In our “post”-pandemic world, marked by climate crisis, inflation and threats of recession, the desire to do more with less seems reasonable – admirable, even.
Within the humanitarian sector, recent data from the UN and the ICRC shows that available funding falls far short of what’s actually needed to meet existing and expected crises, and the needs are growing. An estimated 339 million people will require some form of humanitarian aid in 2023, according to the UN, who are also calling the current global food crisis the “largest in modern history.” Impacts from the climate crisis, ongoing international conflicts, national violence and human rights violations have driven forcible displacement, with over 100 million people forced to leave their homes.
So, here, where the need to do more with less is urgent, we’ve seen a lot of hope that emerging technologies can make for “smarter and faster” disaster response, create greater impact with fewer resources, and even anticipate crises before they strike. In some cases, as we found in our 2021 research on the use of predictive analytics, there is emerging evidence of these benefits. However, it is not straightforward; nor are benefits guaranteed.
We need more evidence on the realised benefits of emerging tech
Over the decade that The Engine Room has been providing support and conducting research, we’ve explored this question of how, in practice, different digital technologies can augment the work of humanitarian organisations.
We examined the potential of biometric technology use in aid – resulting in Oxfam choosing to enter a temporary moratorium on adopting the technology in order to take more time to understand potential risks. We’ve documented the lived experiences of different impacted communities and refugee populations as they engage with the digital ID systems needed to access aid and government benefits. We’ve taken deep dives on predictive analytics, socially assistive robotics and chatbots (forthcoming). We’ve developed ethical frameworks for engaging in technological innovation, documented efforts to deeply ingrain responsible data practices into workflows, and built actionable tools for ethical decision-making.
Our findings paint emerging technologies’ impact as mixed at best – clear-cut cases of benefits “outweighing” harms are few and there’s a clear need for a more accurate accounting of the benefits of these systems versus the costs to fundamental rights.
The research we’ve done, instead, highlights the need to approach each situation holistically, considering sustainability needs, staff and resource constraints, localisation, root-level problems and integration with existing efforts. Each wave of shiny new technology brings additional cost and complexity – more efficiency or improved outcomes are never guaranteed.
What can we do?
We’ve found that adopting an emerging technology, adapting it to a unique and quickly evolving context and implementing safeguards requires serious time, financial and human resource costs.
In their quest for efficiency, then, humanitarians may end up investing a great deal of resources into sometimes unproven solutions without a clear return on investment, redirecting attention and resources from other (perhaps less technologically “cutting edge”) programming that’s more time-tested. Though these costs are not a reason in and of themselves to avoid adopting a new technology, they do point to a need for more research about the actual impacts these technologies have, especially as weighed against the costs and risks.
Learning from our research, several trends emerge. These point to key considerations to keep in mind to ensure tech interventions meet short-term real needs in a sustainable manner. Among other things, we’ve seen:
- It’s always worth digging deeper into the problem that one is trying to address with technical solutions. Tech interventions can masquerade as solutions to root-level problems, while actually being a band-aid. Often, the lowest-resource, highest-impact solution is not high-tech.
- There’s a need to be realistic about the upfront and ongoing human resource cost of adopting a new digital technology or digital process. As would be the case with any new intervention, time and resources are needed for learning, adaptation, iteration and ongoing maintenance. Though we’ve seen this generally recognised on a conceptual level, we’ve heard from practitioners that actual allotment of resources are typically undercounted. In particular, resources tend to be scarce around data protection/digital risk mitigation and ensuring that frontline staff have access to resources and time to learn new skills or adopt new policies.
- We must account for the likelihood of new digital technologies actually adding inefficiencies into project resourcing and planning. As part of our forthcoming research on chatbot use in humanitarian contexts, interviewees shared multiple instances where the new technology had increased both start-up costs and ongoing work (e.g. by introducing a need to review all of the chatbots’ interactions to spot cases that were improperly addressed).
- We must be clear-eyed about the costs of working with third parties, as there may be a mismatch of incentives, expectations and standards that jeopardises long-term sustainability or success. For example, in the case of the IRC partnering with mobile networks across several countries to provide digital cash aid, a commonality across the challenges documented was a misalignment of practices and incentives between the IRC and mobile network operators. These challenges often resulted in delays and reduction in impact.
- As with all new interventions, new digital technologies must be integrated into existing processes. As we saw in our forthcoming research on chatbot use, new technologies can perform badly when understood as a replacement for other forms of engagement, programming or communication. Instead, they should generally be seen as a complementary component of a larger ecosystem of tools.
Keep an eye on our blog for our forthcoming research and more cross-project findings at this intersection. As always, we would love to hear your ideas, research and comments!