This blog is reposted from the Joined-up Data Standards website and written by Bill Anderson, technical lead for IATI who also leads the work of the Joined-up Data Standards project.
Responses to humanitarian emergencies involve the interactions of a wide range of donors, international agencies and local actors. Coordination is critical. Coherent, timely data is a key element of this complex response matrix and at the World Humanitarian Summit in May 2016, 33 of the world’s largest donors and humanitarian agencies committed themselves to “identifying and implementing a shared open-data standard and common digital platform which will enhance transparency and decision-making”. The publishing standard maintained by the International Aid Transparency Initiative (IATI) was recommended as the vehicle for this effort.
In preparation for this ‘Grand Bargain’ IATI has modified its standard to better meet the requirements of humanitarian reporting. These modifications were done in consultation with, amongst others, OCHA’s Financial Tracking Service (FTS), the European Emergency Disaster Response Information System as well as organisations such as the UK Disasters Emergency Committee and the Start Network.
For humanitarian decision-making to be effectively harmonised, data from a variety of sources needs to be joined up. The way in which emergencies are defined and reported is an essential part of this. One improvement to the IATI Standard was thus the inclusion of a new element – humanitarian-scope – which provides a means of accurately defining emergencies and appeals. This is done by specifying a ‘vocabulary’ – a globally recognised list of classifications – which maintains a curated list of codes.
For a standards body like IATI this raises a number of issues. Here are four considerations…
- What recognised authority lists (vocabularies) exist? No global classification system exists that encompasses all humanitarian emergencies. IATI and OCHA have agreed that the best approach for classifying emergencies is to adopt GLIDE numbers, a Global IDEntifier coding system for natural disasters originally developed by the Asian Disaster Reduction Center. OCHA is a participating institution and as such can generate new GLIDE numbers. Furthermore, there is agreement that the system can be extended to include man-made as well as natural disasters.For GLIDE to be recognised as the key global standard there is a need for all participating institutions to advocate for its adoption by all humanitarian-related information systems.
- How are codes accessed? Currently, users of the system are provided with email notifications of changes and can only access the database through manual downloads. This is far from ideal, particularly for data production in the early stages of a sudden-onset disaster.For GLIDE to progress it will have to be dynamically accessible by machines. Investment in a robust API that allows for automated discovery and retrieval of codes will be essential.
- How are codes created and maintained? Whenever a new emergency arises a participating institution can create a GLIDE number. If timely data is required urgently in sudden-onset emergencies there is little time for traditional methods of consultation and agreement between participating institutions. Furthermore, as IATI’s reporting base grows there will be a need for the registration of local emergencies that may fall beneath the radar of the large participating institutions.There will thus be a need for a collaborative, technical framework that allows for rapid consultation and approval of new codes.
- How will users of the data make sense of the codes? Humanitarian financial and operational management has been developed in such a way that emergencies are defined and coded by country and year as well as by the disaster itself − multi-country and multi-year emergencies have new codes generated for each country and year. For a user of IATI data to analyse activities related to the West African Ebola epidemic a search would need to be done across 12 unique emergency codes (three countries over four years). The GLIDE number does also include an event-type code and “EP (Epidemic)” is one of them, but parsing identifiers is not easy.IATI will need to consider how to provide a searchable item to improve access to the data.
IATI is a voluntary standard. It has no power to enforce compliance. It is dependent on the good will of publishers and peer pressure from the humanitarian community as a whole. Improving humanitarian data will require collective efforts from all quarters.