Who Decides, Who Benefits, Who Gets Left Behind in Food-System Change?

Participation has become almost obligatory language in agricultural research and food-system transformation. Yet inviting farmers, communities and local institutions into workshops does not necessarily mean they influence the decisions that matter. A 2026 article, "Toward Genuine Co-Production for Just and Sustainable Agri-Food System Transformation," published in Outlook on Agriculture, argues that co-production can reproduce the very inequalities it is supposed to overcome when power, knowledge and benefits remain concentrated in established institutions.

The article is authored by Hanna Ewell, Diego Augusto Menestrey Schwieger, Maiko Nishi, Vimbayi Chimonyo, Kosal Mam, Meshack Nkosinathi Dludlu, Sarah Freed and Roseline Remans, with affiliations spanning institutions including the International Center for Tropical Agriculture, Wageningen University and Research, CIMMYT, WorldFish, Nihon University and the University of Eswatini. Rather than presenting a new statistical dataset, the authors combine literature with illustrative cases to ask a harder question: when does participation actually become shared decision-making?

The Problem Is Not Participation, but Who Still Holds the Power

Co-production has become a dominant idea in Agricultural Research for Development because it promises to make innovation more relevant, legitimate and responsive to local needs. In principle, researchers, policymakers, practitioners and communities work together to define problems, design solutions and shape implementation. In practice, however, the article argues that participation often stops well before real authority is redistributed.

This is the key paradox the authors identify. Communities may be consulted, researchers may collect local knowledge, and projects may report impressive engagement numbers, yet donors and research institutions can still control budgets, timelines, objectives and final decisions. Under those conditions, participation becomes procedural rather than transformative, creating the appearance of inclusion without changing who determines priorities.

The authors therefore push for a different benchmark. Instead of asking whether stakeholders were present, they argue that projects should ask whether participation altered decisions, redistributed influence and changed outcomes. Genuine co-production, in their framing, requires movement from consultation toward co-decision-making, from simply incorporating local knowledge toward treating different knowledge systems as equally legitimate, and from technology adoption toward deeper structural change.

The shift is significant for development institutions because many current funding and evaluation systems reward visible participation more easily than genuine power-sharing. Counting workshops, participants or technology users is straightforward; measuring who set the agenda, who could veto decisions or whose priorities ultimately shaped an intervention is far more difficult. The article suggests that without this deeper scrutiny, co-production risks becoming another development label that sounds progressive while leaving institutional hierarchies largely intact.

Justice Changes the Test: Whose Knowledge Counts, Who Decides, Who Benefits?

The article integrates three dimensions of justice into the assessment of co-production. Procedural justice asks who participates and whether they have meaningful influence. Recognitional justice asks whether different identities, histories and knowledge systems are respected. Distributive justice examines who receives benefits, who bears costs and whether outcomes are shared fairly.

These distinctions expose weaknesses that conventional participation metrics can miss. A farmer may attend every project meeting but still have little say over decisions. Indigenous knowledge may be documented yet treated as secondary to scientific expertise. An innovation may raise productivity overall while disproportionately benefiting farmers who already have more land, labor, capital or market access. In each case, participation occurs, but justice remains incomplete.

The distributive dimension is especially important in agricultural transformation. The authors note that new technologies can reinforce inequality when wealthier farmers are better positioned to adopt them. Drought-tolerant seeds, for example, may deliver larger benefits to farmers who can also afford fertilizer and labor, while poorer households remain excluded by financial or institutional barriers.

The same problem affects women, youth and marginalized groups, who may face additional constraints related to mobility, literacy, social norms or access to resources. The article argues that researchers should stop treating "farmers" or "communities" as homogeneous categories and instead examine how different forms of disadvantage shape both participation and outcomes. That insight is especially relevant in the Global South, where development interventions often operate within deeply unequal social and economic structures.

Real Co-Production Takes Time and Development Systems Are Not Built for It

If genuine co-production requires trust, negotiation and shared authority, then it is inherently slower and more resource-intensive than consultation. The article identifies three enabling conditions as particularly important: institutional flexibility, adequate time and resources, and a genuine willingness to share power. These conditions are often undermined by short funding cycles, rigid project structures and pressure to deliver easily measurable results.

Illustrative cases show what longer-term engagement can make possible. In Lao PDR, an initially broad stakeholder process evolved into a nine-month iterative approach in which a locally recognized core group repeatedly revisited decisions, debated trade-offs and influenced the final management pathway. The authors highlight this not simply as participation, but as an example of community influence becoming binding rather than advisory.

Examples from Zimbabwe and India reinforce the same point from a different angle. In Zimbabwe, Community Dialogues of Elders placed Indigenous Knowledge Systems at the center of agroecological transformation, including local agricultural practices, governance and intergenerational knowledge. In India, a participatory rice-seed initiative built on traditional knowledge and expanded access to climate-resilient local varieties, with participating villages increasing from five to 30 over a decade.

These cases suggest that equitable transformation depends less on holding a consultation than on building an institutional process that allows knowledge and priorities to shape decisions over time. That creates a difficult challenge for donors and research organizations accustomed to fixed deliverables, predetermined theories of change and short reporting cycles. If transformation requires iteration, then funding systems must be designed to tolerate uncertainty rather than treating deviation from initial plans as failure.

The Bigger Reform Is Institutional: Funders, Researchers and Policymakers Must Change Too

Researchers are encouraged to invest in long-term relationships, analyze power explicitly, recognize intersectional inequalities and take greater responsibility for downstream effects of their work. That means moving beyond publication and adoption metrics toward questions of legitimacy, exclusion and unintended consequences.

Funders face an equally important challenge. The authors argue that trust-building, iterative engagement and problem redefinition should be treated as legitimate project costs rather than inefficiencies. They also call for reporting systems that track representation and benefit distribution across social groups instead of relying mainly on aggregate reach, productivity or adoption figures.

Practitioners are urged to allow end users' needs and aspirations to drive the research rather than consulting communities after key choices have already been made. Policymakers, meanwhile, are asked to create flexible frameworks that recognize local knowledge and support marginalized groups, while coordinating reforms across land tenure, labor, climate and other policy domains that influence food-system transitions.

The authors are careful not to present co-production as a universal solution. They explicitly acknowledge that it can be costly, time-consuming and difficult, and that it will not fit every setting. The article is also limited by its design: it is neither a systematic review nor an empirical impact evaluation, so it cannot quantify how much justice-oriented co-production improves outcomes compared with conventional approaches.

The limitation, however, points directly toward the next research frontier. Future studies need to test whether projects that genuinely redistribute decision-making power achieve better outcomes in resilience, adoption, sustainability and equity, and under what institutional conditions. Better indicators are also needed to measure influence, recognition and distribution without reducing them to another checklist.

Agri-food transformation cannot be judged only by whether innovation spreads. It also matters who shaped that innovation, whose knowledge was treated as valid and who captured the benefits. For governments, development agencies and research institutions, that reframes justice from a social safeguard into a core condition of effective transformation.