Answers are grounded in this paper and cite their sources.
While famine research has matured much since Sen (1981) published his Poverty and Famines, there remains some tension between neo-Malthusian and other positions about the mechanisms of famine, not at least in popular debates (Alemu, 2007; Devereux, 1988; Ellman, 2000; Gráda, 2011; Khandakar Qudrat, 2006; Rubin, 2009). Neo-Malthusian-inspired research tends to focus on macro-level factors, leading often to either a food availability decline (FAD) or food requirement increases (FRI) type of explanation for famines (Daoud, 2010; Malthus, 1826). In contrast, Senian-inspired research relies often on food entitlement decline (FED) explanations (Alemu, 2007; Reddy & Daoud, 2020). Such FED explanations tend to argue that famines can occur despite there being sufficient food at the macro level. Sufficient food means that there is just about enough food to feed the population of interest; abundant food means that there is more than enough to feed that population. Moving beyond these entrenched theoretical perspectives, today, most famine researchers would agree with the statement that famines can occur due to one or several combinations of FAD, FRI or FED (Devereux, 2007; Shaw, 2007). Despite this agreement, there has been a lack of theories that show how exactly Senian and neo-Malthusian can be synthesized in one theory of famines: a theory that accommodates FAD, FRI or FED explanations. In Daoud (2017) ‘Synthesizing the Malthusian and Senian approaches on scarcity: A realist account’, appearing in Cambridge Journal of Economics, I proposed a general theory that provides the necessary conceptual tools to reason about FAD, FRI or FED explanations. In related work (Daoud, 2007, 2010, 2011a, 2011b, 2018), I generalized it to all kinds of scarcity, abundance and sufficiency cases. And with my collaborator, I linked it to social action theory (Bååth & Daoud, 2023), preference formation (Daoud & Puaca, 2011; Puaca & Daoud, 2009, 2011) and reasoning about intangible goods (Söderberg & Daoud, 2012). I will summarize the main components of my theory in the next section, but first, I would like to clarify the purpose of this rejoinder. Peter Bowbrick (2022) has recently published a critique of my work as it appears in Daoud (2017). Although I welcome constructive criticism of my theory, the least one can expect of such criticism is that it engages with the actual theory proposed. His entire focus is on the empirical case—the 1943 Bengal famine—that I used to exemplify my theory. Besides, he seems to have misinterpreted the empirical evidence I have used to develop my theory. In addition, the conceptual validity of my theory does not even hinge on that particular case. To question my theory's validity, one would require a theoretical account of showing how it is conceptually wrong. Yet Bowbrick does not develop such a theoretical account. While I will expose how Bowbrick is mischaracterizing key empirical statements, my reply will primarily focus on the core principles of my theory of famines, and it will demonstrate how my theory generalizes to any case of scarcity. This generalization emphasizes a key point I made in the past about how famine research (and adjacent fields) should focus on the dynamics of scarcity, abundance and sufficiency (SAS), rather than assuming scarcity of resources at the onset of analysis—as is often done in neoclassical economics. That strategy of replying will hopefully be more illuminating for the readership of the Journal of International Development than showing how Bowbrick is misstating historical facts and empirical data. In theory, any famine is a case of resource scarcity because people lack sufficient food and thus die, but the question is what type of scarcity, and in dynamics with what type of sufficiency or abundance (Daoud, 2018). At the macro level (e.g., countries, regions or the planet), scarcity is a relationship between two factors: available resources and the requirements for that resource.1 Holding resources constant, scarcity can arise because requirements are increasing; holding requirements constant, scarcity can arise because resources are decreasing (Daoud, 2007). In other words, a scarce resource differs from a limited resource. A limited resource is just a claim about there being X amount of it; a scarce resource is a resource where there is not enough to satisfy all requirements. For example, there might be only 10 tons of rice available in New York City, but if there is only 1 ton of requirement for rice, then rice is not scarce. It is abundant. Thus, macro scarcity arises always because of either there being too few resources or too many requirements. There is of course a myriad of second-order causes as to why there are too few or too many requirements, from poor governance (Coutts et al., 2019; Daoud, 2015; Halleröd et al., 2013) to natural disasters (Daoud et al., 2016; Shiba et al., 2021, 