What predicts violence in Iraq?
Machine-learning techniques can be effective in identifying areas with higher risk of future violence.
Working with Mercy Corps, the team built a model to predict future violence in Iraq, and identify the variables that are most salient in predicting violence. They find that previous exposure to violence is the most important predictor, and can be a useful rule-of-thumb for stakeholders with limited access to resources. These methods can help humanitarian agencies target their data-collection and improve prediction accuracy.
Delany Brown, KC Brown, Claire Cullen, and Jesus Gallegos Segura
Understanding the improvements in education in Tanzania
Despite remarkable improvements, education service delivery in Tanzania remains low.
Working with the World Bank, the team analyzed Tanzania’s Service Delivery Indicators in education. They find that the quality of education service delivery has improved by a number of measures across Tanzania between 2010 and 2014: Pupils are performing better in tests and teachers are more likely to be in school and teaching. But these changes were not consistent across different contexts within the country. The trends in urban and rural schools have been strikingly different in many cases, as have been the trends between schools located in poor and non-poor regions.
Tristan Lewis, Tasmia Rahman, and Adam Towbridge
Do School Vouchers Improve Learning Outcomes?
Providing vouchers to attend private schools in New Delhi have varied impacts on student performance.
School vouchers have become an increasingly popular tool for increase access to quality education among low-income students. The team reports results of a 5-year randomized controlled trial of a voucher program conducted in New Delhi, India. They find that vouchers have varied impacts on student performance in different subjects: pupils who received the voucher do worse in Hindi, but better in English. Future programs should assure higher-quality private schools are included, and steps should be taken to reduce drop-outs of minority groups from the program.
Filiz Akdas, Lucy Cutting, Anchidtha Roonguthai, and Luis Vicente Sotelo
The Role of Private Pre-Schools in Nigeria and Tanzania
Pre-primary enrollment can be improved by strengthening the private sector.
Enrollment in pre-primary schools in Nigeria and Tanzania falls far below government’s objectives. Working with the World Bank, the team deeper unpacks enrollment trends in these countries, and analyzes a unique data-set of pre-primary providers. The find that the majority of pupils are enrolled in private schools, especially in Nigeria. The poor are less likely to be enrolled in Tanzania. Private schools are far more cost-effective, yet their price remains a prohibitive barrier for the poorest. Nigeria has a high supply of low-cost private providers, but few schools are registered and there is limited oversight of schools. Future government policies can leverage the cost-advantages of private pre-primary provides, whilst assuring high quality standards.
Robert Dean, Sunju Lee, and Ingrid Rojas
How to increase technology adoption in agriculture?
Providing free access to technology can increase adoption in the long run, but price remains a barrier.
The Africa Research in Sustainable Intensification for the Next Generation (Africa RISING), provides small-scale farmers with access to improved agricultural technology: Seeds, fertilizer and farming techniques. Working with the International Food Policy Research Institute (IFRPI), the team evaluated the program and found that use of technology improved due to the program, and the willingness to pay for these technologies also improved over the long run. Use of fertilizer remains low, however, due to prohibitively high market prices.
Alejandra Aponte, Alejandra Arrieta, Rohit Chhabra, and Asad Zaman
Can Partial Subsidies Increase Latrine Ownership in Cambodia?
Subsidies to the poor increase sales, but overall latrine access remains low.
Poor sanitation is one of the largest contributors to child mortality in the developing world. While many factors contribute to low latrine uptake, one major constraint is price. Working with Causal Design, this team evaluated the impact of providing subsidies to poor households, using a randomized control trial. With baseline latrine coverage at only 23% among poor households, the team found that latrine purchase of poor households increases by roughly 2 percent. The impact is larger in villages with more poor households. However, there was no positive impact of purchases to the non-poor.
Bheeshm Chaudhary, Erin Crossett, Lauren McCaw, Cailin Rowe