Image of parched land on crops field
VoxEU Column Climate Change Environment Economic history Politics and economics

Famine, war, and climate change: From the Middle Ages to the Russia-Ukraine War

The Russian invasion of Ukraine brought the spectre of war back to Europe. Famine could be next, not necessarily in the war zone but in poor food-importing countries. All recent food crises have been triggered by war. However, looking at the long run this column finds that most preindustrial famines occurred at times of high population pressure on resources and were triggered by extreme weather events whose occurrence was more probable in phases of rapid climate change. All these factors, plus war and Covid-19, are at work today.

In 2011-12, the world’s last major famine in war-torn Somalia killed tens of thousands of people, maybe as many as a quarter of a million (Maxwell and Najid 2016). More recently, according to the Global Network Against Food Crises, nearly one million people globally were placed at imminent risk from famine between 2016 and 2021. The Network’s annual reports suggest several points. First, aside from 14,000 in Madagascar’s Grand Sud region in 2021, all these unfortunates lived in war zones. The other places involved – north-eastern Nigeria and South Sudan in 2016 and 2017, South Sudan and Yemen in 2018 to 2021, Somalia in 2018, Burkina Faso in 2020, and Ethiopia in 2021 – are very predictable. Second, the periods when famine posed an immediate threat were usually short-lived, and not once did these alerts result in an actual famine. Third, this striking outcome was due mainly to humanitarian aid getting through and/or a cessation of hostilities (Food Security Information Network 2017-2022). Fourth, the Somali famine of a decade ago is a reminder that the world may not always be so lucky.

In this third millennium, given the speeds at which news travels and at which the global humanitarian community can relieve looming food crises, one might expect that famines would be few. And that is broadly the case; for all intents and purposes, famines today only happen in war zones. If we look further back in time, the link between war and famine is highlighted by the huge number of famine victims during WWI and WWII. If deaths in the newly-born Soviet Union in 1918-1922 are included, famines during WWI cost 13 million lives, while during an even more murderous WWII they cost 17-19 million. Many of these famines happened in economically backward places such as Java, Bengal, and Vietnam, where occupation and repression placed intolerable burdens on the poor. But some happened in relatively well-off places by historical and contemporary standards: think of northern Italy, Germany, and Austria during and immediately after WWI, and the Soviet Union, Greece, and the Netherlands during WWII (Wheatcroft and Ó Gráda 2017, Ó Gráda 2019, 2022).

If, then, in modern times war has usually been the trigger of famines, it is easy to understand why the Russia-Ukraine War is a cause for immediate and growing concern. But insofar as ongoing climate change currently risks increasing the ratio population to cultivable land in some of the least developed parts of the world, then the historical trigger of high population density may also be about to reassert itself (Cruz and Rossi-Hansberg 2021). In this sense, both Mars and Malthus are a threat today, and on a global scale – which leads to some worrying reflections.

In a recent article (Alfani and Ó Gráda 2018), we argue that in preindustrial times major famines have almost always occurred in periods of high population pressure on resources. This empirical relationship tended to disappear after ca. 1710 (earlier in England). Up to that point in time, Malthus was broadly right: famines were ‘natural’ events, strictly connected to the tendency of populations to grow beyond the maximum carrying capacity of their environments, that is, beyond the maximum sustainable level given the resources available. A distinction, however, should be made between the occurrence of a famine and its severity: we are not arguing that human agency was incapable of relieving the worst consequences of Malthusian famines. But the claim, associated with Amartya Sen (1981) and others, that famines are mainly about ‘entitlements’ fits more recent historical contexts better than, say, medieval or early modern Europe – or 19th century Asia. In that sense, from about 1710 on European famines became increasingly ‘man-made’ events, that is, the products of poor governance, highly uneven entitlements to food, war, or even direct attempts at causing mass starvation. War, as discussed above, has been the dominant cause of famine in recent decades.

But history has more useful lessons to offer. In Figure 1, three European areas for which particularly good data are available are observed from ca. 1300 until 1850: England, France and Italy. In all three, before the 18th century it appears to have been impossible to exceed a maximum level of population (about 20 million in France, 14 million in Italy, and 5 million in England). The Black Death, which ravaged Europe in 1347-52, reduced populations far below this level, which can be taken as mirroring the maximum carrying capacity of each area. Populations then slowly recovered bar for some area-specific oscillations (for example, in Italy further ravages of the plague in 1629-30 and 1656-57 caused population to decline during the 17th century; see Alfani 2013, 2020). As the figure shows, in all three areas famine years tended to cluster in periods of historically high population. Quantitative analysis reveals that the connection between population and probability of famine is statistically very strong until 1710, with the partial exception of England where it wanes earlier (from ca. 1630).

Figure 1 Famine, population size and climate change in Europe, 1300-1850

Figure 1 Famine, population size and climate change in Europe, 1300-1850

Source: Alfani and Ó Gráda (2018).

Although high population density appears to have been the precondition (or ‘remote cause’) for most preindustrial famines, it was not per se sufficient to start a crisis. For a famine to develop, a trigger (or ‘proximate cause’) was also necessary. In preindustrial times, in the vast majority of cases the trigger was some crop-damaging extreme weather event, usually heavy rain in spring or summer; if this happened for two or more years in a row, so that carry-over stocks were exhausted, a famine occurred. Much more rarely, famines were triggered by drought, especially in southern Europe (Alfani and Ó Gráda 2017).

