Bundling Fertilizer-Reduction Practices into Agri-Environmental Payment Design: A Choice Experiment with Rice Farmers’ Preferences

Calibrating compensation programs for chemical-fertilizer reduction requires reliable evidence on farmers’ preferences across reduction practices and policy instruments. Existing choice-experiment evidence treats reduction as either a single aggregate attribute or mutually exclusive practices, specifies compensation independently of the practice bundle, and underuses the distinction between smallholders and new agricultural operating entities (NAOEs) in heterogeneity analysis. We surveyed 1066 rice farmers across six counties and districts of Jiangsu Province using a multi-binary-attribute choice experiment that represented five fertilizer-reduction measures (soil-testing-based fertilization, deep placement, organic substitution, crop rotation, and straw return) as separate binary attributes, with a cost-proportional compensation rate; estimation used mixed logit. Non-monetary preferences vary qualitatively across measures: soil-testing-based fertilization is intrinsically valued (willingness to pay 24 CNY/mu), while crop rotation imposes the largest non-monetary cost (willingness to accept 152 CNY/mu). Existing Jiangsu flat rates approximate farmer mobilization thresholds for low- and medium-cost practices but under-incentivize rotation. Income subsidies dominate non-income alternatives. Smallholders are more inertial and more compensation-responsive than NAOEs, requiring more compensation for crop rotation and showing a stronger income-subsidy preference. The findings support differentiated compensation by measure and farmer type; the methodological template extends to other agri-environmental contexts in which adoption is genuinely combinatorial.