Economic and demographic influences on household saving in urban Kenya

Abstract

This thesis investigates the impact of social, economic, and demographic influences on the level and composition of Kenyan household saving. While most economists would agree that the size and age structure of the household affect household saving and consumption behavior, few saving studies introduce household size and composition as separate independent variables. Instead, studies generally include saving in either per capita or per adult equivalent form. The demographic factors examined The empirical work used the 1968-1969 Kenyan household in this study include family size, family structure, age, education, and urbanization. Equations were estimated with various demographic specifications: the number of children included in dummy form, children differentiated by age and sex, and various equivalent-adult measures. / budget data which provides comprehensive expenditure and income information on 1,146 urban households. The saving models were fitted using "single-equation ordinary least squares estimation and, when necessary, appropriate adjustments were made for heteroskedasticity. There are five main empirical findings of the thesis. First, income was ~ far the most significant factor in deter- mining urban Kenyan household saving. Saving was an increasing but nonlinear function of income. Second, the Modigliani- Brumberg-Ando life-cycle hypothesis does not apply to Kenyan saving behavior. Third, contrary to most studies of the impact of children on saving and/or consumption which suggest that additional children have negative effects on saving irrespective of family size, it was found that saving is not differentially affected by additional children after the second child. Fourth, the extended-family institution has a significant depressing effect on household saving. Finally, none of the adult equivalency weights yielded an adequate approximation to the impact of family structure on household saving. Concomitant with economic growth, the literacy rate rises, the percent of the population living in urban areas increases, family size decreases, family structure changes from the extended family to the nuclear family, and average incomes increase. The direct effects of these changes on household saving for urban Kenya were explored. The study found that if increased monetized saving is considered an important goal and this is used as a single criterion for determining economic policy, then attention should be focused on increasing family income. Altering family size, improving literacy rates, and increasing urbanization will not directly alter the level of Kenyan household saving significantly. These factors, however, may indirectly have impacts on household saving, particularly through their impact on income. These indirect effects are not revealed in our single-equation model.