Unraveling determinants of inferred and stated attribute non-attendance: effects on farmers’ willingness to accept to join agri-environmental schemes

Attribute non-attendance (ANA) has received very little attention in the context of willingness to accept (WTA), although an increasing number of studies analyze the preferences of ecosystem service providers towards incentive-based schemes. We add to the understanding of ANA behavior by analyzing stated and inferred ANA in a choice experiment investigating farmers’ WTA for participating in agri-environmental schemes (AES) in southern Spain. We use mixed logit models, following Hess and Hensher (2010) for the inferred ANA approach. Evidence is found of ANA behavior for both stated and inferred approaches, with models accounting for ANA clearly outperforming those that do not account for it; however, we produce no conclusive results as to which ANA approach is best. WTA estimates are only moderately affected, which to some extent is consistent with the low level of non-attendance found for the monetary attribute. Stated and inferred approaches show very similar WTA estimates. Additionally, we investigate sources of observed heterogeneity related to ANA behavior by using a sequence of bivariate probit models for each attribute. Overall, our results hint at a positive relationship between ease of scheme adoption and non-attendance to attributes. However, further research is still needed in this field