Personal accounts: managing households during conflict

This thesis examines the impact of political conflict on microfinance engagement to put forth a theory of sparse networks traps. It leverages a natural experiment to distinguish between the effects of conflict on determinants of microfinance efficiency and impact, and includes qualitative evidence from 235 (208 microfinance users and 27 microfinance providers) interviews in the Northeastern Kivu province of the Democratic Republic of Congo. Through a combination of regression analyses and panel data modelling with fixed effects, the research indicates that conflict has a stronger effect on the nature of demand for credit and savings services than it has on the actual performance of financial institutions. By introducing informal financial service providers, including community level rotating savings and credit associations, payday lenders, and moneylenders, the research indicates that the demand for financial services is not greatly reduced during conflict. The reduction in demand reported in the literature is seen in the formal sector, while in the conflict area the demand shifts to the informal sector, resulting in a threefold increase in the likelihood to borrow from an informal source of credit in times of political violence. This shift in user preferences is reflective of an overall decrease in engagement in formal networks and reliance on informal ones, and is reflected in other coping mechanisms such as reduced investment in business creation and increased expenditures in areas that can be considered charitable. The mechanisms by which these choices occur are hyperbolic discounting and reduced trust. In turn, these individual level decisions lead to a sparse networks trap, defined as a fragmentation of the economy into independent enclaves of production and the correlating reduction in interregional interdependence, which may have compounding consequences for post-conflict economic recovery and stability.