Abstract: The importance of risk perception and risk attitude for understanding individual's risk behaviour are independently well described in literature, but rarely combined in an integrated approach. In this study, we propose a model assuming the choice to implement certain risk management strategies to be directly driven by both perceptions of risks and risk attitude. Other determinants influence the intention to apply different risk strategies mainly indirectly, mediated by risk perception and risk attitude. This conceptual model is empirically tested, using structural equation modelling, for understanding the intention of farmers to implement different common risk management strategies at their farms. Data are gathered in a survey completed by 500 farmers from the Flanders region in Belgium, investigating attitudes towards farming, perceived past exposure to risk, socio-demographic characteristics, farm size, perceptions of the major sources of farm business risk, risk attitudes and the intention to apply common risk management strategies. Our major findings are: (i) perception of major farm business risks have no significant impact on the intention of applying any of the risk strategies under study, (ii) risk attitude does have a significant impact. Therefore, rather than objective risk faced and the subjective interpretation thereof, it is the general risk attitude that influence intended risk strategies to be implemented. A distinction can be made between farmers willing to take risk, who are more inclined to apply ex-ante risk management strategies and risk averse farmers who are less inclined to implement ex-ante risk management strategies but rather cope with the consequences and diminish their effects ex-post when risks have occurred.

Abstract: This paper uses a mechanistic frontier approach as a reference to evaluate the ability of conventional parametric (SFA) and non-parametric (DEA) frontier approaches for analyzing economicenvironmental trade-offs. Conventional frontier approaches are environmentally adjusted through incorporating the materials balance principle. The analysis is worked out for the Flemish pig finishing case, which is both representative and didactic. Results show that, on average, SFA and DEA yield adequate economicenvironmental trade-offs. Both methods are good estimators for technical efficiency. Cost allocative and environmental allocative efficiency scores are less robust, due to the well-known methodological advantages and disadvantages of SFA and DEA. For particular firms, SFA, DEA and the mechanistic approach may yield different economicenvironmental trade-offs. One has therefore to be careful when using conventional frontier approaches for firm-specific decision support. The mechanistic approach allows for optimizing performances per average present finisher, which is the production unit in pig finishing. Conventional frontier methods do not allow for this optimization since the number of average present finishers varies along the production functions. Since the mechanistic production function is based on underlying growth, feed uptake and mortality functions, additional firm-specific indicators can also be calculated at each point of the production function.