Responding to Sullivan and Hickel’s recently published research article (in World Development) and an opinion article (in Al Jazeera), Tirthankar Roy, points out how the authors are wrong in claiming that British colonial policies caused several famines in India. All that is fine, except that these articles neither investigate nor come up with any original claim regarding the causes of famines in colonial India. The central claim in their research article is that capitalism did not necessarily result in an improvement of human welfare in the 19th century – contrary to the relatively popular belief that it did. In the opinion piece, they argue the same, but solely with a focus on the negative impact of British colonial policies in India in terms of excess deaths, decline in wages and living conditions. In order to support this distinct set of claims, among other supporting evidence and quantitative techniques, Sullivan and Hickel cite one existing claim (from prior literature) that colonial policies induced multiple famines in India. And yet, as the term colonialism has become a triggering point for Roy in recent years, he titles his shadow boxing exercise as “Colonialism did not cause the Indian famines”. If the intention of Roy is to refute Sullivan and Hickel’s original claim, he fails at it miserably. If the intention of Roy is to weaken Sullivan and Hickel’s set of supporting evidence, one may argue that he does so at least partially, but that’s true only for the opinion piece (and not the research article). However, I will argue in this response why Roy fails to achieve even that! This leaves one to speculate Sir Tirthankar Roy’s real intentions, which is not the task of the current article.

What are the causes of famines? Contesting Theories

Roy’s central argument is that the scholarship linking British colonialism to the Indian famines, is outdated and discarded by the most important research in Economic History. Roy offers five arguments to support this claim and I will investigate all of them, but first, one should note that in academia, particularly in Economics, it is rare for ideas to be ‘universally discarded’. Broadly speaking, on the origin and explanation of famines – there exists several views. 

The Malthusian view would suggest famines are inevitable (and in implication, necessary) checks that would naturally emerge and balance the unrestricted growth of population, and in the process, mitigate the shortage of food. The Entitlement approach, furthered by Amartya Sen, would view famines originating out of failures to map endowments to entitlements, and shortage of food may neither be a necessary nor sufficient explanation for famines. Although Sen’s entitlement approach could possibly be attributed to left-leaning economic thoughts, the Marxist approach to famine is more specific and differs from Sen’s. For example, the entitlement approach does not explain why the Irish masses, in the years leading up to the great famine, became heavily reliant on potatoes (a reduction in their subsistence endowment set), while the Marxist view does. Typically, Marxist historians use a distinct set of analytical tools to examine agrarian and ecological conditions leading up to a famine. In the Irish example, they would argue that a combination of increased extraction of surplus from each worker, and an increase in the number of tenants (through subdivision of land) resulted in limited amounts of land allotted for subsistence production, and in turn, left the Irish population heavily reliant on just one crop.

Further, Cormac O Grada’s work “Famine: A Short History”, engages with different frameworks but refrains from offering a one size fits all argument. He manages to shed light on the different roles that market, natural calamities, wartime rationing and distribution policies may have played in causing and/or worsening most of the famines across the world and over centuries. It exposes both the relevance and restricted nature of the different contending theories of famine, and in my opinion, is perhaps one of the ‘most important research’ on famine out there. While some scholars may fail to acknowledge the enormous potential as well as limitations in all these different takes, for the rest of the world, these debates are far from being settled.

Fact Checking of Roy’s Argument One: 

Let’s cut to the chase. Roy’s first argument is that we don’t really know the frequency and severity of famines before the colonial era, and hence, can’t really compare them with the famines in colonial India. Not really. There exists enormous amounts of documentation on the frequency and severity of famines, for both colonial and pre colonial India. It’s fine if Roy does not want to believe in any of these accounts from pre colonial India, but his beliefs may be rooted in profound ignorance. First, he writes that nationalists such as B M Bhatia claimed that in pre colonial India, severe famines occurred only once in 50 years. False. B M Bhatia never claimed that. He cited these statistics from Alexander Loveday’s work (The History and Economics of Indian Famines, 1914), in which Loveday, a young British scholar at Cambridge himself made that claim, based on a list of famines that he compiled (297 – 1907). Not just Bhatia, but several economic historians, including Cormac O Grada have cited this statistic from Loveday’s work as it is the first academic work to (a) carefully compile a list of famines, after some degree of scrutiny of the sources, and (b) to discuss the economic conditions around these famines. 

