3ie Funded Evaluation, TW4.2.02. A link to the completed study will appear here when available.
The study seeks to evaluate the impact of access to newly constructed or rehabilitated irrigation system in western Madagascar.
Agriculture is the most common livelihood for people of Madagascar, employing 75.3 percent of the rural population and accounting for 28.2 percent of the country’s GDP. The Projet d’Appui au Développement du Menabe et du Melaky (AD2M) implemented irrigation, land titling, and related rural development programmes in 19 communes (a grouping of villages) in western Madagascar.
The AD2M project sought to improve the wellbeing of marginalised farmers facing individual and environmental constraints by implementing a multi-faceted programme that addressed land tenure security, agricultural production techniques and sustainable land and water management practices including expanded access to irrigation, market infrastructure and transportation.
The questions that this impact evaluation will aim to address are listed below.
• Does access to certificated, irrigated land increase rice yields, the total value of irrigated crop production, net revenue from crops and value of crops marketed?
• Does access to certificated, irrigated land enable farmers to increase the number of cropping seasons within the year and lead to crop diversification?
• Did project activities lead to an increase in the adoption of sustainable land management practices at both the community and household-level?
• Given the importance of well-functioning Water Users Associations (WUA) in maintaining irrigation infrastructure and regulating water use, do measures of WUA performance affect the extent to which farmers could realise benefits from their irrigated plots?
The intervention comprises of two components to address land tenure security and increased and sustained income from agriculture. The first component included activities directly related to local land governance and tenure security. The second component focused on promotion of technologies and practices to increase agricultural production, including increasing access to irrigation and finance, introduction of new seed varieties, promotion of cash crops, increasing access to agricultural equipment and increasing access to markets through road construction waterway infrastructure, and microfinance kiosks.
Theory of change
The project targeted five groups of farmers. These include the landless agricultural wage earners with no productive assets and no cattle, households with small, non-irrigated landholdings (less than two hectares) and no cattle, households primarily dependent on fishing, households with small irrigated landholdings (less than two hectares) and a small numbers of cattle (less than 10) and finally households primarily dependent on cattle-rearing. AD2M expected the programme to benefit almost 40 percent of the rural population in Menabe and Melaky regions.
The key outcomes of interest to IFAD and the Ministry of Agriculture of Madagascar are increased access to an efficient and effective decentralised land administration system, and higher and more stable crop incomes through increased access to irrigated land and intensification of irrigated crop production. The key impacts are sustainable increases in higher and more stable annual incomes from all sources of beneficiary households.
The researchers plan to test the impact for more feasibly measured outputs, such as crop yield and diversification, which the theory of change predicts will contribute to the ultimate outcomes of interest.
The evaluation will follow a quasi-experimental approach using propensity score matching. A sample of beneficiary households from 14 out of the 16 irrigated communes of Menabe region will be randomly selected.
The evaluation will replicate the targeting process that was used back in 2007 to identify beneficiaries using the same or similar historic data sources. This would be a combination of observation-based and criteria-based targeting as proposed by Ouma et al. (2007). Observation-based targeting will involve determining where the AD2M programme was adopted, plotting those sites on a map and identifying the common characteristics that the sites share. This information will then be used to determine a group of potential non-programme areas that had characteristics similar to AD2M areas before the programme started.
A survey of 666 treatment and 1,334 comparison rural farm households will be conducted to collect key information and map out the causal chain across inputs, activities, outputs, outcomes, and impacts as well as the underlying assumptions.