As the number of people around the world forced to flee their homes exceeds 100 million, the need for data to underpin effective responses has never been more urgent. However, the transition from situational assessments to high quality socioeconomic microdata on forced displacement produced by national statistical offices will take time. In the interim, can innovative approaches accelerate and enhance the generation and use of socioeconomic data on forced displacement to inform effective action? The consensus at a workshop held recently at UN City in Copenhagen was ‘yes, it can’, and the most fruitful opportunities identified were rooted in complementing and reinforcing traditional approaches. Convened by the World Bank-UNHCR Joint Data Center on Forced Displacement, the workshop brought together 70 specialists from 20 international organizations, governments, and universities, to collectively brainstorm and chart ways forward.
Through bringing people together on better data for sustainable development and innovation and partnerships for more inclusive data, the UN World Data Forum fosters a community that is well-placed to enhance and act on these insights. The UN World Data Forum 2023 not only convened directly relevant conversations, for example, Leaving no-one behind in official statistics: Inclusion of forcibly displaced in national surveys (TA1.18) and Diverse partnerships for inclusive data on migration (TA1.14), but also brought together teams working with innovative methods and data sources with relevant translational potential, for example, insights from the Data for Now Initiative (TA1.03) and Integrating population, displacement and geospatial data in humanitarian emergencies (TA1.07). We are sharing our findings here to invite discussion with the UN World Data Forum community to further develop these ideas into concrete plans for collaborative action.
Three areas of opportunity emerged clearly from workshop discussions, and these map onto the UN World Data Forum 2023 thematic areas of Innovation and partnerships for better and more inclusive data and Maximizing the use and value of data for better decision making. We recommend that these emergent areas are used to guide future investment in innovative approaches to accelerating and enhancing the generation and use of socioeconomic data on forced displacement.
- First, the use of alternative data sources and innovative methods to enhance household survey sampling. For example, the use of gridded population data and building footprint layers to inform robust household survey sampling approaches in data-poor areas affected by forced displacement.
- Second, the curation and integration of diverse data sources to build models for estimating socioeconomic indicators in forced displacement settings. For example, the use of satellite imagery, market price data and household survey data to build models to update socioeconomic indicator estimates between face-to-face survey waves.
- Third, the use of natural language processing to amplify the discoverability of socioeconomic data and evidence relevant to forced displacement operations and programs. For example, the use of tools such as www.nlp4dev.org and the JDC Literature Review Database to allow operations staff and managers to access relevant data and evidence to inform decisions.
Challenges were discussed alongside these opportunities. These included the technical barriers to adapting existing approaches to forced displacement settings, for example, to ensure that their outputs have sufficient spatial resolution to be useful. The question of capability came up in different ways and in different discussions. For example, how organizations can best ensure that skills keep pace with technological development, and how a cross-organizational community of practice could be best supported to accelerate the exchange of relevant data, information, tools, and experience. Important questions of trust and responsibility also arose in discussions of capability. Leaders need to keep up with technological development, not so they can wrangle data into advanced analyses themselves, but so they can use informed good judgement on how to assess and apply their outputs.
A continued focus on data responsibility is key, throughout the whole data collection-to-use pipeline, not only to maintain trust, but also because alternative data sources could introduce new and unknown risks, especially if the sources are private sector systems that are not wholly transparent. The management of data on vulnerable groups comes with specific legal and ethical requirements, and the effective mitigation of newly emergent risks relies on the input of appropriate expertise. New innovative approaches should be stress-tested for trade-offs between the benefit of additional socioeconomic insight and potential harm, and relevant data should be documented and catalogued consistently across organizations.
There was strong consensus that the value of data and evidence must be in its use, and that the operational relevance of any new data innovation should be rigorously and iteratively assessed. This led to usefully frank discussions of what kind of outputs from what kinds of models would be ‘good enough’ to be operationally relevant in different contexts. The statistician George Box is often quoted as having said that ‘All models are wrong, but some are useful’, and many conversations focussed on what would be useful in this sense. George Box may also have said that ‘Statisticians, like artists, have the bad habit of falling in love with their models’, and holding the hammer of new technology, there is always a danger of seeing every problem as a nail. The energy of the workshop discussions did not flow from techno-enthusiasm, though, but rather from a commitment to engage with the complexities of hard-to-use data sources and advanced analytical techniques, driven by a belief in their potential to deliver value by informing responses to forced displacement.
In the JDC, we look forward to working with colleagues across all types of organizations to take this work forward and to help bring together a community of practice as part of the wider UN World Data Forum network on accelerating and enhancing the generation and use of socioeconomic data on forced displacement. Global forced displacement continues to grow, and we should explore every avenue to improve outcomes for those affected by it.