Second, Roy believes that frequency of famines in pre colonial India depended on the frequency of hagiographies. This is an example of how ignorance often leads to false perceptions. Not all the entries in Loveday’s list (which Roy must read or revisit) are based on hagiographies, as hagiographies could not have been written every 5 years – the average frequency of famines in India during 917 – 1907 (according to Loveday). Another resource would be Paul Greenough’s list of Indian famines between 298 BC and 1943-44 which identifies “four famines before 1000 AD, twenty-four between 1000 and 1499 AD, eighteen in the sixteenth century, twenty-seven in the seventeenth, eighteen in the eighteenth, and thirty in the nineteenth.” One can keep adding, but let me end with this one, which again does not rely solely on hagiographies. This list refers (and links) to multiple reports among other historical records, documents 75+ famines between 1500 to 1750. Hopefully these resources would enable Roy to write his next article with more evidence, and claim what he is dying to claim – “famines were as frequent and deadly, if not more in pre colonial India”. Once he does that, I will consider another response but given his citations are all over the place, I am afraid he might end up citing me for making the claim above. I am not even claiming all these resources (and corresponding sources) would check out on face value. I am only doing a fact checking exercise at this point, and my humble submission is just that Roy’s first argument is factually flawed. 

A noteworthy observation is that almost all the pre colonial famines, going by the historical records, hagiographies and reports, were caused and exacerbated by a mix of factors: natural calamities, wars, and poor policy responses. And for that matter, famines could have occurred even as a result of certain economic policies. Maybe Roy would be relieved to know that the British were not the first to induce famines through bad policymaking. For example, some historians argue that the 1291 famine under Alauddin Khilji’s regime, resulted from a set of market policies that fixed the prices of grain. There is no data that can substantiate such a claim, and yet it is a largely accepted one! The point is, studying historical records does not necessarily require analyzing statistical tables, and particularly in the absence of data, depending on historical narratives, accounts, and hagiographies is a widely accepted method. Those who seriously engage with such historical accounts, are almost always careful in making inferences, admissibly due to biases in the narratives, just like a mindful cliometrician would do. In fact, colonial census records are not free from administrative biases and inconsistencies – does that mean Roy should stop using them?

Fact Checking of Roy’s Argument Two: 

In academia, careless and casual lying is known as misrepresentation and willful omission of facts originating from lack of rigorous scrutiny of resources at disposal. Roy’s second argument, “Why did the south have three famines during 1877 to 1899 but not the north?” is a classic example of that. First, the north did witness famines during this time. The famine of 1896-97, which Roy mentions as one of the three major famines to have hit the Deccan, actually started in Bundelkhand and eventually impacted Bihar, United Provinces Central and Western provinces in addition to the Deccan! Second, even the famine of 1877, which was devastating for the Deccan, directly impacted certain parts of north and central India. In fact, all the famines that occurred during the second half of the 19th century, taken together, resulted in excess mortality in almost all the provinces. Third, while it is true that all the major famines during this time came right after severe shortfall of monsoon rains (and subsequent drought), the path from drought to famine to excess mortality passed through several junctions, including that of colonial policies. Even when the severity of drought remained more or less similar across districts for a given famine, there was great heterogeneity at the level of districts in terms of implementation of relief policies, grain exports, attitude of local administration, railway access – to name a few among a multitude of factors that caused a greater heterogeneity of impact (excess mortality, migration etc.) across these districts. Each of the colonial famines, therefore, deserves a discussion of their own, along with a scrutiny of policies before, during and after, but for the sake of time, let’s look at the broad picture. 

