Ongoing Projects

Machine Learning for Peace 

The Machine Learning for Peace (mlpeace.org) project seeks to measure, understand, and predict how civic space is changing in countries around the world. The global trend toward autocratization has brought new attention to the study of political regimes and transitions. Policymakers and practitioners now rely heavily on quantitative data to understand changing political conditions and design effective policies and interventions. Existing data measure these changes annually. As a result, these data are not able to identify sequences of events that could predict changes that will occur over weeks or months rather than years or decades. However, civic space closures often happens abruptly, as governments seize on crises as opportunities to restrict fundamental rights or expand executive authority. Our approach conceptualizes these changes as a the result of specific actions that happen at specific moments in time.

Working with partners in the INSPIRES consortium, MLP produces publicly available, high-frequency data on the occurrence of 42 types of political events in 50 countries and uses these data to forecast future shifts. We accelerate the provision of data to stakeholders by continuously scraping online news published by international, regional, and domestic news sources; using recent advances in Natural Language Processing to measure reporting on civic space events in near-real-time; and using this information to generate accurate monthly forecasts of future trends in civic space. To date, MLP has scraped and processed more than 90 million news stories from more than 300 online news sources in more than 30 languages. Ultimately, we hope our approach will be a useful tool for researchers seeking rich, high-frequency data on political regimes and for policymakers and activists fighting to defend democracy around the world.

Central American Regional Media Project (ReMedios)

Under the ReMedios activity, DevLab is designing and implementing an evaluation of USAID programming meant to strengthen the capacity and security of journalists and media outlets working on issues related to corruption and government transparency. This project will involve a three-wave panel survey of journalists across five countries in Central America, scraping and machine classification of online news, and training a large language model to identify markers of quality in anti-corruption reporting.

Zimbabwe Governance Indicators Analysis

Under the Zimbabwe Governance Indicators Analysis activity, DevLab is developing a rigorous methodology that will be used by US government agencies to evaluate the Government of Zimbabwe’s progress towards its democratic reform commitments.

Local Evaluation and Evidence Support (LEES)

The Local Evaluation and Evidence Support (LEES) activity supports efforts by the Bureau for Policy, Planning, and Learning (PPL) to promote localization of USAID activities by increasing the capacity of local partners and strengthening networks that connect these organizations.