‘Cash transfer’ (CT) programmes are under discussion as an approach to reducing vulnerability to food insecurity in southern Africa. CT programmes have been run in some poorer countries with limited administrative capacity but targeted at easily identified groups, eg pilot programmes in Zambia and Malawi which make a small CT to an arbitrary 10% of the poorest households
(Schubert & Kambewa 2005; Schubert 2006).
The statistical information required to support the design and operation of a national CT programme will depend on the programme objective. A CT targeted at a narrow, easily-identified population group will require very little information (eg an estimate of the number of elderly). But a CT programme giving larger transfers with the objective of poverty alleviation would require a measure of poverty and additional information, for example on household characteristics, which could be used to identify beneficiaries. Without this it will be difficult to set a benefit level to achieve the programme objectives, or establish the number of beneficiaries or the programme cost. In most countries this information is not available from national sources.
The ‘household economy approach’ (HEA) is a simplified method of obtaining household budget data which has been widely and successfully used in south and east Africa for crisis prediction. It has the advantages of reasonable cost and accuracy and several national data sets already exist. Its chief limitation is that the simplification gives estimates only of average values by communitydefined ‘wealth groups’. As this does not give information on variation in income between households within a wealth group, or the characteristics of those households, HEA does not provide the information required for CT programme design.
The study arose from a request by the UK Department for International Development (DFID) for a practical large-area method which could be used to support CT programming. As a large amount of HEA data already exists, HEA was taken as the starting point. An extension of the HEA method was designed (named HEA+) which logically would, with a small increment to the usual HEA data set, allow a better estimate of variability within wealth groups. The purpose of the pilot study, which was conducted in Kazangula District in southern Zambia and supported by the Regional Hunger and Vulnerability Programme RHVP) and CARE Zambia, was to test this approach.