Traceability and Linking in IATI Data

Today I had the privilege to present at the “Big and Open Data for International Development Workshop” at the Centre for Development Informatics of the University of Manchester. In my abstract, I anticipated deep research into traceability of activities in IATI data. We’ve certainly made great strides, and, as one participant of our IATI Learning Workshop of last week remarked, the level of discussion on IATI is high, and although there still are things to fix in today’s data, a lot of it is fine-tuning. So I made a pitch for IATI as a possible field of research.

You can download my slides here (or in PDF if open document format doesn’t work).

It’s an honour and a challenge to be here today.

An honour to be able to present some of the “field work” I am doing, mainly with Dutch NGOs in development cooperation.

And a challenge, as I am working as consultant and developer, with academic aspirations but without an academic track record.

I’d like to do is three things:

• make sure you are all aware of the stream of IATI data that is starting to become available
• show some of the challenges and opportunities caused by that data,
• and pose some questions as possible fields of research using this data.

About IATI

Elise already sketched IATI, the International Aid Transparency Initiative, and the work of Publish What You Fund on the Aid Transparency Index (I’ve been fortunate enough to serve as a reviewer twice) to see how well major donors are doing in publishing information as IATI data. I’ll look into “following the flow” from source to where it is spent.

What makes IATI special is that it combines a multi-stakeholder political approach towards aid effectiveness through transparency, with a technical standard to make sharing information possible.

In a nutshell, the IATI Standard lets you publish information about your “activities”: your units of work, such as programmes and projects, for which you can answer basic who, where, when, what questions:

• What is it? The title, description, what sectors it involves.
• Who is involved? Who funds it, who implements, who is accountable.
• When does it start, when does it end?
• Where does it take place?
• Why? What do you hope to achieve, and what are you achieving?
• How much money is involved, who gets it?

IATI for research?

A few facts and figures about IATI:

• The first version of the technical standard was released in February 2011, and we’re at version 2.02 by now: the standard is evolving.
• Organisations that start publishing their data, create an entry in the IATI Registry to share links to their data sets; there are 472 organisations in this Registry now: all major donors, as well as many NGOs, because donors are starting to require them to publish.
• All data combined now has
  • over 600,000 activities, units of aid,
  • over 440,000 more or less specific locations,
  • over 220,000 distinct links to websites and documents,
  • and my guess: anywhere from 15,000-50,000 actors involved.

So I think it is becoming an interesting body of data for research.

Thanks to websites like d-portal.org and IATI Studio, you can query the data in an easy-to-use format.

But I think it’s real potential is not in aggregating the data, but in linking it, and in combining it with other data sets.

As an example, we created an exercise to combine Oxfam Novib’s data on where they work in poverty alleviation with Global Development Indicator data on poverty levels: are you working in the areas where poverty is highest?

Another example is combining the actor networks from two NGOs, to see which organisations are working with both of them.

IATI in The Netherlands

As of January 1st of this year, grantees of the Dutch Ministry of Foreign Affairs are required to report on their progress through IATI data. No more lengthy documents: you keep your data up to date, and once a year you provide a narrative report of up to 8 pages and a financial accountability statement.

A few weeks ago, Dutch Parliament asked the government to include results in the budget for 2017 onwards, so also the Ministry of Foreign Affairs has to provide these. For the Ministry, being able to see results in IATI data of grantees is instrumental to gather information for their own results reporting towards Parliament.

The traditional data view of the Ministry is this: within the organisation, there is a programme or project with a budget, and one of more grantees contracted to implement it. But what is done by the grantees is hidden in documents and spreadsheets.

The IATI view is that the grantee or grantees publish data about their activities, and link back to the Ministry’s activity. The data view grows. If partners of the grantees also publish, the chain grows further, until ideally we reach individual activities in the field. We can “zoom in”.

This traceability offers better insights in who is working where.

It will enable staff at embassies to have insight into the activities that in sometimes indirect ways receive funding from the Dutch government.

NGOs too can research where they could or should work, or with whom.

I’ve been working on tools to make such structural diagrams, and also start to explore financial flow diagrams. The data on committed and disbursed can be visualised if the linking is right.

The Ministry is also working on their own dashboards, and they are openly available for inspection. They show the multi-tier structure as well.

For instance, they show an overview of the results for a programme in that same multi-tier structure. With results, indicators, and a progress bar for each indicator.

They don’t aggregate the data: it’s simply the list of all indicators on each level right now. It’s simply a view on progress: interpretation and aggregation into higher-level goals or different Theories of Change are not in the data.

Hopefully by now, you already imagine all kinds of ethical and methodological issues, around privacy, security, ownership, data literacy, accountability, and so on.

There are profound changes happening in organisations that embark on this trajectory: cultural, organisational and technical changes.

I won’t go into all of that here and now, but will take a practical angle now.

The guinea pigs right now are 25 Strategic Partnerships, where organisations work together on a 5-year programme. Together, a couple of dozen organisations started publishing data in the IATI format.

There are plenty of practical issues that are being solved now, and these Strategic Partnerships are currently setting up their results frameworks, so there’s not a lot there yet.

But: these Partnerships will continue publishing data for another 4.5 years.

IATI as a field of, or tool for research?

That brings me to the questions I promised at the start. So some questions that come to my mind:

• Does the data help practitioners to communicate about their work?
• Can linking data sources help practitioners in doing their work?
• Does linking activities and data sources lead to better data quality?
• Can the data help study patterns of collaboration, plans and strategies?
• Can the availability of results data help improve studies on impact?

Somewhere last century, I graduated on knowledge engineering. As I mentioned at the start: now that we’re getting close to linked data, I’m feeling “academic itches” again, and would love to see if there are possibilities to explore such questions together.