Study title

Malawi, Swaziland, Mozambique

Household

We undertook a structured household survey to assess the local impact of biofuel feedstock production on livelihoods, ecosystem services, food security and poverty alleviation. This survey was undertaken in four feedstock production sites in Malawi, Mozambique and Swaziland. The selected project sites included:- a hybrid sugarcane project that contains a large sugarcane plantation surrounded by sugarcane smallholders (both rainfed and irrigated) in Dwangwa, Malawi; - a site that contains a large-scale sugarcane plantation and several smaller community-owned plantations in Tshaneni, Swaziland;- a smallholder-based jatropha project in Mangochi, Malawi;- a large jatropha plantation in Buzi, Mozambique.We targeted households with different involvement in feedstock production such as plantation workers and feedstock smallholders (intervention groups) and households not involved (control groups). The household survey included both close-ended and open-ended questions. It contained generic questions to elicit household demographics, livelihood options, agro-economic practices, reliance on ecosystem services and energy access/use patterns. The household survey also contained questions that could be used to calculate indices such as the multi-dimensional poverty index (MPI), a composite measure of poverty. We undertook purposeful sampling. In particular we sampled intervention groups that were involved in feedstock production (i.e. households with plantation workers or feedstock smallholders) and control groups that were not involved in feedstock production (i.e. households that did not contain plantation workers or feedstock smallholders). In each site we sampled different intervention and control groups to represent the unique characteristic of each area and feedstock production mode. Sampling procedures different by group and are explained more extensively in the uploaded material. In total we sampled 1544 households across all sites.In study sites that contain large plantations (i.e. Dwangwa-Malawi, Buzi-Mozambique, Tshaneni-Swaziland) we sampled two control groups, a control group in the vicinity of the plantation and a control group in areas further away that shared similar characteristics (see below). We employed this approach to establish whether living in the proximity of plantations had any positive or negative spillover effects, such as benefits from infrastructure (e.g. roads, clinics, secondary job creation) or disadvantages from the loss of agricultural land and natural areas. To identify the factors that influence the adoption and sustained use of ethanol stoves we undertook a structured household survey in Maputo city. The questionnaire contained mainly perception questions, but also some key demographic and socioeconomic variables.The vast majority of the questions in this household survey were closed-ended questions that were coded appropriately. We focused on areas of Maputo city that experienced substantial promotion of ethanol stoves, and targeted both households that used ethanol and charcoal as the main cooking option.In total 341 households were surveyed from different neighborhoods. Of these, 58 households (17%) had adopted and sustained ethanol use, 42 households (12%) had adopted but discontinued ethanol use and 241 households (71%) never adopted ethanol stoves. The specific study areas within Maputo were selected as they had experienced an extensive adoption of ethanol stoves, as identified through expert interviews with management of NDziLO Company that spearheaded the adoption campaigns. Additionally these areas were also chosen so as to represent households of the predominant socio-economic background in Maputo city. Through this purposeful sampling we tried to capture the factors influencing the adoption and sustained use of ethanol stoves when considering the other prevailing cooking options, rather than provide a baseline study of cooking fuel preferences across the city. Adopters in these areas were randomly sampled through sales records of NDziLO and non-adopters were selected as the neighbours of adopters (see below).