Open Access
Swaminathan, Hema
Graduate Program:
Agricultural Economics
Doctor of Philosophy
Document Type:
Date of Defense:
October 29, 2002
Committee Members:
  • Jill Leslie Findeis, Committee Chair
  • David Gerard Abler, Committee Member
  • Carolyn Elizabeth Sachs, Committee Member
  • David Shapiro, Committee Member
  • Malawi
  • Consumption-expenditures
  • Credit
  • Agricultural household model
  • Bargaining
  • Labor participation
  • Tobit
  • Endogeneity
  • Probit
In recent years, the focus in the study of household behavior has shifted from the household as a homogeneous unit (unitary model) to the different individuals comprising the household (collective model). There are several issues (such as preferences heterogeneity among household members, differential control of resources) that the unitary approach cannot address. The more general models of household decision-making explicitly recognize and model the individualistic elements in the household within a collective framework. Collective models make the assumption of Pareto efficiency in intrahousehold distribution but do not impose a particular solution concept. The concept of bargaining power plays a crucial role in the collective model. The model predicts that the household allocation process and the resultant outcomes will reflect the bargaining power of the individual. Individuals derive bargaining power from multiple sources, many of which correspond to the alternative options available to them in the event of withdrawing from the agreement. The research undertaken in this dissertation focuses on two aspects of household decision making, labor allocation and consumption expenditures, using the framework of the Nash-bargaining model and the agricultural household model. The measures of bargaining power used are access to credit and ownership of land. In this study a distinction is made between access to credit and participation in the credit market. The analyses in this research are based on the data set ‘Financial Markets and Household Food Security, 1995’, that was made available from the International Food Policy Research Institute (IFPRI), Washington D.C. The data are from a household survey of 404 households in 45 villages spread over five districts in rural Malawi. The estimation procedure undertaken in this dissertation follows a two-step approach to correct for choice-based sampling of the survey. In the first step, the probability choices regarding credit program membership status for the household are estimated using the weighted–exogenous–maximum likelihood multinomial logit model. The empirical models estimated in the second step include models to estimate the impact of access to credit on work choices and models to estimate the impact of access to credit and land ownership on consumption decisions within the household. The credit–labor allocation models are analyzed using the random-effects probit model, corrected for endogeneity of access to credit to the labor allocation decision. The effect of access to credit and land ownership on household expenditure shares are analyzed using ordinary least squares and Tobit models, corrected for endogeneity of total household expenditure. The credit–labor allocation models indicate that access to formal as well as informal credit increases participation in off-farm self-employment activities, and reduces participation in own farm work for women in male-headed households. Access to informal credit increases men’s participation in off-farm self-employment activities. Women’s access to formal credit also reduces men’s participation in own farm work. Education in rural Malawi appears to be a differentiating characteristic between those working off-farm and those allocating time to farm work. More education among men in Malawi encourages their participation in off-farm self-employment, whereas those men with less education are concentrated on farms. However, this relation does not appear to be true where women are concerned suggesting they are engaging in low-skill self-employment activities. Locational characteristics were found to be important in explaining participation in self-employment activity. The results from the expenditure models do not unequivocally support the hypothesis that men spend more on items of personal consumption while women are more oriented towards children and household welfare. While women in Malawi do seem to spend more on general household items, health and education, men in Malawi are also spending less on adult goods when they have access to formal credit. Female heads, on the other hand, increase the share of household expenditure on adult goods with access to formal credit. Expenditures on health increase in all the models except for female heads’ access to informal credit, suggesting the importance of health care. Child-related investments are negatively affected by female head’s access to informal credit and women’s (spouses’) share of land. Combined with the results from other equations in the models, it suggests that child-related investments are not yet a ‘priority’ area for the household.