Group Interventions for Agricultural Transformation in Mozambique

1. Financial Literacy and Incentives for Cooperation

The first treatment arm will isolate the impact of a package consisting of capacity building and social encouragement on both short-term contributions towards a common pool resources and long-term operation and maintenance outcomes. For this evaluation, we will use the financial literacy training and matching grant component of the irrigation project, which also requires collective action.

2. Provision of water monitoring information

Two information products will be provided. First, information on plot-specific actual water use and the associated productivity outcomes (in terms of crop per drop) will be discussed with the farmers on a regular basis after the agricultural surveys have established plot-level crop outcomes. Second, for each of the pilot schemes, both sources of individual farmer information will be aggregated and the sub-scheme level and mapped within the scheme. These maps will not only show farmers how much water they use relative to other members of the community leading to an adjustment in their own water use, but also convey information with respect to how their own practices affect the distribution of water to other parts of the scheme.

3. Irrigation infrastructure

The project will support the rehabilitation and development of at least 3,000 ha of irrigated farmland in three central regions of the country.

1. Financial Literacy and Incentives for Cooperation

The financial training will consist of two components: 1) financial literacy which will cover topics such as money management, credit management, financial literacy/management skills, input/output production accounting, and 2) developing a savings plan for a group investment. The group investment will be supported through project’s matching grant window that supports acquisition of equipment (e.g. farm machinery, greenhouse) and value chain development (e.g. product processing, storage, marketing, transportation). The group will be guided on the selection of a reasonable farm investment (such as a cultivator) that the group can achieve through the matching grant component. The participants jointly develop a savings plan and accounting system to guide progress toward saving for the goal. The plan should specify individual contributions and a timeline, taking into account periods farmers will have more cash at hand. Two appointed leaders (i.e. treasurer, lead farmer, producer organization head, or someone chosen by the community) will be trained in financial planning and bookkeeping to keep track of the savings contributions.

In addition to the one-off training, half of the communities will be re-visited an additional 3 times throughout the year to refine and review their savings plan and their progress. The timing of the visits will be aligned with the original savings plan. During these visits, a list will be published of all members contributing towards the investment. Using social encouragement may be a positive way to produce a beneficial outcome; by public recognition of contributions, pride from the community seeing me invest leads to the best outcome being getting the benefits of the communal investments that are a result of my contribution.

2. Provision of water monitoring information

Two information products will be provided. First, information on plot-specific actual water use and the associated productivity outcomes (in terms of crop per drop) will be discussed with the farmers on a regular basis after the agricultural surveys have established plot-level crop outcomes. Comparison with expected water demand for the crops planted by each farmer will help to identify sub-scheme level over- and under-use of water, relative to optimal crop water demand.

Second, for each of the pilot schemes, both sources of individual farmer information will be aggregated and the sub-scheme level and mapped within the scheme. These maps will not only show farmers how much water they use relative to other members of the community leading to an adjustment in their own water use, but also convey information with respect to how their own practices affect the distribution of water to other parts of the scheme. This information tool can be used to assess how water use could be reallocated between sub-schemes in order to achieve higher yields across the scheme.

The exact content of the two information packages is being developed in a pilot study in three schemes within the Manica and Sofala provinces. The pilot will be completed by mid-2016 and aims to decipher the tradeoffs in data collection (cost and precision) to produce the relationship between individual plot yields and water demand/availability, which produces the information content for this treatment arm. The results of the pilot will feed into a refinement of the feedback tool design and required data collection arrangements.


3. Irrigation infrastructure

The project will support the rehabilitation and development of at least 3,000 ha of irrigated farmland in three central regions of the country: Manica, Sofala and Zambézia provinces. Three value chains will be included: (i) Medium scale rice irrigation schemes, (ii) Small scale private irrigated horticulture (fruits and vegetables), and (iii) Small scale outgrower arrangements. In total 42 schemes will be built and are expected to benefit around 2,500 households. Table 1 provides average characteristics of the schemes by value chain. All schemes are expected to be in operation by the end of 2016.

2016-03-01

2018-06-30

Primary outcome indicators:

Outcome name -- Definition -- Measurement level
1. Individual Savings Contributions -- Total annual savings -- Individual
2. Individual achieved savings target -- Binary variable for whether the individual achieved his or her annual savings target -- Individual
3. Yield -- Total Revenue per hectare -- Individual
4. Profit -- Total income received from crop sales net input costs -- Individual
5. Cyclicality of yield -- Variability of NDVI throughout the year -- Communal/individual

1. Financial Literacy and Incentives for Cooperation

The financial capacity building evaluation questions investigate whether inputs and capacity building provided from the project to the farmers groups can stimulate savings and investment. In total 64 groups are identified by the project to receive the basic financial literacy and group savings program. Half of the groups will receive additional follow-up training and information treatment to provide incentives for collaboration. To answer this question, we propose to randomly assign the follow-up interventions aimed at strengthening the impact of the basic training on investment and community savings in the selected groups. The unit of randomization will be farmers groups, so that all farmers within a group receive the same treatment.

2. Provision of water monitoring information

The water measurement evaluation questions will be addressed by randomizing information on water use and revenue per Ha outcomes to farmers within irrigation schemes. Water use and revenue will be measured for each plot in both the treatment and control arms but the information will only be provided to the 21 schemes selected for treatment.

Randomization done in office by a computer using Stata

Randomization will occur at the farmer/ water user association level. Outcome of interest include both cluster as well as individual farmer household level. Treatment will we implemented at the cluster level.

We will stratify the randomization at the level of 3 provinces and 3 value chains (rice cultivation, horticultural cultivation, and outgrowing).

Yes

64 associations

2000 farmers

42 locations will receive the irrigation infrastructure. The financial literacy will be applied in an additional 22 locations. The water information treatment will be implemented in 21 of the irrigation sites. The randomization will be done in such a way that for each particular intervention the proportion of treatment and control of the other interventions is balanced. The resulting allocation is as follows:
- [fin. lit treatment ] + [water feedback treatment ] = 11 organizations
- [fin. lit control ] + [water feedback treatment ] = 10 organizations
- [fin. lit treatment] + [water feedback control] = 10 organizations
- [fin. lit control] + [water feedback control] = 11 organizations
- [fin. lit treatment] + [no irrigation] = 11 organizations
- [fin. lit control] + [no irrigation] = 11 organizations

Our sample size is limited by the number of farmer groups that will be treated and the number of members of these groups. The project will build around 42 irrigation schemes. However the financial intervention will be provided to an additional 22 groups. Given these parameters on cluster size and the number of observations within each cluster, we calculate the minimum detectable effects (MDEs) assuming a power of 0.8 and an alpha of 0.1. For both evaluations the main outcome indicator is yield. Our sampling frame will be based on the random selection of 25 households per scheme. To calculate the MDE of household Maize revenue per hectare (MZN/ha), we take the mean (1051), standard deviations (1186) and ICC (0.10) of yields from the Smallholders Survey collected in an adjacent region of Mozambique (Kondylis, Mueller, and Zhu, 2015). With a baseline and one follow up for yield outcomes, we should be able to detect a 18.7% increase in revenue the financial literacy intervention (64 clusters) and 23.8% in the water monitoring intervention (42 clusters). For the financial literacy intervention, the main outcome indicator of interest is proportion of farmers that meet their saving target after one year. We will have farmer level data on savings contributions for each of the on average 40 members of the farmers groups. We assume currently none of the farmers contribute to a common savings pool. The MDE ranges from 0.08 with an ICC of 0.05 to 0.13 with an ICC of 0.15.