A data science lab to foster public sector innovation and combat poverty

Advances in data science and artificial intelligence are creating new opportunities to improve decision-making and governance. However, their adoption remains limited, particularly due to underinvestment in data infrastructure and reliance on outdated or incomplete data sources.

Across Africa, digital transformation is steadily progressing. In Togo, internet usage rose from 4% in 2011 to 35% in 2021 (World Bank, 2021). Institutions such as the Togo Digital Agency, to which the Togo Data Lab is affiliated, and the integration of digital components in the majority of development projects, reflect the country’s growing commitment to digitalization and data use.

This context presents a unique opportunity to strengthen the role of data science in public service delivery and governance, enabling more informed and effective decision-making processes.

In collaboration with CEGA and the Environmental Markets Lab, the Togo Data Lab is addressing existing gaps through the deployment of the MOSAIKS system (Multi-task Observation using Satellite Imagery & Kitchen Sinks).

The project is structured around three key pillars:

• Program targeting: Identifying needs to improve resource allocation. • Policy support: Providing data to inform decisions on economic opportunities, service access, and living conditions. • Monitoring outcomes: Supporting ongoing evaluations to enhance program effectiveness.

The first application of MOSAIKS will focus on agricultural productivity. This pilot will allow ministry staff to gain hands-on experience with the system. Progress will be tracked through milestones such as team training, tool development, and stakeholder engagement.

FID funding will enable the recruitment of key data professionals, the development of open-source training modules, and the design of predictive models using agricultural data. The initiative is expected to build the capacities of public servants, promote data-driven decision-making, and position the Togo Data Lab as a central actor in public sector innovation.

The project team began by conducting a capacity assessment of ANID’s data collection system, including its structure, resources, and processes. This assessment identified the strengths and weaknesses of the current system for data collection, processing, and reporting, and of the organization’s governance system. Following this evaluation, IPA made two key recommendations for the organization to:

• construct a data collection management system;
• establish a proactive plan for data use.

According to the results, ANID and IPA have jointly created a (preliminary) structure for the laboratory, including its recruitment process, key areas of action, and an institutionalization mechanism. The partner organizations have also laid out plans to establish a network of collaborators and raise funding for the laboratory’s activities.

In addition, they have developed a research and learning agenda to guide priority research work for finance and social protection policy. The team has taken initial steps to ensure the ministry adopts the research agenda.

A capacity-building plan has also been drafted, identifying specific training needs for future laboratory staff. Targeted training programs will be delivered on subjects such as data analysis, the use of impact evidence, and technical tools. With regard to data security, which is essential for the project’s long-term future, ANID has developed a detailed protocol to regulate access and control.

The partner organizations have highlighted the need to build capacity in data visualization, cybersecurity, and integration of data into decision-making processes to improve the country’s use of evidence-based data.

Future prospects

By prioritizing capacity-building activities, including the training provided to ANIL staff by J-Pal on impact assessment for social programs, the teams were able to have constructive and informed discussions on the laboratory’s structure and operations and gain a better understanding of the issues at stake.

In the short term, ANID, which will host the laboratory, plans to conduct initial impact assessments on topics related to cash transfers and social protection. Over the long term, ANID wants the laboratory to become a key instrument in the production and dissemination of evidence-based data, thus facilitating the design of current and future social protection and cash transfer programs, as well as informing political decision-making at the highest level.