Determinants of household activity choice, rural income strategies and diversification - Institutional Constraints to Small Farmer Development in Southern Africa - page 67

the study area with the female heads of households most likely to be poor. Regarding gender
and marital status, Matsumoto et al. (2006) in their studies in Ethiopia, Kenya and Uganda,
found that larger proportions of poor are female households. Compared to women, male
individuals are more willing to explore opportunities, and are more mobile than female heads
of households, thus exhibiting greater propensity for migrating for wage employment.
Physical assets are the key requirements for most income generating activities in most rural
areas. Economic theory of investment supports the view that household endowment in
assets places them in a relatively better position to exploit income from various sources
(Parkin, 2008). Consistent with Holden et al. (2006), we found that physical assets
significantly affect household poverty status. Wealth-differentiated barriers to entry into
non-farm activities, as noted by Holden et al. (2004), have been noted in Burkina Faso,
Côte d'Ivoire, Ethiopia, Kenya, Rwanda, South Africa, and Tanzania.
The results with respect to social capital and physical capital are consistent with findings
elsewhere (Schwarze, 2004), and reflect the crucial role of a strong asset base in wealth
creation. The results with respect to labour contrast sharply with results obtained elsewhere
by Nargis and Hossain (2006) in Bangladesh, and Matsumoto et al. (2006) in Ethiopia,
Kenya and Uganda. This is probably due to the minimal use of family labour in income
generating activities in rural South Africa (Hebinck and Lent, 2007). Although household
size might be large in some instances, few household members contribute to household's
labour supply for economic purposes, suggesting an inherent underutilization of labour in
the system. In terms of model adequacy, despite a low R 2 , the model as a whole seems well
fitted on the basis of the significant F-statistic.
3.7 Conclusion and policy implications
The foregoing results have demonstrated that the gross local income in the survey
communities was derived from both farm and non-farm activities. Besides agriculture, wage
employment, own business, remittances, and pensions/grants were revealed as important
sources of rural income. The clear indication that agriculture is not the sole income source
is quite interesting and substantiates recent doubts about the sector's declining role in rural
livelihoods. It also emerged that the possession of education, social capital, physical capital,
and the availability of labour positively influences income. On the other hand, age and
gender of the head of household were shown to negatively influence income.
The paper therefore argues for a change of direction in the sectoral focus of efforts to
improve rural livelihoods and reduce poverty. Rural income growth and reduction of
poverty requires fundamental reforms that stress the importance of human, physical capital
and social capital for building a strong base for rural income growth and poverty reduction
in South Africa. It is clear, therefore that both farm and non-farm sectors should be equally