Livestock as a buffer against HIV and AIDS income shocks in the rural households of Zimbabwe

Abstract

The objective of this study was to determine the extent to which households use livestock sales to mitigate the impact of HIV/AIDS income shocks. Using survey data from the Muzarabani and Bindura districts of Mashonaland Central Province in Zimbabwe, the econometric analysis indicated that livestock, particularly poultry and small stock, play a significant role in smoothing income fluctuations due to HIV/AIDS. Cattle sales compensate for 71 per cent of income shortfalls due to HIV/AIDS. About 90 per cent of HIV/AIDS-afflicted households, headed mainly by women or children, used poultry and goats as consumption-smoothing strategies when faced with negative income shocks. Government and other stakeholders were advised to support the use of small stock as a strategy for coping with HIV/AIDS economic shock in female- or child-headed households in rural areas.

1. Introduction

There is strong evidence to indicate that HIV/AIDS is having a negative impact on livestock production in sub-Saharan Africa and consequently affecting the livelihoods of the rural poor directly (World Bank, 2002; Food and Agriculture Organisation, 2003; D R Gwatkin & G Deveshwar-Bahl, unpublished draft). As Mogues observed (2006:8)

shocks are often a particularly pernicious phenomenon for the poorest in poor countries. With few good options for shielding themselves from the impact of droughts, death of an economically active family member, or detrimental price changes, in some cases the poorest are said to resort to divesting themselves of their meager assets to stabilize current consumption.

There is widespread acknowledgement in the literature that populations in agricultural and agro-pastoral areas in developing countries draw on their livestock assets to overcome food crises arising from climate and other shocks (Dahl, 1976; Eicher & Baker, 1982; Fafchamps et al., 1998). Of importance in southern Africa is the role of livestock and their products as a principal means by which households mitigate both transitory and non-time varying shocks where informal and formal insurance are incomplete. It is well known in southern Africa that most rural households prefer to sell livestock and purchase other commodities to meet household needs (Sandford, 1983; Swift, 1986).

The role of livestock in providing food directly or as a source of income through sales, which can in turn be used by HIV/AIDS-affected and HIV/AIDS-infected households to buy food and medicinal drugs, cannot be overlooked (Björkman, 2002; International Center for Research on Women, 2002). HIV/AIDS is associated with acute micronutrient shortage, which further compromises the immune system and requires increased dietary supplementation. Animal source foods provide a particular dense combination of critical nutrients (United Nations Children's Fund & United Nations Programme on HIV/AIDS, 2002; Food and Agriculture Organisation, 2003). Importantly, livestock can provide a basis for affected households to rebuild their assets and resource base. Thus, HIV/AIDS-afflicted households have developed coping strategies through keeping livestock. However, the effectiveness of the strategies has not been documented.

Small stock can be an alternative to raising cattle and they provide affected households with a means of improved nutrition and generating income. Chickens, guinea fowl, turkeys, ducks, rabbits, sheep, goats and, where culturally acceptable, pigs offer a wide variety of possibilities for production systems and resource bases. These species are easily managed, affordable and have short reproductive cycles. Moreover, most women in rural areas of sub-Saharan Africa rear poultry and other small stock. According to the International Fund for Agricultural Development (2005), small stock that tend to live closer to the homestead are especially important for women nursing HIV/AIDS patients. They are also appropriate for elderly- and orphan-headed households. Establishing or restocking affected households with these species can provide a significant amount of income from sales and quality animal protein and by-products.

A number of researchers (United Nations Children's Fund & United Nations Programme on HIV/AIDS, 2002; Food and Agriculture Organisation, 2003) have addressed the importance of nutrition in helping individuals in sub-Saharan Africa affected by HIV/AIDS maintain their health. Yet the contribution that livestock income can make to the livelihoods of the poor and its complementary role to other foodstuffs in the household diets have not been fully discussed. The objective of this study was thus to determine the relationship between livestock transactions and HIV/AIDS income fluctuations. The main question to be addressed is do net sales of livestock increase when a household is subjected to HIV/AIDS income shocks?

2. Materials and Methods

2.1 Sampling procedure

The survey that generated the data set used in this study was conducted in the upper Muzarabani district and the Musana communal area of the Bindura district in Zimbabwe. The study was conducted over a period of 52 weeks during the 2004/2005 cropping year. The upper Muzarabani district is composed of 12 wards, of which five are participating in income-generating projects initiated by Hope Humana People to People (a non-governmental organisation). The Musana communal area is divided into nine wards, of which four are participating in the organisation's HIV/AIDS mitigation programmes. A ward is a small administrative unit with approximately 700–1000 households. The sample, representing 30 per cent of the total number of households in the participating wards, was drawn using a stratified sampling method in an attempt to capture a wide range of socio-economic and geographical characteristics. Fifty per cent of the sample was from the HIV/AIDS-afflicted households and the other 50 per cent from the non-afflicted households but in the same wards and with similar soil and climate conditions.

