Information for Policy Design: Modelling Participation in a Farm Woodland Incentive Scheme
研究了如何利用农民和土地所有者调查数据，通过逻辑模型预测他们参与农场林地激励计划的决策，发现仅用普查数据已能较好识别参与倾向，额外调查因成本和小样本反而可能降低预测能力。
Incentive schemes which use farmers and landowners as agents of policy are now the main driving force for modifying the countryside environment. Nevertheless, policy‐makers lack quantitative models that describe farmers' response to schemes, and specifically their entry decisions. This paper examines the benefits from modelling entry decisions with and without additional information generated through surveys of participants and non‐participants. It uses logit models to predict the probability of entry into a farm woodland incentive scheme, initially using only those explanatory variables that are available to policy‐makers from the annual farm census. The model performs well in identifying those factors pre‐disposing farmers to entry. The potential gains from enhancing the information set by incorporating lifestyle and other variables not recorded in the census are examined. The cost of additional surveys, and their small sample size compared with the census, casts doubt on the value of such an approach which may reduce rather than enhance predictive power.