Why Should Privacy Matter in Humanitarian Emergencies?

By The Centre for Non-Traditional Security Studies (NTS Centre)
S. Rajaratnam School of International Studies (RSIS)
Nanyang Technological University (NTU), Singapore

Digital technologies have penetrated our daily lives, and many are growing troubled about what this means for our privacy. But do these considerations change in conflict and disaster settings? This was an important question raised during the “Humanitarian Technology and Innovation: Critical Questions and Implications for Southeast Asia” roundtable held on 11th June by the Humanitarian Assistance and Disaster Relief (HADR) programme here at RSIS.

More data can certainly help improve protection of people’s lives and dignity: the core of humanitarian work. This has created a reflex among humanitarians to collect as much data as possible in order to maximise the chance of identifying operationally relevant details or patterns. This mirrors practices in ordinary circumstances too, with both governments and private sector entities demonstrating the same data-gathering instincts to improve the products and services they offer. There is a crucial difference, however, between the importance of privacy in these two contexts. That distinction comes from differing relationships between privacy and security.

In ordinary circumstances, privacy and security exist in tension with each other. If state surveillance is limited in the name of privacy, then the ability to identify “bad actors” falls, reducing security. This is true both if one adopts a traditional security framework, in which it is the state’s security that is of para-mount concern, or a human security one, in which it is individual security that matters.

Conversely, in humanitarian settings privacy can equal security, meaning that reducing the former actually reduces the latter. Ethnicity, religion, political views, sexuality and gender can be acutely sensitive issues during humanitarian emergencies. These cleavages can underlie the unrest or marginalisation that either produces or exacerbates humanitarian needs in both conflict and disaster settings. Consider the situation of Rohingya people fleeing from Rakhine state to Bangladesh due to racially-driven persecution, or studies showing the correlation of race and class with suffering following Hurricane Katrina in the US. In such situations of bias or persecution, collecting data from which sensitive characteristics can be deduced risks reducing the security of people needing assistance further.

Addressing this is complicated by two further issues. The first is that people defined by those sensitive characteristics often have the most serious needs. As such, aid agencies need to be able to identify them in order to help. In both Hurricane Katrina and Typhoon Haiyan in the Philippines, lower socio-economic echelons – a category that often overlaps with ethnicity, gender and even political affiliation – are documented to have suffered more. This underscores the double-edged nature of this data: it can facilitate both helping and harming the group identified. Second, marginalised groups or those fleeing danger are routinely cast as threats to wider society and therefore require greater surveillance. Consider the media conflation of refugees seeking asylum in Europe with threats of terrorism, and the intensive scrutiny they undergo during the asylum process. As a result, their privacy is all the more likely to be curtailed to the perceived benefit of others’ security and to the detriment of their own.

This highlights the complexity of collecting data in humanitarian settings, and a neglected role for aid groups developing and testing new data technologies. People fleeing conflict, caught in disaster, or otherwise marginalised simulta-neously have a lot to gain and much to lose from greater data collection. However, they have little influence over decisions taken to balance these advantages and disadvantages.

From our roundtable discussion, it is clear governments and aid agencies must adapt their decision-making around collecting, sharing and deploying data to take account of this particular relationship between privacy and security. The default of maximising data collection may often simply not be justified. In determining what is warranted, people’s perceptions of their own best interests must be included.