农村信贷可得性影响因素研究——基于Adaptive Logistic Lasso Regression回归方法

A Study on Factors Affecting Rural Credit Accessibility——Based on Adaptive Logistic Lasso Regression

基于中国家庭收入调查项目(CHIP)2013年度数据的农村住户调查部分,运用Adaptive Logistic Lasso Regression模型实证研究了农村信贷市场中影响农户信贷可得性的因素。研究结果与现有文献并不完全一致,农户动产总额、金融资产总额、当年度非农业经营性固定资产对农户信贷可得性具有显著影响,农户其他的人口学特征(年龄、教育、健康程度等)、工作性质、是否购买保险和拥有土地面积等对可获得贷款不具显著影响;信贷供给端的贷款绩效风险—激励机制、贷款代理人的个人特质及职业素质等很大程度上也会影响信贷可获得性。

Based on the rural household survey data of the Chinese Household Income Project （CHIP） 2013, this paper uses the Adaptive Logistic Lasso Regression model to study the factors that affect rural credit availability in rural credit market. The results of the study are not completely consistent with the existing literature. Farmers＇ total amount of movable assets, total amount of financial assets, and the non-agricultural fixed assets in the current year have significant impact on the credit availability of rural households. Other demographic characteristics （age, educa- tion, etc.）, the nature of work, whether to buy insurance and household land area have no significant impact on the availability of loans; seen from the credit supply side, the risk-incentive mechanism in financial agents, the loan offi- cers＇ personal qualities and professional quality may also affect the credit availability to some extent.