To participate in this research, households were asked to provide sufficient evidence that HIV/AIDS affected them. Individuals who were HIV positive or looking after infected persons in their households were asked to produce HIV test documents from reputable testing centres to confirm their status. Those referred to this organisation by the government Social Welfare Department were mainly foster parents and orphans or people who had been identified by the organisation's project development officers. Other complementary methods, such as focus group discussions with key informants, collecting secondary data from other stakeholders and participatory observations were also used to gather information.

2.2 Data collection

The data were collected on four occasions, during October, January, April and July of the 2004/2005 cropping year. The household characteristics included information on the number of HIV/AIDS patients in the household, the type of household, the main sources of income, shocks the household had experienced (i.e. any income loss due to HIV/AIDS) and whether the household had experienced the death of an economically active family member from HIV/AIDS in the past two years at the time of household listing or had a chronically ill household member. The household survey also collected information on crop production, livestock production and transactions and physical and financial assets.

2.3 Data analysis

To achieve a more precise measure of the effect of HIV/AIDS income shock on livestock transactions the following equation was estimated:

where S _it is the net income from livestock and their product sales by household i and time t of the year and X _it is a vector of household characteristics such as demographic structure and initial asset holdings, which might affect the expected income and precautionary savings and, thus, livestock transaction. Z _it represents the HIV/AIDS income shock. ŷ _it is the mitigation effect of livestock transactions (Shock_it × Livestock transaction) and, thus, β₃ is the effect of interest, which captures the differential impact of the HIV/AIDS income shock among households with different levels of livestock assets. V _it is the livestock assets available to the household. ϵ_it is the random error term. It is anticipated that an HIV/AIDS income shock leads to increased livestock transactions because in most rural areas informal and formal insurance are incomplete.

Livestock assets could be correlated with omitted variables that are associated with livestock transactions such as the household's food production capacity and the extent of its social capital. These factors necessitate the inclusion of household fixed effects in the above model as follows:

where w´ stands for the household fixed effects. Including household fixed effects ensures that the effect of idiosyncratic risks and non-time-varying household risk factors have been factored in. This ensures that the model is comparable to Fitzsimmons's (2003) framework: the household fixed effects fully absorb their measure of idiosyncratic and aggregate risk.

3. Results and Discussion

The results (Table 1) indicate that non-affected households have more cattle than HIV/AIDS-affected households (p < 0.05). The main reason for the difference cited during the survey was the high disposal rate by the HIV/AIDS-afflicted households to meet increased medical costs and general household requirements. The findings also showed a significant difference (p < 0.05) in the mean number of poultry and goats between the HIV/AIDS-afflicted households and the non-afflicted ones. The former had more poultry and goats than the latter. This was mainly attributed to the lower labour intensiveness of small stock and their high reproductive rates, which fitted well into the daily activity profile of many HIV/AIDS-afflicted households. Most of these (90 per cent) were female- and child-headed households, who had less time for and access to income-generating activities. These results are consistent with research findings that recommend poultry and small stock for HIV/AIDS-afflicted households because of their high reproductive rates and lower labour intensiveness (World Bank, 1999; Food and Agriculture Organisation, 2003; United Nations Programme on HIV/AIDS, 2004; International Fund for Agricultural Development, 2005).

Table 1: Livestock characteristics by households

Inventories	Cattle		Goats		Poultry	Average sales per year
	n	mean	n	mean
	HIV/AIDS-affected households in Muzarabani	9	31	70			3	10	32
HIV/AIDS-affected households in Bindura	5	11	52	1	3	27
Non-affected households in Muzarabani	13	26	35	1	3	5
Non-affected households in Bindura	6	8	20	0	1	3

During the survey, sample households were asked to report their motivations for selling livestock. Using the HIV/AIDS income shock estimates, the answers were tabulated according to the severity of the shock the household faced (Table 2). The results provided some evidence to support the buffer stock hypothesis, i.e. that households draw on their assets in times of need (Deaton, 1992). The findings of this study indicated that the main motives for selling cattle for those who face marked negative income shocks were to meet medical and funeral expenses. For those who faced less severe shocks, the main motive was to meet crop production operating costs. Most households (80 per cent) cited consumption (food expenses) as the main motive for selling goats and poultry. The results indicated that small stock, which mainly belong to women, are principally used as consumption-smoothing strategies when households are faced with income shocks. These findings are similar to those of other researchers (Jarvis, 1974; Watts, 1983; McIntire et al., 1992; Reardon et al., 1992), which indicated that most rural households liquidate their livestock assets when exposed to income shock.

