This project will develop a novel analytical framework and quantitative approach based on an integrated assessment of treatment effects via causal-forest (CF) machine learning (ML) to evaluate and…

This project will develop a novel analytical framework and quantitative approach based on an integrated assessment of treatment effects via causal-forest (CF) machine learning (ML) to evaluate and predict the extent to which heterogeneous engagement and capacity-building activities enhance communities? Capacity to progress towards transformative adaptation. The approach will be illustrated using an extensive dataset on ~600 coastal communities in the northeast and mid-atlantic US that have both completed and not completed CRB. These data will be integrated with a CF ML treatment-effect model to (i) estimate heterogeneous, causal treatment effects linked to CRB participation in a nonparametric framework, and (II) predict, for any US coastal community, the extent to which similar engagements with given characteristics will induce particular social, economic, and biophysical adaptation outcomes. This approach allows for the likelihood that similar engagements may have different effects on different adaptation outcomes, and that these effects may vary depending on community characteristics. Results will be used in coordination with partners and stakeholders to provide guidance for effective engagement and capacity-building, targeted to community characteristics. This will be the first rigorous attempt to quantify, using CF ML with data from hundreds of communities, the extent to which engagement strategies induce measurable adaptation outcomes.