Report from one of the sessions at the RSS 2017 Conference. More reports of conference sessions are listed here.
Daria Gromyko of the RSS Official Statistics Section introduced this session at the RSS 2017 conference, organised jointly with the RSS International Development Section.
Hugh Stickland, Deputy Director of Strategy and Engagement at ONS began by noting how ONS' mission statement emphasised ‘Better Data for Better Decisions’ (PDF). Hugh then outlined a range of ONS examples including the state pension age review (where decisions affected employees, the fiscal balance and taxpayers), analysis of suicides by occupation (where the results were used by government and employers to help target efforts aimed at prevention), analysis of the disability employment gap, and analysis to help workless families. Following the Digital Economy Act and the creation of ONS’ Data Science campus, ONS will be collaborating more with others including government departments and academia, and using more administrative data alongside survey data.
Matthew Powell, Oxford Policy Management, presented a range of examples of 'Better Data, Better Lives in the Punjab'. Excitingly, a critical mass of infrastructure, capabilities, and skills to create data systems has built up in the Punjab - and in other areas in the developing world. These systems include both better ways of presenting data, such as maps showing travel time to primary schools, and open web based tools to manage vaccination, health services and disaster relief. These systems have not only improved services and produced financial savings, but are producing data streams such as registers of all children’s vaccination and health status, which overlap with traditional statistical products. However, official statisticians have only had a limited role in these developments and face a challenge to make use of the new sources and avoid being bypassed or becoming less relevant to policy makers.
Tom Wilkinson, head of MI Analytics and Alice Marshall, ODA statistics, presented their work in DfID on 'Democratising Data'. The work is driven by a range of factors including:
- The increased external scrutiny of DfID’s work over recent years,
- The ‘open data’ and ‘democratising data’ agendas,
- The benefits of identifying good practices and which sorts of interventions work best,
- Sharing data across the organisation and with key networks and partners.
A range of scorecards, data tools built on pivot tables and access to more detailed data had been developed, resulting in greater data scrutiny. The work had also resulted in significant improvements to the presentation of statistical outputs.
Deana Leadbeter (Health Statistics User Forum, discussant) thanked all the speakers and identified some common themes, including the need to involve and make use of analysts’ and others’ expertise outside of the official statistics world - both data suppliers and users. She highlighted a key point from Hugh’s presentation about needing the right people in the right roles with right skills and suggested that this needed to cover all analysts working in this area, not just those within the GSS 'family'. She agreed with Matthew’s concern about the lack of engagement between those developing e-solutions and statisticians and other analysts, noted that highlighting the benefits of collaboration for both sides could help facilitate engagement, and invited DfID to consider using expertise and data from more sources, including NGOs.
A Q&A followed, covering a range of topics including how to go beyond slogans and achieve impact, how people make decisions (eg by better press coverage, statistical disclosure control of administrative datasets), how to help users get analytical help with asking questions and using data, and bridging the gap between statisticians and analysts and other experts.
More on the Official Statistics Section on this web page.