Table 2: HIV/AIDS-affected households' motives (%) for selling livestock

	Cattle		Goats		Poultry
Motive/Income shock	Severe	Mild	Severe	Mild	Severe	Mild
Consumption (food expenses)	10	4	80	67	88	82
Medical expenses	13	32	9	11	2	4
Funeral expenses	67	43	4	8	1	0
Crop production expenses	6	15	2	10	5	6
Livestock care	1	2	3	2	1	5
Gifts	0	0	1	0	3	3
Other^a	2	4	1	2	0	0
^aThese include expenses such as lobola (bride price) and traditional ceremonies.

Table 3 shows estimates of regressing net livestock sales on HIV/AIDS income shocks. The results showed that HIV/AIDS had a positive significant effect on household income shocks (p < 0.05). Households affected by HIV/AIDs were 78, 83 and 91 per cent likely to sell their cattle, goats and poultry, respectively, to offset the shock. These results provide evidence to support the literature that states that most rural households use livestock as the asset most easily liquidated to mitigate shocks (Dercon, 1998; Kinsey et al., 1998; Barnett & Grellier, 2003; Beegle et al., 2003). The interaction effect of livestock assets and HIV/AIDS income shock was negative (p < 0.05). The mean values of cattle, goats and poultry offset 71, 53 and 49 per cent of the shock, respectively. The ability of livestock to offset shocks substantially is consistent with the literature in Africa (Jarvis, 1974; McIntire et al., 1992; Kinsey et al., 1998; Food and Agriculture Organisation, 2003). This is also in line with the International Fund for Agricultural Development's (2005) policy implication on mitigatory measures, which suggests using small stock to mitigate the effects of HIV/AIDS on household income because of its high proliferation rate.

Table 3: OLS estimates of the determinants of net livestock sales for HIV/AIDS-affected households

	Cattle		Goats		Poultry
Explanatory variables	PE	t -ratio	PE	t -ratio	PE	t -ratio
Income shock (any income loss due to HIV/AIDS)	0.78**	1.39	0.83	1.34	0.91	1.56
Shock × per capita livestock value	− 0.71**	− 0.85	0.53	− 1.16	0.49	− 1.10
Shock × log per capita physical assets	− 0.89**	− 2.15	− 0.69**	− 1.37	− 0.59**	− 2.41
Shock × per capita cash holdings	− 0. 25*	1.59	− 0.15*	− 1.46	− 0.13*	− 3.37
Shock × crop income	− 0.37**	1.37	− 0.31**	− 2.67	0.24**	− 0.95
Number of chronically ill adults	1.24**	2.41	3.21**	2.36	10.4**	3.07
Mother dead	0.63*	1.67	0.10*	1.31	0.17*	1.25
Father dead	0.18	1.96	0.88**	2.48	0. 91**	3.85
Age of household head	− 0.75	2.34	0.54	1.67	0.78	1.11
Household size	0.77*	0.45	0.68*	1.28	0.81*	2.51
R ^2	0.62	−	0.57	−	0.63	−
Observations	200	−	200	−	200	−
Regressions are at household level: **, significant at 5%; *, significant at 10%. Other regressors included marital status and education level of the head of household and community variables such as access to health services.
PE, parameter estimate.

The results also showed that most rural households have less liquid cash at hand to mitigate these effects. Dercon (2004) cited low time preference for money as the main reason for low cash saving. This is also consistent with the evidence from the survey, where most households were worried about their present consumption requirements and took little heed of future risks, their marginal propensity to save being negative due to harsh macroeconomic conditions. The other reason for low cash savings cited during the survey was inadequate and unfavourable services for saving their cash. The result of the regression model indicated that crop income had a significantly negative correlation with livestock sales (p < 0.05), but did not significantly offset the income shock due to HIV/AIDS. This may be due to the crop–livestock interactions, i.e. crops and livestock competing for the same household resources.

It was found that the death of the father or both parents had a significant positive correlation with the sale of poultry and small stock. Death of the father increased the net sales of goats and poultry by 88 and 91 per cent, respectively. The results indicate that female- and child-headed households depend more on small stock.

4. Conclusion

In this study the econometric analysis indicated that households increase their net livestock sales when faced with income shocks due to HIV/AIDS. The main motives for livestock sales were to meet increasing food and medical costs and funeral expenses. HIV/AIDS-affected households used their livestock assets as a buffer against economic shocks that lower their living standards in times of crisis. The study indicated that the poorer households, particularly the female- and child-headed ones, relied more heavily on poultry and goats. This has important implications for policy makers and organisations involved in designing mitigatory measures against the economic effects of HIV/AIDS. These stakeholders are, therefore, urged to promote projects such as poultry and goat production. These enterprises, given their high turnover and low labour requirements, offer a potential strategy for mitigating the economic effect of HIV/AIDS in rural areas in developing countries.

The authors are grateful to Hope Humana People to People, Zimbabwe, for funding this research project.