Participatory governance of artificial intelligence (AI) agriculture models routinely seats smallholder representatives on advisory councils alongside ministries, technology firms and extension agencies, yet the voting power of those seats is rarely analysed as a coalition-formation problem. The analysis characterises coalition-formation stability as a function of seat allocation, voting weights, preference dispersion among smallholder delegates and the defection costs that bind blocs together. Using a before-after-control-impact design logic applied to computed council configurations, the study evaluates how the intraclass correlation of bloc preferences (0.64) and the between-bloc standard deviation (6.0) condition the durability of winning coalitions across council sizes of 8, 16, 24 and 32 seats and effect magnitudes of 3, 5 and 8 percentage points. The framework proposes that smallholder representation stabilises governance only when bloc cohesion exceeds a threshold that seat-count parity alone cannot deliver, and it derives the minimum detectable effect at 80% power for each council size. The contribution is a formal bridge between weighted-voting game theory and the institutional design of AI agriculture advisory councils, specifying the conditions under which participatory representation translates into durable coalition influence rather than symbolic inclusion.
No takes yet. Share an insight, caveat, or question.
Koffi et al. (2026) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: