A big project challenge

bill_anderson · January 9, 2018, 10:23am

I have been following the development of Ghana’s new national identification system with interest. $50m has been spent on it and I was keen to find out where this came from.

A quick d-portal search (“Ghana identity”) leads me to a $97m World Bank loan for a “GH eTransform Ghana” project and its description tells me the project has four components of which the national id system is the second. The description has some details but the text is curtailed. From a link provided to a Project Appraisal Document I could get a better picture of this project.

That’s a lot of different things which leads to a complicated sector split …

$25m has been disbursed to date, but it isn’t clear which components have received funds. (Interestingly in the OECD CRS the activity is split and reported in separate rows per sector so there is more useful information on spend than in IATI.)

At this point I’m none the wiser on where the $50m Ghanaian spend has come from. The documents linked on IATI give me no clue so I go to the World Bank site where there are a load more documents and I find in the latest Implementation Status & Results Report that

In light of the Government’s decision to finance the digital ID component from its own funds, all activities related to the digital ID were cancelled.

Took me three hours to get there…

A few questions:

Are big complex projects manageable from an accountability point of view?
Should IATI be doing a better job structuring information about what is actually happening on the ground?
If an IATI activity is always going to provide only a partial picture should we be bothering about trying to make the standard more comprehensive.

ariag · January 10, 2018, 4:19pm

As someone who has played this game many times, I am frustrated and annoyed for you. From my work, it seems the best way to make complex projects manageable from an accountability point of view is to do a better job structuring (and reporting) information about what is actually happening on the ground. I think this is the most important piece, the open ag study and mine both highlight the need to know what is happening sub-nationally so being able to do that is critical. As a follow on to the first question if big complex projects aren’t manageable for accountability we are failing, these are the projects that must be accountable, but maybe less who is funding did money get spent and more did this work happen, where, and does that match with what was supposed to happen. (Side note I think this community in the short time I have been a part of it has made so much progress and I firmly believe we are getting to a point where large project information can be used for accountability, and I’m really excited to keep tweaking and pushing to cross the ever moving finish line.) As for the third question, maybe this is where we need to be not just a standard but a helpful platform that links up to other data so that there is a one stop shop to find all the information so the different pieces of the picture are all there, but the standard house the piece of information that are unique to IATI, so instead of more complex we streamline the standard to what is most needed and relevant (This has to be driven by what is relevant to users not what is easy to publish). I think getting users to use IATI (and I mean d-portal because nobody is going to use xml) and explain what they want, what was missing what was hard to find what else would work will help us get there.

markbrough · January 11, 2018, 4:24pm

Nice post. Rather than any issue with IATI itself, this suggests to me a couple of improvements to recommend the World Bank make to their own publication:

Publish all documents, not only a subset (what are the criteria for the existing subset?)
Publish components of projects in addition to the projects themselves, if possible (is this information captured in World Bank systems in a structured way?)

There are also other challenges with World Bank data, including absence of trust fund projects and timeliness (though it should be noted that the Bank also has some cool features in their IATI publication like pretty detailed results data).

What is the best way of conveying these kind of requests for improvements to publishers?