Background Expansion of agro-industrial approaches has raised concerns over occupational and environmental exposures, for example through intensive agrochemical use. Agricultural production decisions are influenced by assumptions regarding unit-specific criteria of ‘productive efficiency’ (revenues and yields), with limited attention to the association between costs and consequences and broader health determinants (sustainability and health effects). This study applies a population health perspective to investigate the occupational and environmental consequences of production options by incorporating a comprehensive range of criteria. Specifically we investigate, in partnership with producers: what is the ”best” agricultural production method for producing bananas in Ecuador and blueberries in BC?

Methods Two MCDAs per jurisdiction (Ecuador and BC) are used to calculate aggregate scores to rank production methods (agro-industrial, agro-ecological, and mixed-methods). The first MCDA is an ‘actual’ model, representing real-world decisions (constrained producer choices). The second MCDA is a ‘preference’ model representing no constraints (producer preferences). Additionally, discrete choice modelling is used to simulate hypothetical scenarios of components (e.g. policy instruments) that would sway producers towards their preferences, with sensitivity analyses to consider the implications.

Results If agro-industrial production is not the highest rank, a case can be made for more sustainable agriculture. The sensitivity of how decisions could move towards sustainable solutions that produce less health consequences and policies to facilitate such pursuit are assessed.

Conclusions As producers express greater concern for sustainability and certification that recognise that ”good practices” are applied, MCDA suggests a way that evidence can be collected and analysed to support decision-making, transparently and comprehensively.