An Application of Partial Least Squares to the Construction of the Social Institutions and Gender Index (SIGI) and the Corruption Perception Index (CPI)

This paper discusses households’ food insecurity among low income, poor urban households in and around the City of Tshwane, South Africa’s capital city. Using systematic random sampling with sampling interval of three, primary data were collected from 900 selected households, though only data from 827 households were analyzed following a rigorous coherence tests. The survey was conducted in Attridgeville, Soshanguve, and Tembisa. In the process, the study employed the use of two-way analyses of variance to explain differences between actual and expected household food security perceptions and those of severe, moderate and mild food insecurity. A favourable (adverse) variance could be interpreted to imply that means for achieving household food security are lower (higher) than predicted or that food security is higher (lower) than expected given the same level of main determinants. The observed variance is partitioned into components attributable to different sources of variation. ANOVA provides a statistical test of whether or not the means of several groups experiencing favourable (adverse) variances are equal. The main findings are that variances in the population means of households’ experiences of food insecurity vary by income class of the head of household, engagement in formal or informal income sources and by categories of social grants received. Poorer households that depend largely on cash income for food purchases experience highest food security variances and those receiving state pension. As such, timely receipt of household income under conditions of unimpeded access to social grants will improve urban poor households’ food security. The level of educational attainment has a very strong impact on a household’s food security. Those with “no schooling” have the lowest level of food security. Experiencing high variances in access to child grants, and low incomes, younger female household heads experience the highest degree of variances in food security and should be particularly targeted in an effective food security policy plan. Negative food security variance among these categories of South Africans could be devastating.