Nonfarm diversification and the well-being of rural farm households in developing countries: Evidence from Ghana using new dataset

Abstract This paper examines the impact of participation in both farm and nonfarm activities on both household consumption expenditure per adult equivalent and household per capita income, in rural Ghana. The objective is to ascertain whether the results are sensitive to the choice of well-being measure. We use a nationally representative dataset on 8,059 rural farm households collected in 2012/13. In order to account for potential selectivity and endogeneity biases, which previous studies failed to correct for, we adopt the endogenous switching regression (ESR) estimation technique. We find diversified households to be systematically different from their undiversified counterparts in terms of socioeconomic and demographic centeracteristics, thus justifying the empirical method used. Our results indicate a higher observed mean consumption for the diversified sub-sample compared to its counterfactual, implying that households participating in nonfarm enterprise activities in addition to farming have greater mean consumption compared to households engaged solely in farming. Similar conclusions are reached when income instead is used as the well-being indicator. Our findings, thus, indicate that the well-being implication of farm-nonfarm diversification is insensitive to the choice of well-being measure.