An organisation has several systems in place to collect and process data internally
They deploy several dashboards and monitoring processes to manage what this data means - and ensure they can remedy any issues in a timely manner
They may also openly publish their data, or provide websites / dataviz to their stakeholders
The publication of IATI then becomes a “simple” matter of:
– Selecting a subset from their wider dataset
– “Save as” IATI XML
– Publish / update
Hence, in these types of scenarios, an organisation uses their data extensively. But, the IATI XML is rather useless internally - it’s just a format of some of the information they spend a great deal of resource to deliver.
I think this might challenge some perceptions around data use, particularly the notion that those-that-publish-must-also-use. Of course, this observation focuses on using their own IATI data - the wider access to other publishers is a different matter.
However, I wanted to share - as it seems an increasing likelihood that publishers will output data in the IATI format, but find no need to then reuse it themselves.
That’s how I’ve always interpreted the statement “use your IATI data”, @stevieflow - less like eating your own dog food and more akin to serving your guests water from the same tap that you and your family drink from (same source, but obviously not the same glass of water). The trick is getting organizations to install internal plumbing when they claim they can’t afford the investment but dutifully buy a crate of bottled water (i.e. spreadsheets, manual effort, spinning plates) every day.
…fine, I’ll drop the metaphors!
At InterAction were trying to shift from a narrative of “publish your data, we can help” to a narrative of “improve your internal data systems and infrastructure, then you won’t need our help.” But obviously that’s fraught with issues - NGO starvation cycles, lack of technical human resources, failed systems integration investments, etc. Exceptions abound, but even some large and/or well-funded organizations struggle to wrangle the numerous projectized and requirement-driven (not needs-driven) systems that proliferate.
And then IATI just becomes one more thing, one more plate to spin. The technical answers may be clear to some but the political and organizational answers are more difficult. Thoughts?
My point with this observation, however, was that not all IATI publishers will use their own IATI data, as for them, it might just be an output from all the processes they’ve got in place. I make this observation in the context of data use being a key focus. We should be prepared that some organisations might not need to use their IATI data.
I’m hoping @YohannaLoucheur will in turn make the point that using open aid data is something practitioners in organisations should do much more of, however…
Can we think of any examples of organisations for whom using their own data in IATI format / via IATI is preferable to using their own data as they did before IATI / in the native format? It might tease out a few more specfic IATI use cases…
The first one I can think of is when an organisation wants to compare their data with that of another organisation, and IATI is the common format. But perhaps that is cheating a bit, are there any others?
Devtracker? It would be good to confirm why using IATI format was easier than just hooking it into Aries?
It would be difficult for most to use their own IATI data as most organizations who report to the standard have far more complex business processes than the schema can capture (not to mention we have spent fortunes developing and buying systems to manage them), and without that information, the context is missing and significantly weak for decision making. So in the end, for many of us, I suspect, this ends up being an output populated from a different point in the information lifecycle than really is relevant to us as an organization. We already have this information in systems where we report from, but what is interesting is to bring in data from others. IATI could end up being a data source to get data on partners . . . I say this from the perspective of funders and speaking from our own experience. From partner countries, this could be used and it is why I feel that the use of data being showcased usually centers around partner countries and the success in that field.
We designed a new version of Global Affairs Canada’s Project Browser earlier this year. I insisted (fought for, hard) it run on our IATI data for 2 main reasons: so we’d see ourselves what we were putting out, and in anticipation of combining it with data from other organizations.
Our Project Browser has been around since 2004 and was already being used by staff quite a bit - mostly because the info is presented in a much more user-friendly way than in our internal systems, and it’s more accessible from the field. Yes, we have internal tools to pull information from systems, but often the Browser is enough for what you need to do - and you can eg download a csv file of your search results, or send a partner a link to the project profile.
Using IATI data enabled us to show more details on the Project Browser (eg transactions), and helps improve it (eg spotting mistakes etc). However, if you asked GAC staff if they use IATI data, they would all say “no” with a great degree of certainty (actually, mostly they’d say “What’s IATI”).
