We spoke with Cristiano Rossignoli from WorldFish and who is based in Florence and Mooghdo Mahzab from the International Food Policy Research Institute (IFPRI) in Dhaka to explore their insights from their participation at the Asian Evaluation Week (AEW) 2024, their work, and their perspectives on CGIAR’s evaluation systems. They highlighted the importance of strengthening Monitoring, Evaluation, Learning, and Impact Assessment (MELIA) processes and shared how evaluations can better align with strategic objectives, inform policy, and foster collaboration across CGIAR centers and country programs. While the lead of CGIAR’s Evaluation Function, Svetlana Negroustoueva, was unable to attend in person, she contributed remotely.
What We've Learned
- Strengthening MELIA: MELIA should be integrated into project design to align objectives with measurable outcomes. Early collaboration between MELIA teams and program managers is crucial.
- Country-Level Evaluations: Evaluations at the country level help ensure CGIAR programs are cohesive, responsive to national priorities, and aligned with strategic goals.
- AI and Evaluation: AI and machine learning present opportunities for improved evaluations but require high-quality data inputs and attention to environmental costs.
- Collaboration Across Centers: Cross-center learning sessions can enhance MELIA systems, promoting collective effectiveness and broader application of evaluation insights.
- Recommendations: Evaluations should focus on aligning research with national priorities and include practical examples and workshops to improve understanding and application.
Background and Relationship with CGIAR
Cristiano Rossignoli: I have been part of CGIAR since 2018. I am an agricultural economist with a PhD from University of Pisa in Italy, and my work has always focused on using science to drive meaningful change. My background in sustainable aquaculture brought me to WorldFish, where I lead the MELIA research work—Monitoring, Evaluation, Learning, and Impact Assessment—and the CGIAR Aquatic Food Initiative. These roles allow me to combine my scientific expertise with evaluation, ensuring that our programs generate impact and learning outcomes. I currently divide my time between Malaysia, where I am based, and Italy.
Moogdho Mahzab: I joined CGIAR more recently, in early 2024, as a development economist with IFPRI. I completed my PhD in Economics at the University of Virginia and a postdoc at Stanford University, in the USA. My work at IFPRI focuses on policy-driven research, particularly food security and agricultural mechanization, with a strong focus on Bangladesh and South Asia. One of our recent major contributions was the completion of a food security and nutrition assessment for Bangladesh, which has had a strong policy influence.
Insights from the Asian Evaluation Week (AEW): Perspectives on Impact Assessment and Evaluation Methodologies
Cristiano Rossignoli: I delivered presentations on the complexities of impact assessment in CGIAR. I emphasized the challenges of assessing long causal chains and the need to consider both intended and unintended consequences of our interventions. I used a case study for this, so I presented the evaluation of the GIFT tilapia program in Bangladesh. Another presentation focused on the sustainability impacts of small-scale aquaculture in Bangladesh, highlighting the importance of evaluating trade-offs and outcomes beyond singular dimensions.
Beyond assessing impacts, strengthening data-driven decision-making is key to maximizing research effectiveness. Real-time data, digital tools, and participatory evaluation methods provide critical insights that inform policymaking and investment decisions. By integrating innovative evaluation methodologies, CGIAR can enhance the scalability and application of research, ensuring that findings translate into actionable strategies. My second presentation at AEW focused on the sustainability impacts of small-scale aquaculture in Bangladesh, emphasizing the need to assess trade-offs and multidimensional outcomes rather than focusing on single impact indicators. These comprehensive evaluations can provide decision-makers with the necessary evidence to guide future investments and interventions.
Moogdho Mahzab: My presentation centered on the Bangladesh Integrated Household Survey de(BIHS), a nationally representative, publicly available dataset that has become a valuable resource for development economists globally. I discussed how BIHS enables rigorous research and policy analysis, contributing to global research efforts. I also shared experiences using natural experiments to address causal questions in development contexts where randomized trials are not feasible.
AEW also provided a valuable opportunity to engage with other CGIAR centers, such as WorldFish, CIMMYT, and IFPRI, fostering discussions on cross-center collaboration. Strengthening partnerships between centers could enhance research synergies and maximize impact by leveraging complementary expertise. At the country level, Bangladesh has been a leader in several CGIAR initiatives, providing a model for knowledge exchange and scaling up successful approaches in other developing regions. Sharing lessons learned from Bangladesh’s experience in research-to-policy translation could help inform best practices for similar contexts worldwide.
Additionally, modern impact evaluation methods were a recurring theme at AEW, with discussions highlighting the limitations of traditional approaches. While conventional evaluation frameworks remain important, there is growing recognition of how RCTs, machine learning, and other advanced techniques can improve the accuracy and effectiveness of impact assessments. As CGIAR advances its evaluation efforts, integrating innovative methodologies will be key to ensuring that impact assessments remain robust, adaptive, and policy-relevant.
AI and Machine Learning in Evaluation: Opportunities and Challenges
Cristiano Rossignoli: Many sessions focused on the use of artificial intelligence (AI) and machine learning (ML) for evaluation. While these tools offer significant opportunities, I highlighted concerns about the importance of data quality. Despite the effort, quality of the data available from evaluation seems affective the AI tools consistently. Ensuring high-quality inputs should be prioritized.
Moogdho Mahzab: I noted the emphasis on AI and ML but raised concerns about the energy costs associated with running these models. While these technologies are being widely adopted, the environmental trade-offs are not sufficiently addressed. Future discussions need to consider the broader sustainability implications.
