Understanding the Average Impact of Microcredit Expansions : A Bayesian Hierarchical Analysis of 7 Randomized Experiments WORKING PAPER
Understanding the Average Impact of Microcredit Expansions : A Bayesian Hierarchical Analysis of 7 Randomized Experiments WORKING PAPER
复制标题
了解小额信贷扩张的平均影响：7 个随机实验的贝叶斯层次分析 工作论文
DOI：
--
复制
发表时间：
2016
期刊：
影响因子：
--
通讯作者：
Rachael Meager
中科院分区：
文献类型：
--
作者：
Rachael Meager
我对来自7个小额信贷随机试验的证据进行了贝叶斯分层分析，以评估对家庭结果的总体影响以及这种影响在不同地点的异质性。在所有结果中，结果表明，小额信贷的影响是积极的，但与对照组的平均水平相比很小，不能排除零影响或负面影响的可能性。这些研究的标准合并指标表明，治疗效果平均有60%的合并，这表明特定部位的效果具有实质性的外部有效性。交叉研究的异质性几乎完全是由在小额信贷扩张之前经营企业的27%家庭的异质性效应造成的，对这一群体的影响总体上似乎要大得多。用岭回归方法评估特定地点的协变量和治疗效果之间的相关性表明，剩余的异质性与经济变量的差异密切相关，但与研究设计方案的差异无关。平均利率和平均贷款规模与处理效果的相关性最强，且均为负相关。凝胶编码：C11、D14、G21、O12、O16、P34、P36∗麻省理工学院(研究生)。联系人：rmeager@mit.edu†我感谢埃丝特·杜弗洛、阿比吉特·班纳吉、安娜·米库什娃、罗布·汤森德、杰夫·哈里斯、维克多·切尔诺朱科夫、安德鲁·赫尔曼、本·奥尔肯、曾傑瑞·豪斯曼、拉尔斯·汉森、希拉·米切尔、基里尔·布尔萨亚克、科里·史密斯、乔纳森·哈金斯、瑞安·乔达诺、塔玛拉·布罗德里克、阿里安娜·奥纳吉、格雷格·霍华德、尼克·哈格蒂、约翰·费尔斯、杰克·利伯森、彼得·哈尔、马特·洛、雅罗斯拉夫·穆钦、齐苏亚·卡吉、小王、阿卡隆·波斯特，以及2015年芝加哥-麻省理工学院学生会议、麻省理工学院经济发展研讨会、麻省理工学院经济学人午餐和耶鲁大学午餐/劳工研讨会的与会者和忠告。我还感谢我在分析中使用的7项研究的作者，以及发表这些研究的期刊，因为他们公开了他们的数据和代码。所有剩下的错误都是我自己的。这是一份工作文件，请将批评和更正发送到rmeager@mit.edu
I perform a Bayesian hierarchical analysis of the evidence from 7 randomized trials of microcredit to assess the general impact on household outcomes and the heterogeneity in this impact across sites. Across all outcomes, the results suggest that the effect of microcredit is positive but small relative to control group average levels, and the possibility of a zero or negative impact cannot be ruled out. Standard pooling metrics for the studies indicate on average 60% pooling on the treatment effects, suggesting that the site-specific effects have substantial external validity. The cross-study heterogeneity is almost entirely generated by heterogeneous effects for the 27% households who previously operated businesses before microcredit expansion, and impacts on this group appear to be much larger overall. A Ridge regression procedure to assess the correlations between site-specific covariates and treatment effects indicates that the remaining heterogeneity is strongly correlated with differences in economic variables, but not with differences in study design protocols. The average interest rate and the average loan size have the strongest correlation with the treatment effects, and both are negative. JEL Codes: C11, D14, G21, O12, O16, P34, P36 ∗Massachusetts Institute of Technology (Graduate Student). Contact: rmeager@mit.edu †I thank Esther Duflo, Abhijit Banerjee, Anna Mikusheva, Rob Townsend, Jeff Harris, Victor Chernozhukov, Andrew Gelman, Ben Olken, Jerry Hausman, Lars Hansen, Shira Mitchell, Kirill Boursayak, Cory Smith, Jonathan Huggins, Ryan Giordano, Tamara Broderick, Arianna Ornaghi, Greg Howard, Nick Hagerty, John Firth, Jack Liebersohn, Peter Hull, Matt Lowe, Yaroslav Mukhin, Tetsuya Kaji, Xiao Yu Wang, Aaron Pancost, and the participants of NEUDC 2015, The Chicago-MIT student conference 2016, the MIT Economic Development Lunch Seminar, MIT Econometrics Lunch Seminar and Yale PF/Labor Lunch Seminar for their suggestions, critiques, and advice. I also thank the authors of the 7 studies that I use in my analysis, and the journals in which they were published, for making their data and code public. All remaining mistakes are my own. This is a working paper, so please send critiques and corrections to rmeager@mit.edu