Evaluating the impact of ecosystem service assessments on decision-makers

Ecosystem services support human livelihoods and economies but are declining in many places. Ecosystem service assessments estimate the benefits that nature provides to people and can be used to evaluate trade-offs in impacts and changes resulting from land use decisions. Such assessments can affect the capacity of decision-makers to make sustainable land use decisions, but the actual impact of such projects on decision-maker attitudes is almost entirely unstudied. We addressed this knowledge gap by evaluating the impact of an ecosystem service assessment on decision-makers in California. We asked how decision-makers’ understanding of and attitudes about ecosystem services changed “pre-” and “post-” assessments and between treatment groups where ecosystem services were assessed and a comparison group where ecosystem services were not assessed. Mixed methods included regression models to estimate the treatment effect of the assessment (using a difference-in-differences approach), as well as interviews and direct observations to further understand how decision-makers responded to the assessment. Regression results showed small increases relative to the comparison group in decision-maker understanding of ecosystem services and perceived relevance of ecosystem services to their work. Interviews confirmed that decision-makers learned specific ways that they could use ecosystem services in conservation and development decisions and believed that doing so would improve outcomes. These results demonstrate how ecosystem services assessments can facilitate a conceptual shift in the minds of decision-makers, which is a necessary ingredient for subsequent policy impact. Impact evaluation studies of this type − that estimate a counterfactual and explore rival explanations for observed outcomes − are needed to truly understand whether ecosystem service projects impact decision-makers and, ultimately, produce outcomes for environmental and human well-being.