Discrete choice experiments (DCE) have been in-creasingly applied to assess farmers’ preferences for alternative designs of hypothetical agri-environmental schemes. We extend the DCE approach by asking farmers not only to choose among alternative conservation contracts but also to choose the land area they would put under contract. This extended approach allows us to formulate optimal contracts (in terms of management prescriptions and payment levels) taking explicitly account of the effect of contract design and farm characteristics on the participation area. We use a multinomial Heckman model. In its first stage, we estimate the probabilities of specific contract types being chosen and the corresponding marginal willingness to accept (WTA) figure. In the second stage, we identify the factors affecting the land area offered under contract by means of an OLS regression which takes account of the sample selection bias from first-stage choices. Based upon the estimates from both stages, contracts for a hypothetical conservation scheme are optimized such that environmental benefit is maximized given alternative program budgets. The results show that the optimal design of conservation contracts is sensitive to the program budget.