2022) to policy interventions (Conklin et al., 2018; Daoud et al., 2019; Daoud & Nandy, 2019; Nandy et al., 2016; Ponce et al., 2017), and disentangling those causes is the target of research. Malthusian-inspired theories about famines tend to focus on macro-level scarcity, as their target is often the rapid population increase and the lack of production capacity (Daoud, 2010, 2017). According to these theories, when either food production fails (that is a FAD explanation) or the population increases (that is an FRI explanation) too fast, a famine will arise. Nonetheless, Malthusian theories have a hard time explaining how famines can arise in situations of macro sufficiency or abundance, that is, when there has been neither a production failure nor a rapid population increase (Daoud, 2010, 2011a). Macro-level theories are important in explaining scarcity, and thus famines, but they need to be complemented by micro-level theories about how famines can arise despite there being sufficient or abundance of resources at the macro level (Daoud, 2011a). At the micro level (i.e., persons, households and groups), scarcity has one additional driving mechanism beyond available resources and human requirements at the macro level: entitlements. Entitlements refer to ownership relations and the legitimacy under which they can be exchanged between individuals (Bååth & Daoud, 2023; Sen, 1981, p. 1).2 That is, even if there is macro sufficiency of food (where available resources are just about equalling requirements) or macro abundance of food (when there are more than enough available resources than requirements), some individuals' entitlements are insufficient to enable them to access available food resources (Daoud, 2011a). That insufficiency of entitlements can arise due to many possible reasons, such as unemployment, discrimination or disability (Reddy & Daoud, 2020). That point is one of the critical contributions that Sen made to famine research through the concept of entitlement failure (that is FED explanations). Senian theory does not preclude the fact that famine could arise because of macro factors but emphasizes that micro factors are important contributing factors explaining famines (Osmani, 1995). In fact, in some cases, FED can be the main or only important factor explaining famines, as Sen shows in his Poverty and Famines. Nonetheless, although the Senian theory is critical to our understanding of the causes of famines, its emphasis on entitlements makes it hard for this theory to carry the entire burden of a general theory of famines (Daoud, 2007, 2011b, 2017). In a series of contributions, I developed a theory of SAS that serves to bridge macro and micro theories of famines (Daoud, 2007, 2010, 2011a, 2011b, 2018). Figure 1 shows the key concepts and their relationships. This figure describes a stratified reality (Bhaskar, 2008; Lawson, 1997). Events are what empirically manifests. Researchers explain how that event is possible by referring to the necessary generative mechanisms that produce it. In turn, those mechanisms may be explained by referring to underlying causal structures—the ‘myriad of second-order causes’ that I referred to previously. Thus, the goal is to explain how an event of scarcity (or abundance, and sufficiency) can occur. In the Daoud (2017) article, the event of interest is the 1943 Bengal famine. My SAS theory refines existing scarcity definitions by defining a scarcity event as always consisting of a relationship among three generative mechanisms, shown in the second level of Figure 1: requirements (R), entitlements (E) and available resources (A). The mechanisms A and R measure available resources and requirements, respectively, in the system of interest.3 The mechanism E measures whether an individual, in that system, has enough capacity to acquire sufficient resources to satisfy his or her needs. Logically, then, there can only be seven combinations of first-order causes to why famines arise: (1) R is increasing; (2) R is increasing and E is failing for some social groups; (3) R is increasing and E is failing for some social groups and A is decreasing; (4) R is increasing and A is decreasing; (5) E is failing for some social groups; (6) E is failing for some social groups and A is decreasing; and (7) A is decreasing.4 In Daoud (2007), I named these scarcity types based on which combination they fall into. For example, the following combination, ‘(5) E is failing for some social groups’, I named quasi-scarcity, because there is sufficient or abundant food at the macro level, but a social group is still starving to death.5 Those combinations show that scarcity has different types of causes, leading to different explanations and policy responses. Using these combinations in explaining famines enables a more precise language about the causes of famines than merely referring to FAD, FRI or FED. As Malthusian-inspired famine theories focus mainly on FRI or FAD, they