Last week I was at the annual conference of the Centre for the Study of African Economies (CSAE) in Oxford. In all there were about 259 papers presented, which may make it the biggest development economics conference in the world (NEUDC, the largest development economics conference in North America had 130 papers in 2023). You can read the full CSAE programme here, but here’s my brief analysis.
First, papers were presented on 43 unique countries, mostly in Africa, but also included papers on Bangladesh, India, Pakistan, Nepal, Brazil, China, Indonesia, Malaysia, Thailand, Philippines, Peru, Venezuela, Jamaica, Kyrgyzstan, as well as all of the multi-country studies. The countries with the most studies on them were Ethiopia (24 papers), Nigeria (19), South Africa (17), Kenya (16), and Ghana (15).
Next, I looked at papers by sector. Note that these are my subjective allocations of papers to sectors, based partly but not wholly on the conference structure, and bearing in mind that there is a fair amount of overlap among categories (for example, papers looking at the relationship between agriculture and education). The largest categories were agriculture (37 papers), firms (30), education (29), social protection (24), and macroeconomics (20).
I’m not going to attempt to summarise all 259 papers, or any of the panels, but I will do a round-up of the education papers. Of these 29, drafts are online for 27. Ten of those are RCTs. Note that I list the author listed as the presenter on the programme, and the method in parentheses, where RCT stands for randomised control trial, DiD for difference-in-difference, RD for regression discontinuity, IV for instrumental variables, and OLS for ordinary least squares.
The education papers
Exposure to violent conflict worsens children's health, nutrition, and schooling, even several years after a conflict (Sakketa et al., DiD)
Disappointing follow-up results from a landmark RCT on gender norms in India. The curriculum changed norms in the short-term, but after six years had no impact on women's schooling, employment, or age of marriage (Dhar et al., RCT)
Algorithmic reallocation of teachers across Senegal could improve test scores, but increase inequality (Alidou et al., OLS)
Management quality at district education offices is associated with school performance in Ghana (Mansoor et al., OLS)
Training for school leaders in Rwanda did not improve learning (Lauterbach et al., RCT)
A text-message outreach campaign in Côte d’Ivoire had no impact on enrollment in a youth employment programme (Rouanet et al., RCT)
Children in Brazil's Bolsa Familia programme scored better on high-stakes secondary school exams if their family received a payment in the days before the exam. This had persistent impacts on college enrolment (Larrinaga et al., RCT)
Iterative A/B testing of a technology-enabled tutoring programme in Botswana led to large gains in efficiency (Cullen et al., RCT)
Adaptive edtech software on tablets increasing learning for children with disabilities in Kenya (Odhiambo et al., RCT)
There are wage gains for Indonesians in their mid-20s from better test scores at age 10 (Crawfurd, OLS)
A light-touch teacher training programme in Mozambique had no impact—big effects need big inputs (Karachiwalla et al., RCT)
In hundreds of Malawian schools there is only one qualified teacher for every 190 pupils in lower primary. In this context new teacher recruitment did improve test scores (Asim et al., DiD)
School grants in Ethiopia helped children in historically disadvantaged regions catch up (Araya et al., DiD)
Negative economic shocks can actually increase education, when the shock is to livestock in pastoralist families, which reduces demand for child labour (Noritomo, IV)
Overlap between the school and farming calendar reduces grade attainment in Malawi (Allen, IV)
A large-scale education stipend programme for girls in Bangladesh reduced preference for sons (Priyanka et al., IV)
High temperatures in Ghana reduced children's attentiveness, but not their test scores (Vidogbena et al., IV)
International PIRLS data shows that average grade 4 reading fell by 0.29 standard deviations after COVID in South Africa. Inequality also widened, with no fall for the richest fifth of schools (Böhmer et al., OLS)
Encouragement to listen to educational radio had no impact on learning in Sierra Leone, but home-based workbooks did (Buchel et al., RCT)
Building pre-schools and improving primary school are substitutes in Côte D'Ivoire (Michel, RCT)
New university openings reduced fertility in Ethiopia (Yigermal, DiD)
When schools charge fees parents focus on one child to the detriment of the others. When school is free they can support them all (Sandholtz, DiD)
Attending double-shift (half-day) schools in Gambia has negative effects on learning (Jallow, IV)
Ghana increased secondary school from 3 to 4 years in 2007, with substantial positive effects on test scores and university enrollment (Awadey, RD)
More schooling in Senegal had no impact on attitudes to domestic violence or female genital mutilation (Perez-Parra, IV)
A five-day training on management and socio-emotional skills for unemployed young people in Tanzania had no lasting impacts on employment outcomes (Kipchumba et al., RCT)
The wage returns to investment in schooling in DRC are highest at secondary school (Matungulu et al., IV)
CGD blog posts reflect the views of the authors, drawing on prior research and experience in their areas of expertise.
CGD is a nonpartisan, independent organization and does not take institutional positions.
