This study evaluates a transfer'programme'in Sierra Leone to determine the impact of unconditional transfers'and aid-for-work'programmes'on livelihoods and forest conservation.
Due to hunting and logging rights imposed in the Gola Rainforest National Park, located in south-eastern Sierra Leone, communities in the vicinity of the park benefit from the Community Development Fund (CDF). The CDF consists of a one-off grant proportional to the village size. The CDF resembles a windfall to communities who receive them in that they are unconditional transfers. These CDF transfers aim to promote sustainable land management in forest areas beyond the reserve boundaries, where the legal restrictions concerning land use do not apply.
This impact evaluation examines the short-term impacts of a payment scheme on livelihoods and conservation near the Gola Rainforest National Park (GRNP) in eastern Sierra Leone.
The impact evaluation will assess the impact of the programme on conservation outcomes, particularly reduction in deforestation caused by land conversion for agriculture, logging and mining activities.
The objective of the study was to test the impact of receiving aid, distributed by the CDF programme through three treatment arms.
Treatment 1: uses a 'business as usual' scenario where aid was distributed to communities via traditional governing structures, the local chief.
Treatment 2: aid was handed out to each household.
Treatment 3: an aid-for-work programme in which individual households receive aid conditional on participating in a public road improvement programme.
Finally there was a control group of randomly selected villages who did not receive any aid.'
Within the programme, no cash was distributed. Instead communities received vouchers with which goods could be ordered from a pre-specified list of over 40 consumption, investment and public goods. The goods menu reflected local market prices. These lists contained not only prices and item descriptions, but also pictures so that those who are illiterate can comprehend the items easily.
Theory of change
The causal pathway in the project is as follows: (i) communities received aid vouchers from the implementing agency, (ii) decided whether to allocate aid for private or community use, (iii) and if for private use, they decided on whether to consume or invest. If invested they decided whether to invest in off-farm employment, or agricultural intensification, or clear more land for agriculture. When the aid is given through the group to the chief, there may be less individual benefits and more public goods. This reduces the income effect compared to the individual treatments. When individual aid is made conditional on work, this reduces the reciprocity motive relative to the individual unconditional aid.
The villages were chosen for their close proximity to the border of the national park. Treatments were allocated randomly, with stratification at the chiefdom level: 24 villages received aid via the chief, Households in 21 villages received aid directly, 24 villages enrolled in the aid-for-work treatment, and 22 villages served as a control group. Within each village up to 30 heads of households were interviewed. Baseline sample consisted of a total sample of 2,379 households and the endline sample of 2,251 households.
Outcomes were measured across a range of social, economic and land use indicators. The key intermediate outcome of interest is the use of the aid vouchers.
The results show that the way in which aid is distributed'communal versus individual and windfall versus earned'has a significant effect on how the aid will be used. Earned aid given directly to the individual leads to more consumption with little attention given to public goods. Windfall aid given to community leaders leads to more public goods that are better managed. There is no significant impact on economic, social and conservation outcomes. This may in part be explained by the high sense of community we find in these villages'over 60 per cent of aid allocated to community projects. Much of the aid being spent on a community project does significantly dampen the potential individual-level impact. In addition, there is also the possibility that the per capita amount of aid (USD 15) may have been too low to impact households significantly.