马旺林 等 | 发展中国家农业合作社的增产效应：基于Meta分析的实证研究

Wanglin Ma, Sanghyun Hong, W. Robert Reed, Jianhua Duan and Phong Quoc Luu (2023). Yield effects of agricultural cooperative membership in developing countries: A meta-analysis. Annals of Public and Cooperative Economics. first online

原文链接（点击“阅读原文”跳转）：

http://doi.org/10.1111/apce.12411

摘要：

本文利用Meta分析法研究了农业合作社参与对作物和牲畜产量的影响。在具体分析中，本研究收集了19个发展中国家42项研究所估计的158个农业合作社的增产效应。研究发现，已有相关文献存在“正出版偏误”（positive publication bias），即研究者和期刊更偏好于发表发现农业合作社显著增产效应的研究。在修正“出版偏误”后，研究发现农业合作社的增产效应较弱且不显著。Meta回归分析显示：在已有研究中，农业合作社产量效应的变动主要由具体研究的异质性引起，例如已有研究在样本类型（是全样本还是子样本）、会员比例、计量方法（基于工具变量的参数法，非参数法，抑或最小二乘回归）、影响效应的类型（针对处理组的平均处理效应、平均处理效应或回归系数）、农业产出的类型（谷物或其他）和气候区（热带地区或非热带地区）等的不同。

Abstract

This study uses a meta-analysis to synthesize the effects of agricultural cooperative membership on the yield of crops and livestock. It collects 158 estimated yield effects from 42 studies, covering 19 developing countries. Our analysis finds evidence that there exists positive publication bias in the empirical literature, confirming that researchers and journals have a preference to publish articles that report positive and significant results. After correcting for publication bias, we find that cooperative membership has a small-sized and insignificant effect on the yield. The meta-regression analysis reveals that variation in the reported yield effects can be largely explained by the study attributes such as the sample type (full sample vs. subsample), membership ratio, econometric approaches (instrumental-variable based parametric approach, non-parametric approach or ordinary least square regression), effect size types (average treatment effects on the treated, average treatment effects, or coefficient), agro-product type (grain or others), and climate zones (tropical or non-tropical).

转自：“三农学术”微信公众号

如有侵权，请联系本站删除！