Marxists and nationalist historians revived that story, some calling these events genocide or holocaust, slaughter on a mass scale engineered by colonialism. Mike Davis’ book Late Victorian Holocaust belongs in this scholarship. Davis rightly stressed the unusual climatic conditions of the last nineteenth century and the element of surprise that it caused. But his explanation of the famines followed the colonial-mentality theory.

Tamoghna Halder too follows the “colonial-mentality theory” of famine causation. His piece cites the word “attitude” as many as four times. He thinks the British colonial rule in India intentionally withheld relief from suffering Indians. “Attitude” begs in the racism and brutality of British colonial rule in India.

There are five reasons that make the “colonial-mentality theory” of famine a bad theory. First, the attitudes and biases of the colonial state cannot be properly researched because the British Indian state did not work according to an announced official policy statement. No matter what some postcolonial historians may think, we cannot get inside the heads of colonial officers to check what they were thinking. We can only see actions and writings. Individual officers had their private beliefs, but they did not always articulate these or speak in one voice and rarely had influence in policymaking. So, the question – did the administration wilfully withhold relief – can never be answered.

Second, the question has an insidious effect. It obstructs open debate. Did indifference drive the imperialists? This question tricks the historian into answering with a “yes” every time because to answer “no” sounds like saying the imperialists were not so evil after all. Answering “no” sounds like a defence of the empire. It is a question that does not permit disagreement. We can, of course, say that it is a bad question because we cannot test it. That is exactly what it is.

Third, the colonial-mentality story does not fit two facts. First, the late-nineteenth-century Deccan famines were regional, whereas colonialism was pan-regional, suggesting that local geography mattered. Second, the death rate from famine and related epidemics fell sharply and permanently from around 1900 (if the flu pandemic is ignored). India was still under colonial rule. That the mentality of the colonial rulers changed radically is a bad argument. There must be another explanation.

Fourth, the colonial-mentality theory overlooks that human agency is always constrained. State capacity to intervene in a tropical famine in the 1870s was constrained by poor information, distorted information, limited money, limited knowledge of causation, and conflict among stakeholders. Closer to our times, the Covid pandemic illustrates the play of all these things that limit the state’s capacity to cope. So did the Deccan famines. The theory of wilful refusal to help is a bad model because it overlooks the possibility that the state could not help.

Fifth, as I said, Indian society carried discriminatory sentiments, which also contributed to famine deaths. An overly state-oriented theory will never be enough to explain why famines ended.

State capacity theory

These anomalies lead some to discard the colonial-mentality model and adopt a different one to understand famines. Call it the “state-capacity theory.” One asks whether the state and society had the capacity to stop severe droughts blowing up into a famine facing an arid environment. The answer is they did not have the means to cope with disasters until the end of the nineteenth century. State and society acquired the means to cope from 1900.

What were these means? As for state intervention, the means included the famine relief codes, canals, sanitation, epidemic control, and collection of weather, crop, and water data. The state enabled the railways. The works of Michelle McAlpin in the 1980s and recently Robin Burgess and Dave Donaldson suggested that the end of famine had owed to faster and cheaper transportation of food via the railways. If so, inadequate communication and transport, and insufficiently integrated markets, were why droughts triggered famines in 1876, 1896 and 1898.

Food production also increased since the late nineteenth century. Farmers like those in Punjab produced, and Indian merchants traded more food. Epidemic research, and movements to gain equality in water access, like B.R. Ambedkar’s in Mahad, played parts. The effort to gain the capacity to cope delivered a sharp fall in death rates from 1901 (the flu pandemic excluded). My book Monsoon Economies tells this whole story of mitigation (MIT Press 2022).

Bad data

Halder uncritically accepts a flawed explanatory model, colonial-mentality, which allows him to ignore a good explanatory model, state capacity. Instead, his critique of my piece tries to salvage the colonial-mentality model with dodgy statistical work.

Disputing my claim that the late-nineteenth-century Deccan famines were regional whereas colonialism was pan-regional (and therefore, colonialism did not matter as much as we think), Halder says, “the north did witness famines.” This is a completely wrong reading of the census data. Population rose in the North Zone in 1871-1901 for every census year in these 30 years. Population fell in the famine-hit South and West Zones in 1872-1881, and famine-hit West and Central Zones in 1891-1901 (Leela and Pravin Visaria in the Cambridge Economic History of India). There were some excess deaths in the North, but small compared to the Deccan, and due to diseases like malaria, unrelated to food and water scarcity.

Halder then appears to recycle an Indian nationalist claim that famines were less frequent and handled better before British rule. If that were indeed the case, we might conclude that the British wilfully withheld relief. I believe that droughts and famines were a part of the history of India’s drylands. We will never know whether these were more frequent or less frequent before British rule and were handled better or worse before. We do not have the right kind of data.