tend to cover explanations of famines that fall into combinations ‘(1) R is increasing’, ‘(4) R is increasing and A is decreasing’ and ‘(7) A is decreasing’. And thereby, they tend to overlook the other types of explanations. Although Senian explanations of FED tend to prefer one particular combination, that is, ‘(5) E is failing for some social groups’, they would embrace other explanatory combinations, as long as E is involved in the explanation. Thus, they tend to downplay combinations (1), (4) and (7). At the end, which explanatory combinations one endorses should depend on the empirical data of the case of interest. Thus, my theory of famines should be regarded as a framework for generating hypotheses and thinking about the causes of famines. My theory invites more research to evaluate the usefulness of my theory across many cases. As Bowbrick stands completely silent about the validity of my theory of famines, I am inclined to assume that he actually agrees with it. It seems that the source of Bowbrick's aggravation is solely the empirical case I used, yet here, he is still mischaracterizing my empirical work. Thus, next, I address two central mischaracterizations that Bowbrick leveraged. Bowbrick claims that ‘Daoud states repeatedly that there was plenty of food in Bengal’ (Bowbrick, 2022, p. 4), yet nowhere in my manuscript do I claim that there was plenty (or macro abundance in my terminology) of food in Bengal. My analysis finds that it is likely that there was macro sufficiency of food in Bengal, which is completely different from macro abundance. Based on summarizing what previous literature has found (including the Famine Inquiry Commission), I wrote ‘it is difficult to be entirely conclusive about the food supply data’ (Daoud, 2017, p. 10). To develop a sense of how uncertain the data are, I show the variation of food-production data, gathered from mainly four sources: the Famine Inquiry Commission (1945), Goswami (1990) and Islam (2007), and Sen (1981). Starting by the beginning of the year 1941, I find that the Commission estimates that there was 13% more food in 1943 than in 1941 (rice estimations only), yet no famine occurred in 1941. Sen adjusts this estimate downwards from 13% more food to about 11% (rice and wheat estimations). Goswami estimates that there was 3.7% less food available in 1943 than in the comparison year (1941) (rice only), and Islam finds that this number was only 2% less (rice and wheat). Based on these and other studies, I concluded that ‘it does seem relatively clear that a situation of severe food scarcity (or systemic abundance of food) can be ruled out’ (Daoud, 2017, p. 10). Thus, those estimates lead the analysis to the only logical option left: The most likely scenario is that there was macro sufficiency of food. Yet I argued that it is hard to explain why millions of people died6 by merely referring to macro sufficiency. To explain why so many died, I account for various underlying causes of FAD, FRI and FED. Based on a synthesized analysis of Sen and others, I conclude that the entitlement failure of petite farmers in rural versus urban areas was one of the main explanations for the Bengal famine. But that failure could not have happened without a mixture of events: the dire context produced by the World War II, the priorities of the British authorities and the local Bengali government. This mixture led to food prices soaring (partly also affected by food speculation) and rural farmers becoming unemployed in masses. In other words, several factors converged leading to the catastrophic onset of the famine that caused the deaths of millions. Nonetheless, Bowbrick erroneously asserts that ‘Daoud claims a Malthusian cause of this famine, that Bengal's population grew so fast that the province could no longer feed itself’ (Bowbrick, 2022, p. 22). As I previously delineated in my theory of famines, a population can increase while food availability can still be sufficient. The Bengali population increase did pump up the pressure on available food resources, but that increase does not have to lead to macro scarcity, let alone the catastrophe that followed, leading to the deaths of millions. These types of erroneous assertions lead me to the next set of problems in Bowbrick's criticism. Historical data are inherently uncertain. Virtually all historical research must yield to the fact that data are limited, and when they are available, those data are often imprecise (King et al., 1994). This uncertainty is inflated when the historical event of interest occurred during a wartime period. The 1943 Bengal famine occurred during the World War II, which had raged for almost 4 years, conditioning the British interest in South Asia (Mansergh & Lumby, 1973). By the end of 1941, Japan and India were officially at war. The first air strikes on the Burmese city of Rangoon occurred in January 1942. Later in February, Singapore came under Japanese control, and