Hi David,

Thanks for the feedback.

The lack of sectors really comes out as downturn if we compile all of that data in a single data-store (such as openaidsearch.org for example). I could make the case more elaborate to show the actual downturn, because I see IATI data in a single data-store lacking trustworthiness in the context of research (academic, media etc). Not being able to fully and completely retrieve for example:

• Show me ALL activities in regions Asia + Africa in 1998 in sectors Privatisation +Water Sanitation will simply render a incomplete answer to the user

Just based on this example IATI will become unsuitable for research of aid development data. So from my point of view we should not just depend on feedback (like mine) but rather make the case for mandatory namespaces in IATI.  If we were to decide to just provide guidance on that, the process of having a complete IATI overview will take a long time or even worse, it may just never happen. Which would be a real shame and missed opportunity. 

Siem

Thanks Siem

I think this proposal would cause backwards compatibility problems (but please feel free to challenge on that). It may be that through feedback from users of the data, such as yourself, that providers will start to improve their data.

I'm not really in a position to make this call, but it may be that improved guidance to providers comes out of this request for now.

However, this suggestion is also past the deadline for the 1.03 process, so I'm moving it into the general Modifications, Additions, Improvements forum for continued consideration.

Thanks

Siem

Just to note that there are other reasons why that query will not produce complete data whihc I'm sure you are aware of:

- DAC region codes are both  incomplete and overlapping

-  multiple incompatible region vocabularies

- confusion over when region codes should be include - only when there is are no countries, when some countries are not in a region, both countries and their region.

- lack of clarity about which dates to use

and of course the ultimate problem, that IATI data is a subset of unknown coverage of all Aid activity.

While we agree with making the sector element mandatory, we cannot make the use of CRS Purpose Codes mandatory: we can however improve our guidance to strongly recommend the use of CRS purpose codes.

Strongly agree that at least one sector should exist. However, this would not solve the problem on openaidsearch.org because that’s more an issue about DAC/CRS purpose codes not being used, rather than no sectors being published at all. Recommend altering guidance to strongly encourage use of CRS purpose codes; recommend agencies' own sector codes should also be published where available.

Wonder about the extent to which DAC/CRS purpose codes are appropriate classifications for the projects of different sorts of organisations (traditional, DFIs, south-south, humanitarian, etc.) - a review of this could be very useful in considering changes to guidance or requirements to use a single sector vocabulary and thereby ensure comparability of data.