This website uses cookies.

By clicking the "Accept" button or continuing to browse our site, you agree to first-party and session-only cookies being stored on your device to enhance site navigation and analyze site performance and traffic. For more information on our use of cookies, please see our Privacy Policy.

  1. Home
  2. ASSA Annual Meeting
  3. 2023
  4. Program
  5. Stemming Human Capital Losses Due to COVID-19 School Closures: Experimental Evidence Across Contexts
« Back to Results

Stemming Human Capital Losses Due to COVID-19 School Closures: Experimental Evidence Across Contexts

Paper Session

Sunday, Jan. 8, 2023 1:00 PM - 3:00 PM (CST)

Hilton Riverside, Grand Salon B Sec 9 & 12
Hosted By: American Economic Association
  • Chair: Eliana La Ferrara, Harvard Kennedy School

Building Resilient Education Systems: Evidence from Large-Scale Randomized Trials in Five Countries

Noam Angrist
,
Oxford University and Youth Impact
Micheal Ainomugisha
,
Building Tomorrow
Saipramod Bathena
,
Alokit
Peter Bergman
,
University of Texas-Austin and NBER
Colin Crossley
,
Youth Impact
Claire Cullen
,
Oxford University and Youth Impact
Thato Letsomo
,
Youth Impact
Moitshepi Matsheng
,
Youth Impact
Rene Marlon Panti
,
IPA
Shwetlena Sabarwal
,
World Bank
Tim Sullivan
,
NewGlobe

Abstract

Even before the COVID-19 pandemic, more than 2 billion people lived in countries affected by conflict, climate shocks, or disease. During these emergencies, schools often close, necessitating interventions that can complement or substitute traditional schooling. Ideally, these interventions can promote learning during a crisis, scale rapidly, induce high levels of engagement across different geographies and models, and personalize instruction to children of varying levels and cultural backgrounds. However, rigorous evaluation in emergency humanitarian settings is challenging and rare, especially across multiple countries. We present results from large-scale randomized trials across five countries (India, Kenya, Nepal, Philippines, and Uganda) and 16,000 students working with NGOs and governments delivering phone-based, remote instruction to children and their caregivers during COVID-19. We find high engagement across sites. Despite highly heterogeneous contexts, we find consistently large and robust effect sizes on learning ranging from 0.10-0.40 standard deviations. In a subset of trials, we randomized whether the intervention was provided by NGO instructors or government teachers. The results show similar effects, indicating scalability within government systems.

Apart but Connected: Online Tutoring and Student Outcomes

Michela Carlana
,
Harvard University
Eliana La Ferrara
,
Harvard Kennedy School

Abstract

The demand and supply of online tutoring has surged in response to the COVID-19 outbreak to address learning gaps. This paper evaluates the effectiveness of an Tutoring Online Program (TOP) implemented in the Spring 2020 during the COVID-19 outbreak and then replicated in the Spring 2022. University students volunteered to help disadvantaged pupil: they were randomly assigned to middle school pupils, from a list of potential beneficiaries compiled by Italian school principals. Using original survey data collected from students, parents, teachers and tutors, as well as administrative data, we find that the program substantially increased students' academic performance (by 0.27 and 0.21 SD on average in 2020 and 2022, respectively). TOP significantly improved students' socio-emotional skills, aspirations, and psychological well-being, only when implemented during school closure. The intervention had a positive impact on tutors enhancing their measured empathy.

Behavioral Nudges to Prevent Learning Losses and Dropout Risk during the Pandemic: Experimental Evidence from Brazil

Guilherme Lichand
,
University of Zurich
Julien Christen
,
University of Zurich
Eppie van Egeraat
,
University of Zurich

Abstract

Using a cluster-randomized control trial with 18,256 high-school students across 87 schools in the State of Goiás, Brazil, this paper documents that behavioral nudges, sent through text messages to students or their caregivers during remote learning, significantly increased proficiency levels in standardized assessments conducted with high-school seniors in the following year (a 0.19 s.d. increase in the summary measure of math and Portuguese standardized test scores; 90% CI: [0.03,0.35]), partially mitigating dramatic learning losses in the context of COVID-19. Impacts were positive across the entire test score distribution, but nudges increased inequality in test scores. Impacts on dropout risk were not statistically significant, although there was substantial heterogeneity in treatment effects – significantly higher for the students most at risk of abandoning school. Additional experiments that varied the content of nudges at the student level illustrate the opportunities and challenges of using behavioral insights to motivate students in the context of remote learning.

Telementoring and Homeschooling During School Closures: A Randomized Experiment in Rural Bangladesh

Abu Siddique
,
King's College London
Asad Islam
,
Monash University
Hashibul Hassan
,
Monash University
Liang Choon Wang
,
Monash University

Abstract

Using a randomized experiment in 200 Bangladeshi villages, we evaluate the impact of an over-the-phone learning support intervention (telementoring) among primary school children and their mothers during Covid-19 school closures. Following the intervention, treated children scored 35% higher on a standardized test, and the homeschooling involvement of treated mothers increased by 22 minutes per day (26%). We returned to the participants one year later, after schools briefly reopened, and find that impacts on learning gains and homeschooling had persisted. Academically weaker children benefited the most from the intervention that only cost $20 per child.

Discussant(s)
Dave Evans
,
Center for Global Development
Philip Oreopoulos
,
University of Toronto
Paul Glewwe
,
University of Minnesota
Shwetlena Sabarwal
,
World Bank
JEL Classifications
  • I0 - General
  • O0 - General