@IATI-techteam, the link from this Dashboard total-expenditure element to the Standard page (presumably describing the element) returns a 404 error.
@bill_anderson any updates on this? The number of publishers without (2014) coverage reference data now far exceeds those with. The knock on is that the coverage-adjusted score is calculated with the arbitrary 20% for many many publishers. This just looks strange, particularly to those new to IATI.
The IATI technical team did a manual collection of reference spend for 2014 and 2015 from all IATI publishers two years ago. The process involved emailing publishers, collecting data manually and storing in a public Google sheet. This is not a sustainable method for collecting coverage data and the IATI tech team has not done reference spend collection for any new IATI publishers over the past two years. This means that the coverage data and then coverage-adjusted scores on the Summary Statistics are not up to date for all IATI publishers.
As a result, the current plan is to remove the coverage and coverage-adjusted columns from the Summary Statistics page on the Dashboard, keeping the coverage page only for reference, until the process for collecting reference spend is automated and up to date. In the meantime, we strongly encourage that all IATI publishers include in their organisation file total-expenditure.
How about putting a statistic about which publisher includes total expenditure in the org file on the statistics page? It would highlight whose coverage could theoretically could be calculated (as imperfect as it may be) and may provide an incentive for others to include that data in their org files so we get closer to actually having an automated way of calculating coverage. Assuming people still agree that coverage is important.
Thanks for the comments and ideas @Michelle_IOM
We will be emailing all IATI publishers next week to inform them about the planned changes of removing coverage-adjusted score and would strongly encouraged all to use
total-expenditure in their organisation file.
total-expenditure in the summary statistics methodology, we don’t have an ETA yet for when that will be incorporated in the publisher statistics, but would keep in mind the idea about highlighting publishers that have already included
total-expenditure in their file.
Have publishers been emailed?
I assume that the methodology to use the
total-expenditure would run (at least) a year “behind”. For example:
- In 2018, coverage data would be based on 2016 expenditure
- In 2019, coverage data would be based on 2017 expenditure
- In 2020, coverage data would be based on 2019 expenditure
Because - it seems unrealistic for publishers to have their
total-expenditure published (and all their activity data, on the 1st January each year. For many, getting this data together might take a few months. Therefore, it would seem the IATI dashboard methodology would need to accommodate this, in line with regular financial process & timescales.
Great to see this move forward, nevertheless. Thanks.
(At least some) publishers have been emailed, yes.
The core message is as follows:
"In the future we plan to compare the value of the element published in your organisation file with the calculated total spend (expenditure + disbursements) for a given year published in your activity files.
To prepare for this please ensure you are publishing an organisation file which contains the element for at least all years that you have been publishing to IATI."
I’m not sure I understand this last part: “for at least all the years that you have been publishing to IATI”. This seems like a fairly tall order, and not necessarily a very useful one for publishers who have (very partial) data going back over many years.
I assume coverage will be assessed for recent years (probably a year behind, as suggested by Steven). Could the Tech Team confirm what is the proposed approach?
Perhaps another mini-conversation at TAG?
Yes, the TAG would be a good opportunity to make sure we all understand what is being proposed.
However, it would be great if the @IATI-techteam could share more details ahead of time so we can prepare.
Going over the thread again, I’m not sure we fully answered some queries.
Could the @IATI-techteam provide clarifications on the questions above - in particular regarding the period for which coverage would be calculated ie 1) what time lag (2 years, as suggested by Steven) and 2) how far back in time.
I suppose the detail around this bit is also important, since it seems the methodological complexity has been shifted to here.
Thanks for the follow-up and apologies for the delay in getting back to you. Following the tech audit last August, there was no immediate action to work on the IATI dashboard and we have not looked into a new methodology for the coverage assessment.
