32 New Findings from the Global Education RISE Conference 2021: Parents, Politics, and the Pandemic—Plus the Education Interventions People Would Trash

Last week was the annual conference for the Research on Improving Systems of Education (or RISE) program, a large scale, multi-country research program developed to answer the question: “How can education systems be reformed to deliver better learning for all?”

You can read the full conference program and you can watch videos of all the sessions. But here, I’ve broken down the key findings and takeaways from each presentation.


  • “Scores across math items drawn from different exam years—when taken by an identical group of students on the same day—closely track fluctuations in Ghana’s national pass rates over time." (In other words, it's not the students!) –Abreh et al. (paper; blog post; key slides)

  • “Citizen-led assessments of learning have addressed accountability and participation in education systems" in at least three ways: better data, parental empowerment, and evidence-based policy. –Saeed et al. (paper; key slides)

  • Communities have limited power to monitor teacher activities (they don’t pay their salaries!) so Pratham focuses on communities supporting teachers, while also teaching communities *how* to monitor learning outcomes for themselves. –Bano (key slides)

Parental involvement and the demand for education

  • In Mexico, providing small grants to parent associations did not boost learning outcomes, but providing information about ways to support children’s learning did reduce dropout rates and disciplinary actions in schools. –Nakajima et al. (paper; key slides)

  • The first primary schools in Nigeria (really, the first ones, in the 1840s) had a large impact on social mobility, according to preliminary findings. –Nyéki, Okoye, and Wantchékon (key slides)

  • In the Gambia, parental aspirations translate into higher educational investments and—"when complementary inputs are present"—higher test scores, but higher “standard deviation” changes in test scores don’t mean much improvement in skills when starting from a low base. –Eble and Escueta (key slides)

  • A package of interventions in Madagascar—activities to target learning to students’ learning levels, assessment, and community engagement—boosted test scores and reduced dropout rates. –Maruyama, Igei, and Kurokawa (paper; key slides)


  • Indonesia’s “problems with education quality and learning have … stemmed from the continued political dominance of predatory political, bureaucratic, and corporate elites… Efforts to improve learning outcomes … are unlikely to produce significant results unless there is a fundamental reconfiguration of power relations.” –Rosser, King, and Widoyoko (paper; key slides)

  • Why does Vietnam do so well in enrolment and learning? It’s because of political commitments to education and high levels of societal engagement, which are working to enhance accountability within the system. But countervailing forces hold it back from doing even better! –London and Hang (paper; key slides)

  • Nepal undertook a major decentralization of its education system in 2017. The result? Local leaders want to be in charge of hiring teachers, but they don’t want to discipline them (when needed). –Sabarwal et al. (paper; key slides)


  • Training school heads on violence prevention in Peru increased reporting of violence and reduced transfers from schools. (It didn’t affect test scores, but come on, not everything has to improve test scores. Let’s just keep the kids from getting hurt!) –Smarrelli (key slides)

  • A school governance reform in Tanzania that shifted focus from school inspections to school support had little impact BUT adding low-cost measures to increase follow-up from ward education officers modestly boosted learning. –Cilliers and Habyarimana (paper; key slides)

  • In a survey of 900+ officials in 36 countries, government education officers recognize the existence of a global learning crisis but underestimate its extent in their own countries, often dramatically overestimating the proportion of ten-year-olds who can read. –Crawfurd et al. (key slides)

Equity and choice

  • Distributing large cash grants through school councils to public schools in rural Pakistan led to learning increases in both the public and private sectors. Learning increases at private schools, which didn’t receive the grants, appear to have been driven by competitive pressure. –Karachiwalla et al. (paper; key slides)

  • In Indonesia, making public junior secondary schools *less* selective led to lower learning gains than expected but the gap narrowed between the top and bottom quintiles. –Beatty et al. (key slides)

  • Ethiopia invested in measures to boost learning (more classrooms, teachers, textbooks, and support for teachers), but learning outcomes actually declined. Could  be because of a decline in preschool / school readiness;  increased enrolment among the most disadvantaged students; or factors external to the education system, like conflict? –Tibebu Tiruneh, Rolleston, and Oketch (key slides; presenter note)

  • Providing a five-day science, technology, engineering and mathematics (STEM) bootcamp in Tanzania led to more progressive gender attitudes among girls (but not boys). It also boosted girls’ weekly hours of studying in STEM-related subjects. –Ahn, Hahn, and Yoon (paper; key slides)

What education interventions would you trash?

In this session, experienced education researchers and practitioners highlighted the worst widespread education interventions.

  • Leonard Wantchekon would trash donor interference in education! He’d redirect aid to focus on incentivizing agency by government officials, promoting stakeholder involvement, and involving communities in decisions and monitoring. (key slides)

  • Rukmini Banerjee would get rid of the age-grade curriculum—in other words, the system that teaches children based on their age rather than their level of learning. “Sometimes when I look at these [textbooks], I feel it’s designed to ensure that many children get left behind.” (key slides)

  • Jishnu Das would trash vouchers. “80% of the [voucher] money goes to children who would have gone to private schools without the vouchers… Since taxes can be quite regressive, it’s taking money from the poor to give to the rich.” (key slides)

COVID-19 learning loss

  • Big losses in South Africa: “In 2020 grade 2 students lost between 57% and 70% of a year of learning relative to their pre-pandemic peers. Among a grade 4 sample, learning losses are estimated at between 62% and 81% of a year of learning.” –Wills, Kotzé, and Ardington (paper; key slides)

  • While schools were closed during COVID-19 in Botswana, “SMS messages and phone calls with parents to support their child” improved “learning by 0.12 standard deviations (SD).” Now being adapted in several other countries! –Matsheng, Angrist, and Bergman (paper; key slides)

  • “Over-the-phone mentoring and homeschooling support delivered by volunteers” during pandemic school closures in Bangladesh “improved the learning outcomes of treated children by 0.75 SD and increased homeschooling involvement of treated mothers by 0.64 SD.” –Siddique et al. (paper; key slides)

  • We didn’t present this at the conference, but Crawfurd et al. (including me) also have a paper out on tutoring phone calls during COVID school closures, this one in Sierra Leone. (paper; blog post)

Instructional coherence and differentiated instruction

  • Researchers rated coherence across three key aspects of instruction—primary curriculum standards, national examinations, and actual teaching delivered in the classroom—in Uganda and Tanzania. Coherence was low. –Atuhurra and Kaffenberger (paper; key slides)

  • Experiments to improve instructional coherence found that shortening teacher guides (in Kenya) had a positive impact, particularly for the lowest performing students. Teaching students to set goals boosted math and science test scores for girls (in Uganda). –de Laat, Gray-Lobe, and Sullivan (extended abstract; key slides)

  • Scaling up Mindspark (a computer-based adaptive learning platform) in schools in Rajasthan had sizeable effects on math and Hindi for students in grades 3-8 (but not 1-2). –Singh and Muralidharan (key slides)

  • A catch-up program for students in Zambia, implemented by teachers after school hours, is associated with sizeable gains in reading ability among primary school students, with some evidence of gains increasing as scale increases. –Pershad et al. (key slides)

Teachers and teacher effectiveness

  • Training teachers in Uganda to ask sharp questions, use evidence, frame specific hypotheses (i.e., think like scientists) increased the pass rate in the primary leaving exam by 24 percentage points! –Nourani, Ashraf, and Banerjee (paper; key slides)

  • Estimates of teacher value-added for academic skills in Vietnam are relatively low, which is surprising given Vietnam’s high level of performance. –Krutikova et al. (key slides)

  • In Ghana, engaging school managers improved the effectiveness of training teachers in differentiated instruction. –Beg, Fitzpatrick, and Lucas (paper; key slides)

  • Many believe India has a major teacher shortage, with too high pupil-teacher ratios. But removing fake student enrollment suggests a surplus of 300,000+ teachers. A hypothetical rule improving teacher allocation to schools increases it further! –Kingdon and Datta (paper; key slides)

Importantly, both the highlights from the papers and the key slides are, of course, my own choice. There is much more to each of these research projects, and if you’re interested, I encourage you to read the papers (where available) or to watch the full presentations.

The RISE conference consistently showcases a wide range of novel education research, demonstrating that sustainable improvements in education systems means much more than a single intervention in a single part of the education sector. It also demonstrates how much we have yet to learn.

Note: In these summaries, I’ve listed the presenter first and—if there are other authors—indicated that. This may not be the same order of authors on the paper.


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.