Tony Atkinson and Maria Ana Lugo have explained why strong GDP growth has not led to a dramatic drop in headcount poverty in Tanzania. They argue that this is due, inter alia, to the way in which the proceeds of growth have been distributed as well as due to measurement issues. The apparent dissonance between GDP growth and changes in poverty is of interest in other countries as well. Explaining why these data can appear inconsistent sharpens the focus on the role of growth as a foundation for poverty reduction, as well as the importance of spreading the fruits of growth widely.

This paper examines one aspect of the historical income distribution in Tanzania – the shares of top incomes – using information published as a result of the administration of the individual income tax.Although tax data were used in earlier studies of developing countries (see for example, Okigbo, 1968, who uses data for Nigeria), they have tended in recent years to be rejected as a source. In one sense, this is not surprising. Income taxes only cover a part, sometimes a very small part, of the population. The resulting data cannot provide a picture of the overall distribution.

Christopher S. Adam (Oxford University and IGC), Pantaleo J. Kessy (Bank of Tanzania and IGC), Johnson J. Nyella (Bank of Tanzania) and Stephen A. O'Connell (Swarthmore and IGC) develop an econometric model of the demand for money in Tanzania - using quarterly data from 1998 to the present - and discuss the policy implications of the results, including their relevance to the velocity-forecasting exercise that plays a key role in the Central Bank of Tanzania’s policy framework. This paper is the outcome of research collaboration between staff of the Bank of Tanzania and the IGC.