In the Standards Day at the 2017 IATI TAG in Dar es Salaam, we agreed to deprecate the BudgetIdentifier codelist and provide communications materials to publishers on implementing aid and budget alignment.

This guidance explains how the methodology now works. It stems from work conducted by the IATI TAG Working Group on Aid and Budget Alignment, and reflects decisions made by the WP-STAT and IATI Steering Committee. In summary:

  1. Use the capital-spend field
  2. Use detailed CRS purpose codes
    a) don’t use very broad purpose codes like 43010 - Multisector aid
    b) use the more detailed “voluntary” CRS purpose codes.

1) Publish the capital-spend field

For activities at the commitment stage, the percentage of an activity that is capital expenditure should be published using the capital-spend field. The definition of capital expenditure follows the IMF GFS definition:

Capital spending is generally defined as physical assets with a useful life of more than one year. But it also includes capital improvements or the rehabilitation of physical assets that enhance or extend the useful life of the asset (as distinct from repair or maintenance, which assures that the asset is functional for its planned life). Capital includes all aspects of design and construction that are required to make the asset operational. Source: D Jacobs, 2009, Capital Expenditures and the Budget, IMF PFM Technical Guidance Note No 8. (IMF, Washington)]

This definition was approved by WP-STAT under the written procedure in February 2016.

2) Publish detailed CRS purpose codes

(a) Don’t use very broad codes

Broad “multisector aid” or “sector not specified” purpose codes are not mappable to budgets and are too aggregate to tell you much that is useful about the activity. Publishers should not use these purpose codes if they want activities to be mappable to budgets.

List of multisector codes that should not be used (source: IATI 2012 Study on Better Reflecting Aid Flows in Country Budgets, Annex 4 - codes that are “not classifiable - multisector”):

  • 43010 - Multisector aid
  • 43050 - Non-agricultural alternative development
  • 43081 - Multisector education / training
  • 43082 - Research / scientific institutions
  • 52010 - Food aid/food security programmes
  • 99810 - Sectors not specified

(b) Use the more detailed “voluntary” CRS purpose codes, not “parent” codes

In a fairly broad sample of 35-40 countries, various studies found that more disaggregated (“voluntary”) codes under 15 existing CRS purpose codes were needed to map to budgets. Publishers should use the voluntary codes instead of the “parent” codes if they want their activities to be mappable to budgets.

The CRS purpose codes, including the more detailed voluntary codes, are listed in the DAC and CRS codelists Excel sheet, which has also been provided in a more accessible format by [~336].

The additional codes were approved by WP-STAT under the written procedure in February 2016. The IATI Steering Committee approved the use of the more detailed CRS purpose codes (instead of using the “Common Code" codelist) in March 2014.


Feedback on this guidance is welcome.

Comments (33)

Yohanna Loucheur

Mark, thanks for sharing this.

For clarification, the SC approval of the use of the more detailed CRS purpose codes was in March 2014 (the document’s url is misleading, and for some reason the paper doesn’t have a proper header - perhaps something the Secretariat could rectify? [~545] ?).

It’s document 4c on the agenda of the March 2014 meeting: http://www.aidtransparency.net/governance/steering-committee/steering-committee-documents/2014-meetings

Mark Brough

I’m not sure whether this needs to be part of the decimal upgrade process to become Guidance? I sort of presumed not, given that it is just re-stating existing decisions, but if [~367], [~522] and others think it does need to go through the decimal upgrade process then I’m proposing it for inclusion in 2.03. (I am also a bit unsure whether guidance needs to go through the decimal upgrade process in general?)

NB deprecating the BudgetIdentifier codelist was accepted at Standards Day on the basis that guidance / comms to publishers would be provided on how to now implement aid and budget integration.

Petya Kangalova - IATI Secretariat

[~469] Yes, I checked with [~522] and you are correct that amending the Guidance does not form part of the upgrade process. I have now moved the post to Standard Management-> Modifications, Additions, Improvements

Petya Kangalova - IATI Secretariat

[~336] While guidance does not need to be approved as part of the upgrade, we have included a number of guidance proposals that were discussed at the TAG into the 2.03 upgrade index so that they can be added alongside the upgrade process, if not earlier. I will need to check with [~522] about specific deadline for adding them.