How can MELIA Systems in CGIAR be Strengthened?
Cristiano Rossignoli: MELIA should be integrated into the project design phase rather than being addressed later as an isolated activity. By embedding evaluation processes at the design and inception of programs, we can better align objectives with measurable outcomes. Increased collaboration between MELIA teams and program managers is necessary to achieve this.
Moogdho Mahzab: Cross-center learning sessions can play a significant role in strengthening MELIA systems. Regular sharing of experiences, challenges, and solutions among centers would enhance the collective effectiveness of MELIA processes. I believe that greater collaboration would ensure that insights from evaluations are applied more broadly.
Country-Level Evaluations and Strategies
Cristiano Rossignoli: Country level evaluations, covering both of impact and of performance and process, would be crucial for aligning CGIAR activities across centers. This approach would ensures that programs are strategically cohesive and responsive to national priorities, ultimately enhancing their effectiveness.
By fostering joint learning from both successes and challenges, this approach can strengthen accountability to funders and drive more results-oriented decision-making. At the same time, it is essential to maintain center-level autonomy to preserve innovation and flexibility in implementation.
Moogdho Mahzab: Country-level evaluations would improve consistency and allow successful examples, such as those in Bangladesh, to inform work in other regions. A systematic approach is needed to balance alignment with flexibility, ensuring that centers can address country-specific needs effectively.
Engagement with CGIAR’s MELIA Framework
Cristiano Rossignoli: I am familiar with aspects of the CGIAR’s MELIA framework and have reviewed their development and operationalization. Going forward, it would serve as a valuable reference for guiding evaluation processes. However, more efforts are needed to demonstrate its practical application through case studies or interactive sessions. Such efforts would help bridge the gap between high-level frameworks and center-level implementation.
Moogdho Mahzab: I have briefly engaged with the framework through proposal writing and initial discussions at IFPRI. While the framework provides essential guidance, practical examples and targeted workshops would help improve understanding and adoption. This would be particularly useful for scientists working across centers with varying evaluation priorities.
Mechanisms to Strengthen Engagement with Evaluation and Recommendations
Cristiano Rossignoli: Enhancing engagement with evaluation requires embedding MELIA processes from the outset of project development. Early collaboration between MELIA teams, program managers, and evaluators—particularly during proposal development and the inception phase—can ensure alignment with evaluation objectives and strategic goals. Additionally, establishing regular review mechanisms will facilitate continuous progress tracking, enable evidence-based decision-making, and support timely course corrections for greater impact.
Moogdho Mahzab: In addition to early integration of evaluation processes, creating platforms in CGIAR for regular knowledge sharing at the country and center levels would be beneficial. Sessions that focus on sharing results, evidence-based challenges, and opportunities could strengthen collaboration and improve programming through enhanced evaluation outcomes.
Cristiano Rossignoli: I have engaged with recommendations from independent evaluations of CGIAR, particularly those addressing the alignment of programs with national strategies. These recommendations are highly relevant and underscore the importance of creating coherence between research activities and country priorities. I also support recommendations emphasizing the realistic timelines required for achieving research outcomes. I have not been involved in developing Management response actions, which could further clarify the role of Centers in their implementation.
Moogdho Mahzab: I have reviewed the recommendations to SG evaluations and found several areas of alignment with ongoing work at IFPRI. While we have not yet had detailed discussions on all recommendations, their relevance to country programs, particularly in Bangladesh, is clear. A more structured synthesis of recommendations applicable at the center level would strengthen their integration and facilitate the implementation of relevant actions.
Final Recommendations
Cristiano Rossignoli: Evaluations at the country strategy level are essential for creating a unified approach while maintaining the flexibility required for effective implementation. Greater integration of MELIA into project design and increased collaboration across centers will strengthen CGIAR’s impact assessment efforts.
Moogdho Mahzab: Strengthening evaluations requires systematic planning, knowledge sharing, and alignment with national priorities, and internal coherence within CGIAR as well. Country-level evaluations and collaborative platforms would ensure that lessons learned are applied effectively across regions, enhancing CGIAR’s contributions to development outcomes.
About Cristiano Rossignoli
Dr. Cristiano Rossignoli is an Agricultural Economist with a PhD in Agricultural Economics (Animal Health Economics) and an MSc in Animal Production Science from the University of Pisa (Italy), and an MSc in Sustainable Aquaculture from Stirling University (UK). With over 15 years of experience, he has expertise in Impact Assessment and M&E, working on international projects aimed at enhancing livelihoods, nutrition, and food security. Cristiano has held key positions at the University of Pisa, FAO, and Oxfam, and has contributed to food security projects in countries like Afghanistan and Thailand. Cristiano leads Africa-Asia BluTech Superhighway Project, the CGIAR Initiative on Aquatic Foods and acts as the MEL and Impact Assessment Research Lead for WorldFish.
About Dr. Moogdho Mahzab
Dr. Moogdho Mahzab is an Associate Research Fellow with the Poverty, Gender, and Inclusion Unit, based in Dhaka, Bangladesh. He is a development economist with research interests in environment, health, and political economy. Prior to joining IFPRI, he was a Postdoctoral Fellow at Stanford Graduate Business School and Stanford Woods Institute for the Environment, Stanford University. He received his PhD in Economics from the University of Viginia. He also received the U.S. Fulbright Scholarship from Bangladesh to pursue a master’s in Policy Economics at Williams College.
