An Exploratory Study of Trust and Material Hardship in Ghana

We explore associations among interpersonal (thick and thin) and institutional (legislative, executive, and judicial) trust and material hardship outcomes in Ghana. We use data from the 2008 Afrobarometer survey. Material hardship is conceptualized in terms of frequency of going without five basic necessities/consumptive deprivations, each of which a separate outcome (food, water, medical care, cooking fuel, and cash income). Five multinomial logistic regression models are estimated. Models also include relevant socioeconomic and cultural factors. The stepwise forward entry method was used to identify important variables for each model, and common and unique predictors of each outcome are discussed. Results suggest that deprivations are trust-specific in Ghana. Interpersonal and institutional trust types matter differentially (for example, trust in parliament/national assembly is a unique food deprivation predictor; trust in courts is a unique water deprivation predictor; and trust in police is a unique cash income deprivation predictor). Common trust predictors across models are thin trust (trust in other Ghanaians) for predicting food and cooking fuel infrequency outcomes and executive trust for predicting water, medical care, and cash income infrequency outcomes. Some of the non-trust variables, like education and ethnicity, were prominent predictors of deprivation outcomes across all models. This is one of the beginning studies to explore micro-level data in an important area of trust and material hardship in Sub-Saharan Africa. The study also attempts to conceptualize material hardship within the context of a developing country and provides insightful results that could be beneficial to other research investigators and further probed in future studies building on this work. The study draws important implications for intervention policies based on analyses of material hardship composite parts which provide insights with great detail into specific target groups on specific component-outcomes.