Agricultural cooperative loan repayment in Karnatka, India: a discriminant analysis

The primary objective of this study was to develope a quantitative function which could discriminate between defaulters and non-defaulters of crop production loans from primary agricultural credit cooperatives in Mandya District, Karnataka, India. This function was to be used to develope descriptive typologies of defaulters and non-defaulters which could be used to target potential defaulters for some form of supervision.

The second objective was to test six hypotheses as to why individuals are defaulters. The first was that borrowers with strong political influence are more likely to default than borrowers with little political influence. The second hypothesis was that better educated borrowers are more likely to default than less well educated borrowers. The third was that smaller farmers are more likely to default than larger farmers. The fourth was that borrowers with high average propensities to consume are more likely to default than borrowers with lower average propensities to consume. The fifth hypothesis was that borrowers with a high ratio of short-term loans to expenses are more likely to default than borrowers with lower ratios. Finally, it was hypothesized that borrowers with high debt-to asset ratios were more likely to default than borrowers with lower debt-to-asset ratios.

The function was tested using an F ratio. This test indicated that the function could not discriminate between defaulters and non-defaul ters at a reasonable level of significance. Partial F tests were used to test the contribution of each variable to the discriminant function given that all other variables were already present in the model. The household average propensity to consume was the only variable which aided the model's ability to discriminate. Given that the model was not significant, no descriptive typologies could be developed.