Knowledge co-production for decision-making in human-natural systems under uncertainty

Abstract

Decision-making under uncertainty is important for managing human-natural systems in a changing world. A major source of uncertainty that challenges decisions is rooted in their multi-actor settings, i.e., the poorly understood societal actors with diverse values, complex relationships, and conflicting management approaches. Despite general agreement across disciplines on co-producing knowledge for viable and inclusive outcomes in multi-actor settings, there is still limited conceptual clarity and no systematic understanding on what co-production means in decision-making under uncertainty and how it can be achieved. Here, we use content analysis and clustering to systematically analyse 50 decision-making cases with multiple time and spatial scales across 26 countries and in 9 different sectors in the last decade to serve two aims. The first is to synthesise the key recurring approaches that underpin high quality decision co-production across many cases of diverse features. The second is to identify important deficits and opportunities to leverage existing approaches towards flourishing co-production in supporting decision-making. We find that four general approaches emerge centred around: promoting innovation for robust and equitable decisions; broadening the span of co-production across interacting systems; fostering social learning and inclusive participation; and improving pathways to impact. Additionally, five key areas that should be addressed to improve decision co-production are identified in relation to: participation diversity; social learning; power relationships; governance inclusivity; and transformative change. Characterising the emergent approaches and their key areas for improvement can help guide future works towards more pluralistic and integrated science and practice.

Entry points for accelerating transitions towards a more sustainable future

“…This diversity and the differences between these cases required us to use a hybrid (systematic and integrative) search process for case selection, suitable for each scientific and grey literature. Hybrid searches often seek to address the inherent limitation of one search approach with another search approach 87 and have been used in previous research 66,88 .…”

“…A secondary co-author then grouped all the shortened insights based on their similarity in two hierarchical steps of entry points (as general processes) and their specifics (as the context-specific means of implementation). This was a fully iterative process where the shortened insights were grouped and regrouped several times to minimises the variance within each entry point whilst maximising the meaningful difference between different entry points (a standard balance often sought in other clustering examples 66,88 ). Eventually the clustering of coded materials from cases resulted in a set of 9 entry points, each with 3 specific means of implementation (Supplementary Data 4).…”

Quantifying resilience under deep uncertainty

“…The process of building the simulation model underlying the generated scenarios may also contribute in this regard, serving as a repository of knowledge (Alexandra et al, 2023) or competing perspectives (Gotts et al, 2019) about the system, and help kick-start dialogue and joint reasoning among stakeholders about their shared system. With appropriate guidance (Moallemi et al, 2023), this may help create shared perspectives, and eventually enable a quantitative approach as demonstrated in Chapter 4, with all the benefits that brings.…”