Jen perssob writes about public engagement & ethics, data sharing and digital services.
Humanitarian data: balancing openness and ethics
An hour after the April 2015 Nepal earthquake, information on the number of displaced persons and their location was available online. Crowdsourced open data helped to provide knowledge needed to support the humanitarian response to a changing situation in near real time. But the collection and use of this kind of data also presents some risks. Jen Persson looks into the pros and cons.
Francis Irving speaking at the Open Knowledge London meet up in April about his work with the Humanitarian Data Exchange (@humdata) at the United Nations, described how through collaboration we can create and access licensed open data (download the presentation here or watch on YouTube here.).
The case study was from the UN’s Office for the Coordination of Humanitarian Affairs (OCHA) which allocates international funding in a natural disaster or other emergency. Their work often relies on accurate and timely data—using information about refugee movements, the landscape, and NGO capabilities – to make better informed decisions.
Image: Screenshot of Humanitarian Data Exchange
OCHA has hundreds of country offices, and works with many partners, such as the Red Cross. Many of these partners gather data during a crisis, storing data in spreadsheets on their own systems but without interoperability or a way to share timely information with others.
The Humanitarian Data Exchange aims to improve this, by enabling collaboration through an open source software tool designed to achieve collective benefits of accurate, timely, local and trusted data.
Turning open data into trusted knowledge
Having a means to collect accurate, timely, local and trusted humanitarian data that can be acted on is especially important when the usual communication channels have been compromised or are missing. – as is often the case in crisis situations.
The data in the Exchange are created collaboratively with a process easy enough to not deter data mappers, but sufficient to ensure the needed accuracy. The crowdsourced data are then validated through a sign-up process using verified trusted third parties – a process that addresses criticisms levelled at some other crowsourcing platforms about the accuracy and robustness of their data.
The type of information collected is often similar across humanitarian contexts, which means the sector is well suited to benefit from open data approaches. Commonalities between data types can be exploited to enable learning across different contexts provided that the appropriate shared tools, infrastructure and standards are applied to make the data shared usable and encourage engagement.
As an example Francis walked us through the ‘Who, what, and where data’of Somalia that uses a template spreadsheet for data collection. As long as the user enters standard data types, they see a map of the country completed with the data, such as the names of the NGOs that are working in the region, places where most humanitarian aid is being spent, and spending on what theme. This tool creates an easy-to-use view of typical metrics for response evaluation, created by simply entering the data and without the user having to design a visualisation by themselves.
However whilst these new open approaches and technologies clearly have potential some thorny questions remain about how the data is used, what should be available, who owns it, who can access it and, perhaps the central question, who decides?
Successfully answering these questions will be fundamental to determining the true value of these approaches in the humanitarian and development contexts and beyond.
Actions and ethical actors
The public benefit of open data technologies often focuses on its value to end users without recognising the costs – particularly to the communities that generated it.
"The technological ease with which these platforms can be put in place means that marginalised groups might also be viewed simply as data sensors, cheap sources of hard to get information", suggests Evangelia Berdou, (Vulnerability and Poverty Reduction Research Team at the Institute of Development Studies) in the report Mediating Voices.
Tools like Ushahidi, an information crowdsourcing software program that was created to collect and disseminate information by Kenyans in the violent aftermath of their elections in 2007, and OpenStreetMap, were two examples she concluded are creating positive opportunities for coordination, for collective action, and for advocacy. Ideas that are at the heart of what it means to be a citizen in the 21st century.
How some of the thornier principles are handled in practice however - autonomy, data sovereignty, indigenous rights, language barriers, privacy and surveillance - need greater attention.
Some of these subjects were considered at the 2015 International Open Data Conference in Ottawa and captured in the report, The open data revolution: an international open data roadmap. Many will doubtless be addressed again in Madrid later this year.
"In addition to examining the nature of partnerships that can realise the potential of open ICTs for understanding and action, we need to consider more carefully the character of actors that drive their production and implementation. For example, we know very little about how social technology entrepreneurs that espouse the values of the open source movement operate, how they position themselves in relation to larger players, and how they seek to influence the policies of major institutions."
The power that open data and the open data principles offer, to set the agenda, to initiate and guide approaches and long term interventions, are not without risk, and need to be used wisely.
As the authors of Shooting our hard drive into space and other ways to practise responsible development data’ say, "the nuanced ethical implications of how data changes our relationships with stakeholders and partners in country are harder to track."
These challenges place an even bigger responsibility in the hands of intermediaries, who need to be aware of the opportunities and risks of the data that they are working with, and embed sensitivity and responsibility in their data handling practices.