Rangoon was occupied at the beginning of March. During those months, the governments of India and Bengal implemented the denial policy, that is, to destroy any Indian agricultural resources or public infrastructure that could likely fall under Japanese control (Bhatia, 1967, p. 326). In the context of these events, Bengal officials and other actors had a hampered capacity to collect reliable data. This unreliability makes data on the Bengal famine inherently uncertain, and that uncertainty necessarily permeates my and many other scholars' argumentation. Like other studies, I thoroughly discuss how data uncertainty restricts my analysis (see, e.g., section 3.1 in Daoud, 2017). Conversely, Bowbrick is making a series of contradictory statements. On the one hand, he writes, ‘The raw data were known to be meaningless’ (Bowbrick, 2022, p. 9) (see in particular his section 6.3 titled The availability estimates are also meaningless). On the other hand, he is absolutely certain about his claims that ‘there could not have been plenty of food’ in Bengal. How does he know that ‘there was no plenty of food’, given that he also knows that the data are meaningless? The contradictions in Bowbrick's criticism do not stop there. Although he devotes about 24 pages to dismissing virtually all viable causes proposed by well-renowned scholars, he presents no alternative about why the 1943 Bengal famine occurred. The logic of Bowbrick's argumentation is structured along the following lines, culminating with an absence of an alternative. First, he put forth that ‘the 1943 Bengal famine is highly emotive and politically charged’ (Bowbrick, 2022, p. 1) and, thus, then presents a statement that famine research is fraudulent and builds on myths (see his sections 1 and 2). Second, he then (falsely) claims that Daoud (2017) is stating that there was plenty of food (macro abundance) in Bengal (see his section 3). Third, he argues that there could not have been any extra consumption (e.g., due to military needs or hoarding) in Bengal (see his section 4). Fourth, he argues that food speculation (causing abnormally high food prices) could not have caused the famine. Fifth, he argues that ‘the availability estimates are also meaningless’ (Bowbrick, 2022, p. 12). Then, he adds that any crop forecasts are meaningless, including that derived from crop production, crop forecasts and the estimated impact of the fungus outbreak in 1943 (see his section 6). Sixth, Bowbrick moves on to stipulate that price statistics were meaningless by writing that ‘Nobody knew how much was sold at any price’ (Bowbrick, 2022, p. 14), or ‘Price statistics are notoriously inaccurate’ (Bowbrick, 2022, p. 14) (see his section 7). Seventh, he states that British colonialism did not suppress Indian interests and that there was no political bias against Indian rural farmers (see his section 8). Eighth, he claims that there was no inflation or that inflation did not contribute to explaining the famine—despite that Bowbrick just refuted the measurement error of price statistics (see his section 9). Ninth, Bowbrick argues that the Malthusian approach does not apply in the Bengal case since the Bengali population could not have increased rapidly (see his section 11). Consequently, a natural reaction to Bowbrick's nine steps of argumentation is ‘but how does Bowbrick explain the Bengal famine then?’ Yet Bowbrick leaves the reader wanting an alternative explanation. To reiterate, observe that nowhere in his argument does he mention my theory of famines—which is about how to synthesize Malthusian and Senian theories of famine—and developing that theory is the core contribution of Daoud (2017). Scientific progress and integrity rely on at least two criteria that help counter occurrences of falsehoods or myths. First, scholars should supply constructive criticism of existing work; second, they should openly and critically reflect on the shortcomings of their work. Bowbrick fails on both criteria. Regarding the first principle, while he provides ample criticism to my empirical statements, he does not engage with my theory of famines at all—the key contribution of my article Daoud (2017) ‘Synthesizing the Malthusian and Senian approaches on scarcity: A realist account’ appearing in Cambridge Journal of Economics. With regard to the second principle, although Bowbrick correctly states that historical statistics are inherently uncertain, somehow, he seems to believe that this uncertainty does not apply to his argument. By failing on both principles, Bowbrick ends up formulating arguments that are beset by contradictions. His contradictions have already been addressed by Sen but on That failure in principles the purpose which Bowbrick set to occurrences of falsehoods and myths in famine research. is not to this article as no data were or in this
No takes yet. Share an insight, caveat, or question.
Adel Daoud (2023) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: