| 摘要: |
| 目的 作为小农户衔接现代化农业的重要途径,农业生产性服务在突破资源环境约束,助力农业可持续发展方面具有重要作用。探讨农业生产性服务对温室气体排放的影响及内在机理,对中国实现“双碳”目标至关重要。方法 基于东北三省894份玉米种植户的实地调研数据,文章采用生命周期评价法测算农业温室气体排放并构建了双重机器学习模型框架,实证检验农业生产性服务对农业温室气体的影响。该模型克服了传统因果推断模型诸多严格特定条件,解决了“维度诅咒”的问题。结果 (1)农业生产性服务能够显著抑制农业温室气体排放,采取多种稳健性检验后,该结论依旧成立,其主要通过促进土地流转的规模效应、采纳保护性耕作的技术效应以及农业劳动力的替代效应等间接路径实现“减排效应”;(2)就地理区域而言,农业生产性服务对东北三省的中西部地区的减排效应更显著;就粮食作物收入水平而言,农业生产性服务对粮食收入水平较高的农户温室气体减排效果更显著;(3)从不同的生产服务环节来看,耕整地、播种及施肥等环节对降低温室气体排放更显著。结论 基于此提出农业生产性服务在我国农业温室气体减排行动中的重要作用,这为继续加强农业生产性服务建设和推动农业可持续发展提供了科学的政策启示。 |
| 关键词: 农业生产性服务 温室气体排放 双重机器学习模型 生命周期评价法 绿色发展 |
| DOI:10.7621/cjarrp.1005-9121.20260505 |
| 分类号:F323 |
| 基金项目:国家社科重大项目“‘双碳’目标下农业绿色发展体系创新与政策研究”(22&ZD083);国家自然科学基金项目“基于分工视角的不同生产服务主体响应行为与农业生产性服务高质量发展:作用机理、效果评估及政策优化”(72373101);中国博士后科学基金面上项目“基于生命周期视角的奶牛养殖业粪污资源化利用环境回弹效应的量化研究”(2022M712202) |
|
| THE IMPACT OF AGRICULTURAL PRODUCTION SERVICES ON GREENHOUSE GAS EMISSIONS AND MECHANISM——CAUSAL INFERENCE BASED ON DOUBLE MACHINE LEARNING |
|
Wang Lulu, Lyu Jie, Zhang Junyan
|
|
College of Economics and Management, Shenyang Agricultural University, Shenyang 110866, Liaoning, China
|
| Abstract: |
| As an essential pathway for smallholder farmers to connect with modernized agriculture, agricultural production services play an important role in breaking through resource and environmental constraints and contributing to the sustainable development of agriculture. Exploring the impact of agricultural production services on greenhouse gas emissions (GHGs) and their underlying mechanisms is crucial for China to realize the “double-carbon” goal. Based on the field research data of 894 maize farmers in the Three Northeastern provinces, we used the life cycle assessment method to measure agricultural GHGs and constructed a double machine learning model framework to empirically test the impact of agricultural production services on agricultural greenhouse gases, and the model overcame many strict specific conditions of the traditional causal inference model and skillfully solved the problem of “curse of dimensionality”. The results found that: (1) Agricultural production services could significantly curb agricultural GHGs, and this conclusion was still valid after various robustness tests, and the “emission reductions” were mainly achieved through indirect pathways, such as the scale effect of promoting land transfer, the technological effect of adopting conservation farming, and the substitution effect of agricultural labor; (2) In terms of geographic regions, agricultural production services had a more significant mitigation effect on the central and western regions of the Three Northeastern provinces; in terms of food crop income levels, agricultural production services had a more significant effect on GHG mitigation for farm households with higher food income levels; (3) In terms of different production services, ploughing, sowing and fertilizer application were more significant in reducing GHGs. Based on this, the important role of agricultural productive services in China's agricultural GHGs reduction actions is proposed, which can provide scientific policy insights for continuing to strengthen the construction of agricultural productive services and promoting sustainable agricultural development. |
| Key words: agricultural production services greenhouse gas emissions double machine learning modeling life cycle assessment approach green development |