Credit Constraints, Risk Sharing, and Household Welfare: The Case of Indonesia

Abstract

This thesis studies household welfare and financial markets and in particular empirically examines access to finance, human capital, saving and risk sharing group formation using Indonesian households as a case study. Inefficient financial markets in developing countries lead to inefficient resource allocation, economic inequality, and high transaction costs. Households who are marginalised from financial systems find themselves unable to access financial services and smooth their consumption.

The first thing to consider is how credit constraint exists and how to identify it. Credit constraints may arise from market mechanisms: demand for loans and loan supply. In order to assess credit constraints, I use Direct Elicitation Methodology (DEM) and then examine the gathered information and other household characteristics using multinomial logit model. Using Access to Finance (A2F) survey, I find that Indonesian households are likely to experience supply-side rather than demand-side constraints. I also find that financial literacy plays vital role in accessing services from formal financial institution. Moreover by elaborating several types of constraints, the welfare loss is estimated: the constrained households due to risk-related reasons experience loss in terms of annual income between Rp. 16 millions and Rp. 19 millions.

In the second empirical study, I investigate the impacts of earnings risk on schooling and saving. I borrow Basu and Ghosh's model (2001) to develop a theoretical framework of two-period model, which depicts the relationship between earnings risk, schooling and saving. Using the Indonesia Family Life Survey (IFLS) data set, the decision to enter schooling is motivated by earnings risk which is measured by occupational earnings risk and earnings range or the variability between maximum and minimum earnings level across the IFLS wave. This study finds that education decrease variability over future income. Given the results that the pure risk effect is more dominant than utility smoothing effect, it can be said that to some extent saving is inadequate to anticipate the declining of household income due to earnings risk. The results also show that earnings range is close to Basu and Ghosh’s predictions.

Another issue related to financial markets is the barrier to insurance for households, which also limits their capability to manage life risk. As a result, alternative risk coping mechanisms emerge to provide these households with different ways of securing insurance arrangements and in particular as risk sharing groups. In this third empirical research, I investigate the risk sharing group formation where the group is characterised by barriers to insurance. I use several tests to examine full risk sharing hypothesis, borrowing-saving hypothesis, limited commitment, moral hazard, and hidden income. Using the IFLS data set, this study provides evidence of the failure of the full risk sharing hypothesis, which is mainly due to limited commitment and moral hazard problem. Furthermore, I show that the endogenous group formation emerges within IFLS households.