Our shared understanding of what IATI is, and is not,

Thanks Rolf Kleef and colleagues - great post

rolfkleef:

Attempts to make IATI into a detailed financial report or monitoring and evaluation framework can result in the creation of a top-heavy, complex and burdensome framework.

Just so I understand clearly - are you referencing “attempts” that are within the current set of proposals? If so, which? If not, are there wider attempts documented elsewhere?

(I’m guessing this is concerns about the current proposals, given that the post is in the 2.03 category)

Thanks

A most welcome intervention. I think this fits in with the emerging yet elusive desire to simplify the standard.

Generally …

• Is our problem the fact that we have many different stakeholders with different needs and priorities and the standard is trying to be all things to all people? In other words reaching a common understanding of what the standard is not as easy as it sounds (but necessary nevertheless)
• Can we learn from the Open Contracting standard which promotes the use of extensions in order to keep its core simple?
• Can we all agree on a set of guidelines which allow us to assess how trivial or disruptive proposed changes are?

Specifically …

• It would be helpful if you could turn your general concerns into specific comments on all those 2.03 topics that you feel need addressing.

Hi all,

A very pertinent conversation, and well-timed considering the amount of proposals relating to Results currently in discussion.

The changes that have been proposed in the 2.03 Upgrade come from consultation with NGOs who feel that the IATI Standard does not allow them to accurately describe their work. In all honesty, I believe this is the first time I have heard the phrase ‘snapshot’ with relation to publishing using the IATI Standard! (Although I admit, I was away for 14 months prior to April 2017, and it may have come up in my absence).

The key issue, from conversations that I have been involved in, is that donors have made IATI a reporting tool. For example:

Dutch MFA Reporting Guidelines:

“The Netherlands Ministry of Foreign Affairs has made progress reporting using the IATI standard mandatory from the 1st of January 2016 onwards for all ODA activities it is funding and which have a financial envelope of more than 250.000 Euro” pg.7

“The primary means to assess the progress of the activity is the result element of the IATI publication.” pg.10

Clause from contract with Dutch MFA:

“Type of Report: Narrative progress report, consisting of: 1. Periodic updates on progress, drawn up in accordance with the standards described in the Guidelines”

“the narrative reports you provide on the activities financed must comply with the International Aid Transparency Initiative (IATI) standards.”

DfID:

Representatives of DfID recently suggested (in a consultation about the updated AGA template) that DfID will accept narrative reports in an IATI format. This confirms what was implied in the DfId Civil Society Partnership Review:

"DFID will work collaboratively with both Bond and other donor governments to explore the possibility of reducing DFID reporting requirements where CSOs proactively publish their data through IATI. "

Sida:

Sida are also planning to replace traditional reporting with requiring partners to publish IATI.

If donors have made IATI a reporting tool then NGOs need to know what the limits are, especially when donors encourage organisations to publish using as many of the Standard elements as possible:

Dutch MFA Reporting Guidelines:

" the ministry strongly encourages partners to also publish any other, non-mandatory fields for which they have data. The more information available, the more useful IATI publications become, and the better the insights to improve development outcomes."

DfID CSPR:

“At a minimum, by the end of 2017, centrally funded CSOs will be required to meet the full International Aid Transparency Initiative (IATI) standard on all their funding, and ensure that all the organisations in their delivery chain also meet the standard.”

To suggest that IATI is designed to only provide a snapshot is contradictory to a lot of material that NGOs receive. Clarification on what is actually expected is pivotal. NGOs are concerned that they will not receive funding if they do not fully comply with donor requirements.

We also need to consider the fact that by publishing IATI data, an organisation is making themselves open to assessment, analysis and criticism to the outside world. Does a ‘snapshot’ mean only publishing the good stuff, or the data that is easy to quantify? How does this match the perspective of IATI making the sector more transparent and accountable?

If NGOs are expected to publish all their transactions and results (to demonstrate VfM etc), then the Standard should be adapted to add as much data as possible to ensure an organisations can accurately portray the richness of their work. The NGOs take the risk of otherwise misrepresenting their work, and facing consequences, either through anti-aid media or withholding of funding.

If NGOs are only expected to use IATI to provide a snapshot, fantastic. This would save a huge amount of work! But please make that clear in the guidance and reconsider the move towards IATI as a reporting tool.

One more tiny note, if IATI is about a snapshot, does it still provide usable, comparable data across the sector?

Many thanks,

Daniel

OK, here are my thoughts (think maybe the tags should be changed on this thread tho - see Steven’s comment).

251 organisations based in the UK use IATI to share data on their development activities (IATI Registry, 1 June 2017). Many have been doing so since 2012.

We find that that the IATI users we work with don’t have a problem with the standard evolving, as they recognise this is necessary to keep it relevant. The main issue for IATI users is the speed that the updates occur (see below) and not being clear of the value of the updates - whether they are a) to fix things so that the data is more reliable and usable or b) bring in new users/extend to new areas (Rolf’s significant changes).