Thanks @Mhirji and @YohannaLoucheur - thanks, I think some great pointers about what the donor transparency / sharing use case is i.e. not in competition with detailed internal business process type discussions but clearer, quicker, and easier to share. Also the best format for combining with data from partners.
The takeaway for me is that both of these use cases support the need to slow down a bit on making IATI fill more roles, and double down on making sure that it is easier to combine and visualise together IATI data from different sources e.g. can we make displaying joint projects which are separately reported by both funders work without manual intervention, or can we solve double counting issues for when we want to combine funder and implementer IATI data to make sure that we get all the data e.g. if the funder is reporting the full value, but only the implementing partner is reporting the sub-national locations. At the moment, the recent DI report highlights how most of the visible IATI data use is just one publisher at a time, without a lot of manual intervention e.g. to equalise use of hiearchies, remove double counting etc.
I think this is very important to start addressing what may be the most important use case on donors’ side: using the data in country, for eg donor coordination discussions. Both aspects are important: avoiding double-counting, but also showing the real/full origin of funds (joint funding and upward traceability).
(Note that solving this, and getting user-friendly tools to display the data, would also be useful for other users such as government officials, local stakeholders etc).
Absolutely agreed - how do we get started? I have seen a lot of work already one this, off the top of my head by @rolfkleef and @stevieflow which always makes me surprised that we haven’t cracked this yet, maybe we have (using other org activity IDs for recieved funds, and ‘related activity codes’?), and the standard/data quality has not caught up - in which case we need to document what we are waiting for.
My previous idea (which failed due to me not making time) was to find a set of test cases / activities that all the various people interested in this could use as examples to test different approaches. Can you think of some Canadian projects which include multiple funders that we could start with?
I think we’re waiting for data (how many publishers use other org IDs and related activity codes consistently?), but also need to work out the presentation aspect.
Happy to help with test cases. Unfortunately we don’t have an easy way in our systems to identify jointly funded projects. Chances are that these would be more visible in AIMS; wonder if you or @markbrough would have a way to search for these in the AIMS you work on? If we find an initial list of projects, we could reach out to the other orgs involved and ask their help.
That being said, I wonder how Perspective would be solving this, in terms of presentation?
We’ve been doing the shifted narrative for a while now, and have found evidence of the following:
The IATI data standard is a useful, relatively neutral framework to follow if the organisation’s data management history has been problematic or there are political (small p) reasons why different parts of a federation don’t share data between themselves.
We’ve seen that over time (5 years now), organisations with a higher proportion of institutional funding will generally be more willing to invest in changes to accommodate IATI data production, but on their own timescales and within a wider systems update/change. However, some funders aren’t always willing to wait for the data, which is why you get some organisations trying to implement an interim solution such as AidStream or the IATI2CSV Convertor.
We have funders now recognising that the production of IATI data requires investment, and willing to help fund that investment (within proportion).
We have organisations in the UK that are willing to produce IATI as a byproduct of their own data use (Stephen’s point) because they see IATI data as a global public good for international development/humanitarian response. I think this remains a strong reason for engaging with IATI as producers.
So perhaps the maturity of the UK IATI scene means that we’ve got past some of the issues that the US may just be starting to grapple with.
However, I think that Stephen’s point is important. If organisations are not necessarily publishing to use their own data, why would they engage over the longer term?
The following is speaking as myself, not Bond or IATI Board:
I feel that funders’ requiring organisations to use IATI has been a double-edged sword. It’s been useful to require organisations to produce open data on international development projects to support public scrutiny - and IATI at its most elemental provides a good framework to do this.
It’s been far less useful for government funders to see IATI as a means to their own digital innovation and frame it in this context to their partners (funded organisations). And on reflection, maybe it hasn’t been helpful in return to ask funders to justify their own use of their partners’ data as it just reinforced this.
Maybe a better solution would have been to stress IATI as a global public good. Then we would start looking at it from a global network level/scale rather than an individual organisation or government level. One reason why: At the Members Assembly, there were multiple requests from partner countries for help with using the data, and yet there are still only a limited number of ‘pilots’ and technical providers/funding governments helping them. This kind of ‘drip’ effect is too slow, and there other solutions on the horizon that already have the potential to outscale IATI in the medium term.