First, involvement depends on actual opportunities for supplementary activities. Local opportunities determine, as exogenous factors, individual involvement in supplementary activities. A descriptive analysis of the nature and extent of supplementary activities should give insight into the problem of excess capacity for many supplementary activities. Opportunities for supplementary activities depend on present comparative advantage and on the dynamics of technological and institutional development. Second, individual and household characteristics affect participation in supplementary activities. Some additional assumptions are required to interpret these determinants as exogenous factors. In contrast to farming, which is essentially a household endeavour, supplementary activities are mainly individual undertakings. Also when people work together in supplementary activities, it is common that costs are incurred by and benefits accrue to individuals. Individuals can generate personal income by supplementary activities without
giving up household farm income. Conversely, I do not account for second-level effects between individual supplementary income, household expenditures, and farm husbandry. Third, the individual nature of supplementary activities calls for a closer look at individual diversification behaviour. While the second research question concerns determinants of individual involvement in supplementary activities, the third research question relates to individual diversification behaviour. Because diversification is commonly associated with risk behaviour, an assessment of attitude to risk has been made. Multiple measures of risk aversion provide insight into how personal differences shape individual livelihoods. Fourth, household supplementary income may alter crop productivity, production efficiency and input demand. It is the aggregate household supplementary income rather than individual income that brings about this effect. Since aggregate household supplementary income is the result of a random composition of the household, it can be used as an independent variable to explain crop husbandry characteristics. Fifth, aggregated household supplementary income is closely related to farm income due to labour and capital substitution effects. Therefore income decomposition relies on descriptive
analyses and does not seek to identify causality relations between farm and supplementary income. The focus is on coherence between farm and supplementary income components, income distribution and poverty incidence.
The following sections provide a description of the research setting. It shows that livelihood is a matter of a household as well as an individual enterprise and it assesses technical and institutional features of prevalent economic activities.