Back to History

Fields Changed

Registration

Field Before After
Abstract Almost 1 billion adults worldwide are unable read or write in any language, with an estimated cost of US$ 5 billion annually to low-income countries (UNESCO 2020). This is of particular concern in West Africa, which has some of the lowest education indicators in the world. In Niger, the subject of our study, adult illiteracy rates are estimated at 70% (USAID 2023). Illiteracy is particularly striking in rural areas and among women. Niger’s ranking in terms of educational attainment is comparable to other Sahelian countries, namely Burkina Faso, Chad and Mali (UNDP 2017). While policymakers and academics alike have focused on primary and secondary education programs, such programs will only have impacts in the medium to long-term. Adult education programs have the potential to bridge these gaps and yield more immediate returns, but are often “characterized by low attendance, high drop-out, limited skills attainment and rapid skills depreciation (Aker et al 2024). This is potentially due to high opportunity costs for adults, the irrelevancy of such skills in daily life, poorly-adapted pedagogical approaches and teacher absenteeism. The growth of information technology has offered new opportunities to address some of these constraints, either by providing learners with greater flexibility during the learning process (Ksoll et al 2023, Angrist et al 2022), allowing students to practice outside of class or affecting the returns to education (Aker et al 2012). Our solution is a mobile phone-based adult literacy program. It will build upon platform originally developed for native Spanish-speakers in the US (Cell-Ed). The platform uses voice, audio and SMS to deliver literacy lessons (micro-modules). Each micro-module has three components: 1) a short audio lesson on a particular concept (letters, syllables, words); 2) a SMS reinforcing the voice lesson; and 3) a SMS-based interactive quiz on the lesson. To activate the program, participants send a SMS to a pre-arranged phone number. Students can access the program 24 hours a day, 7 days a week, allowing them to learn when, where and how they wish. This research will test the impact of this innovation in Niger, targeting low-literate adult men and women in the Dosso region. We will test the platform in Hausa, one of the primary indigenous languages in Niger. Across 100 villages, we will stratify by geographic area and randomly assign villages to have access to the platform (which involves training on how to use it, and access to a facilitator) or none. We will also cross-randomize a subsidy, provided via mobile phone credit, to determine if and how liquidity constraints affect access. Almost 1 billion adults worldwide are unable read or write in any language, with an estimated cost of US$ 5 billion annually to low-income countries (UNESCO 2020). This is of particular concern in West Africa, which has some of the lowest education indicators in the world. In Niger, the subject of our study, adult illiteracy rates are estimated at 70% (USAID 2023). Illiteracy is particularly striking in rural areas and among women. Niger’s ranking in terms of educational attainment is comparable to other Sahelian countries, namely Burkina Faso, Chad and Mali (UNDP 2017). While policymakers and academics alike have focused on primary and secondary education programs, such programs will only have impacts in the medium to long-term. Adult education programs have the potential to bridge these gaps and yield more immediate returns, but are often “characterized by low attendance, high drop-out, limited skills attainment and rapid skills depreciation (Aker et al 2024). This is potentially due to high opportunity costs for adults, the irrelevancy of such skills in daily life, poorly-adapted pedagogical approaches and teacher absenteeism. The growth of information technology has offered new opportunities to address some of these constraints, either by providing learners with greater flexibility during the learning process (Ksoll et al 2023, Angrist et al 2022), allowing students to practice outside of class or affecting the returns to education (Aker et al 2012). Our solution is a mobile phone-based adult literacy program. It will build upon platform originally developed for native Spanish-speakers in the US (Cell-Ed). The platform uses voice, audio and SMS to deliver literacy lessons (micro-modules). Each micro-module has three components: 1) a short audio lesson on a particular concept (letters, syllables, words); 2) a SMS reinforcing the voice lesson; and 3) a SMS-based interactive quiz on the lesson. To activate the program, participants send a SMS to a pre-arranged phone number. Students can access the program 24 hours a day, 7 days a week, allowing them to learn when, where and how they wish. This research will test the impact of this innovation in Niger, targeting low-literate adult men and women in the Dosso region. We will test the platform in Hausa, one of the primary indigenous languages in Niger. Across 105 villages, we will stratify by geographic area and randomly assign villages to have access to the platform (which involves training on how to use it, and access to a facilitator) or none. There be two types of treatments, described below. We will also cross-randomize motivation messages.
Last Published January 20, 2026 06:22 AM February 21, 2026 06:19 AM
Primary Outcomes (End Points) Adopt and usage of the platform, per day/month Amount of spending on the platform Reading levels (based upon an in-person reading test) Mobile phone knowledge and usage Attitudes towards education Children's enrollment and attendance in school Adopt and usage of the platform, per day/month Amount of spending on the platform Reading levels (based upon an in-person reading test) Mobile phone knowledge and usage Attitudes towards education Children's enrollment and attendance in school Self-esteem and self-efficacy Household production, income-generating activities and ability to respond to shocks WTP for the technology
Experimental Design (Public) We will use a village-level cluster RCT, stratifying by sub-region and randomly assigning 100 villages to treatment or control. After stratifying by treatment status, villages will be randomly assigned to three different subsidy levels, in the form of airtime credit. Within all villages, we will first identify eligible participants, defined as those who are illiterate. Amongst eligible participants, we will stratify by gender and randomly select 12 men and 12 women per village, which conforms with Ministry norms. We will also have a spillover sample of 6 participants per village. We will use a village-level cluster RCT, stratifying by sub-region and randomly assigning 105 villages to treatment or control. After stratifying by treatment status, villages will be randomly assigned to treatment (76 villages) or control (29 villages). Within all villages, we will first identify eligible participants, defined as those who are illiterate. Amongst eligible participants, we will stratify by gender and randomly select 8 men and 8 women per village, which conforms with Ministry norms. We will also have a spillover sample of 4 participants per village.
Planned Number of Clusters The proposed sample size is 104 villages, with 20 students per village. The proposed sample size is 105 villages, with 16 students per village and a spillover sample of 4 per village.
Planned Number of Observations We will choose 16 students per village, plus 4 spillover students, for a total of 20 students per village and 2,100 students. We will choose 16 students per village, plus 4 spillover students, for a total of 20 students per village.
Sample size (or number of clusters) by treatment arms We will have 75 treatment villages and 30 control villages We will have 76 treatment villages (divided into two treatments) and 29 control villages
Pi as first author No Yes
Back to top