Food systems play an important role in shaping our diets. Recently, there has been growing attention to food systems transformation, which involves changing different components of global food systems to make nutritious foods more accessible to consumers. At the 2021 UN Food Systems Summit, more than 100 countries pledged to transform their food systems, a clear recognition of the importance of healthy diets in preventing all forms of malnutrition worldwide.

While the potential of food systems transformation is promising, the process is complex and unpredictable. To ensure interventions produce intended outcomes and contribute meaningfully to food systems transformation agendas, strong evidence on best practices is needed. However, due to the complexity of food systems and limited guidance on how to conduct rigorous evaluations, evidence is only slowly starting to emerge.

To help address this knowledge gap, a team of researchers from the Global Alliance for Improved Nutrition (GAIN), IFPRI, the UN Food and Agriculture Organization (FAO), and other institutions synthesized insights from program evaluations of six food systems interventions implemented in Africa and South Asia. The interventions aimed to increase access to nutritious foods by influencing both supply and demand. Some programs targeted specific age groups, while others focused on micro, small, and medium enterprises (MSMEs). The synthesis, published in the Journal of Nutrition, revealed key challenges and formulated recommendations for the design and evaluation of future food systems interventions. 

Methodological challenges with evaluating food systems interventions

Evaluating the impact of food systems interventions presents several methodological challenges. One notable obstacle is the limited evidence on the processes through which food systems interventions are expected to achieve desired impacts. This means that researchers need to make many assumptions when developing the theory of change, which is the roadmap that explains how various activities and inputs within an intervention are expected to bring about change.

Since food systems are dynamic, interventions intended to influence them may require adjustments throughout the implementation period. The interconnected nature of food systems means that changes at any level within a food system can influence an intervention’s effectiveness. These food system characteristics create challenges with respect to clearly defining the intervention and identifying the outcomes that need to be assessed.

Another consequence of the interconnectedness of food systems is the difficulty of isolating the impact of intervention outcomes from external factors. Traditional experimental designs, where participants are randomly assigned to a treatment or a control group, are often not feasible for food systems interventions. The inability to control or manipulate different variables within a food system makes it hard to determine what would have happened in the absence of an intervention.

Data acquisition presents another hurdle. Interventions generally involve private-sector companies. Obtaining timely and accurate data can be a significant obstacle, whether due to poor recordkeeping or proprietary information. Another data-related challenge relates to tracking the way food travels through the food system, as many points in the supply chain are untraceable.

Finally, a single outcome—or a limited number of outcomes—rarely suffices to reflect the results of an intervention. Determining the most suitable outcomes for measuring impacts is thus a challenge. Focusing only on nutritional outcomes and neglecting other factors like food affordability, accessibility, and consumer behavior does not account for the complexity of food systems interventions, leading to an incomplete understanding of intervention effectiveness.

Recommendations for food systems interventions

To help address these challenges, the synthesis proposed several recommendations for conducting rigorous evaluations:

1. Strong theory of change: Evaluations should always be guided by a strong theory of change, developed in close collaboration with the team implementing the intervention. 
2. Range of outcomes: A range of outcomes should be assessed to understand the full impact of interventions at multiple stages. Outcomes should be chosen to comprehensively capture trade-offs, synergies, and unintended consequences.
3. Combination of methods: To assess a wide range of outcomes, a combination of evaluation methods should be used. This includes quantitative and qualitative methods with triangulation across them.
4. Diverse and adaptive methods: Evaluators should consider methods that are not traditionally used in the evaluation of nutrition programs. These include contribution analysis which determines how an intervention may have contributed to the result based on a careful analysis of changes along the theory of change.
5. Adaptive and flexible design: Adaptability and flexibility are needed to accommodate and capture changes during the implementation phase.  Documenting such adaptations are fundamental for maintaining evaluation rigor.
6. Transparency is essential: Research protocols, which include details on the research questions and methods, should be registered before the evaluation begins. Likewise, changes in the program and subsequent adaptations to the evaluation methods need to be thoroughly documented and explained.

By acknowledging the challenges and implementing the recommendations outlined above, researchers can design and conduct more rigorous evaluations of food systems interventions. Collaborating across disciplines, and with implementing teams, is essential to develop and implement innovative methods for capturing complex outcomes. This will ultimately lead to a stronger body of evidence, guiding decision-makers to effectively transform food systems to deliver healthy diets for all.

Jef Leroy is a Senior Research Fellow in IFPRI’s Nutrition, Diets, and Health Unit; Lynnette Neufeld is the Director of the Food and Nutrition Division (ESN) at the Food and Agriculture Organization of the United Nations (FAO); Stella Nordhagen is a Senior Technical Specialist with the Global Alliance for Improved Nutrition (GAIN); Sydney Honeycutt is Communications Consultant.