This is currently on hold so I cannot confirm what period the coverage assessment would cover. The recommendation for the Dashboard was to revisit re-building the Dashboard as we have updated the community in our last blog: https://iatistandard.org/en/news/iati-technical-team-our-detailed-quarterly-update/
In the meantime, I would still encourage using total-expenditure in the organisation file .
Hi IATI Technical Team:
Following up on this I wanted to let you know that we have been looking into this issue at Publish What You Fund.
We looked at the total-expenditure field of the 45 publishers assessed in the 2018 Aid Transparency Index and compared these with the total of expenditures and disbursements for that year. This is what we found:
- Most publishers are not using the total-expenditure field to publish annual expenditure. Of the 45 publishers in the Index, 4 have published 2018 total-expenditure data, 7 have published 2017 data and 8 have published 2016 data.
- The figure used is often just the sum of disbursements + expenditure recorded in a publisher’s IATI activities, so does not help to provide a comparison with which to measure coverage.
In order for this proposed methodology to work we would need to encourage publishers to consistently add their data to the total-expenditure field and to have some form of verification, in order to address point 2. Linking to an annual report or official figure elsewhere is one suggestion of how this could be done.
Hi Alex. Many thanks for providing this useful analysis from the work you have done.
In order for this proposed methodology to work we would need to encourage publishers to consistently add their data to the total-expenditure field and to have some form of verification, in order to address point 2.
Yes, I agree that we need to encourage more organisation to be using the
total-expenditure field. I believe the ‘verification’ part will be something you will be doing for the Aid Transparency Index. One way will be to see if organisations have added a relevant
document-link in their organisation file.
I suspect the “verification” part may also become relevant for the IATI Coverage methodology, as part of the intention to:
In fact, I suspect much of the complexity of the new coverage methodology has been shifted into this bullet point.
Hi Petya - thanks for getting back to me.
To avoid confusion I suggest we focus the discussion in this thread on IATI’s coverage statistics rather than on the Aid Transparency Index. My suggestion is that the total-expenditure element should have some form of verification to address my point 2:
One way of doing this would be to require a published document such as an annual report, audited accounts or other public statement which corroborates the figure entered in the total-expenditure field. I’m not sure whether this could/should be done using the existing document-link in the org file, by adding a new code to the document-link codelist or as an additional subelement in the total-expenditure element.
It would be good to hear views from those more involved in the technical detail of the standard as to whether this is a workable proposal and what may be the best way to implement it.
We know that the B codes in the DocumentCategory codelist are to be used for organisation level documents. Of these, B01 - Annual Report seems the most likely (in fact, the only?) existing category for budgetary/spent information ?
I’d be very wary of mandating a document-link of a certain type, as that would a) be in place for everybody (and for some publishers there may be good reason not to include it, and b) as it involves a codelist value it would have to be implemented by a new ruleset
The document-link block being added within the total-expenditure element is interesting - in the same way a document-link could be relevant to a specific transaction or location (for example). I think that possibility takes us into a new area of discussion - around the future flexibility of the standard - so maybe not on the table for addressing this more immediate need.
@alexjtilley thank you for following up. As @stevieflow has summarized, currently there is the option to add a
document-link at organisation level, specify the type and link to relevant document, such as Annual reports. However, this is not a mandatory field in IATI.
One way of doing this would be to require a published document such as an annual report, audited accounts or other public statement which corroborates the figure entered in the total-expenditure field.
I would agree that the above would be quite a strong requirement. Advising organisations and recommending that they add/ refer to a relevant document is beneficial for users of the data, can be one way forward.
One potential issue is where an organisation has a mix of activities (hence expenditure) and some of these are not relevant to IATI. The figure in the total-expenditure element may be the result of calculations that will not necessarily appear in a clear way in the annual report. For instance, total-expenditure for IATI may be “pillar 1C + all of pillar 2 + pillar 3A”. Not insurmontable, but we’d need to think of ways (and provide guidance on how) to communicate what total-expenditure represents if it’s less than 100% of an organization’s budget.