Makandwe Nyirenda et al.
Employment, government cash transfers and household living arrangements in old age: Implications for intergenerational support in rural South Africa
Makandwe Nyirenda (Africa Centre/Southampton), Maria Evandrou (Southampton), Jane Falkingham (Southampton), Victoria Hosegood (Southampton/Africa Centre)
I enjoyed reading this paper because it is very much in line with similar analyses we are doing at Agincourt using AHDSS data. Like you, we are trying to disaggregate elderly status by “dependent” and “productive” using living arrangements as a lens. Unlike you, we are not using direct measures of income but rather pension eligibility. We find that large multi generations household with “productive” elderly members are increasing over time which I think is consistent with your results of “increasing downward resource flows.”
One question.. when you examine living arrangements and resource flows, how do you handle endogeneity issues..i.e. factors that may contribute both to hh composition and income generation patterning?
Nice to see that two DSS sites are examining similar issues..
Thank you for taking the time to read our paper. It is indeed gratifying to see similar work, albeit from a different perspective, being done at our sister DSS site.
You raise a very important issue in your question. One approach we adopted to try and account for this was to restrict the analysis to individuals who are co-resident with the older person in the surveillance area. But then the focus of our paper was not so much on income generation, rather on the implications of financial resource transfers within a household given who at a particular moment in the household holds an income.
I would be curious nonetheless to hear from you on how you may have dealt with it in your analysis.
I wonder if part of the increase in downward flows is a changing perception of children’s needs. That is…are children increasingly perceived as being more costly to raise? Such perceptions seem to accompany increasing emphases on formal education for example. Grandparents might be one convenient source of such support for alloparenting, especially in societies where states do not do so.
I would say rather than perception, increase in downward support is explained by increasing accessibility to social pensions by older people on one hand and increasing adult unemployment levels on the other. Unless the unemployed adults move away from the household, such structural changes would only entail the older person increasingly providing for the younger people.
Thank you to the discussant Jacques Emina for the useful comments to our paper. To start with this is work in progress, so we will definitely take into consideration your suggestions going forward.
Regarding you comment about those aged below 18 being in marital unions and/or having children, first with regard to being in marital union this should not have significantly affected our analysis. Marriage rates in our study population are very low with age at first marriage relative high, late twenties for women. There is though a fair bit of child bearing among teenagers, particularly girls. But then this should not change our results. But we will have a look at those children with a child support to see to what extent it is influenced by having children among the under 18s.
Regarding your concern on the length of the paper, combining some sections and the introduction, we shall look at that going forward.
Thank you once again for taking time to read the paper.