BE² Blog series – Equipping the government to use data and evidence through Embedded Evidence Labs

Equipping the government to use data and evidence through Embedded Evidence Labs

Governments are key actors in tackling the global education crisis (Piper, Benveniste, and Angrist, 2025). They design and implement policies and programs at scale that impact millions of lives and involve billions of dollars in public spending. This gives them a unique opportunity to improve people’s lives by ensuring that both the design and large-scale implementation of these policies are informed by data and evidence. Yet despite this potential, many governments struggle to systematically integrate evidence and data into their policymaking processes. Common challenges include:

• Monitoring and evaluation efforts by policymakers that focus more on accountability than on future policy improvement.
• Limited technical capacity, with government staff lacking the technical knowledge and skills for evidence use.
• Weak information systems that fail to provide timely and relevant data.
• Institutional cultures shaped by political objectives that offer few incentives to incorporate evidence (Dercon, 2025 [1]).

The Uganda Ministry of Education and Sports Embedded Lab Team, and the IPA Philippines Embedded Lab Team for the Department of Education of the Philippines exchange on their Lab design, successes, and challenges. © IPA 2024

Embedded Evidence Labs in Education

Over the last years, IPA has worked with governments to overcome these challenges by deploying Embedded Evidence Labs (Embedded Labs) [2].

An embedded Lab is a dedicated team embedded within government that works to align sector priorities with data and evidence, supporting both decision-making and implementation. These Labs follow a three-stage cycle to achieve impact on public policy:

1. Policy Challenge Mapping: The Lab assists in identifying the main issues within its sector and translating them into learning and research questions.
2. Data and Evidence Generation: The Lab launches learning and research strategies tailored to the nature of the questions, such as descriptive analysis, predictive analytics, impact evaluations, or Monitoring, Evaluation, and Learning (MEL) systems for effective implementation.
3. Applying Data and Evidence to Policy: The Lab connects results with the right people at the right time and assists decision-makers in using evidence to improve policy.

Figure 1. Embedded Evidence Lab learning cycle

Embedded Labs employ diverse strategies for navigating complex evidence-use environments

Embedded Labs operate within complex environments involving numerous actors operating at various levels of the education system, each with distinct incentives and technical capabilities, e.g., national governments, provincial administrations, school leaders, etc. These actors are embedded within organizations characterized by evolving cultures, processes, priorities, and resources. Moreover, decision-making is not a linear process; it is often messy, political, and shaped by competing factors, e.g., labor market needs, election priorities, difficult economic environments, ideology. To effectively navigate this complexity, Labs employ a combination of three complementary approaches to evidence use: linear, relational, and systems-based (Best and Holmes, 2010) [3].

From a linear perspective, Embedded Labs co-design research agendas and research designs with system stakeholders (e.g. government units that design and implement policy), implement evidence generation and learning strategies, and carry out strategic dissemination through user-friendly and actionable formats (e.g., policy briefs) and diverse communication channels (e.g., websites or workshops).

This approach is enhanced through relational strategies, in which Embedded Labs build partnerships with key actors involved in evidence generation and use, supporting them in developing the skills and knowledge needed for effective engagement. For example, for a specific project, Labs create collaborations between academic researchers and education policymakers and implement capacity-strengthening activities, such as training in evidence-generation methodologies, ensuring that partnerships are well-equipped to co-design, decide on, and implement evidence-informed policies. These projects are co-designed, focusing on policy-relevant questions identified with the policymakers and incorporating their knowledge and goals into both the design and implementation.

Additionally, Labs adopt a systems approach that acknowledges the organizational context in which evidence is generated and used. Recognizing that these actors function within institutional frameworks, Labs address organizational conditions necessary for implementing evidence-informed policy. Being teams within the government allows them to understand not only the government’s processes and timelines but also the power relations and narratives that influence decision-making. For example, when scaling an evidence-informed intervention, e.g., teacher training, a Lab team may support the development or enhancement of government agency information systems to enable monitoring of policy implementation, thereby ensuring fidelity and effectiveness at scale. They may also help the government units advocate for political support and changes in legal frameworks or operation manuals among other activities.

—————————————————————————————————————————

Box 1. Case study – the Rwanda Education Embedded Lab

An example of how Labs combine linear, relational, and systemic approaches is the case of the Rwanda Education Embedded Lab. This Lab operates under the Ministry of Education (MINEDUC), in close coordination with the National Examination and School Inspection Authority (NESA) and the Rwanda Education Board (REB).

The Lab is supporting the scale-up of performance-based teacher contracts. The program, which was rigorously evaluated in collaboration with researchers from Georgetown University and IPA, showed improvements in both teacher performance and student learning. Following these positive results, the Ministry asked the Lab to scale up the program while testing different delivery methods to identify the most effective approach, using randomized controlled trials (RCTs) – a key example of a linear approach to evidence uptake.