I think everyone would agree that changes relating to a) is crucial to the integrity of the Standard and the data produced. A good example of this is when the Schema was changed at 2.01 to bring in a clearer structure. This change was welcomed by those using their own systems as it cleared up the ambiguity that had existed previously.

The frustrations that arise from updates tend to be around:

• slowness of updates to tools and services that people rely on

• the speed these changes can be incorporated into organisational systems

• pressure from donors requiring funded organisations to use the latest version of the Standard (and not recognising the constraints of the above points)

A quick review of the IATI upgrades shows that there have been 6 upgrades since Dec 2012, with 4 of these occurring across 2014 and 2015. The Open Contracting Data Standard appears to have had 8 upgrades since June 2014. Maybe this is the pace for new standards, but as Rolf points out, it’s very difficult for organisations to keep up with this pace of change.

In contrast, ISO standards tend to have a slower pace of change, with perhaps 5 years between changes, and 2 years to incorporate those changes. This seems more realistic and perhaps something to aim for.

Point b) also needs more thinking about. I agree with Rolf that what the IATI data standard aims to achieve (and what it does not) should be clearly defined. We also talked about that in our research with UK and Dutch CSOs on the Results elements - report here: https://discuss.iatistandard.org/uploads/default/original/1X/4680e9d88392df8b529f76c8f8bd0daf4711d230.docx

Relevant to b) I also agree with Dan. For many organisations, IATI is not optional (see previous posts on this on IATI Discuss) and has lost its value as an open and voluntary initiative. Whilst I absolutely believe in ‘publish once, use often’, donor governments who request IATI data as programme reporting and then lobby for changes to the Standard to make it fit their requirements (eg. some of the changes to 2.02) are blurring the lines between IATI as useful open data and IATI as a full-on reporting tool. I’d like to see this addressed.

Thanks for all the responses, a few more remarks:

• We felt the need to articulate this statement together in the context of the current upgrade process. (I’ll be adding my own thoughts to several of the proposals still, others hopefully too…)

Personally, I feel that several proposals add clutter or unclarities, or have alternative options.

As an example: “adding document links to results” brings up issues that can already be resolved by adding documents of type A08, or possibly also A07 or A01 to an activity.

• I think working with “proven user needs”, perhaps in the form of extensions or demonstrable problems in adding or using actual data, is a good way forward. My impression is that HXL is also somewhat on this path.

The activity in IATI is the unit of aid, and so I think we should promote that to data users as well. For results, some of the quantitative parts of a results framework end up in <result>, <indicator> etc, some of it will be available just as text for now as <document-link>.

For me, it is key right now to focus on blocking issues and improving documentation, and letting organisations integrate things in their processes and improve data quality.

Through creation of a Python Library, we have attempted to develop a reference implementation of the letter (rather than the undocumented ‘spirit’ (that varies from person to person)) of the ‘Standard’ (starting at 2.02 before extending to other versions). In doing so, we have determined that as it stands, it is fundamentally not possible to do due to contradictions and lack of clarity in components that may or may not make up the IATI Standard.

The historical / current approach to upgrades, taking proposals piecemeal rather than as a combined whole, means that over time small changes have been added and added without stepping back to look at the resulting Standard. There are placeholder definitions, a mess around Guidelines, a SSOT that doesn’t act as a Single Source of Truth, Rules and Rulesets that are optional, mandatory, and defined in multiple ways all at once, contraditions between definitions due to not following the DRY principle, and more. Some of these could be fixed at or between decimals by simply stepping back and fixing silly mistakes / bugs, though others cannot.

V3(.0.0 - semver and patch numbers would be useful in several ways) is something that needs proper consultation and design (based on actual use cases, ideally from as blank a sheet as possible) that cannot fit within the proposed upgrade schedule.

Until that point, we can either continue making best-effort piecemeal additions as has occurred historically, or ensure that new changes have a clear need (so are essential, rather than nice-to-have modifications) and fit within the wider scope of IATI. How to define the wider scope, however, is (as mentioned by others) unclear.

What action would you like to see IATI take to further discuss this? Do you think it should be handled through the TAG, or would you like to see it on the agenda for the next Members Assembly (MA) in October? Rolf Kleef Roderick Besseling John Adams Daniel Mackenzie

If you want to go the MA route, I will ask for it to be included in the agenda as long as TAG members can commit to preparing a paper by early September, clearly stating the problem and proposing solutions for the MA to discuss.

We propose that significant changes to the standard are only made based on experience and learnings gained from organizations publishing to the standard over the next two years. Making significant or several changes to the standard now will risk taking away the energy and capacity from publishers, users and tool builders.

Rolf Kleef : do you propose to defer the adoption of the IATI 2.03 upgrade?