Institute for Development Studies, University of Nairobi

This paper seeks to examine the impact Marketing Board operation has had on the level of cotton producer prices as well as the success of marketing policy in stabilising these prices. A model of the Ugandan cotton sector is presented in a relationship between a set of endogenous policy target variables (i.e. interseasonal and intraseasonal producer price variability) and a set of exogenous variables and parameters (i.e. producer prices, ginners' allowance, export tax rates, etc,) some subset of which are potential instruments of policy. Within these relationships, we analyse the interrelatedness of unit export revenue, ginners' allowance, ginners' prices as well as producer prices. We come to the conclusion that in the price determination exercise performed by the Marketing Board, producer prices are a "residual" itm. Furthermore, we analyse and quantify the extent to which the Marketing Board has reduced the variability of producer prices, we conclude that prior to 1965, the level of prices set was frequently inoptimal and this encouraged intraseasonal fluctuations. However, we find that the Board has absorbed 15% of the total interseasonal instability, but because of the residual nature of producer prices in the disposal of unit export revenue, the price variable could not have been a "policy instrument" in the models. Therefore, prices could not have been positively adjusted so as to minimise their variability. Export tax rates do not seem to have been varied for the purpose of stabilising prices either. Considering that this tax is unjustifiable from the point of view of world demand elasticity for Uganda cotton as well as the fact that the tax is resource distortive, we are then left with only one justification for this tax: a second-best solution to the government's fiscal needs. Finally, we observe that the Board's administrative costs have expanded out of proportion with the Board's level of operation due to internal inefficiency and political pressure on the Board to expand its bureaucracy as employer of "last resort". Nevertheless, we are led to conclude that a neutral tax such as a land tax with less "announcement effects" could be used to obtain xhe necessary fiscal revenue that might then be deployed to employ people in more productive occupations than marketing the cotton crop.