Evaluating the impact of alternative interventions to improve adoption of digital innovations by smallholder farmers in Egypt.

Last registered on December 21, 2023

Pre-Trial

General Information

Title

RCT ID

AEARCTR-0012723

Initial registration date

December 19, 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

December 21, 2023, 8:01 AM EST

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

Locations

Country

Egypt

Region

Middle East and North Africa (MENA)

Primary Investigator

Name

Fatma Abdelaziz

Affiliation

International Food Policy Research Institute & Wageningen University

Contact Primary Investigator

Other Primary Investigator(s)

PI Name

Erwin Hendricus Bulte

PI Affiliation

Wageningen University

Contact Investigator

PI Name

Kibrom Abay

PI Affiliation

International Food Policy Research Institute (IFPRI)

Contact Investigator

Additional Trial Information

Status

On going

Start date

2023-07-01

End date

2024-11-30

Keywords

Agriculture, Behavior, Welfare

Additional Keywords

Digital agriculture training, Smallholder farmer, Digital literacy, Edutainment

JEL code(s)

Secondary IDs

Prior work

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

Abstract

Digital innovations could facilitate access to markets and smallholder commercialization through the following ways: (1) reduce communication and information costs; (2) improve farmers’ knowledge and know- how about market options and prices; (3) improve access to input and output markets; (4) enable and build social networks and connections; (5) facilitate the delivery of other services associated with agricultural markets such as credits and finance; (6) improve management of input and output supply chains; (7) increase communication linkages with other stakeholders involved in agricultural marketing. Despite these advantages that digital innovations could offer, their adoption remains low and heterogenous, especially in Africa, where agricultural markets remain underdeveloped (e.g., Abate et al., 2023; Aker and Cariolle, 2023). Despite the proliferation of digital tools targeting smallholder farmers in recent years in Africa, the vast majority of these remained at pilot stages, with limited evidence of successful scaling and limited impacts to transform agricultural markets.

There exists several demand and supply-side factors that may explain the low adoption of digital innovations and associated heterogeneities across smallholder farmers in Africa. From a supply perspective, this low adoption could be due to several factors, including but not limited to insufficient public and private investment in complementary infrastructure, unsustainable business models, and asynchronous pace of change (Abate et al., 2023). The demand side factors may include: lack of digital literacy, lack of context-specific needs assessment, and digital divide and more importantly accessibility and usability as well as user trust and confidence. However, we lack empirically grounded evidence on alternative and cost-effective interventions to improve adoption and scale-up of digital innovations in various settings. In particular, empirical evidence on most effective strategies to improve access and usability of digital agricultural innovations to smallholders with limited level of literacy remains scarce.

Taking Egypt as a case study, the Egyptian market has an array of innovative digital agricultural tools that offer different services to farmers such as sourcing inputs, providing post-harvest and logistical support, to accessing market information and selling crops online. However, the uptake of these technologies has been quite low, as Egyptian farmers lack the awareness of the benefits of digital tools and can easily get excluded from this agricultural digital revolution due to a lack of accessible training (OBG, 2022; AGBI, 2023). This project aims to test alternative interventions to promote the adoption of digital agricultural innovations in Egypt.

External Link(s)

Registration Citation

Citation

Abay, Kibrom, Fatma Abdelaziz and Erwin Hendricus Bulte. 2023. "Evaluating the impact of alternative interventions to improve adoption of digital innovations by smallholder farmers in Egypt.." AEA RCT Registry. December 21. https://doi.org/10.1257/rct.12723-1.0

Sponsors & Partners

Experimental Details

Interventions

Intervention(s)

The alternative interventions we launch and evaluate in this project are part of the German Corporation for International Cooperation (GIZ)’s Agricultural Innovation Project (AIP). The AIP seeks to promote the use of digital mobile apps to smallholder farmers to increase their incomes. Accordingly, the AIP project of GIZ designed various digital promotion strategies to promote the adoption of digital innovations. These interventions are meant to encourage and enable smallholder access and use of agricultural digital mobile applications. The AIP project and hence these interventions focused on two governorates in Egypt: Minya and Benisuef .
The International Food Policy Research Institute (IFPRI) partnered with GIZ to redesign some of these digital promotion strategies and eventually evaluate them using rigorous impact evaluation methods.
1- Digital literacy training through Trainers of Trainees (TOT) sessions
2- Pre-recorded community play videos
3- Real Live Community Play

Intervention (Hidden)

Three alternative digital promotional strategies were chosen and implemented in collaboration with the local farmer organizations in Minya and Benisuef:
1- Digital literacy training through Trainers of Trainees (TOT) sessions: This involves an instruction-based training that focuses on 3 existing successful agricultural digital apps in Egypt that provide market advisory services to farmer
2- Pre-recorded community play videos: This involves a prerecorded video of a community play that was held before in Minya and Benisuef and was organized by GIZ. The pre-recorded video we will offer in our intervention are meant to provide additional educational material and information to framers in an entertaining way. The community play videos are going to be offered as additional intervention together with the digital literacy TOT.
3- Real Live Community Play: This involves an actual play held in the vicinity of 6 different communities across Minya and Benisuef. A tent and an actual theatre-style set up, with 100 chairs seating arrangement will be put in place. The real live community play intervention is going to be offered as additional intervention together with the digital literacy TOT. This intervention serves two purposes: (i) helps to identify whether the salience nature of the community play and associated edutainment services can encourage farmers participate in the digital literacy training: (ii) evaluate the potential of edutainment on adoption of digital innovations.

