This paper attempts to address a recurring theme in agricultural resources and environmental economics – the low adoption rates of pro-environmental agricultural practices in many developing countries. By improving the Norm-activation model, this paper incorporates external incentives and social norms into the framework and employs multivariate probit and ordered probit models to explore how environmental conscience, external incentives and social norms influence rice farmers’ adoption behaviours and intensive use of pro-environmental agricultural practices. The case study, involving 954 household-level data of rice farmers from rural Hubei province, China, reveals that the adoption rates of certain practices are very low, and that only 6.5% rice farmers adopt three or more pro-environmental agricultural practices. Results show that straw returning and soil testing and fertilizer recommendation are complementaries, and that environmental conscience, external incentives and social norms all positively affect the adoption behaviours, while the adoption intensity is significantly influenced by awareness of consequences, perceived efficacy, external incentives and descriptive norm. These findings underscore that policy interventions to improve rice farmers’ environmental conscience, to provide well-designed external incentives and to activate social norms are needed to enhance the adoption of pro-environmental agricultural practices in developing countries. GRAPHICAL ABSTRACT

Purpose: The purpose of this paper is to estimate shadow prices of agricultural carbon emissions produced by agricultural inputs, rice paddy and burning crop residue, and to explore the impact of cropping pattern on marginal abatement cost (MAC). Design/methodology/approach: The shadow price of agricultural carbon emissions is estimated by applying directional distance function and non-parametric methods. Findings: The estimated shadow price of agricultural carbon emissions ranges from 6.78 to 557.83 yuan/ton, and the average value is 62.50 yuan/ton (or $10.18/ton). The MAC value varies in different provinces and years. The regional difference of MAC shows a decreasing trend during the investigation period. Cropping pattern shows a significant negative impact on agricultural MAC. A 1 percent decrease of rice proportion leads to a 0.31 percent increase in MAC value. This implies that the higher the proportion of rice is, the lower the economic cost to reduce agricultural carbon emissions would be. Practical implications: It is feasible to draw up appropriate mechanisms for the allocation of emission reduction responsibilities according to conditions in various regions, with emphasis on the local cropping patterns. There is a trade-off between reducing carbon emission and increasing crop yields. Originality/value: This study calculates agricultural MAC by using the shadow price approach, taking agricultural carbon emissions as undesired environmental output. The study also provides a reference emission right price and provides guidance to make use of cropping structure adjustment and optimization for exploring the emission reduction strategy.