DAC WP-STAT group redefining purpose codes (sector codes) to align with SDGs


(SJohns) #1

WP-STAT, the group that defines the codes used by donors for reporting on overseas development assistance recently discussed a proposal to redefine the DAC purpose codes (which are used in IATI as the main source of sector codes) to better align with the Sustainable Development Goals: http://www.oecd.org/officialdocuments/publicdisplaydocumentpdf/?cote=DCD/DAC/STAT(2015)9&docLanguage=En

Does anyone know if any IATI members (through also being donors who are members of DAC) have been or will be part of these discussions? It would be good to follow this process as it will have a direct and hopefully positive impact on how we report and tag our development projects against the SDGs.


IATI and SDGs - untangling threads
(Wendy Rogers) #2

Thanks for highlighting Sarah. Although we (the IATI Tech Support Team) are not members of the DAC we will of course update the DAC Sector codelists (eg http://iatistandard.org/202/codelists/Sector/) that are available to IATI publishers as soon as we are notified by the DAC of any changes as and when they are made.

Also I just wanted to remind publishers that as and when they are able to move to publishing their data at V2.02 of the IATI Standard they will also have the option of directly referencing their activities to the SDGs by using the new vocabularies that have been added to the list of Sector Vocabularies - see codes 7 to 9 of http://iatistandard.org/202/codelists/SectorVocabulary/


(SJohns) #3

Is it possible within the standard for organisations to assign more than 1 sector code vocab to an activity (at activity level)? ie so if an organisation wanted to use both the DAC code and the SDG Indicator, could they do that within the rules of IATI? I should probably know this but have brain freeze …


(Wendy Rogers) #4

Hi @SJohns and you can unfreeze your brain because it is allowed to specify more than one sector vocabulary. The only rule is that if the percentage attribute is used then it must of course add up to 100 for each vocabulary used see http://iatistandard.org/202/activity-standard/overview/classifications/

For information there are currently already a number of publishers who currently include DAC sector codes (@vocabulary = 1) but who also publish their own internal sector codes as well (@vocabulary = 99)


(Yohanna Loucheur) #5

Hi Wendy
In relation to the above, someone in our department signaled that the IATI version of the DAC sector codelist has not been updated with the new energy codes that were added some months ago following a WP-STAT decision.

This leads me to believe that the much bigger recent modification to the sector codelists, where over 40 codes were added to align with country budgets (as IATI’s request), has probably not been reflected yet in the IATI codelist either. As we hope to see these codes being used soon, it would be good to update the codelist as quickly as possible.

Beyond this, I wonder whether it would be worth creating a regular watch of WP-STAT decisions related to these codelists so that any change on the DAC side can be reflected in the IATI lists in a timely manner?


(Wendy Rogers) #6

Hi @YohannaLoucheur just to let you know that we have started the process for adding the new codes to align with country budgets to the DAC Sector codelist (please see here)

Also, we were aware that there had been some other previous updates to the DAC purpose codes that we don’t appear to have received any notification about? We usually use the email notification that we receive regarding updates to the DAC lists as our trigger for updating the codelists. We are therefore looking into this issue to avoid missing any updates in the future. In addition, we have also scheduled a full cross check audit of all our codelists derived from the DAC codelists so that we can bring the two lists back into step ASAP.


(Yohanna Loucheur) #7

Stumbled on this thread again today as I try to clarify the rules around using non DAC sector codes.

In the quote above, Wendy implies that the percentage attribute is not mandatory (“if the percentage attribute is used”). However, the guidance on sector codes is not quite so clear:
“Note that if multiple sector codes are used in multiple vocabularies, then each vocabulary’s percentages should add up to 100%” and also “All reported sectors from the same vocabulary MUST add up to 100%” http://reference.iatistandard.org/202/activity-standard/iati-activities/iati-activity/sector/

On the other hand, the schema does show the percentage attribute as optional: http://reference.iatistandard.org/203/schema/downloads/iati-activities-schema.xsd

The way I understand the above, if multiple codes are reported in a non-DAC sector vocabulary (e.g. SDGs, COFOG or any of the other vocabularies in the SectorVocabulary codelist) without a percentage attribute, users will assume an even split across the codes - rather than using them as flags. Is this what others understand as well?


(Andy Lulham) #8

@YohannaLoucheur there’s a parallel discussion regarding sector percentages over here:


(Yohanna Loucheur) #9

Thanks Andy for reminding me of that other discussion! Having re-read it, I think they are related but slightly different.

My question is rather whether percentages are treated as if they were mandatory - in the sense that even if percentages are not provided, users/platforms will interpret the codes as having percentages, and will split the project value over them.

The alternative would be that if percentages are not provided, users/platforms do not attempt to plit the project value over the sector codes - treating them as simple flags.

The exemples you provided in the other discussion all seem to demonstrate that indeed, sectors are assumed to come with percentages, irrespective of whether these are specified or not.


(Yohanna Loucheur) #10

Perhaps I should share where that question comes from:

The WP-STAT recently approved the creation of a new CRS field to report SDG targets (and goals, temporarily) as flags. https://one.oecd.org/document/DCD/DAC/STAT(2018)41/en/pdf

(TL;DR) Here are the central elements of the proposal:

  • The Secretariat’s proposal is to track the SDG focus of development co-operation at the target level and to allow the possibility of reporting at the goal level for a transitional period.
  • the Secretariat’s proposal is to allow multiple entries in the SDG field, with a maximum number of 10 reportable targets
  • the Secretariat’s proposal is to allow the SDG field information to be recorded with a qualitative flag, indicating to which SDG targets the activity is contributing.

The SDG goals and targets are already non-embedded code lists in the IATI SectorVocabulary element. However, if the Sector element is used as having implicit percentages, it is not the right element to publish SDG targets in a way that will align with CRS data. It does not look like the Policy Marker element would work either, since the CRS SDG target field does not have significance levels.

Would be very interested in thoughts from IATI community.


IATI and SDGs - untangling threads
(Mark Brough) #11

Yes, I think this is correct when using DAC codes. In Bangladesh, if percentages are not provided (or do not add up to 100%) then we automatically give them percentages (e.g. dividing 100 by number of activities or rescaling if they don’t add up to 100).

I would guess that the usage of the SDG targets etc. is still relatively limited in IATI data, so there could be a rule that only sectors using the DAC CRS purpose code vocabulary should be assigned % figures.

However, I think a much better way to do this would be to use the new (since 2.03) tag element. I guess we need a couple of new vocabularies to be copied over from the SectorVocabulary codelist to the TagVocabulary codelist to cover SDGs. Additionally: should we also not be trying to keep more consistency between these two codelists anyway? I can see situations where non-statistical “tagging” of CRS codes could also potentially be useful for example.

Updated: @andylolz just pointed out that a version of the tag proposal on Discuss was approved, so I’ve updated my comment.


(Yohanna Loucheur) #12

Thanks Mark! Had forgotten about the tag element in 2.03 (since GAC hasn’t moved to 2.03 yet - this would actually be a great incentive to finally do it).