Understanding the famines under colonial rule:

One can broadly divide the timeline between the Battle of Plassey till India’s independence in three parts: 1757 – 1858 (takeover and expansion of EIC rule, and multiple severe famines), 1858 – 1900 (British Raj with multiple severe famines) and 1900 – 1947 (British Raj with almost no severe famines, except in 1943). In the first part, which witnessed the Great Bengal Famine of 1770 (among several others), the EIC initially left the matter of disaster management to the goodwill of private individuals, while simultaneously ensuring increased amounts of land tax collection during and after the famine. In the case of the subsequent famines, the EIC attempted to act, but with no real intentions of preventing loss of life. As Alexander Loveday remarked, “If the mortality of 1783 was less, it was mainly due to the lesser intensity of the famine; and though the expenditure was greater in 1832 and 1837, its benefits were counterbalanced by an almost equal incompetency”. 

The second half of the 19th century saw some change in the attitude of the British as the crown took over, but several of the administrators were still high on Malthus. Enough has been written on Richard Temple’s erratic switch from a relatively generous mode of relief (Bihar famine, 1874) to a strictly conservative one in 1877 (The Deccan famine). His policy of providing only a pound of grain a day, known as ‘Temple Ration’, compounded the severity of the deccan famine. On top of it, the British saw the process of famine relief as a window to exploit labor – the already struggling masses had to earn the relief by working in large scale construction (roads, among other public projects) in both rural and urban areas. In exchange they were paid at best subsistence wages, also known as the famine wages. This measure of ‘anyone able to work should earn their relief’ continued even in the case of the famines during 1896-97 and 1899-1900, and excess mortality too, remained a reality, but these famines also saw a relatively greater expenditure on relief as well as somewhat uniform implementation of ‘The Famine Codes’ (drafted during the 1880s) at the district level, albeit not at a province level.

In the third period (1900 – 1947), there was no severe famine other than the 1943 Bengal Famine, which possibly makes Roy so satisfied that he goes on to claim that colonialism did not cause the famines, rather it mitigated those. He attributes this success to the rapid expansion of railways during the late 19th century and afterwards, which integrated the domestic markets faster than ever, and he is correct – but only partially. Railways were not the sole channel through which famines were mitigated. As several scholars including Cormac O Grada notes, there was a gradual shift in attitude of the policymakers and administrators, away from ‘hard line Malthusianism’. There is ample text based evidence, in reports after reports that the British administrators wrote (including the policy responses), where voices such as Richard Temple’s gradually faded away, and that of principled Smithians who also wanted to save some lives, rose to prominence. They did not compromise on the principles of the free market and open trade policies, but certainly moved away from the argument of Malthusian checks and balances. 

And let’s talk about this railway business for a second. The railways were not set up as a policy response to mitigate famines. The impact of railways in mitigating or preventing famines in the late 19th and early 20th century, is simply a by-product, not the objective of the railways’ existence. The policy failures that multiple scholars blame for the famines, were either direct measures pursued in response to famines, or measures related to land, infrastructure and trade policies in the years leading to (as well as during) the famines. The role of the former receives more prominence in academia but even the latter, mostly furthered by the scholarship from left historians, raises several questions regarding the British policies and their impact on Indian agriculture and ecology. Core argument is that certain areas were left so vulnerable that any weather shock could easily translate to famine (or famine-like) conditions. Areas that depended on floods for farming, were left vulnerable to floods by mindless infrastructure projects. Cultivable lands were left uncultivable in several parts of the doab. In some places they pursued irrigation projects and messed up the water distribution, while not pursuing such projects in areas that needed it. One should be mindful that none of these are uncontested claims, but the objections to these are not full proof either – leaving certain questions open and worthy to investigate.

Roy’s third argument: lies and damned lies, without statistics

In his third point, Roy criticizes Sullivan and Hickel for relying on Allen’s work on living standards on the grounds that the “India dataset on wages … is of doubtful value.” and of course, he does not cite any supporting evidence. Perhaps, careful citation is a task left only for the leftist scholars. Roy goes on to claim that Sullivan and Hickel should not have cited this data, for two reasons – (a) “almost all the data that Allen and others use come from the indo-gangetic plains, which did not see famine in the nineteenth century”, and (b) “there is nothing comparable—in fact, nothing at all—for the regions where the Deccan famines broke out.” Let’s count how many things Roy threw out of the window in these two sentences. 

