Duration analysis has been used to examine the impact of time‐varying and time‐invariant variables on the speed of adoption of fertiliser and herbicide by smallholder farmers in East and West Shewa in the central highlands of Ethiopia in the 25 years preceding 1996. The estimated models suggest that economic incentives were the most important determinants of the time farmers waited before adopting new technologies; traction power in the form of oxen and infrastructural factors (in particular proximity to markets) also appear to have been important influences, but less so than prices. Other agricultural inputs (area of farmland, labour, credit), extension services and farmers' personal characteristics (education, gender, age) appear to have had little, if any, effect on adoption behaviour. There is also evidence that the speed of adoption of herbicide, on both tef and wheat, was slower than that of fertiliser.

No abstract

This paper evaluates alternative designs for contracts between a regulator and an agricultural producer to increase the supply of environmental public goods. Contract design, based on the principal‐agent model, takes into account an asymmetry of information between the regulator and producer whereby the regulator is unable to observe precisely the producer's compliance costs. An example is included of contracts designed for nitrate abatement.

This note analyses the design of agri-environmental schemes for risk-averse producers whose input usage is only observable by costly monitoring. The scheme penalises producers in proportion to input use in excess of a quota. A striking result is that if the scheme is designed in such a way that producers always comply with the quota, risk aversion is not relevant in determining the level of input use.

A model of adverse selection and moral hazard in agri-environmental schemes is developed based on the input quota mechanism of Moxey et al. ( Journal of Agricultural Economics, Vol. 50, (1999) pp. 187-202) and Ozanne et al. ( European Review of Agricultural Economics, Vol. 28, (2001) pp. 329-347), rather than the input charge mechanism of White ( Journal of Agricultural Economics, Vol. 53, (2002) pp. 353-360), but the variable fine of the latter rather than the fixed fine assumed by Ozanne et al. ( European Journal of Agricultural Economics, Vol. 28, (2001) pp. 329-347) is used. Incentive-compatible contracts, including the optimal probabilities of detection (and, therefore, monitoring frequencies and costs) for more and less efficient farmers, are identified. It is shown that the input charge and input quota approaches lead to identical outcomes - in terms of abatement levels, compensation payments, monitoring costs and probabilities of detection - confirming the equivalence of input quotas and input charges under asymmetric information. It is also shown that the optimal contracts are independent of the risk preferences of farmers with regard to being caught cheating. Copyright 2007 Blackwell Publishing Ltd.