Randomized allocation of smallholder irrigation scheme and market-oriented agricultural extension intervention in Zambia.

Last registered on August 16, 2023

Pre-Trial

General Information

Title

RCT ID

AEARCTR-0011757

Initial registration date

August 16, 2023

Initial registration date is when the trial was registered.

It corresponds to when the registration was submitted to the Registry to be reviewed for publication.

First published

August 16, 2023, 11:59 AM EDT

First published corresponds to when the trial was first made public on the Registry after being reviewed.

Locations

There is information in this trial unavailable to the public. Use the button below to request access.

Request Information

Primary Investigator

Name

Satoshi Shimizutani

Affiliation

University of Tokyo

Contact Primary Investigator

Other Primary Investigator(s)

PI Name

Yasuyuki Sawada

PI Affiliation

University of Tokyo

Contact Investigator

PI Name

Hitoshi Fujiie

PI Affiliation

Japan International Cooperation Agency

Contact Investigator

Additional Trial Information

Status

On going

Start date

2022-05-01

End date

2024-05-01

Keywords

Agriculture, Behavior, Labor

Additional Keywords

infrastructure, smallholder irrigation, market-oriented agriculture, smallholder farmers, RCT, Zambia

JEL code(s)

Q16 R&D • Agricultural Technology • Biofuels • Agricultural Extension Services

Secondary IDs

Prior work

This trial does not extend or rely on any prior RCTs.

Abstract

The impact of infrastructure development is difficult to measure because of its innate nature of non-randomness. We examine the impact of smallholder irrigation scheme by assigning farmer groups randomly to make a causal inference on livelihood of smallholder farmers. Moreover, a half of farmer groups which are randomized assigned among the treatment group receive market-oriented agricultural extension too. This study reveals the relative impact of infrastructure (hardware) and agricultural extension (software). This study stands out as an exceptionally rare evaluation of an infrastructure project based on randomized controlled trials (RCTs). As far as we know, it's also the pioneering RCT conducted for an irrigation project.

External Link(s)

Registration Citation

Citation

Fujiie, Hitoshi, Yasuyuki Sawada and Satoshi Shimizutani. 2023. "Randomized allocation of smallholder irrigation scheme and market-oriented agricultural extension intervention in Zambia. ." AEA RCT Registry. August 16. https://doi.org/10.1257/rct.11757-1.0

Sponsors & Partners

There is information in this trial unavailable to the public. Use the button below to request access.

Request Information

Experimental Details

Interventions

Intervention(s)

The primary intervention of this study is to diffuse the construction technic of small-scale irrigation schemes covering both simple weirs and canal to farmer fields, which is called "E-COBSI" (Community-Based Smallholder Irrigation Development Project) approach in central regions of Zambia. The approach has been developed more than 10 years by Japan International Cooperation Agency (JICA) in Malawi. The approach is a package of training components and expansion strategies integrating small scale irrigation construction, water management, operation, and maintenance of irrigation facilities, farming, marketing, gender and nutrition improvement.

Intervention Start Date

2023-04-01

Intervention End Date

2023-08-31

Primary Outcomes

Primary Outcomes (end points)

Farming income (horticultural, non-horticultural), Non-farm income, Household consumption

Primary Outcomes (explanation)

Secondary Outcomes

Secondary Outcomes (end points)

School enrolment, Time use (all household members)

Secondary Outcomes (explanation)

Experimental Design

The goal of this study is to measure the impact of the E-COBSI approach. The approach is a full package of a variety of components and we decompose the E-COBSI approch into hardware component (construction of small-scale irrigation facilities) and software component (market-oriented agricultural extension) to measure the relative imporance of intervention. Thus, we set up two treatment groups. Treatment group A receives a hardware component only while Treatment group B receives a full package of E-COBSI approach (both hardware and software). Control group does not receive any intervention. The gap between Treatment group B and Control group captures the impact of the full package of the E-COBSI approach. The difference between Treatment Group A and Control group measures the impact of the hardware component and that between Treatment group A and Treatment group B does the impact of the software component. Thus, we can understand the impact of the full package and the relative importance of the hardware and software component.

Experimental Design Details

Not available

Randomization Method

We conducted randomization in office by a computer.

Randomization Unit

The unit of randomization is a farmer group.

Was the treatment clustered?

Yes

Experiment Characteristics

Sample size: planned number of clusters

93 farmer groups (31 for Treament group A, 31 for Treatment Group B and 31 for Control group).

Sample size: planned number of observations

1,773 farmer households.

Sample size (or number of clusters) by treatment arms

31 farmer groups to receive a full package of the E-COBHI approach (both small-scale irrigation construction and market-oriented agricultural extension), 31 farmer groups to receive a partial component of the E-COBSI approach (only small-scale irrigation construction) and 31 control groups without any intervention. The sample size of farmer households is estimated to be 1,773 in total.

Minimum detectable effect size for main outcomes (accounting for sample design and clustering)

The main outcome of this study is horticultural sales. The average of horticultural sale in Zambia is estimated to be 1,300 Zambian kwacha and the standard deviation is 2,600 kwacha. The percentage is estimated to be 30% per year. Alpha is set to be 0.05, Beta 0.2 and Power 0.8. The minimum detectable effect size for main outcomes is 698 for each treatment and control group. This study holds two types of treatments and the minimum detectable effect size is estimated to be 2,094 and our sample size of 1,773 exceeds the minimum.

Supporting Documents and Materials

IRB

Institutional Review Boards (IRBs)

IRB Name

The University of Tokyo

IRB Approval Date

2022-11-10

IRB Approval Number

22-253

Analysis Plan