Household Decision Making Under Stress: Three Essays on Agricultural Production in Southern Africa

Anna Leigh Josephson, Purdue University


Rural households in Southern Africa face numerous economic challenges, arising from a range of idiosyncratic and covariate risks. This research seeks to understand strategies employed by households to cope with different types of shocks, in the context of Zimbabwe and Malawi. I focus on several techniques used to cope with risk and shocks in an uncertain environment, including intra-household resource, land, and crop, as well as labor allocations. To provide insight on household decision making, I expand and adapt theoretical models. This helps me to better understand household decision making in the complex and risky environments in which they reside and to more broadly grasp the dynamics of small households in the developing world. This research is comprised of four chapters, including an introduction and three distinct essays. I consider three topics relevant to agricultural households: (1) intra-household dynamics, (2) crop choice decisions, and (3) labor allocations. I consider these three topics, and the way in which households treat them in their increasingly risky, complex, and changing environments. In the first essay, I extend the typical considerations of intra-household dynamics by examining joint income: income that is earned and shared by men and women working together. Using data from 693 households in Malawi, I empirically test the assumption that all household income is pooled, when joint income, as well as income earned individually by men and women, is accounted for in the analysis. I develop an intra-household collective model which explicitly includes joint and individual relationships to explain income allocation. To evaluate changes in income and expenditure, I use rainfall variation, as different levels of rainfall will have different impacts on the production by different households members. Ultimately, I reject the hypothesis of complete income pooling and full insurance within the household. However, I find evidence that household members pool income and insure one another for expenditures on essential goods and thus I conclude that there is strategic income pooling behavior with respect to particular types of expenditure, resulting in partial insurance for the household. In the second essay, I investigate the decision to grow and allocate area to sorghum, a long-promoted, drought resistant crop in Southern Africa. Sorghum, though tolerant to extreme weather conditions, is not considered to be delicious or worth eating to many individuals in Zimbabwe. Using a panel of 355 households located across Southern Zimbabwe, I employ a behavioral model, which explicitly accounts for households' own preferences and characteristics, as well as the perceived and inherent traits of crops. In a context in which households are largely producers and consumers of their own food, I find that own household preferences dominate the decision to allocate area to sorghum. The inherent drought resistant traits of sorghum are relevant when households are faced with variable and low average rainfall. Additionally, participation in off-farm labor drives planting behavior, serving as an alternative strategy for households facing drought and the complexities of cultivating crops in a more arid environment; households which have members working off-farm are less likely to plant sorghum, suggesting that they purchase food with income earned off-farm, should their maize crop be insufficient. These results indicate that preferences must be considered when examining households' planting behavior. Finally, in the third essay, I investigate ex post labor reallocation as a household strategy for coping with unanticipated events. The study spans a period of extreme unrest and stress: I measure the responses of 355 households to death and rainfall shocks in Zimbabwe, during the country's long-term economic crisis with data from 2005 and 2013. In this strained environment, shocks compound existing stresses. I find that, as relative returns to agriculture fall, households are pushed out of the farm sector into off-farm work and shocks tend to exacerbate household responses to these pushes. There is also strong heterogeneity in responses, in part related to household characteristics. Wealthier and larger households are more likely to return to on-farm labor following a shock than their poorer and smaller cohorts, who instead diversify (or are forced to diversify) the household labor portfolio through migration and reliance on off-farm earnings. Taken together, these results suggest that household strategies differ depending on household characteristics and individual traits. Central within all of these essays is the examination of questions about how households cope with unanticipated events, experienced risk, and unexpected shocks. When their conditions change and new situations arise, households must find a way to adapt, within the constraints imposed by the circumstances of their context. My ultimate ambition with this research is to better understand these contexts and constraints, and thus the microeconomic challenges facing rural households in developing nations in Southern Africa.




Shively, Purdue University.

Subject Area

Agricultural economics

Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server