An empirical analysis of nature, magnitude and determinants of farmers’ indebtedness in India

Purpose - According to the 70th round of the National Sample Survey published by the Government of India in 2014, the incidence of indebtedness among households in the rural areas of Telangana state, India, is twice that of rural all-India. Around 59 per cent of rural households are indebted in Telangana as against 31 per cent all-India. The purpose of this paper is to examine the extent and magnitude of indebtedness among rural households in the Medak district of Telangana state. Further, the authors wanted to identify the sources of credit to these households and for what purpose the loans were utilised. Design/methodology/approach - To achieve the objective, the authors conducted a primary-level household survey in one of the distressed districts in newly formed state. The authors applied the Bayesian and the Lasso regression methods to identify the factors that impact indebtedness of a household. Findings - The OLS results based on the Lasso regression results show that among all the explanatory variables, principal occupation, use of modern technology, the rate of interest, household medical expenditure and source of loan are significant, indicating that these variables significantly affect the loan taken by the farmers in the study area. The study shows that alternative sources of non-farm income and promotion of modern technology in agriculture can reduce the incidence of farmers’ indebtedness in India. Originality/value - The paper contains significant information with regard to indebtedness. It focusses on the issue troubling the authorities the most. It provides the ground realities of the incidence of indebtedness in Medak, one of the most distressed districts of Telangana, a Southern Indian state. There have been very few similar studies done in the newly formed state. The paper has employed an advanced statistical technique, i.e. Heckman’s selection regression technique, to study farmers’ indebtedness in India. It provides a means of correcting for non-randomly selected samples, which otherwise can lead to erroneous conclusions and poor policy.