Essays on Risk and Time Preferences, Vulnerability to Shocks and the Joint Choice for Financial Products in Ghana

This thesis consists of three empirical papers on Ghana, focusing on risk and time preferences, vulnerability to shocks and the choice for savings, insurance and credit among household members. The first paper discusses the vulnerability of households in the three northern regions, Upper East, Upper West and Northern, to shocks. Though the country has impressively reduced the level of poverty over the years, achieving the then millennium development goal on poverty well ahead of schedule, regional disaggregation points to the three regions having more than twice the national poverty incidence. Households in these areas are largely engaged in agriculture which is subject to persistent shocks and may affect their future poverty risk as well as the general level of welfare depending on the coping strategies adopted. We contribute to the literature on household level shocks by developing a shock index which helps to classify households into vulnerable groups, which to the best of our knowledge is the first of a kind. We are thus able to utilize the relative information value of shocks to identify the level of vulnerability of households. Subsequently, we examine the coping strategies adopted by these classes of households. The dataset used for the analysis is the Ghana Africa Rising Baseline Evaluation Survey data, collected by the International Food Policy Research Institute (IFPRI). We find that over a five-year period, about 35.8 percent of the households can be classified as highly vulnerable to shocks with 40.6 percent being moderately vulnerable and 23.6 percent as less vulnerable. While the less vulnerable group experienced no shock, the moderately vulnerable group experienced fewer shocks than the highly vulnerable group. When faced with shocks, households rely largely on their savings supplemented with crop stock and livestock sale as well as support from relatives. Comparing moderately vulnerable households to highly vulnerable households, it was observed that the former group were more likely to use savings than the latter. It was further noted that highly vulnerable households do not have access to support mechanisms to cope with shocks. Though government support and support from NGOs or religious institutions were not significant coping for the two vulnerable groups, moderately vulnerable households were more likely to use support from relatives unlike highly vulnerable households. Highly vulnerable households were therefore more likely to engage in religious sacrifices or prayers to respond to some of the shocks that affect them which was not the case for moderately vulnerable households. In the second paper, we investigate the joint choice for savings, insurance and credit using the Ghana Living Standards Survey 7 data. As individuals and households in developing countries are exposed to various shocks but with limited formal insurance to insure them, they may combine available financial products to reduce the impact of shocks on their welfare. This study deviates from earlier works that focuse on the head of the household as a representative agent of the household, by examining the choices of individual household members and comparing them to the head of the household. We employ the trivariate probit model to compare the choices of the various categories of household members (head of household, Spouse, children and other members) and also examine how risk and time preferences affect their decisions. We find a complementarity between the three products as well as significant differences in the choices of household members, especially for savings and credit. What more, risk preferences significantly determine savings, confirming the existence of self-insurance, while time preference does not affect savings. These two observations give credence to the importance of precuationary measures over impatience in the decision to save in developing countries. We also find that ownership of mobile phones significantly affect the utilization of savings and insurance products; an indication that mobile money is helping to reduce the supply side contraints of financial services especially to the poor. The third paper considers the determinants of risk and time preferences, introducing the role of trust in the family. Risk and time preferences have been identified as important elements in the decision-making process of economic agents. Studies on the two have usually focused attention on other institutions but the family. There is growing interest on the importance of the family in shaping economic preferences of individuals. We introduce trust in the family as a possible determinant of preferences of family members. Estimation results show that though trust in the family has a positive relationship with patience it also reduces the risk tolerance of individuals. This may suggest that trust in the family increases collectivism as against individualism which has been found to increase optimism and risk tolerance. The negative effect of trust on risk tolerance does not necessarily mean trust in the family is not beneficial, since it makes people more patient. The implication however is that as families build trust among members, they should also have a more open discussion on the relevance of self-reliance which can generate among members an enterprising spirit. In addition, family trust should increase the creation of insurance networks to engender risk tolerance.