|
Field
Trial Status
|
Before
in_development
|
After
on_going
|
|
Field
Last Published
|
Before
April 02, 2026 01:41 PM
|
After
September 25, 2026 11:54 PM
|
|
Field
Intervention (Public)
|
Before
The intervention combines three components: credit, extension services, and contract farming. We test whether credit products aligned with agricultural seasonality improve uptake and enable investment in adaptive technologies. Extension services, delivered via in-field agronomists and ICT tools, seeks to equip farmers with timely, actionable guidance for managing climate variability, an area with limited existing evidence. Contract farming improves farmers’ market access and supports quality certification, allowing farmers to obtain better prices for higher-quality products. This mechanism aims to address price distortions that otherwise disincentivize investment in quality-enhancing practices.
|
After
The intervention combines three components: credit, extension services, and contract farming. We test whether credit products aligned with agricultural seasonality improve uptake and enable investment in adaptive technologies. Extension services seek to equip farmers with timely, actionable guidance for managing climate variability, an area with limited existing evidence. Contract farming improves farmers’ market access and supports quality certification, allowing farmers to obtain better prices for higher-quality products. This mechanism aims to address price distortions that otherwise disincentivize investment in quality-enhancing practices.
|
|
Field
Primary Outcomes (Explanation)
|
Before
1) Take-up of treatment will be measured by individually considering take-up of microcredit and take-up of contracts.
2) Knowledge will be a standardized index based on a vector of knowledge questions. Adoption will be a separately standardized index based on a vector of adoption and practice questions, including acreage planted to commercial crops. For our statistical tests, we plan on using the high powered weighted index from Anderson and Magruder (2024), though we will also report results from a flat index following Kling et al (2007). If that index rejects, we anticipate separately testing the component variables.
3) Agricultural productivity, farm profits, and household income will be standardized and tested in an index. If that index rejects, we anticipate separately testing the component variables. Once again, we plan on using the highly powered Anderson and Magruder (2024) index for our primary tests though we will also report results from an unweighted index.
|
After
1) Take-up of treatment will be measured by individually considering take-up of microcredit, extension contact, and take-up of contracts.
2) Knowledge will be a standardized index based on a vector of knowledge questions. Adoption will be a separately standardized index based on a vector of adoption and practice questions, including acreage planted to commercial crops. For our statistical tests, we plan on using the high powered weighted index from Anderson and Magruder (2024), though we will also report results from a flat index following Kling et al (2007). If that index rejects, we anticipate separately testing the component variables.
3) Agricultural productivity, farm profits, and household income will be standardized and tested in an index. If that index rejects, we anticipate separately testing the component variables. Once again, we plan on using the highly powered Anderson and Magruder (2024) index for our primary tests though we will also report results from an unweighted index.
|
|
Field
Randomization Method
|
Before
Randomization done by random number generator on computer
Randomization will be stratified by governorate, survey round, and whether the cluster-level share of tenant farmers is above the median across all clusters.
|
After
Randomization done by random number generator on computer
Randomization will be stratified by governorate, survey round, village, and cluster size.
|
|
Field
Randomization Unit
|
Before
Randomization unit is the farmer cluster, where one cluster consists of 10-12 farmers.
|
After
Randomization unit is the farmer cluster, where one cluster consists of 2-5 farmers.
|
|
Field
Planned Number of Clusters
|
Before
226 clusters of farmers
|
After
525 clusters of farmers
|
|
Field
Power calculation: Minimum Detectable Effect Size for Main Outcomes
|
Before
With clustered randomization at the zone level (10 farmers per cluster, an ICC of 0.1), the study can detect the following minimum detectable effects (MDEs) with 80% power and a 5% significance level:
- Contract farming take-up: 7 percentage point increase in take-up (based on 1,950 farmers, 20% of whom have prior contract farming experience at baseline)
- Downstream effects of contract farming: MDE of 0.35 standard deviations (assuming 50% take-up of contract farming among 1,960 farmers)
- Downstream effects of loans: MDE of 0.27 standard deviations (assuming 75% take-up of the loan treatment among 1,940 farmers)
|
After
With clustered randomization at the zone level (5 farmers per cluster, an ICC of 0.1), the study can detect the following minimum detectable effects (MDEs) with 80% power and a 5% significance level:
- Contract farming take-up: 6 percentage point increase in take-up (based on 1,955 farmers, 20% of whom have prior contract farming experience at baseline)
- Downstream effects of contract farming: MDE of 0.30 standard deviations (assuming 50% take-up of contract farming among 1,960 farmers)
- Downstream effects of loans: MDE of 0.23 standard deviations (assuming 75% take-up of the loan treatment among 1,974 farmers)
|
|
Field
Secondary Outcomes (End Points)
|
Before
1) Welfare indicators (mental health, perceived improvements in quality of life)
2) Food security (FIES)
3) Soil degradation, including soil salinity (measured via soil tests by enumerators through a simple technique and soil moisture (measured by enumerators conducting a “squeeze test”)
4) Irrigation water use (measured using various methods, including number of hours of using irrigated water)
5) Strategies adopted to adapt to climate-related events
6) Sensitivities of farm outcomes to climate-related events
7) Loan repayment rates
8) Selection into contract farming, including differences in selection with and without loan conditionality. We plan on evaluating covariates that predict selection in each case, and also examining differences in treatment effects which may be attributable to selection.
|
After
1) Welfare indicators (mental health, perceived improvements in quality of life)
2) Food security (FIES and HDDS)
3) Soil degradation, including soil salinity (measured via soil tests by enumerators through a simple technique and soil moisture (measured by enumerators conducting a “squeeze test”)
4) Irrigation water use (measured using various methods, including number of hours of using irrigated water)
5) Strategies adopted to adapt to climate-related events
6) Sensitivities of farm outcomes to climate-related events
7) Loan repayment rates
8) Selection into contract farming, including differences in selection with and without loan conditionality. We plan on evaluating covariates that predict selection in each case, and also examining differences in treatment effects which may be attributable to selection.
|