One, in saying that the indo-gangetic plains did not see any famine in nineteenth century, he disregarded the death of millions of people that died from the Upper Doab Famine (1860-62), casualties from the Bihar famine, and several hundred thousand more from direct and indirect impact of the 1896-97 famine – which, as mentioned earlier, started in Bundelkhand and impacted large parts of United Provinces, Bihar (in addition to other places that don’t belong to the indo gangetic plains). How on earth does he keep repeating the lie that 19th century northern India (and the indo gangetic plains) did not witness any famine? 

I am now tired of this business of counting, but, two, he completely disregarded the work of both Parthasarathy as well as Broadberry & Gupta, who would disagree with each other on certain accounts, but both provide long term series of wages in southern India. Broadberry and Gupta, in fact, estimates northern and southern Indian wages to be at similar levels, both in terms of grain wages as well as silver wages, and show that the southern Indian wage data is suitable even for international comparisons. If Roy does not find that exercise a useful one, he should have offered a counter argument, but instead he chose to dismiss these previous works altogether by misrepresenting and wilfully omitting….

Roy’s fourth argument: (mis)using Ravallion

Several scholars have argued how exports compounded the severity of famines in colonial times. One is free to agree or disagree based on evidence. What Roy does in his fourth argument, however, may cause Martin Ravallion to roll his eyes from the grave. In a haste to defend colonial trade policies in the face of famines, Roy spins the story from Ravallion (1987) and claims that Ravallion showed how “food exports did not expose the countryside to a food shortage” and that “trade stabilized domestic consumption rather than reducing it”. Ravallion’s original claim is way more nuanced. Yes, he does remark: “Adam Smith’s influential followers in India appear to have been right in their assessment of the main qualitative effects of trade during famines.” But observing the small change in short run, he also comments, “external trade was a less-than-perfect consumption stabilizer and export sluggishness and output lags considerably diminished its short run performance”, and concludes, despite his theoretical support of the Smithian principles, “… trade’s quantitative effect in insulating domestic rice consumption from output uncertainty does not support an optimistic assessment of the potential for short-run stabilization by these means. And the extent of famine relief by trade is likely to have depended heavily on its short-run effects.” The other works that Roy cites, do not really establish any direct causal channel between trade openness and mitigation of weather shocks. Lack of causality is not necessarily a demerit, but that does not settle the discussions either.

Roy’s fifth argument, or an incoherent rant

Roy’s fifth argument is a rant against Mike Davis. Rants are often fun, engaging and even enriching to read, but this one is sheerly incoherent and incomprehensible. Yes, weather shocks happened throughout the period of 1900-1947, and those did not translate to famine – and yes, part of it is attributed to the railways but parts of it, as explained earlier in this response, is also because of a sharp change in the attitude of the policymakers. His remark that there is no evidence that colonial apathy ended in the 1900s, is a denial of the series of reports that document these changes, and I have discussed this at least twice in this response already!

And there goes his five arguments, written with  utmost disrespect and disregard for almost all previous work on the topic, generously sprinkled with ignorance and half baked truths. I am no particular fan of either the left or right-leaning scholarship on Indian famines, and personally for me, much remains desired from the left, as several of their existing claims have long been contested, but the ground of contestation has to be valid, and rooted in facts – neither distortions, nor intentional omissions are going to help Roy’s academic credentials. However, Sir Tirthankar Roy may indeed succeed in reaping whatever fruits he is expecting out of his mission of defending colonialism.

Tamoghna Halder is an assistant professor in Economics at Azim Premji University. His primary research interests lie at the intersection of Economic History and Economics of Identities. 

Acknowledgement: Thair Mohhamed, a research assistant working with me has helped in checking some of the facts used in this response.

Photo: The Bengal famine of 1943 by Chiitoprasad Bhattacharya.