Introducing the panel, Ms Katharina Höne (Researcher in Diplomacy and Global Governance, DiploFoundation) argued that there are vast opportunities related to data and that our ability to measure the current state of development is paramount to improving the lives of many. The conference as a whole discussed many useful applications of data, and in particular new forms of data such as social media and geo-spatial data. Höne argued that while there are frequent and strong calls for statistical capacity development, equally strong attention needs to be placed on developing the capacities of policy makers and diplomats for the data driven era. Hence, a number of question arise: How do we develop these capacities? What are the needs? What are possible approaches? And, what and who should be included?
Ms Maria Fasli (UNESCO Chair in Analytics and Data Science; Director at the Institute for Analytics and Data Science, University of Essex, UK) who teaches data science as part of her professorship and engages in capacity building and work with policy makers, began by stressing that more data does not equal more knowledge and argued that data is only valuable if it is put in the service of better decision-making. She introduced various levels of data analytics: (a) descriptive, (b) diagnostic, (c) predictive, and (d) prescriptive analytics – ranging from hindsight to insight and ultimately foresight. However, she cautioned that as we move to insight and foresight, the level of complexity increases and the tools available to us in these areas are still in their early stages. She cautioned that policy-makers need to be aware of the limitations of the various types of analytics and mindful of unrealistic promises. Fasli also highlighted a number of challenges of big data analysis and machine learning. Namely, there is often a problem of explainability and interpretability of machine learning algorithms. Algorithms might be biased, given the limitations of the data sets they have been based on, and there is the danger of violating existing codes of practice or ethical standards. Looking at capacity development, she highlighted that not all skills can be found within one individual and that it is paramount to build good teams of people with a number of different skills. Such skills should also include story telling skills, negotiation and communication skills, and the ability to critically question the data and its analysis.
Mr Graham Nelson (Head of the Open Source Unit, UK Foreign and Commonwealth Office (FCO)) explained that the Open Source Unit is based on the idea to fuse data, in particular open source data, and diplomacy. The unit has been operating for more than two years now and has grown from 10 to 20 employees. It is multidisciplinary and brings three specialisms together: data science, open source intelligence, and behavioural science. The aim is to address policy challenges and to provide better evidence-based, insightful and reflected decisions. Nelson provided a number of examples which included an interactive tool that maps the Brexit conversation in online media based on content, themes, and mood and clusters conversation according to similarity. Another example included using open data to understand online extremism. A final example was a tool that supports crisis response activities by scraping news websites, with a machine reading the content, and pulling out key themes, which are accessible via an app that maps the global conversation. Nelson also mentioned the importance of using data-driven insights to counter disinformation online, to provide a better evidence-base for policy decisions, and to map the UK FCO’s own activities.
Ms Grace Mutung'u (Kenya ICT Support Network) described her work as focusing on fostering ICT policies and their reform in Kenya and beyond. She highlighted that there is a lot of talk about the need for more data on Africans in sub-Saharan Africa. Governments are asking mobile network operators for more data and require citizens to increasingly use online services and provide biometric, housing, and ethnographic data. However, she cautioned that we need to wonder to what extent this perceived lack of data should be taken at face value. It is important to question to what extent existing data could be linked better, used better, and what other interests are linked to this push for more data. She also argued that there is a push for citizens in Kenya and other countries to use governmental online services. Yet, there is still a huge gap in Internet penetration and digital literacy standing in the way of using these services. In addition, while there is growing digitisation in various sectors, there is lack of safeguards such as data security and privacy. Mutung’u also pointed to positive changes due to appropriate regulations, and the impact of the EU’s General Data Protection Regulation (GDPR) and emerging national policies. In conclusion, she advocated for a more people-centric data economy, which puts people at the centre.
Mr Javier Teran (Statistician at the Centre for Humanitarian Data, UN OCHA) explained that the mission of the Centre for Humanitarian Data, which opened one year ago, is to work with partners to foster the positive impact of data and humanitarian work. He highlighted the centre's four areas of work: (a) offering data services that connect relevant data sets with users such as governments and intergovernmental and non-governmental organisations, (b) supporting data policies to ensure the responsible use of humanitarian data, (c) fostering data literacy to increase the data capacities of individuals and organisations in the development community and beyond, and (d) network engagement to build a community around data for development and work with communities and relevant stakeholders. Teran showed a number of examples of data visualisations and interactive tools provided by the centre. The topics covered included mapping conflict-related displacement in Afghanistan, Rohingya refugee camps, energy consumption of refugees and displaced people, and education in conflict. He explained that various organisations share data with the centre such as UNOSAT, the International Federation of the Red Cross and Red Crescent Societies, and the UN refugee agency. In conclusion, he stressed the importance of safeguarding privacy and working with the appropriate level of data aggregation.
Asked about the current biggest need in capacity development for data, Fasli emphasised that a better understanding of how to turn data science into action and how to close the feedback loop from data to policy and back was needed. Nelson argued that a shift towards a more data-centric thinking in policy-making needs to take place. Mutung’u stressed that, speaking from her context in Kenya and sub-Saharan Africa, capacity development needs to empower individuals and organisations to move from being consumers of data and data insights, to becoming producers through appropriate and early education. Teran pointed to the importance of being able to use the findings of data scientist appropriately.
Further questions from the audience touched on the potential mis-use of data by policy-makers, the need to also train statisticians to work with policy makers effectively, and the short-term political perspective that stands in contrast to a more long-term perspective needed to invest in the appropriate technology.