Colonialism did not cause the Indian famines

By Tirthankar Roy. He is a Professor of Economic History at the London School of Economics.

"Accusations that Britain drained India of resources and starved Indians to death are nothing new in some circles, but at the moment they are being spread with religious zeal. Al Jazeera, the media conglomerate funded by the Qatari government, has recently joined in. But this accusation has long been disproved, and the truth is that British rule in India mitigated famine.

In an Al-Jazeera piece, Dylan Sullivan and Jason Hickel[1] claim that British rule inflicted ‘tremendous loss of life’ in nineteenth-century India by causing the devastating Deccan famines. Imperial policies, they say, made famines ‘more frequent and more deadly.’[2] They cite the economic historian Robert Allen to suggest that Indian living standards declined in the nineteenth century, because the British rule ‘drained’ India of money and food. Indians were starving when the famines hit them. Colonialism triggered the genocide. Social media posts spread similar messages with religious zeal. Many blogs and sites sharing such sentiments have reprinted Sullivan and Hickel’s piece.

Sullivan and Hickel are not doing anything new. In a 2001 book, Mike Davis said Britain’s apathy towards her Indian subjects caused mass deaths in late nineteenth-century Deccan.[3] Indian nationalists argued a hundred years ago that food export supported by a free trade policy and the British-built railways left the Indian countryside with too little food. B.M. Bhatia said that ‘in the earlier times a major famine occurred once every 50 years,’ whereas ‘between 1860 and 1908, famine or scarcity prevailed in … twenty out of the total of forty-nine years,’ implying that colonialism made famines ‘more deadly’.[4]

Economic historians have examined and discarded every one of these assertions. Sullivan and Hickel are salvaging an outdated idea by ignoring the most important research done on the subject. I will offer five grounds to show why their version of history is incorrect and biased.

Five errors of a biased argument

First, the claim that colonialism caused famines cannot be verified against previous experience because there is no evidence that famines were less frequent or less deadly before. Whereas the government statistical system recorded the colonial-era famines, the precolonial data came from hagiographies and travelogues. These dissimilar datasets cannot be compared. The frequency with which famines occurred in earlier times depended on the frequency with which hagiographies were written. If this was once in fifty years, we would conclude that famines happened once in fifty years, as Bhatia did. The authors of these hagiographies praised the relief work because the kings paid them to do that. We might conclude that precolonial kings were more caring than the British. None of this makes any sense.

Second, colonialism cannot be accused of causing famines everywhere. The famines that Davis called late-Victorian holocausts happened in the Deccan, 1876-77 in the Rayalaseema region west of Chennai (Madras), and 1896 and 1898 in the Deccan Traps closer to Mumbai (Bombay). All along, in the Indo-Gangetic Basin, the population rose. The Raj ruled over the north and south. Why did it spare the north?

Here is the answer. The British Raj did not start the famines. Geography did. 1877 was the driest year in over a century (1871-1978) for which rainfall data exists. The average rainfall that year was 30 percent short of the long-term level, and a 25 percent shortfall developed again in 1896 and 1899. Monsoon failure of such an order can cause distress by drying up all accessible water sources. The effect was disastrous in the Deccan Plateau because it was normally much drier than the north and did not have rivers fed by the Himalayan snow, unlike the north. Monsoon rain was the primary source, and when that failed, cultivation stopped, and distressed people lived on infected water to die of cholera. Colonialism had nothing to do with the reason they died.

Third, the living standard research that Sullivan and Hickel cite has a problem. Allen’s work on living standards is pathbreaking, but the India dataset on wages in northern India compiled by early-twentieth-century historians is of doubtful value. These historians left too few details on who earned these wages for what works, or on the labour market, or whether they were men or women, or the contracts involved. These are just numbers without context. A trend drawn by comparing such numbers over the long run may mean that the context of work changed, not the living standard. For example, there was an enormous expansion of casual-wage-based hiring of individual workers in the nineteenth century, whereas earlier employers often hired families or groups and paid them customary fees. The kind of work for which the Mughal household or urban silk factories or the East India Companies hired people did not exist anymore in 1900. Earlier wages were often job-based, later wages were daily or hourly. For all these reasons, later wages could look smaller, but that would not mean that the living standard had fallen.

That Sullivan and Hickel rely on weak data is the least of their problems. The evidence is not relevant at all. Almost all the data that Allen and others used came from the Indo-Gangetic Basin, which did not see famine in the nineteenth century. There is nothing comparable—in fact, nothing at all—for the regions where the Deccan famines broke out.

Fourth, the nationalist criticism of food exports has long been discredited. The economist Martin Ravallion showed in a 1987 article that food exports did not expose the countryside to a food shortage. Food exports rose when Indian prices fell below world prices or after a good harvest and fell when there was a bad harvest.[5] In this way, trade stabilized domestic consumption rather than reducing it.

Fifth, that colonialism caused famines is based on flawed logic. Dryland Deccan famines disappeared after 1900, though weather shocks did not. The significance of the end of dryland famines was momentous for India. It led to a permanent fall in death rates. From 42-50 per thousand in 1911-21, the rate fell to 33-38 in the next decade, 30-32 in 1931-41, and 25 in 1941-51.

Any good theory must explain the end of famines and their occurrence jointly. Colonial apathy cannot do that because there is no good way to show that apathy ended around 1900. It is a nonsensical idea and an unverifiable one. This is why Mike Davis’s work cannot be trusted. He avoided showing how mentality changed, yet the end of famine demanded that he should.

Why did the famines really end?

Why did the famines really end? Michelle McAlpin in the 1980s and recently Robin Burgess and Dave Donaldson suggested that the end of famine was owed to the railways.[6] McAlpin examined the experience of the Bombay Presidency, where the second and third of the three Deccan famines had happened, and showed that markets and railways improved food distribution enough to reduce the impact of harvest failure. Burgess and Donaldson show that access to railways reduced local food price instability. In short, statistical research confirms that the railways caused the end of famines and delivered the gift of life to generations of Indians born after 1900. I have made a similar point about well water in the Deccan countryside.[7]

These works suggest a better theory of why the famines happened. The capacity of the states and the markets to provide food and water to the needy was small against the scale of the natural disasters. All large natural disasters reveal such a syndrome. They show that the capacity of the people in charge of relief can be constrained by poor information, distorted information, limited money, limited knowledge of causation, and conflict among stakeholders. Covid illustrated the play of all these things that limited the state’s capacity to cope. But Covid also showed how governments learned lessons and moved on. So did the Deccan famines.

The Raj learned lessons. The three Deccan famines generated data and research under government sponsorship on a scale not seen before. The results were famine codes (a blueprint for relief), canal construction, railways, sanitation of water bodies, cholera control, and collection of weather, crop, and water data. The effort to gain the capacity to cope delivered a sharp fall in death rates from 1901.

Droughts are, and were, common in tropical monsoon geographies. The start of a peacetime drought had nothing to do with states. But states could learn lessons and be better prepared. The Raj learned how the government could get better at managing massive weather shocks, the type to hit the Deccan at the end of the nineteenth century. Sullivan and Hickel dare not tell that story, for that would dilute their partisan message."