Andy Lulham

Just spotted something, [~469]:

markbrough:

Publishers should use the voluntary codes instead of the “parent” codes

This line is very important – I think it would be clearer to say this in the first line of 2b. So instead of:

(b) Use the more detailed “voluntary” CRS purpose codes

you could say:

(b) Use the more detailed “voluntary” CRS purpose codes, not “parent” codes

Mark Brough

Thanks for flagging this [~336], agree it would be a good addition. [~353] - I cannot edit my post above anymore, so please can you add this in? Also, what is the progress on adding this guidance (I presume it would be to the IATI Standard website)? I am not really sure what needs to happen next.

Petya Kangalova - IATI Secretariat

[~469] I have amended the text in the proposal. As with other guidance (see point 3 in the index), we will be following the same timetable for the upgrade process and will add to the IATI standard website.

Steven Flower

HI

markbrough:

deprecate the BudgetIdentifier codelist

I’m interested in how this will happen. The rest of this thread seems to be about adding / changing guidance around the standard. If the actual BudgetIdentifier codelist is to be acted upon, then something needs to happen?

For reference / more widely, I don’t think we should do similar to OrganisationIdentifier “codelist”, which has some narrative to declare that it is deprecated, but is still “included” in the standard.

Steven Flower

Other thing.

[~472] mentions

Obviously, CRS purpose codes are published in Sector element. Country-budget-item no longer needed - but kept for potential future needs

(see full thread):

Am I correct in thinking that this proposal misses out this point. Does it need to be made explicit that publishers should use sector rather than country-budget-item ?

(In turn, I think this then starts to raise complexities in terms of multiple sectors and, even, secondary sectors, potentially. Sorry)

Mark Brough

Thanks for the feedback on this [~433]! On deprecation, I do not know how this is supposed to happen (I am not aware of any particular process around deprecation). But yes, this thread is very much focused on providing guidance.

stevieflow:

Does it need to be made explicit that publishers should use sector rather than country-budget-item ?

Yeah perhaps. Maybe the point 2. in the draft guidance should be changed to:

markbrough:

Use detailed CRS purpose codes in the <sector> element

However, I am not sure what you mean by complexities in terms of multiple or secondary sectors. You use these sectors / more detailed purpose codes in the same way as you would use any other CRS purpose codes.

Mark Brough

Thanks for the feedback on this [~433]! On deprecation, I do not know how this is supposed to happen (I am not aware of any particular process around deprecation). But yes, this thread is very much focused on providing guidance.

stevieflow:

Does it need to be made explicit that publishers should use sector rather than country-budget-item ?

Yeah perhaps. Maybe the point 2. in the draft guidance should be changed to:

markbrough:

Use detailed CRS purpose codes in the <sector> element

However, I am not sure what you mean by complexities in terms of multiple or secondary sectors. You use these sectors / more detailed purpose codes in the same way as you would use any other CRS purpose codes.

Steven Flower

Great, thanks Mark

markbrough:

However, I am not sure what you mean by complexities in terms of multiple or secondary sectors. You use these sectors / more detailed purpose codes in the same way as you would use any other CRS purpose codes.

Was just thinking of scenarios whereby publishers might want to maintain a sector code as a primary classification, but then also utilise the voluntary codes for potential “non-statistical” classifications. This might be a tangent though - so better to focus on the other parts of this thread!

Steven Flower

Im trying to figure out if anything around this has happened via 2.03.

Should the BudgetIdentifier codelist have been deprecated?

Any idea [~353] [~469] [~472] [~336] ?

I can see that the voluntary purpose codes are now in both the 2.03 5-digit sector list, and also the source OECD DAC XML ([~336] can confirm) - so that might now affect this:

markbrough:

The CRS purpose codes, including the more detailed voluntary codes, are listed in the DAC and CRS codelists Excel sheet, which has also been provided in a more accessible format by [~336].

Andy Lulham
stevieflow:

I can see that the voluntary purpose codes are now in both the 2.03 5-digit sector list, and also the source OECD DAC XML ([~336] can confirm)

I can indeed confirm this. In fact, the OECD DAC XML adds the “voluntary” purpose codes as codes (i.e. consistent with the way they’re included in the IATI codelist), and just flags the voluntary bit in the status (it says “vonlontary” but that’s a typo).

stevieflow:

Should the BudgetIdentifier codelist have been deprecated?

Searching back through this forum, it appears the answer to this is yes it should.

Andy Lulham

Bumping this. I found the relevant bit in the standards day notes (rows 102 and 103):

Proposal: Deprecate the BudgetIdentifier codelist
Outcome: Accept
Discussion: Consensus as codelist has been replaced by expanded CRS Purpose codes

Proposal: Remove reference to the BudgetIdentifier codelist in the definition of country-budget-items/budget-item/@code
Outcome: Accept
Discussion: We would deprecate at 2.03 and remove in integer, although it was noted that there is no process on the deprecation of individual elements. Comms would need to be provided to publishers about this, perhaps as separate guidelines.

