Ó Gráda sets the scene by asking how famine is to be distinguished from dearth, when and where famines have occurred, and whether famines (rather than epidemics) are the ultimate check on population growth.
His introduction also touches on the use of oral history. This leads into a survey of the horrors commonly associated with famine: crimes, slavery, infanticide and child abandonment, and cannibalism. In most cases the evidence for these, or their extent, is limited, with accounts that may be inspired as much by literary tropes as experience.
Traditional ways of coping with famine include resort to famine foods, borrowing money, and migration.
"Certain tribal or ethnic groups developed their own 'niche' preventative strategies against famine. For the nomadic Fulbe people of the Sahel the niche was a form of transhumant pastoralism; for speakers of the Shari-Nile languages, it was a variety of sorghum that was particularly resistant to conditions of both extreme drought and extreme flooding."
Demographic analyses of famines consider how many died, the gender and age distribution of deaths, and missing births and subsequent demographic rebounds. The causes of death during famines are slightly counter-intuitive, with actual lack of food dominating in famine areas with modern infrastructures, such as Leningrad under siege or the Netherlands Hunger-winter, but deaths from infectious diseases dominating in areas with poor public health infrastructures. Famines may also have long-term impacts on mental health and disease susceptibility.
Popular sentiment during famines often blames hoarding, and a common response of governments is to put controls on food prices and/or trade. In contrast, economists tend to argue that markets help to prevent and alleviate famines and should be left alone. Ó Gráda looks at how markets have worked in practice, finding different conclusions in different cases.
"the long-run gains from better spatial and inter-temporal arbitrage are clearly evident in ... the markets for grain in Pisa, Rome and England. ... The reduced variability in the series — in Pisa and England from about 1600, and Rome from about 1700 — implies reductions in the cost of transport and holding carryover stocks. These must have significantly reduced the vulnerability of the Italian and English poor in the early modern era."
"During the Orissa famine in 1867, the balance of trade responded too slowly and weakly to mitigate the damage done. At the time the pro-free trade Calcutta Review argued that in the event of a poor harvest, trade offered a cushion, since the affected region could import more and export less, and thus ensure itself against famine. Though fine in principle, in practice this mechanism worked too sluggishly. In an earlier era prohibitions on exports (and distillation too) would have been allowed, and probably would have provided some temporary respite against famine."
The longest chapter is devoted to the famine in Bengal in 1943, where Ó Gráda evaluates Amartya Sen's entitlement analysis.
"Sen and others have described the famine as the product mainly of bureaucratic bungling and accompanying market failure. I see it instead as largely due to the failure of the British authorities, for war-strategic reasons, to make good a genuine food deficit."
Action against famines by governments has taken different means: maintenance of granaries, attempts to purchase food, and public works schemes. Private action against famine has always had a place, but is now dominated by the work of global NGOs, where problems include exaggeration of claims, poor forward planning, and lack of coordination. And with state aid there is a risk that the dumping of food may harm local agriculture.
As examples of famines resulting from deliberate government action, Ó Gráda considers the Soviet Union, China in 1959-1960, and Ethiopia and North Korea.
And looking to the future, a final chapter considers agricultural trends, climate change and desertification, and some of the regions where famine risks are highest, with Niger in 2005 as a brief case study.
Ó Gráda is too aware of the problems with definitions, varying political contexts, and irregular and inconsistent data sources to try to shoe-horn famines into any kind of quantitative model. There isn't a single regression in Famine. There are some tables and charts — "Nominal agricultural wages and the price of rice in Bangladesh, 1972-76", for example, or "Year-to-year variation in the price of wheat, England 1260-1914" — but these are used to illustrate specific points.
While mostly based on secondary sources, Famine makes some use of quotations from newspapers, diaries and memoirs; it has a good mix of detail and overview. As an accessible multidisciplinary survey, it may be useful to specialists — historians, social scientists, development workers, economists — as well as lay readers.