Skip to content

Enhancing evidence use in government social programs

Briefing rural communities on the advantages of the Unified Beneficiary Registry in Malawi. Photo shared by Matthias Rompel. November 2021

Government leaders are often required to make many high-stakes decisions every day. In an ideal world, all these decisions would be based on evidence but there are a number of real-world constraints that often get in the way of using data for decision-making. IDinsight’s embedded learning partnerships provide timely, flexible, and responsive support to government leaders by answering important, urgent questions and helping to build a culture of and system for data and evidence use.

In this blog, I share with you my experience working with the government of Malawi through a multi-year Learning Partnership where IDinsight was embedded within the Ministry of Gender, Children, Disability, and Social Welfare, to enhance the social impact of several government programs.

First, here is the context within which government learning partnerships operate:

  • Evolving government priorities: Government needs and interests are dynamic and always evolving. Issues are given pre-eminence and focus based on political priorities and the interests of different supporting partners and stakeholders. As we began our partnership with the Malawi Government, our Theory of Change had to be as broad. We needed to build enough buffer in our implementation plans and budgets to accommodate any unforeseen changes.
  • A complex stakeholder ecosystem: More than one ministry and/or agency could have a stake in making decisions about the programs being supported. In our case, multiple donors/funders had varying goals and interests, civil society organizations were pushing for a certain level of accountability, and of course, bureaucrats within the Ministries needed to respond to political needs. We faced significant delays in decision-making as officials were obliged to get consensus before taking action.
  • Stretched government staff: The government staff have many competing urgent tasks which limits their capacity to dedicate time to certain streams of work. This leads to changing timelines and priorities that shift more often.

From implementing the Malawi Learning Partnership, we discovered five things that can set you up for success working in a long-term partnership with governments. Here are our top takeaways:

1. Context is key

Proposed solutions must be grounded in local realities. There isn’t a “one-size-fits-all” approach. IDinsight’s engagement was made more effective by:

  • Having a local team. Having teammates of Malawian/African origin at all levels – including leadership – resulted in a deep understanding of the local stakeholder landscape and policy environment and made it easier to build trust, share concerns, and come up with context-relevant recommendations.
  • Setting expectations. Spending a reasonable amount of time to understand the government’s motivations – what they cared about, their vision, their needs – and jointly defining what success would look like for them before beginning the partnership enabled us to stay on track and provide relevant, timely services and support.
  • Involving the entire program ecosystem. Acquiring input and buy-in from other influential partners that work closely with the government helped IDinsight provide more relevant and practical recommendations. Providing additional evidence and context to program partners as well as informal advice and light touch consulting built trust and provided an avenue for seamless partnerships. It also created external accountability for decisions and ideas that the government committed to implementing. In some instances, partners were able to pool additional financial and technical resources to enable the success of impactful ideas or recommendations that had no initial funding. 
  • Being embedded. Working from the Ministry premises enabled us to not only have access to decision-makers but also to act as thought partners for both formal and informal decisions. This helped to build trust and camaraderie between our team and other relevant stakeholders.  
  • Adapting our communication channels. We realized that better communication occurs via informal ways including Whatsapp messages and corridor conversations, with instant follow-up emails to document and formalize what you discuss or agree on.
2. Evidence alone is not enough to influence decisions

Governments often operate under resource and political constraints. It is important to understand the political economy, the influence of funding institutions, and the strategic direction of your government partner (especially in the face of national leadership changes). Sometimes, subtle advocacy is required for other stakeholders to help drive the impact forward. For instance, after we worked with the government to pilot and evaluate a CashPlus Linkages initiative – providing add-on social services at cash transfer collection points – in one district (Thyolo), we presented our results to all stakeholders involved in the Social Cash Transfer Program. We saw an increased interest among funding organizations which led to the rollout of this impactful Linkages Project to two additional districts, reaching a further 24,000 households.

3. Re-considering measures of success: small wins aggregate into impact

A Learning Partnership is unique in the way progress is experienced. Unlike projects with a fixed scope and timeline, pathways to impact in long-term flexible Learning Partnerships evolve and more often than not, are multi-dimensional. The success of a Learning Partnership should therefore be measured beyond reach and ultimate decisions influenced. The time it takes to get ideas and recommendations considered, accepted, incorporated, and applied is usually much longer, but once ideas are adopted and owned by the government, the impact follows quickly. Small and soft successes along the way compound to improve processes and systems. The Malawi Learning Partnership required several years to show meaningful results in system improvements and policy change and of course complete adoption and use beyond the life of the Learning Partnership. To keep the team motivated, we celebrated, documented, and reflected on the small wins regularly. We realized the learning partnership had achieved intended and unintended effects and considered the progress as part of the impact.

4. Significant impact arises from going beyond influencing policies and supporting delivery 

For lasting impact in a Learning Partnership, there is value in generating evidence for policy making. However, a higher impact can be achieved if support is extended into program implementation. Policy questions are often longer-term. Policy implementation, though, can be influenced in the shorter term in a more significant way. For example, as the government was figuring out Covid-19 policy responses and guidance, IDinsight’s technical support on the Covid-19 Urban Cash Initiative helped the government to refine its approach to designing rapid responses to social protection during emergencies by effectively strengthening data quality. Some of the quick wins interventions such as the Light Touch Add-Ons pilot have gone to enhance longer-term policies. The Malawi Learning Partnership managed to strike this kind of balance by contributing to broader policy actions and strengthening the delivery and implementation of programs under the policy.

5. Embedded relationships should be deliberate in transferring skills

As IDinsight implemented its workstreams, there was an emphasis on transferring skills and ownership to the government right away. To be effective, capacity building should be embedded from end to end: from the process of co-creating to generating the evidence to engaging users. Training should be a core part of the Learning Partnership and must be delivered both formally and informally with the understanding that you will hand over ownership and support. This has been effective in the Malawi Learning Partnership where the Department of Planning and Research (DPR) and IDinsight worked alongside each other to design rapid assessments during emergencies to help with quick improvements to the Covid-19 Urban Cash Initiative. After a joint, post-distribution study, the DPR took ownership and completed subsequent post-disbursement monitoring (PDM) exercises, but also implement insightful regular (quarterly) PDMs for its normal social cash transfer program.


Understanding how your government partner functions and adapting internal operating procedures to their ever-changing environment is what increases your impact potential and results in building a culture of evidence-based decision-making. Recognizing the broader ecosystem within which your partner operates should also inform your approach and anchor your recommendations in their reality – increasing the probability of implementation, and therefore, enhancing your social impact.


I’m deeply grateful for the contributions, recommendations, and reviews of my colleagues at IDinsight who made this blog post possible. I am especially thankful to Frida Njogu-Ndongwe, Krishna Ramesh, and the IDinsight Strategic Communications team.