Proposal to add a new code to ``related-activity`` codelist to declare an ``iati-activity`` that is from another publisher, but talking about the "same" activity.
See initial discussion: https://github.com/IATI/IATI-Schemas/issues/83
Proposal to add a new code to ``related-activity`` codelist to declare an ``iati-activity`` that is from another publisher, but talking about the "same" activity.
See initial discussion: https://github.com/IATI/IATI-Schemas/issues/83
Moving to the 1.05 upgrade forum for consideration.
I am not sure this is needed. IATI publishers should only report those activities for which they are responsible, and not the activities from another publisher (publish once only!).
In the case a donor has an activity which funds exactly one activity of an implementing organisation, I would consider that as two distinct activities, each with their own iati-identifier:
- A funding activity of the providing organisation (with its own set of atrtibutes). The providing organisation is donor is responsible for reporting this activity
- A activity of the recipient organisation also with its own set of attributes
- Linked by one or more incoming fund (IF) transactions.The IF transactions allows to refer to the funding activity.
The same mechanism could be used when one donor activity fund multiple activities of the recipient organisation.
So there seems not to a need to introduce a new code. Are there any examples that one activity being reported by more than one publsher reflects a business (in contrast to reporting technical) need?
Hi Herman
Thanks for the feedback
This originated from an observation that the schema descriptions for iati-identifier and other-identifier were contradictory (https://github.com/IATI/IATI-Schemas/issues/83)
It seems that "other-identifier" was described as being there to provide reference to other activities published by other organisations, yet it was actually used for internal references that the reporting-org may want to disclose.
Hence, this proposal was for the function to enable references to other iati-activities outside of the corpus of a single reporting-org. It could be useful for one org to make reference to the fact that the same activity is published elsewhere...
Granted, that traceability via transactions (IF is one way, but surely any transaction can create a link?) is a very useful way to make these links, but this code would also enable organisations to reference activities that are not necessarily a direct chain/link.
Of course, this all depends on take-up, and knowledge of other publishers/activities - but it could be a useful means to create further linkages?
Hi Steven,
Shouldn't there be guidelines how to link up activiteies in the chain? Because there so many possible ways to do this, it will make the use of the IATI data for chain transparency hard. Therefore I am a bit reluctant to endorse this functionality: this is yet another technical possibilty to create linakges.(apart from transactions and participating organisations). Do we really need this? Maybe we do, but where does it fit in the broader picture.
Hi Herman
Thanks for the post - appreciated.
As you indicate, this could be an additional mechanism to talk about traceability, but without the financial granularity presented by transaction/provider-org/@provider-activity-id for example.
As mentioned about - this new code could be a means to reference an activity that is within a chain, but not linked via transactions. An implementation agency could reference an activity published by a donor (for example) two or three links up the chain - whereby funds are not transferred directly. Does that make sense? Of course, this could just be one instance. How about activities that are published across federated organisations, for example - they may want to express a relation, but not necessarily financial.
As far as I can see, participating-org is a less precise method to link actual activities, as all that can be declared is the org reference. There is no possibility to link two activities with the participating-org element (adding an activity-id attribute to participating-org would be a separate suggestion, and not for 1.05)
Hence, for any kind of linkages between different publishers, we could envisage the following methodology:
1 - linkage of activities via funds/transaction
Enables precise linkage
2 - linked of activities via related-activity
Can express general linkages, and/or where funds are not linked
3 - linkages via participating-org (preferable with @ref)
Expresses organisations are involved in this activity, but no link to funds or actual activities
Agreed, that fuller guidance would be needed around this - but hope this furthers the discussion
The change as implemented is:
4 - Multifunded - A multifunded, or co-funded activity. The identifier should be globally unique and shared by all reporters of this activity.
Becomes:
4 - Co-funded - An activity that receives funding from more than one organisation.
And we add:
5 - Third Party - A report by another organisation on the same activity (excluding activities reported as part of financial transactions - eg. provider-activity-id - or a co-funded activity using code = 4)
This has been moved to 2.02 Upgrade Proposals, as it is now most relevant to the forthcoming decimal upgrade.
This proposal has been discussed amongst the IATI Technical Team following the end of the initial suggestion phase of the v2.02 upgrade.
As noted above, the IATI Technical Team highlight that this proposal has already been implemented into version 2.01 of the IATI standard. Therefore, no further action will be taken.
This proposal has been archived as it is not being taken forward in the version 2.02 upgrade of the IATI Standard.
10 Comments