22 Apr 2011

Mapping for Results - what does this World Bank project tell us about Tanzania?

A lot of people have been pushing recently at the link between mapping and accountability. Whether it's detailed local maps of reported crime in the UK or East Africa's own Ushahidi platform, the internet and mobile phones are enabling new map-based ways of collecting, visualising and sharing information that can potentially be used to hold decision makers to account.

The most recent example comes from the World Bank. They recently published their Mapping for Results site, which presents (on a map, of course) details of 1250 current World Bank-financed projects in over 16,000 locations in 79 countries. Each location has a marker that can be clicked to reveal more details of the project and its location.

For Tanzania, the map shows 40 financed activities in 524 locations:

Mapping for Results - Tanzania map (click to enlarge)

The accountability potential is obvious. Look at the map above, for example, and you'll see relatively few project locations in the west of the country - Kigoma, Tabora and Rukwa regions - compared to much more dense concentrations of projects in the north and east. We can use this to start asking questions about why this is the case? Is it that these three regions are a long way from Dar and relatively inaccessible? Does distance matter because it makes projects more expensive to implement or more time consuming for busy Bank staff to prepare, or simply because it makes problems in those areas less obviously visible?

Being able to meaningfully ask such questions depends on the data being accurate. This being a World Bank initiative, they've taken care to add disclaimers like "not all data has been validated by TTLs, may be updated, and be subject to change" and "the precision of each project depends on the quality of the publicly available project documentation.".

But they do seem pretty confident in the reliability and detail of the maps: "The maps display the number of activities of a given World Bank project. For example a project in health may have different hospital locations; the project activities layer will show all these locations on the map."

However, there seems to be some lack of consistency in this - the site also states that national and budget support projects are not included on the map, so different locations for national projects are not shown.

For the water sector in Tanzania, for example, three bank projects are listed - the Lake Victoria Environmental Management Project, the Dar es Salaam Water Supply and Sanitation Project, and the Water Sector Support Project (known to everyone outside the Bank as the Water Sector Development Programme). The first two, which are much smaller projects, are linked to a number of locations and displayed on the map, while the much larger project - the WSSP/WSDP - has just one marker on the map (highlighted in the image above), placed seemingly randomly somewhere near Chalinze.

The reasons for this are pretty obvious - that mapping 10 villages in every district that had been selected for rural water supply projects with WSDP support would be a major undertaking and probably impossible for a team in Washington. And that's before you get to urban water supply or water resource management projects. But to omit the biggest projects (and I'm sure there are other similarly dominant projects in other sectors that are also effectively omitted), is a pretty serious weakness for the map. Besides making the map misleading, this omission also has the potential to undermine moves towards improved donor coordination since many national projects will also be multi-donor projects.

I think it's worth thinking back to Owen Barder's recent posts on aid transparency - his main point being a call for transparency to be citizen-centred rather than donor-centred. For maps, that means disaggregation to the lowest level possible, giving some sense of the size of a project. And to be honest, to most people in Tanzania it don't really matter who's funding a project - a map showing projects by all donors (and even those funded by local revenues) would be far more useful to us than a map showing only WB-financed projects.

I should not put too much weight on these criticisms - after all, the World Bank's moves towards greater transparency are positive. This is also a first effort, based on information from Project Appraisal Documents, etc. that are explicitly donor-centric documents - the information wasn't collected with this purpose in mind. And the maps are designed, at least in part, to show other donors what is possible. I'm sure they will improve over time. 

If future maps can break down bigger projects to a more meaningful level and be extended to include non-WB projects, they will become a very useful tool. Until then, it will be most useful as a demonstration.