不同农地经营权流转方式的空间依赖性分析

Analysis on the Spatial Dependence of Different Ways to Transfer Farmland Management Rights

研究目的:探究不同农地经营权流转方式的空间依赖性。研究方法:基于2015年省级截面数据,对转包、出租和入股三种农地经营权流转方式进行全局与局部空间自相关检验,结合空间计量模型进一步探讨不同流转方式下空间依赖的机制及强度。研究结果:(1)全局层面,经营权流转空间收敛特征显著,转包和出租方式表现为空间正自相关,入股方式则呈现随机分布特征;(2)局部层面,"高—高"聚集和"低—低"聚集为各省区主要空间聚集模式,热点区域主要分布在中部和东部地区,冷点区域主要分布于中部、东部和西部地区;同一省区、不同流转方式下,空间聚集模式存在显著差异;(3)针对不同流转方式,基于传统最小二乘法回归的残差项检验均拒绝无空间依赖关系的原假设;进一步使用SLM和SARAR空间计量模型展开分析,结果表明,整体层面的经营权流转、转包以及出租等流转方式均存在显著的空间依赖性。研究结论:不同农地流转方式具有经济行为上的空间规律,不应简单地视流转为一个抽象概念,应将具体的流转行为从空间依赖和空间异质层面纳入政策制定过程,更为精准地制定具有区域化与省域化特征的农地流转政策。

The purpose of the paper is to explore the spatial dependence of different ways to transfer farmland management rights. Based on a provincial cross-sectional data set in 2015, the spatial auto-correlation tests are employed generally and partially and the spatial econometric models are adopted in analyzing the mechanisms and magnitudes of the spatial dependence under the three channels of farmland transfer, i.e. subcontracting, leasing and investing. The results show that： 1）on the general level, the transfer of farmland management rights shows the significant spatial convergence. In the spatial auto-correlation tests, subcontracting and leasing show a positive relationship, while investing tends to distribute randomly. 2）On the partial level, the High-High clustering and Low-Low clustering are the main patterns of spatial agglomeration. Hot-spot regions mainly distribute in central and eastern China, while the coldspot regions mainly distribute in central, eastern and western China; the patterns of spatial aggregation are significantly different in terms of various ways of farmland transfer under the same provincial unit. 3）In terms of the ways of farmland transfer, the traditional regression models with OLS estimation method show that the null hypothesis, i.e. the spatial dependence does not exist, is rejected. Furthermore, the spatial econometric models, SLM and SARAR are estimated and the results show that the spatial dependence is significant when taking the transfer of farmland management rights as a whole and subcontracting or leasing as a dependent variable separately. In conclusion, different ways of farmland transfer imply the spatial regularities from the perspective of economic behaviors, so it＇s unreasonable to treat farmland transfer as an abstract concept. Hence, the specific behaviors of farmland transfer should be included in the policy-making process from the perspectives of spatial dependence and spatial heterogeneity, which will help make the targeted farmland transfer policies fitting more regional and provincial features.