作者：郑宏运，马旺林，王芳，李谷成

题目：互联网使用提高了中国香蕉生产的技术效率吗？来自一个矫正选择性偏误的分析

摘要：尽管许多研究已经分析了互联网使用对农业绩效和农村发展的影响，但对互联网使用与作物生产技术效率之间的关系仍知之甚少。为了填补这一空白，本文分析了互联网使用对中国香蕉生产技术效率的影响。本文结合使用倾向得分匹配与Greens’(2010)的随机前沿模型来处理选择性偏误问题，并使用分数回归模型与无条件分位数回归模型估计互联网使用和其他控制变量对技术效率的同质性和异质性影响。在控制由可观测因素和不可观测因素造成的选择性偏误后，我们发现，互联网使用者和非使用者的香蕉生产技术效率值分别为0.571和0.537。分数回归模型估计结果表明互联网使用对技术效率有同质的正向影响。与之相反，无条件分位数回归模型的结果证实互联网使用对技术效率存在异质性影响，互联网使用对技术效率值处在较低分位点的香蕉种植户的效率提升作用更大。本文的发现强调了在农村地区促进互联网使用以提升农业生产绩效的重要性。

Zheng, Hongyun, Ma, Wanglin, Wang, Fang, and Li, Gucheng. (2021). Does internet use improve technical efficiency of banana production in China? Evidence from a selectivity-corrected analysis. Food Policy. In Press

全文链接： https://doi.org/10.1016/j.foodpol.2021.102044

Abstract:

Although many studies have analyzed the Internet use effects on farm performance and rural development, little is known about the relationship between Internet use and technical efficiency (TE) of crop production. This paper fills the gap by analyzing the impact of Internet use on the TE of banana production in China, using an innovative approach that combines a propensity score matching technique with Greene’s (2010) selectivity-corrected stochastic frontier model to address selectivity bias. We employ both the fractional regression (FR) model and the unconditional quantile regression (UQR) model to, respectively, investigate the homogeneous and heterogeneous TE effects of Internet use and other control variables. After correcting for selectivity bias stemming from both observed and unobserved factors, we show that the TE scores of Internet users and nonusers are 0.571 and 0.537, respectively. The results of the FR model show that Internet use exerts a homogeneous positive impact on TE of banana production. In contrast, the UQR results reveal that TE effects of Internet use are heterogeneous, and the Internet users at the lower TE quantiles receive higher TE than their counterparts at higher TE quantiles. Our findings underscore the importance of promoting Internet use in rural areas to boost farm performance.