Deprecate OrganisationRole code for “Accountable” (excluded 2.03)

This proposal is part of the 2.03 upgrade process, please comment by replying below.


**Schema Object** iati-activity/participating-org/@role

**Type of Change** Deprecate code

**Issue** There is no commonly agreed definition of an “Accountable” organisation and the use of this classification is therefore confusing.

**Proposal** Deprecate value “2 - Accountable” from [OrganisationRole]( codelist

**Standards Day** No information available

**Links** N/A

From the procedural point of view this is a breaking change, so it should be proposed for the next decimal upgrade.

I would also like to do another attempt to clarify the definition of accountable: the accountable organization can be hold accountable for the execution of the activity. When you only publish your own activities, you are accountable for those activities. When you publish activities on behalf of other organizations, those other organizations are accountable.

Just want to flag that we had a similar discussion on this a few years ago.

The original intention for the definition of “Accountable” was the country institution with whom an aid agreement has been signed. For example:

  • Ministry of Agriculture implements the project => Implementing
  • Ministry of Finance signs the agreement/MoU => Accountable

This field was originally proposed in the context of enabling alignment with country budgets - see 2010 Steering Committee paper. There is some discussion on it in the 2012 Study on Reflecting Aid Flows in Country Budgets:

There are also cases where the accountable institution – i.e. the institution signing the aid agreement – may not be the institution benefitting from the funds.

IATI Guidance from 2016 also points in the same direction.

If we follow this definition, there would be a limited number of projects that this is related to, and it would be particularly focused on projects funded by official donors. Following the finalisation of the methodology for aligning aid with country budgets, it is worth considering whether this field is still needed. However, even with that methodology, in the example above, the information that “Ministry of Finance signed the agreement” would only be visible if we retained the accountable organisation. I guess the question is: how valuable is that information?

In any case, I think this field could definitely do with a clearer definition than the one currently published on the IATI Standard website.

I agree with Herman and Mark that the description needs updated and we need to establish a common understanding.

The role description “An organisation responsible for oversight of the activity and its outcomes” does not mention anything about country institutions.

I haven’t been in a situation where I have worried about the definition - I have understood it to include Financial accountability, as a relevant code-value to use to tag the role of your organisation in a particular Network of partners. Of course, the general assumption would be that any organisation that receives any Funding will become accountable for that, but if you are situated e.g.3 levels downstream from the Funding organisation it would be worthwhile if you use this tag to declare that you arry the (financial) accountability towards the partner at level 2, on behalf of the implementing partner at level 4 - and even though the partner at level 2 is bound to be equally accountable to the Funding organisation at level 1, it would only present itself as ‘extending’ from your perspective. I have found that this makes sense, and is an important element in any attempt to provide guiding principles on the nature of data expected to be delivered by different partners - depending on their role. I would be happy to get back with more info on that ambition, but fear that it will be after deadline for this upgrade.

The above discussion proves the level of confusion that exists about this particular org role and would support its deprecation.

1 Like

I think I would probably also be in favour of deprecation as long as we are sure that this is the case and don’t have to add this (or something similar) back in in future:

I have seen that such information is captured in several systems but I wonder if others can speak to this – how useful is it, and how likely is it that data provided according to this definition would actually be any good?

We understand there to be consensus to modify the definition of “Accountable” and NOT to deprecate this code.


Notes from consultation calls w/c 3rd July

The proposal was reviewed by those on the call and there was no objection from the group about keeping organisation role code “Accountable”, however, a modification of the description was put forward as follows:
“An organisation that can be held accountable for the overall execution of the entire activity. If no accountable organisation is named, it is assumed that the reporting organisation is accountable for the activity.”

There was some discussion to what extent the organisation role code “accountable” is being used by publishers and data visualisation platforms

In the latest update on 2.03 upgrade process, this modification is flagged in yellow (unclear on state of consensus).

Reading the discussion yet again, I see:

  • a lot of confusion on the meaning of this code;
  • some support for deprecation;
  • concerns that perhaps we’ll realize later on that in fact we needed it, hence should proceed with caution;
  • evidence that the few publishers using this code use it to signal very different things;
  • no convincing (any) use case for this org role.

The proposed new definition “that can be held accountable for the execution of the activity” is a big departure from the original definition (which Mark documented early in the thread). It’s also the definition of an implemeting agency (role 4), not necessarily the reporting organization. And it does not align with for instance OJ’s understanding of what “accountable” means.

Let’s try again: what is the use case for this org role? What needs to be tracked that is not achieved via proper use of the other roles + traceability? A clear answer to this may help come up with the right definition (or achieve consensus on deprecating it).

I haven’t followed this thread, so please forgive me if I have missed something or repeat something.

The attached document is the way I interpret it at the moment.

This is not necessarily the way we are implementing it at SC UK, at least not completely. When we receive awards through a managing agent we report them as the extending org. We always report ourselves as the accountable org because we are accountable to the originating donor for what is delivered to the beneficiaries. In my interpretation (which may be wrong), any organisation between the funding organisation and the implementing organisation is also an extending organisation.

I stress that this is my personal interpretation – I don’t claim that it is a SC UK official position or that it is accurate with respect to IATI’s intentions. It would be very useful to all concerned, I think, if IATI would publish a diagram similar to the attached to make this easy to grasp on a single page. (Was it Dijkstra who said you should be able to fit any thought structure on a single page?)

Bob Walkden
Business Architect
Save the Children UK


Apologies all. In trying to maintain a traceable history of posts this has got very confusing.

  • The original problem we were trying to solve was that it is difficult for data users to identify the ‘owner’ of an activity published by a secondary publisher. (ie where did the publisher get the data from). (see Standards Day Notes Row 94)
  • A proposal was subsequently introduced to add “Responsible” as a role.
  • In addition a proposal was made to deprecate “Accountable” as there were a variety of interpretations as to what it meant.
  • During consultations there was consensus that “Responsible” was unnecessary as the secondary publisher problem could be solved by using “Accountable” with a revised definition (and we therefore shouldn’t deprecate “Accountable”)
  • It was then pointed out that this new definition was inconsistent with the original definition.

So where I think we are is:

  • No one wants a new “Responsible” code
  • There is currently no consensus as to whether we should deprecate “Accountable” or not
  • There is currently no consensus on a coherent definition for “Accountable”

Personally I recommend that either we focus our minds on an acceptable definition of “Accountable” or leave things as they are and return to this at a later date. I do not recommend deprecating “Accountable” (unless the discussion on a definition takes us there).

Discussions on deprecating “accountable” predate the issue of secondary publishers. This role has been totally confusing from the beginning; few publishers use it, and those who do have simply decided on their own definition rather than using IATI’s.

I suppose we could come up with a clearer definition, but would ask again about the actual, concrete use case. Unless we can agree on a use case, I suspect publishers will continue to use this code with their own definition.

I say leave things as they are for now. If a clear use case emerges, a definition will likely emerge as well and we can deal with it then.

That would be my preference, too.

@bill_anderson @YohannaLoucheur Agree, if we change something it should address a real problem.

Yeah agree with leaving things as they are for now, and try and fix this next time around, once we’re clearer about what information this role should be trying to convey

I agree to leave it for now to avoid further confusion.

But we should flag “Roles and their meanings” for a discussion at TAG level. Should we break this out to another thread and start again?

I also think that with all the discussion of traceability, we need to remember that role definitions are for this activity, not one further up or down the chain.