Date of Award


Degree Name

Doctor of Philosophy



First Advisor

Lahirir, Dr.Sajal


This study examines empirically the transition, persistence, and loan demand in the rural credit market using panel data. The data was collected for seven rounds (1994-2009) from 15 villages in Ethiopia. The sample size is about 1500 households for each round. . Chapter one examines the determinants of simultaneous borrowing and lending. We also investigate why some households in rural Ethiopia simultaneously borrow and lend. Who are these households? Panel logit model is estimated for the sub-sample of borrowers and lenders. The result suggests that households that simultaneously borrows and lends are relatively better-off households. The probability of being a simultaneous borrower and lender is higher for households with strong village level networks. Moreover, households that are affected by common shock are more likely to be a simultaneous borrower and lender. Chapter two examines the dynamics and persistence in the rural credit markets in Ethiopia. It also examines the determinants of dynamics and persistence in borrowing and lending. Duration, dynamic probit, and dynamic multinomial logit models are estimated. We control for unobserved heterogeneity and initial condition. The result reveals the existence of positive duration dependence in both only borrowing households and simultaneously lending and borrowing households. The longer the duration as a borrower, the more likely to exit from borrowing. The longer the duration out of borrowing, the more likely to re-enter to borrowing. Off-farm work, fertilizer use, household size, and storing crop are an important determinant of the probability of exit from borrowing. There is also true state dependence in lending, borrowing, and simultaneously borrowing and lending households. This means the probability of being a borrower in the current period is highly correlated with being in the same state in the previous period. Poverty status, flood, labor sharing, membership in mutual help association, total oxen owned, storing crop, off-farm activities, and fertilizer use are an important determinant of the probability of being a borrower. Chapter three examines the determinants of demand for credit in rural Ethiopia. Bias due to data truncation, variation of the interest rate, and using loan data from a single source are the challenges in estimating demand for credit in the context of rural credit market. This study captures data truncation by estimating a panel Tobit model. The variation in the interest rate is also controlled by using village dummies and their interaction with the source of the loan. Total loan obtained from multiple sources is used as a dependent variable. The result reveals that initial endowment proxied by the value of assets, household size, the age of the head of the households, transitory income, and real per capita consumption are the most important determinants of demand for credit.




This dissertation is only available for download to the SIUC community. Others should contact the
interlibrary loan department of your local library or contact ProQuest's Dissertation Express service.