I’m not sure what happened with these, but [~433] is right in pointing out that neither has been actioned in v2.03.

Relatedly, the aid and budget alignment guidelines were bumped to v2.03, but haven’t been added yet.

Steven Flower
andylolz:

I’m not sure what happened with these, but [~433] is right in pointing out that neither has been actioned in v2.03.

What’s the best way to remedy this [~453] ?

Yohanna Loucheur

One concrete consequence of this not being actioned is the recent report by Cohen and Mekuria that is based on the Budget ID codelist rather than the expanded CRS codelist. That’s a missed opportunity to see what IATI can tell us about aid flows to a specific country.

Bill Anderson
YohannaLoucheur:

One concrete consequence of this not being actioned

This is not correct.

If you read the acknowledgements in the report you will see that the methodology was written by Simon Parrish. This would have been before the agreement by WP-STAT of new CRS purpose codes.

Yohanna Loucheur

The methodology was written by Simon, but the report itself is more recent, having been published in January 2018 (thus presumably written in 2017). Though it does refer to the “common code” as a recent proposal, which is very confusing.

In any case, clearly marking this codelist as deprecated, as decided by the TAG a year ago, and adding the guidelines, would help avoid further confusion for both publishers and users of IATI data.

Bill Anderson

As part of the Version 2.03 upgrade the Budget Identifier Vocabulary codelist was changed from being embedded to non-embedded.

The plan has always been to implement the deprecation of Code 1 in this list as part of the standard procedures for making changes to non-embedded codelists.

If I could add a personal note on this: It seems to me that a number of comments on this issue are getting close to the limit of what might be called “polite impatience”. Give us a break folks!

Andy Lulham

Gah… Polite impatience is exactly the tone I aim for. Apologies [~522].

bill_anderson:

The plan has always been to implement the deprecation of Code 1

Great! I didn’t see a github issue for this, so I’ve created one:
github.com/IATI/IATI-Codelists-NonEmbedded Mark BudgetIdentifierVocabulary code 1 as withdrawn (to indicate deprecation) opened 01:23PM - 23 Feb 18 UTC closed 03:57PM - 15 Aug 18 UTC andylolz andylolz

This comment on IATI discuss confirms that BudgetIdentifierVocabulary code 1 should be deprecated. I think as far as codelists are concerned, withdrawn...

I’ve also created the related pull request, and requisite discuss post.

I don’t think this addresses the accepted proposals from standards day mentioned above, though. Or the agreed plan to add the aid and budget alignment guidance, that this thread is about.

UPDATE: I think this pull request ought to fix the second accepted proposal mentioned above.

Mark Brough

I also want to bump this thread. I am really trying to be polite and patient but we will be celebrating this thread’s first birthday in one month – would be great to know how we can move this guidance forward. Is there a reason why it is taking so long to add this to the IATI website?

[~453] [~353] [~522] would you like a pull request for adding this guidance the IATI website? I am happy to provide. Please let me know (also if there is anywhere particular you would like it to go).

Andy Lulham

I’ve written a blog on this topic:

Publish What You Fund – 26 Apr 18 It's time to act on aid and budgets - Publish What You Fund

UPDATE 30/04/18: We are encouraged to see that IATI have committed to publishing their guidance on how to implement aid and budget alignment this week.   Information on aid that is aligned with country budget classifications is a key demand...

It includes a graph of all (more than 1 million) IATI activities, grouped by their 5-digit DAC sector code:

38

A large number of activities use sector codes that can’t be aligned to recipient country budgets. A good first step to addressing this would be to get this guidance published.

Mark Brough

This guidance has now been published by [~453] live on the IATI Standard website:

http://iatistandard.org/203/activity-standard/overview/country-budget-alignment/

Now that this guidance is finalised and live, it would be great if:

  • publishers can take a quick look at the guidance (it is pretty short and clear) and see what steps they could take towards making their data more useful as part of the budget process at country level;
  • organisations using the data can look at how they could stand to benefit from using this data, and how they can begin to integrate into their own systems.

Many thanks to [~353] and [~453] for getting this on the IATI Standard site, and to all of the organisations involved in the many years of work to get us to this point, particuarly Canada and [~472] .


Please log in or sign up to comment.