At the same time, the Lab applies relational and systems approaches. For example, the systems approach focuses on developing information systems such as the Comprehensive Assessment Management Information System (CAMIS) and the Teacher Management Information System (TMIS), as well as by building capabilities to ensure quality implementation at scale. The relational approach includes coordinating across government agencies—such as the Rwanda Education Board and the National Examination and School Inspection Authority (NESA)—and advocating to maintain the prioritization of scaling efforts at all levels within these agencies. Together, these approaches resulted in the incentive scheme reaching 4,762 teachers and 207,523 primary school students by 2023.

—————————————————————————————————————————–

How to embed Evidence Labs to effectively impact education policy and systems

Promoting the adoption and implementation of evidence-informed policy may require addressing systemic factors, as illustrated by the example of Rwanda. When establishing a Lab within a government agency, it inherently involves making changes across several dimensions of the organization to ensure the Lab’s effectiveness and long-term sustainability. To achieve this, IPA aims to institutionalize Labs by embedding them as an integral part of the agency’s processes and structures. This is done through a co-design process with the government, focused on four key dimensions:

• Process: Define the learning cycle that links policy challenges with data, evidence, and decision-making processes.
• Structure and Governance: Establish the Lab’s location and organizational setup, and determine how it will interact with other technical and political stakeholders to ensure alignment and effectiveness.
• Capacity: Design and implement a strategy to strengthen system capacity. This includes enhancing the skills and knowledge of individuals—both within the Lab and across departments that engage with it—as well as improving the functionality and integration of information systems.
• Resources: Develop and execute strategies for securing and allocating financial resources to sustain Lab operations and personnel. This also involves reviewing legal frameworks to formally integrate the Lab within the agency’s structure and deploying approaches to ensure political and ecosystem-wide support. Additionally, Labs work to shape organizational culture by promoting the value of evidence use and cultivating demand for evidence-informed practices.

—————————————————————————————————————————–

Box 2. Case study – MineduLab, the Innovation Lab of the Ministry of Education of Peru

MineduLab was officially launched by Minister of Education Jaime Saavedra in 2016. Since then, the Lab has lasted through seven presidential transitions and sixteen changes of ministers of education, demonstrating the success of its institutionalization efforts.

To institutionalize the Lab, the Ministry—supported by IPA and J-PAL—embedded a team of impact evaluation experts within the Office of Monitoring and Evaluation (OSEE, by its Spanish acronym). Through annual calls for proposals, this team systematically builds partnerships with researchers from local and international universities, as well as with implementing units within the Ministry, to propose innovations that are evaluated using randomized controlled trials (RCTs).

Additionally, a legal framework was created that granted the Lab an official mandate as an innovation unit. This framework consists of a directive that outlines the Lab’s structure and functions in a way that preserves its original purpose, while also allowing it to access the Ministry’s budget to fund its team.

Consequently, the Labs not only influence policy by integrating data and evidence into decision-making, but also have a systemic impact, as they modify the processes, structures, capacities, and resources of a government agency, facilitating the institutionalization of data and evidence use. This dual impact is illustrated by the theory of change for the Embedded Evidence Labs.

—————————————————————————————————————————–

Figure 2. Embedded Evidence Lab Theory of Change

Embedded Evidence Labs show a clear path for governments to successfully integrate evidence in their policies and implementation. In sum, Embedded Evidence Labs are a strategy that enables governments to incorporate data and evidence into policy design and supports the scaling and effective implementation of evidence-informed policies at the national level. By integrating a linear, relational and system approach, Embedded Labs are well-positioned to address various barriers to evidence use, fostering an environment where evidence-informed policymaking becomes a routine and institutionalized practice within the government agency. In a global context of scarce resources, this strategy makes it possible to leverage existing resources through the institutional strengthening of governments, enabling them to improve education systems for lasting social impact.

Author and contact information

Authors:

Juan M. Hernández-Agramonte (Senior Director Embedded Labs, IPA);

jmhernandez@poverty-action.org

William Sims (Senior Manager Embedded Labs, IPA).

wsims@poverty-action.org

At IPA, we support over 24 Labs across 14 different countries and 8 sectors – including 10 Labs in education. To achieve this, we provide technical assistance to these government teams to establish Labs and strengthen their strategies, identifying lessons from different contexts, and promoting exchanges between Labs to foster peer learning and the adoption of best practices. By supporting these networks of government officials with the ability to influence their institutions, we help drive institutional changes that embed data and evidence into decision-making—ultimately improving education policy and student learning outcomes.

About this blog series

This blog is part of a blog series: “How can we improve the use of evidence for policy, planning and implementation in education”? The Blog Series is hosted by Building Evidence in Education (BE²) and Global Education Evidence Advisory Panel Secretariat. 

This blog series will cover identified trends and practices that increase evidence use in education. Where available, blogs will also discuss the impact that increased evidence use has.

The content of this blog represents the opinions of the authors, and do not necessarily reflect the opinions of BE² or GEEAP.