2000 THE CHINESE FAMINE OF 1959-61 14l price of grain to other commodities had limited power in explaining starvation in China because grain was rationed in urban areas and the grain market in rural areas during a famine was too thin to be meaningful. Third, because interprovincial grain trade by private traders was prohibited and the central verne nt did not have the capacity to deliver relief to rural areas during the period of a sharp decline in grain output nationwide, the subsistence of farmers in a province during that period depended on the food production in that province. These specific institutional arrangements provide a convenient way to tes he validity of FAD or entitlement as an appropriate approach for under standing famines. The per capita grain output in a province can be used as the proxy for food availability in that province. If FAD is the appropriate approach food availability would be the only variable responsible for the famine deaths In the Chinese setting, a persons entitlement to food was legally determined by his/her residence status. The proportion of rural population in a province represented the proportion of population in that province who did not h legally protected rights to food. It is a proxy for the degree of urban bias in that province. If Sens entitlement approach is the right framework to under- stand the causes of a famine, the urban bias might also be a significant variable in determining famine deaths in addition to food availability To be specific, the hypothesis we seek to test in the regression analysis can be summarised as follows In a famine in China, the death rate in a province is positively related to the proportion of rural population in that province (urban bias hypo- thesis) and negatively related to per capita grain output in that province (food availability hypothesis) We will use the serious famine that occurred in 1959-61 in china as our ase study. In the next section, we document grain production declines, issues of procurement, and the resulting famine. In Section 3, we conduct a statistical assessment of the relative effects of food availability and urban bias on the observed deaths from this human disaster 2. Collectivisation and Demographic Crisis: 1959-61 Chinas agricultural collectivisation started in 1952. The farming institution was changed from household farms to mutual aid teams, to elementary operatives and then to advanced cooperatives that consisted of about 150 households. The advanced cooperative was the major farming institution by 1957. Agricultural output increased continuously in the period 1952-7 with an average annual growth rate of 4.6%. Encouraged by this success, the Commu- nist party decided to adopt a bolder approach to mobilise surplus labour to re a household's property and exchange c Royal Economic Society 2000price of grain to other commodities had limited power in explaining starvation in China because grain was rationed in urban areas and the grain market in rural areas during a famine was too thin to be meaningful. Third, because interprovincial grain trade by private traders was prohibited and the central government did not have the capacity to deliver relief to rural areas during the period of a sharp decline in grain output nationwide, the subsistence of farmers in a province during that period depended on the food production in that province. These speci®c institutional arrangements provide a convenient way to test the validity of FAD or entitlement as an appropriate approach for under￾standing famines. The per capita grain output in a province can be used as the proxy for food availability in that province. If FAD is the appropriate approach, food availability would be the only variable responsible for the famine deaths. In the Chinese setting, a person's entitlement to food was legally determined by his/her residence status.15 The proportion of rural population in a province represented the proportion of population in that province who did not have legally protected rights to food. It is a proxy for the degree of urban bias in that province. If Sen's entitlement approach is the right framework to under￾stand the causes of a famine, the urban bias might also be a signi®cant variable in determining famine deaths in addition to food availability. To be speci®c, the hypothesis we seek to test in the regression analysis can be summarised as follows: In a famine in China, the death rate in a province is positively related to the proportion of rural population in that province (urban bias hypo￾thesis) and negatively related to per capita grain output in that province (food availability hypothesis). We will use the serious famine that occurred in 1959±61 in China as our case study. In the next section, we document grain production declines, issues of procurement, and the resulting famine. In Section 3, we conduct a statistical assessment of the relative effects of food availability and urban bias on the observed deaths from this human disaster. 2. Collectivisation and Demographic Crisis: 1959±61 China's agricultural collectivisation started in 1952. The farming institution was changed from household farms to mutual aid teams, to elementary cooperatives and then to advanced cooperatives that consisted of about 150 households. The advanced cooperative was the major farming institution by 1957. Agricultural output increased continuously in the period 1952±7 with an average annual growth rate of 4.6%. Encouraged by this success, the Commu￾nist party decided to adopt a bolder approach to mobilise surplus labour to 15 We do not need to worry about how to measure a household's property and exchange entitle￾ments as these two variables were not important determinants of the rights to food in the speci®c socialist setting in China. 2000] 141 THE CHINESE FAMINE OF 1959±61 # Royal Economic Society 2000