Multiple Credit Constraints and Borrowing Behavior of Farm Households: Panel Data Evidence from Rural Ethiopia

Promoting an inclusive rural credit market in developing countries is a reemerging and pressing development agenda, given its importance in the poverty reduction and economic growth process. Existing literature mainly focuses on the supply side of the market with little or no attention given to demand aspects. This paper analyzes both the demand and supply side factors affecting credit constraints and borrowing behavior of farmers. Two waves of survey data, which included about 1,200 randomly selected households from four zones of the Amhara region in Northern Ethiopia, were used for the analysis. The Generalized Linear Latent and mixed model (gllamm) was employed to account for unobserved heterogeneity and potential correlations across credit constraint categories. The results show that the probability of quantity rationing increased in the study area between the years 2011 and 2013. Exposure to climatic shocks, age, and lack of education were found to increase the probability of being constrained while female and married heads were relatively less constrained. The results further indicate that borrower's perceived probability of rejection due to strict lending policies and institutional rigidities; the transaction cost of borrowing; and risk aversion behavior of farmers highly reduced the probability of borrowing from the formal credit market. Compared to North Shewa, farmers living in South Wollo zone were found to be discouraged and hence did not prefer borrowing from the formal sector. However, farmers in West Gojjam were less discouraged and had a higher probability of participating in the formal credit market, signifying zonal variation in credit constraints and borrowing behavior. This suggests the need to work on more innovative lending approaches by giving attention to context-specific factors to build demand-driven, climate-smart, and inclusive rural credit market.

在发展中国家推动包容性农村信贷市场（inclusive rural credit market）是一项重新受到重视且紧迫的发展议题，因其在减贫与经济增长进程中发挥着关键作用。现有研究大多聚焦于信贷市场的供给侧，却极少关注需求层面。本文同时剖析了影响农户信贷约束与借贷行为的供需两侧因素。本研究采用两轮调查数据开展分析，数据样本来自埃塞俄比亚北部阿姆哈拉地区（Amhara region）四个分区的约1200户随机抽取家庭。为处理不可观测异质性（unobserved heterogeneity）与信贷约束分类间的潜在相关性，研究采用了广义线性潜在混合模型（Generalized Linear Latent and mixed model, gllamm）。研究结果显示，2011至2013年间，研究区域内的信贷数量配给（quantity rationing）概率有所上升。遭遇气候冲击（climatic shocks）、年龄偏大以及受教育程度不足，均会提升农户面临信贷约束的概率；而女性户主与已婚户主的信贷约束概率相对更低。进一步分析表明，借款人因放贷政策严苛、制度僵化而感知到的拒贷概率、借贷交易成本，以及农户的风险厌恶（risk aversion）行为，均会显著降低其从正规信贷市场（formal credit market）获得贷款的可能性。与北舍瓦（North Shewa）相比，南沃洛（South Wollo）分区的农户更易产生借贷抑制，因此不愿向正规信贷机构申请贷款；而西戈贾姆（West Gojjam）分区的农户借贷抑制程度更低，参与正规信贷市场的概率更高，这体现出不同分区在信贷约束与借贷行为上的异质性。上述结果表明，有必要结合各区域的具体情境因素，探索更具创新性的放贷模式，以构建需求导向、气候智能型且包容性的农村信贷市场。