The rural economic setup in developing countries is customarily dominated by primary production activities, mostly in the agriculture sector. While rural areas have been shown to experience high poverty rates, livelihood diversification is recommended as a measure to help reduce poverty. This is can be done by bolstering household income portfolio through supplementing nonfarm income, than solely depending on income from agriculture activities. This paper observes determinants of rural livelihood diversification using the extended panel data of the Tanzania National Panel Survey. Two measures represent livelihood diversity in the study: number of livelihood activities households engage in, and household share income spread. The Panel Poisson and Tobit models are used to estimate the determinants of livelihood diversity. General factors influencing diversity include household wealth, experiences to shock (drought/floods, fall in prices of crops), and household demographic characters (number of working age individuals and age of household head). An analysis of the determinants by wealth status indicates less wealthy and wealthy households diversify the most with respect to assets they possess, while access to finances gives contrasting results depending on sources of finance. Policy implications relate to promoting policies that support sustained asset accumulation, increasing access to rural financing, and establishing safety net programs that minimize risks associated with shocks.
Keywords: livelihood diversification, number of income sources; share income spread