Market participation, innovation adoption and poverty in rural Ghana

Agricultural commercialisation via increased market participation and innovation adoption has been widely argued to reduce poverty. However, empirical evidence suggests that both of these are persistently low in developing countries. Recent analyses suggest that different types of transaction costs and social capital may influence both market access and innovation adoption decisions. This thesis investigates these two factors in agricultural commercialisation and poverty reduction. Using data from three GLSS survey rounds, Chapter 1 investigates the determinants of the decision to sell as well as the decision of how much to sell, focusing on the role of transaction costs. The empirical analysis is carried out at household level and for a specific crop (maize). A Heckman two-step model is used to control for self-selection into market participation, using measures of fixed transaction costs as identifier variables. The overall results, although generally consistent with previous literature, show an unexpected positive relationship between remoteness and market participation, which might reflect peculiarities of Ghanaian crop marketing systems. Chapter 3 investigates the relationship between social capital and innovation using primary data on 305 Ghanaian farmers collected during field work in 2012 (described in Chapter 2). The chapter analyses innovation (the decision to adopt, its timing and intensity) at crop level, focusing on a non-traditional cash crop, exotic varieties of mango. The analysis investigates the role of different types of social capital, both in disaggregated and aggregated forms. The results suggest that social capital should not be overlooked in the innovation process, supporting recent evidence that there exists a positive relationship between the “know-who” and adoption dynamics. Finally, Chapter 4 investigates the impact of innovation adoption on objective and subjective measures of poverty. Matching techniques are used to estimate the Average Treatment Effect on the Treated, using primary data. The results show that adoption does not impact objective poverty but it does have a significant positive impact on self-perceived poverty status.