Demand and supply factors constraining the emergence and sustainability of an efficient seed system - three experiments in Uganda

Last registered on November 21, 2022

Pre-Trial

General Information

Title

RCT ID

AEARCTR-0006361

Initial registration date

August 25, 2020

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 25, 2020, 1:55 PM EDT

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

Last updated

November 21, 2022, 4:02 AM EST

Last updated is the most recent time when changes to the trial's registration were published.

Locations

Country

Uganda

Region

Busgoa

Primary Investigator

Name

Bjorn Van Campenhout

Affiliation

Ifpri

Contact Primary Investigator

Other Primary Investigator(s)

PI Name

Caroline Miehe

PI Affiliation

KU Leuven

Contact Investigator

PI Name

Robert Sparrow

PI Affiliation

Wageningen University

Contact Investigator

PI Name

Prossy Renzaho Ntakyo

PI Affiliation

NARO

Contact Investigator

PI Name

Fredrick Bagamba

PI Affiliation

Makerere University

Contact Investigator

PI Name

David Spielman

PI Affiliation

IFPRI

Contact Investigator

PI Name

Geoffrey Otim

PI Affiliation

ISSD

Contact Investigator

Additional Trial Information

Status

Completed

Start date

2020-09-15

End date

2022-08-01

Keywords

Agriculture

Additional Keywords

seed systems, information clearing house, learning failures, information, input quality, agricultural technology adoption

JEL code(s)

O13; Q12; Q16; D82; D83

Secondary IDs

Prior work

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

Abstract

Agricultural technology remains under-adopted among smallholder farmers in Sub-Saharan Africa. We investigate how the quality of an agricultural technology – improved maize seed – affects its adoption. The research entails three hypotheses that will be tested in a series of randomized controlled trials among agro-input dealers and smallholder farmers in Uganda. In a first hypothesis, quality concerns that constrain uptake are caused by information inefficiencies at the level of the agro-input dealer, who is assumed to lack knowledge about proper storage and handling. An intensive training program is expected to increase improved maize seed quality and subsequent adoption by farmers. A second hypothesis conjectures that information asymmetry between seller and buyer with respect to the quality of seed – a classic lemons technology – leads to under-adoption. We implement a crowd-sourced information clearinghouse similar to yelp.com to test this hypothesis. This hypothesis targets the interaction between farmers and input dealers. A third hypothesis targets farmers directly, as sub-optimal adoption is assumed to be caused by learning failures: Farmers might attribute disappointing outcomes to poor input quality, while in reality many input dimensions like the time of planting, weeding and fertilizer application co-determine outcomes. An ICT-mediated information campaign that stresses the importance of paying attention to all input dimensions is implemented to test this hypothesis.

External Link(s)

Github repo for the project

Registration Citation

Citation

Bagamba, Fredrick et al. 2022. "Demand and supply factors constraining the emergence and sustainability of an efficient seed system - three experiments in Uganda." AEA RCT Registry. November 21. https://doi.org/10.1257/rct.6361-3.0

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)

3 interventions will be implemented in a 2x2x2 factorial trial. A first intervention consists of an intensive agro-input dealer training to improve seed handling and storage at this level. A second intervention consists of a crowd-sources information clearing house that will be set up at the catchment area level.

Intervention Start Date

2021-02-02

Intervention End Date

2022-03-01

Primary Outcomes

Primary Outcomes (end points)

quantity of quality seed sold by input dealers, adoption rates of quality seed by farmers

Primary Outcomes (explanation)

quantity (kg) of improved seed varieties of maize (hybrids and OPV) sold by the input dealer over the last agricultural season; indicator value =1 of farmer used improved seed (hybrid of OPV) obtained from an agro-input dealer on any maize plot

Secondary Outcomes

Secondary Outcomes (end points)

input dealer skill/knowledge, farmers' seed quality perception, farmers' input dealer quality perception, farmer skill/knowledge, yield/profit/income/consumption

Secondary Outcomes (explanation)

Experimental Design

To test the three hypotheses, we will implement three interventions that are combined in a field experiment where various treatment and control groups are randomly assigned to either a treatment or control condition. The randomized control trial (RCT) will take the form of a 2x2x2 factorial design, with each intervention corresponding to one hypothesis. To test the first hypothesis, a random sub-sample of input dealers will receive training on proper seed handling and storage. To test the second hypothesis, a rating system will be set up among a random sub-sample catchment areas of input dealers catchment areas, and farmer and input dealers will receive feedback on the ratings before the start of the planting season. To test the third hypothesis, a video that points out the importance of combining improved seed with other inputs and careful crop management will be shown to a random subset of farmers.

Experimental Design Details

Randomization Method

The randomization of the interventions are the catchment area level were done by a computer based on a census of input dealers. The randomization of the farmers will be done through se

Randomization Unit

the first and second treatment are implemented at the catchment area level. For the third treatment, treatment status will also be determined by the computer, but a geographic sampling approach will be used where enumerators are supplied with gps coordinate from where they have to find the closest household to include in the study.

Was the treatment clustered?

Yes

Experiment Characteristics

Sample size: planned number of clusters

112 catchment areas

Sample size: planned number of observations

3200

Sample size (or number of clusters) by treatment arms

in 56 catchment areas, treatment 1 will be implemented and in 56 not; in 56 catchment areas, treatment 2 will be implemented and in 56 not; 1600 farmers will receive treatment 3 and 1600 farmers will serve as control.

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

159 kilograms sold (standard deviation of 454 kilograms); MDE for adoption is < 6.5% (from a base of63 percent)

Supporting Documents and Materials

IRB