This paper examines the challenge of achieving\ud a balance between the implementation of centrally\ud designed pro-poor policies and the decentralization of\ud responsibilities to local governments in many African\ud countries. It analyzes the implementation of the Rural\ud Water Supply and Sanitation Program in Tanzania. Key\ud mechanisms for planning and allocating resources are\ud analyzed at ministry, district, and village levels. Results\ud show that a mixture of policy incoherencies, technical\ud shortcomings and political influence determine that only a\ud small proportion of funds reaches the underserved areas.\ud We argue that a greater connection between the bottom-up\ud and top-down planning mechanisms, and a sharp increase\ud of downwards accountability are needed before decentral-\ud ized decision-making result in better resources allocation.\ud Meanwhile a bigger intervention from central government\ud is needed.Peer ReviewedPostprint (published version

This paper reports the findings of three water point mapping (WPM) studies carried out in three rural districts of Tanzania covering 3,363 water points. The methodology was designed to estimate the influence that consideration of the quality of the water supplied and the year-round functionality of each water point would have on each district's coverage figures. To this end, the study included measurements of basic quality parameters (692 analyses) and characterisation of the year-round continuity of service, in addition to the data collected in standard WPM campaigns. Both the quality and year-round continuity results were analysed in a disaggregated form by water point technology. The results show that 22% of the improved water points analysed had more than 10 CFU (colony-forming units)/100 ml of water and 19% were seasonal. Moreover, water service coverage in the districts studied fell 40% on average when quality and year-round continuity were considered. There is a strong need to include these basic factors in the minimum standards to be delivered and monitored in rural areas. Simple and efficient methodologies for including quality and year-round continuity measurements in the information routines, such as the one presented here, are a necessary step to this end.Peer ReviewedPostprint (published version