The occurrence of extreme weather events is obviously uncorrelated with population levels – but many have argued that heightened meteorological instability is connected to climate change and to phases of inversion of the long-term trend from warming to cooling, or vice versa (Bradley and Jones 1992, Parker 2013). These historical phases are captured in Figure 1 by the Celsius degrees of variation relative to the 1961-1990 average. Interestingly, the maximum density of famines (in the 1580s-1600s) coincided with what is sometimes identified as the nadir of the Little Ice Age (Mann et al. 2009; Collet and Schuh 2017), although it bears noting that successive years of weather adversity do not require a Little Ice Age (Kelly and Ó Gráda 2014). The point here is not that low average temperatures affected crops directly (the main grains are fairly resistant to intense cold), but that phases of inversion of the long-term climate trend or, more generally, phases of quick climate change, may have increased the likelihood of crop-damaging extreme weather events. If these coincided with a state of precarious equilibrium between population and resources, itself the product of demographic growth, a famine developed.

Our findings for the past are, unfortunately, potentially relevant for the present. For a few years now, the Food and Agriculture Organization (FAO) in its yearly reports has been warning that the global fight against hunger has stopped making progress. The prevalence of undernourishment worldwide, which had remained stable in 2015-2019 at about 8%, jumped to 9.3% in 2020 and 9.8% in 2021. This sudden increase was the direct consequence of Covid-19, through the increase in food prices caused by the ensuing disruption to production and distribution of goods, and by the measures put in place to contain the pandemic (FAO 2022: 10 and 47). And yet, the FAO also warned that Covid-19 simply compounded the negative effects on food production of climate change, which was the direct culprit of a long series of localised environmental disasters such as droughts, storms, and floods, capable of compromising agrarian production across vast areas.

Based on historical experience, how can we describe the current situation? A decade or two ago, there was reason to be optimistic about the future prospect of a famine-free world. Moreover, the experience of recent years suggests that we can still just about cope with the threat of imminent peacetime food crises. But global warming has compromised our previous ability to cope with continuous population growth and the threat of famine. Rapid climate change, itself boosted by anthropogenic factors, will continue and can be expected to make the occurrence of extreme weather events rather probable. To this add shocks like Covid-19, whose impacts on food security are partly ‘natural’ (due to the appearance and the spread of the new pathogen) and partly ‘man-made’, due to the pandemic containment policies applied worldwide. Finally, there is the Russia-Ukraine war, entirely the consequence of (misguided) human choices and involving major food producers. Will this war be the straw that breaks the camel’s back – the trigger, or proximate cause, of major famines? This question has no easy answer. Although, in the current situation, all the major factors associated with past famines appear to be present simultaneously, much depends on future and unforeseeable developments. One thing, however, is clear: the risk of famine is today much higher than it has been for many decades. It must not be underestimated.


Alfani, G (2013), “Plague in Seventeenth Century Europe and the Decline of Italy: An Epidemiological Hypothesis”, European Review of Economic History 17(3): 408–430.

Alfani, G (2020), “Pandemics and asymmetric shocks: Lessons from the history of plagues”,, 9 April.

Alfani, G and C Ó Gráda (eds) (2017), Famine in European History, Cambridge University Press.

Alfani, G and C Ó Gráda (2018), “The timing and causes of famines in Europe”, Nature Sustainability 1: 283-8.

Bradley, R S and D Jones (eds) (1992), Climate since A.D. 1500, Routledge.

Collet, D and M Schuh (2017), Famines During the ‘Little Ice Age’ (1300–1800): Socionatural Entanglements in Premodern Societies, Springer.

Cruz, J-L and E Rossi-Hansberg (2021), “Unequal gains: Assessing the aggregate and spatial economic impact of global warming”,, 2 March.

Food and Agriculture Organization (FAO) (2022), The State of Food Security and Nutrition in the World.

Food Security Information Network (2017-2022), Global Reports on Food CrisesNY: World Food Program.

Kelly, M and C Ó Gráda (2014), “Change points and temporal dependence in reconstructions of annual temperature: did Europe experience a Little Ice Age?”, Annals of Applied Statistics 8(3): 1372-94.

Mann, M E et al. (2009), “Global signatures and dynamical origins of the Little Ice Age and Medieval Climate Anomaly”, Science 326: 1256–1259.

Maxwell, D and N Majid (2016), Famine in Somalia: Competing Imperatives, Collective Failures, 2011-12, Hurst.

Ó Gráda, C (2019), “The famines of WWII”,, 2 September.

Ó Gráda, C (2022), Innocent Bystanders: The Civilian Victims of WW1 and WW2, forthcoming.

Parker, G (2013), Global Crisis: War, Climate Change and Catastrophe in the Seventeenth Century, Yale University Press.

Sen, A (1981), Poverty and Famine: An Essay on Entitlement and Deprivation, Clarendon Press.

Wheatcroft, S G and C Ó Gráda (2017), “The famines of World Wars I and II”, in Alfani and Ó Gráda (eds), Famine in European History.

0 Reads