Halder seems to side with the nationalist line by saying that “there exists enormous documentation on the frequency and severity of famines, for precolonial India.” I am not entirely sure where he is going with this. He does not go as far as to say that famines were less frequent and handled better before colonial rule but must imply that, otherwise, I do not see the point of this statement.

“Enormous documentation”? About what? Numbers who lived? Numbers died? Rainfall? Food supply? Population? Public spending on welfare? Death rates? Causes of death? The answer to all these questions is we have nothing of that sort. Without that data, severity and frequency over a long period cannot be measured. These sources do not give us demographic data because they were not created for that purpose. Most sources were travelogues or hagiographies, and very few came from public records. Their purpose was not to collect demographic data. Famines were mentioned incidentally.

On “enormous documentation”, Halder’s ally is Alexander Loveday. Halder cites this author repeatedly as if I have unjustly suppressed the voice of a ground-breaking thinker. I have discussed and rejected Loveday’s work in The Paradox of the Raj: How British Rule Changed India’s Economy (Palgrave 2019). Loveday, a Cambridge scholar in 1914, did not do original research but prepared a list of known famines since the beginning of the Common Era. He had no demographic data because his sources did not yield any. Scholars ignore the book because it is unimportant, amateurish, and lightweight.

We can do better than that with precolonial history. Scholars like Tim Dyson, Sumit Guha, Kingsley Davis, and the Visarias have tried to reconstruct the precolonial population trend. These scholars disagree on the average growth rate. But they do agree on two things: low (near-zero) population growth rate (or high death rate) in precolonial India and the onset of a sharp fall in death rates around the 1920s. Combining the two suggests an obvious conclusion: British rule could not have mattered as much as we think in causing famines. Halder does not cite these findings.

The drawback of Halder’s colonial-mentality theory is exposed most starkly when explaining the end of the dryland famine. My theory is that the onset of a demographic transition in India from 1900 (excluding the flu pandemic) happened because the state and society gained the capacity to cope with such events. Halder’s explanation is: “The second half of the 19th century saw some change in the attitude of the British as the crown took over”. I do not know what “some change in attitude” means, but it sounds like a very awkward start to an explanation for India’s demographic transition.

Bengal 1943

I conclude with a brief comment on the Bengal Famine of 1943. There is a tendency among scholars, and Halder seems to fall in that trap, to treat this famine like the Deccan famines. These were completely dissimilar events. They were dissimilar because different things caused them. Food procurement during World War II caused the Bengal famine. Climatic shock did play a role, but probably not the main role. Without the War, Bengal might see a severe shortage but no famine. Because the state was an agent in causing the crisis, it is not credible to say that the state lacked the capacity to provide adequate relief. Mentality and attitudes did matter more on this occasion than state capacity.

But whose attitude? Whose mentality? “Colonial-mentality” does not make any sense in 1943 Bengal. India was not governed by the British in 1943. Only the federal government with a limited set of duties and in control of the army were British responsibilities. Elected Indian legislators governed provinces that usually performed welfare and relief duties.

These elected legislators running the Bengal government in 1943 did not ask for relief, did not declare the famine as a famine, and worked on the hunch that the food shortage was due to merchant speculators but did not test the hunch. Any biography of the key figures involved will tell us they did that not because the British told them to do it. They were not puppets. This was a democratic government, meaning that the politicians knew that their power flowed from the ballot, not the governor’s or Viceroy’s desks. The task of regulating markets had been devolved to this government and did not lie with London or Delhi. The British needed the Bengal politicians as much as they needed the British, Bengal being the eastern frontier of the War. It is not credible that Bengal politicians acted in fear. They did not act at all. For crucial months they were in denial. Most historians of the Bengal famine seem afraid of asking why Indian mentality proved so unhelpful in 1943.

If their attitude is a puzzle, so is that of the provinces with plenty of food. There was no famine in most regions, and Punjab had a surplus. There may have been trading restrictions, but these could be negotiated. Why did interregional trade fail to rescue Bengal?

Going back to the Deccan famines, what lessons do they offer? While nature or war triggers natural disasters, the scale of deaths and destruction depends on the capacity to cope and the many factors that impair that capacity. Politics is a potential factor. There were many other factors, most importantly, knowing the causes.

Tirthankar Roy is a professor of economic history at the London School of Economics and Political Science and the author (with Leigh Gardner) of The Economic History of Colonialism (Bristol, 2020).

Headline photo by Colomb, J. C. R.: Map of the world showing the extent of the British Empire in 1886.