Rural income dynamics in post-apartheid South Africa: implications for reduction of poverty and income inequality - Institutional Constraints to Small Farmer Development in Southern Africa - page 75

The analyses employed here to measure the effect of differentiated sources of income on
inequality was suggested for similar studies by Adams (1999), Piesse et al. (1999) and
Leibbrandt et al. (2000). Decomposability of income inequality allows inequality to be
partitioned either over subpopulations or sources (Adams, 1999). In this technique total
inequality is divided into a weighted sum of inequality by various income sources (for
example, non-farm and agricultural income) and it encompasses source decomposition
of the Gini coefficient. The Gini coefficient is frequently used for analysis of distribution
of income because it can be decomposed by income source, illustrating the effects of
alternative income sources on total income equality. In their study, Van de Berg and Kumbi
(2006) used a similar approach to obtain estimates of the contribution of selected sources of
income on inequality in Oromia, Ethiopia. Their analysis follows the common expression
for the Gini coefficient (G) for the distribution of total income within the group. See the
Appendix at section 2 for the steps involved in carrying out estimations for the baseline-
related study presented in this volume.
4.3 Results and discussions
This section presents the results of the analysis with respect to rural income distribution
and the application of Gini decomposition technique in the limited context of one rural
and one peri-urban settlement. The policy implications of these results are then examined.
4.3.1 Socio-economic characteristics of survey households
The characteristics of the sample households were analyzed by means of descriptive statistics
(previously shown in Table 3.1). Characteristics included gender, age, marital status, main
source of income and education of household head interviewed, in addition to the size of
the household's field. From the evidence, the average household size was more or less the
same in both study communities although the Mlungisi households seemed marginally
larger. On average, household heads in the rural community were older than those in the
peri-urban location, reflecting the tendency for individuals to move back to their villages
as they grew older. There were correspondingly more widowed persons in the rural area,
where household heads were relatively older, less educated, and cultivated relatively larger
fields than those in the peri-urban location of Mlungisi (see Table 3.1).
4.3.2 Sources of rural income
Households in the two communities derive income from several activities, both farm and
non-farm activities (see Table 3.2). Most households pursued at least one of the following
activities: crop farming, livestock rearing, own business, wage employment, and augmented
income by means of pensions, grants and remittances. The average total income for the 12
months recorded for each community was computed. According to the data, households