Intervention Start Date

2023-09-18

Intervention End Date

2024-01-06

Primary Outcomes

Primary Outcomes (end points)

1. Participation in digital literacy training
2. Awareness of digital agricultural tools
3. Uptake of digital tools
4. Digital literacy knowledge test scores
5. Adoption and utilization of digital agricultural tools for marketing purposes
6. Market participation and commercialization behavior and associated outcomes

Primary Outcomes (explanation)

1. Participation in digital literacy training. This will be constructed based on comparing the different compliance rates to the 3 treatment arms, to evaluate which treatment was likely most attractive for farmers to participate in and attend.
2. Awareness of digital agricultural tools. Respondents will be asked to report on their awareness of the 3 agricultural digital apps, in the baseline (before they receive any digital literacy training), and in the midline (after they receive digital literacy training) and later in the endline.
3. Farmer uptake of digital tools. This will be constructed based on downloads data available in the backend database of the 3 different digital apps that farmers will receive instruction on, this data will be shared by the mobile app developers. Self-reported downloads in the baseline, midline and endline data will also be compared.
4. Digital App literacy knowledge test scores, this will be constructed based on a 1 month recall of the training material covered during the training sessions. A test score will be constructed based on the farmers responses to the set of questions on the different features of the 3 digital apps farmers received instruction on in the midline survey. The research team will generate binary indicators of pass or fail test scores.
5. Adoption of digital agricultural tools. This will be constructed using the midline and later, the endline survey data; respondents will be asked to report whether they have attempted to sell their crops/ actually sold their crops / attempted to make an order and purchase an agricultural input / actually ordered and received an agricultural input via any of the 3 digital apps. The reported number of attempts will be used to construct an adoption variable.

Secondary Outcomes

Secondary Outcomes (end points)

Secondary Outcomes (explanation)

Experimental Design

The intervention will follow a randomized control trial design. The randomization takes place at the farmer/ individual level. The invitation of eligible farmers to participate in different kinds of digital support schemes in each farmer organization will be led by the active 30 farmer organizations in Minya and Benisuef, that GIZ have partnered with under the GIZ AIP project.
Following the farmer data received from each FO, a baseline phone survey took place between July 1st-Sepetmeber 7th 2023, a total of 4000 surveys were covered in Minya and Beni Suef. After ruling out respondents that do not assume farming as their main occupation, we were left with a total of 3350 respondents; 2400 of which have access to smart phones and the remaining 950 do not.
For the 2400 farmers that own smart phones, they will be eligible to receive a package of interventions to encourage the use and uptake of existing digital tools. The 2400 farmers from the different farmer organizations will be randomly assigned into three treatment arms, based on the type of treatment/ type of digital promotional strategy they receive. These farmers will be interviewed at midline and endline. These interventions/ dissemination methods include:
 Control Group: no intervention.
 Group A: Digital App Literacy training sessions
 Group B: Pre-recorded community play videos + Digital App Literacy training sessions (Training sessions will be provided, but here trainees will see video of play before the training)
 Group C: Real Live Community Plays + Digital App Literacy training sessions (Training sessions will be provided, but here trainees will see the actual play before the training)

Experimental Design Details

To exploit power gains from repeated observations, we will collect three rounds of data, baseline, midline and endline data (e.g., McKenzie, 2012). The baseline phone survey was conducted before laying out the three different digital promotion strategies. The information collected from the baseline include pre-intervention outcomes, and key characteristics expected to generate potential heterogeneities in the impact of the different promotion strategies. The baseline data was also used to exclude respondents that do not assume farming as their main occupation or do not have smart phones, implying that they cannot possibly benefit from the different kinds of digital promotion strategies that will be laid out in our set of interventions.
A midline phone survey will be collected with the set of eligible farmers chosen for the interventions, after about two-three months from launch the above interventions/ digital promotion strategies. An endline survey will re-survey the same households 6-9 months later after the midline to measure the medium-term impacts and effects of the different digital promotion strategies

Randomization Method

The randomization was done at the farmer level using the baseline data and after excluding those farmers without smartphone and those whose main occupation is not farming.

Randomization Unit

Farmer

Was the treatment clustered?

Experiment Characteristics

Sample size: planned number of clusters

Treatment was not clustered, but randomization was stratified.
Number of clusters by treatment arm:
Control: 573 Farmers
T1 (Group A): 601 Farmers
T2 (Group B): 578 Farmers
T3 (Group C): 649 Farmers

Sample size: planned number of observations

2401 farmers (based on baseline survey data)

Sample size (or number of clusters) by treatment arms

Number of clusters by treatment arm:
Control: 573 Farmers
T1 (Group A): 601 Farmers
T2 (Group B): 578 Farmers
T3 (Group C): 649 Farmers

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

Our power calculations aim to achieve the standard and widely adopted 80 percent power at a significance level of 5 percent.

Supporting Documents and Materials

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

Request Information

IRB