Download
1 / 39
390 likes | 605 Views
Paying the Piper: The High Cost of Funerals in South Africa. Anne Case, Princeton University Anu Garrib, Africa Centre UKZN Alicia Menendez, University of Chicago Analia Olgiati, Princeton University. Institutions that evolve over a long period of time often do so for many reason
E N D
Paying the Piper: The High Cost of Funerals in South Africa Anne Case, Princeton University Anu Garrib, Africa Centre UKZN Alicia Menendez, University of Chicago Analia Olgiati, Princeton University
Institutions that evolve over a long period of time often do so for many reason • Funerals, for example, • Pay respect to those who have died • Console the grieving • Mark the social status of the dead and his or her household • Knit social fabric within extended families and within communities • Redistribute some of the deceased’s resources
Large funerals may not put households at financial risk, when people die in old age, and so such funerals may be sustainable for a very long period • However, large funerals may have very different consequences when people begin to die in large numbers in prime age • AIDS crisis has knock-on effects brought about by the fact that institutions developed largely to bury people in old age are now being applied to a great number of prime-aged deaths
The financing of burials will affect a household’s ability to improve their members’ life chances • to maintain a stock of productive assets • to stake migrants in urban areas until they find work • to finance schooling, and more broadly • to provide adequate nutrition and a healthy environment for children
In this paper • Document the cost of funerals in South Africa • Explore how households make decisions on funeral spending • Discuss issues associated with reducing funeral spending, and possible ways forward
This paper documents funeral costs and financing for deaths that occurred between 2003 and 2005 in the Africa Centre DSA • Specifically, we analyze funeral arrangements following the deaths of 3,751 people who died between January 2003 and December 2005
What do funerals cost? • On average, households spend the equivalent of a year’s total expenditure on food and groceries, measured at median household expenditure in the DSA • Approximately one-quarter of all individuals had some form of insurance, which helped surviving household members defray some fraction of funeral expenses. • However, an equal fraction of households borrowed money to pay for the funeral.
How do households determine appropriate spending for funerals? • We build a model in which households respond to social pressure to bury their dead in a style consistent with the observed social status of the household and that of the deceased. • Households that cannot afford a funeral that meets social expectations must borrow money to pay for the funeral. • The model leads to empirical tests, and we find results consistent with our model of household decision-making.
Data • Funeral costs • A model of household decision-making • Structural estimates from our model, and additional tests using our data • Concluding thoughts
Africa Centre for Health and Population Studies, UKZN • Since 2000, approximately 11,000 households (~100,000 people) in the Umkhanyakude District in northern KwaZulu-Natal have been under demographic surveillance • The surveillance site includes both a township and a rural area administered by a tribal authority • At six month intervals, demographic and health information is collected on all household members • Individuals may be resident in the Demographic Surveillance Area (DSA), or may be non-resident members of households that claim them as members • Approximately two-thirds of all persons under demographic surveillance are resident in the DSA at any one time.
Data sets • Verbal autopsies • Illness and Death (IAD) Survey • Covering deaths January 2003-December 2005 • 3751 deaths • Household Socioeconomic Surveys (HSE1 2001) and (HSE2 2003/04)
Burial societies/funeral policies • 28 percent of the deceased had some form of funeral policy, or belonged to a burial society • For 20-30 Rand a month, these policies pay out upon death • Participation highly correlated with old-age pension receipt
Number of assets owned at HSE1 .3 .2 Density .1 0 0 5 10 14+ HSE1:number of assets
Total funeral spending and number of assets owned at HSE1 (2001) 8,000 6,000 mean of total funeral spending 4,000 2,000 0 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14+
Fraction borrowing money and number of assets owned at HSE1 (2001) .3 .2 Fraction borrowing for funeral .1 0 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14+ Number of assets owned at HSE1
Conditional on borrowing, fraction borrowing from a money lender .8 .6 .4 Fraction borrowing from money lender .2 0 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
Summary of preliminary findings • Funerals are expensive • And often leave households vulnerable • How are decisions made on funeral spending?
A model of household decision-making Let = characteristics marking an individual’s status and = community and extended family perception of household “income” (resources) at the time of the death The community and extended family form an opinion about the appropriate size of the funeral, F*, according to the deceased’s status and that of his household at the time of the death: . The funeral expenses we observe in our data are the desired spending plus an idiosyncratic error: .
A model of household decision-making Community and extended family do not observe household income (resources). Instead, they observe a vector of household and individual characteristics X2 that are correlated with income, which they use to form an expectation of household income. Households that experience an unobserved income shortfall will be less able to meet social expectations with respect to the size of the funeral, without borrowing money. The probability that the household will need to borrow (B=1) to finance a funeral of size F* can be written: .
This provides us with several checks, and a formal test, of our model. First, characteristics associated with lower individual status will have different predictions for spending and borrowing than do characteristics associated with lower household income. Characteristics of the deceased associated with lower individual status that is, with lower values of should reduce both the size of the funeral, as in (1), and the probability of borrowing, as in (4).
In contrast, any information available to the community that causes them to revise downward their estimate of household income, should reduce the size of the funeral, as in (1), but increase the probability of borrowing for the funeral. We examine these in turn.
Formal tests of the model The model suggests patterns of coefficients that should hold between the spending and borrowing equations. Writing We see
Table 8. Testing Predictions of the Model Table 8. Testing Predictions of the Model
Recent economic work has suggested, if the crisis results in lower population growth, that AIDS could “endow the economy with extra resources which … [will] raise the per capita welfare of future generations.” (Young, 2005). • This earlier research, however, assumes a constant savings rate over the life of the crisis, in order to focus on the effect of a potential fertility decline. • To the extent that productive resources are diverted into expensive funeral celebrations, earlier predictions that the pandemic will benefit future generations economically are less likely to come to pass.
Conclusions • Whether due to honoring an ancestor, or to the desire to please extended family or community, funerals in South Africa are elaborate and expensive • Cost appears to be dictated by the status of the deceased and observable household resources • Changes to this institution may be difficult without social coordination that everyone will spend less (e.g. Swaziland) • Given that this entails agreement not only between those living locally, but also extended family coming from afar for the funeral, this may be difficult
