1. the future of Statistical Production CSPA

2. We need to modernise We have a burning platform with: rigid processes and methods; inflexible ageing technology; increasing cost of traditional data collection methods; inability to quickly respond to emerging information needs; slow to harness new and alternative sources of data (such as sensor, satellite); difficulty in attracting and retaining skilled staff in the competitive labour market. In an increasingly digital and data rich environment statistical organizations are struggling to remain relevant.

3. Statistical organisations already participate in many international engagement activities that facilitate sharing. All statistical organisations need to modernise and standardise their systems and processes. The marginal cost of doing this in a way that supports collaboration and complies with CSPA is relatively low, but the potential savings enabled by such a standard approach are high. Key Messages

4. Toolbox

5. A vision for an aligned and collaboratively led community. Allows all of us to benefit from collaboration and sharing. Four main principles: −Openness −Flexibility −Participation −Pragmatism The Statistical Modernization Community

6. Extending our existing sharing practices to the next level. Heads of Organizations agree in principle to sharing and collaborating in the community. Organizations contribute as much as they can, in what form they can (capability, funding, etc.) Sharing in the Community

7. Vision of an aligned and collaboratively led community. An active community seeking to leverage Common Statistical Production Architecture (CSPA) will allow statistical organisations to contribute to achieving the vision of the High Level Group (HLG). Individual organisations voluntarily identify the nature of their contributions with the support of a global community. Together, the community will create a robust and scalable business driven statistical production platform. The role of HLG is to provide stewardship and to assist in steering the community to deliver on the shared goal in an efficient manner. They recognise the right of individual organizations to determine their own contributions based on their own priorities. HLG are the holders of the vision and they are influential supporters of the work done to help realise this goal. Benefits of the Community Approach The community approach is a more effective approach to delivering statistical services, by lowering the costs of overall production and minimising duplication by working towards a shared goal. Statistical organizations are able to optimise their level of contribution based on their current capability and resourcing constraints. The engaged community working on project of mutual interest strengthens understanding, design, implementation and adoption – ensuring statistical services are more fit for purpose scalable and robust. The community can broker the exchange of information within the community, linking members who share a common interested. The community openly shares learnings and implementation issues, and seeks to overcome. Sharing to maximise participation The community encourages participation and sharing at every level. It is only through active sharing that the community will deliver on HLG’s aspirations. Collaboration and licencing guidelines must neither restrict nor discourage participation within the community. The community should aim to gather requirements for a new statistical service as widely as possible. Multi- organization collaboration during requirements gathering ensures that statistical service can be widely adopted for a marginal cost. Licencing arrangements should enable services to be used widely within the community. While there are a range of available licencing arrangements, the owner of the service must clearly communicate the licencing agreements. At a minimum these arrangements should supporting bounded usage with the community. Principles For this model to successful, community members commit to working in alignment with CSPA and the following key principles: Pragmatism Community members contribute by doing the things they care about and delivering a Minimal Viable Product. Collaboration is about collective engagement for the common good and is the faster route to project success. Participation Members will actively contribute either time, code, funds to make a tangible contribution to the community. Members share their thoughts, comments and evaluations with each other. Flexibility The community is self-organised and operates free from central control. New innovations are encouraged to emerge quickly. Projects can be initialised and progressed as community members identify common interests. Openness Community welcomes new members to contribute and share within Statistical Organisations. All communications about contributions are open and transparent

8. Deciding what to invest in as a community We need to expose investment plans where we are willing to collaborate Tailored level of contribution allows every country to participate

9. Discussions of Need identified a great deal of common need. CandidateCountry Probabilistic record linkingLead: Canada Collaborators: New Zealand Implementers: Australia Classification ServiceLead: Norway Validation rules specification (rules engine)Lead: Australia Cloud based.statLead: OECD (provisional) Microdata access (Confidentialised analysis of microdata) Collaborators: Australia, Finland Test: New Zealand Implementers: Canada Coding / using machine learningCollaborators: Canada, New Zealand, Italy, UK Web scrapingCollaborators: Australia, New Zealand, Italy, UK Edit and checking - BANFFImplementers: Australia, Canada, New Zealand FAME wrapped capabilitiesCollaborators: Australia Geospatial visualisationCollaborators: some indication from Poland Imputation – CANCEISCollaborators: UK Wrap BLAISE to surface capability services Interviewer workload management Admin data classification (e.g. for scanner data)

10. A tool to capture\update the current state of statistical organisations’ portfolios and highlight the areas where investments are planned. A portfolio management tool The process of synthesizing investment plans uses the CSPA portfolio management tool to create an overall picture of the CSPA community’s ICT portfolio. This unified view provides the CSPA community with: A view of investment coverage and gaps Areas of likely duplication or overlap Areas of vendor engagements A consolidation of the types of investments to be undertaken

11. How do I find out more? The CSPA Catalogue

12. CSPA Wiki http://www1.unece.org/stat/platform/display/CSPA How do I find out more?

13. There is a knowledge brokering benefit to this work over and above the building of re-usable connections. This could lead to other forms of collaboration being facilitated. We think that technical integration issues have been solved (enough), the challenge now is whether we can develop ways of working together to create mutual benefits. What is the alternative ? The premise is that international statistical agencies are going to have to become more inter-operable so why not with each other. Reflections

14. This is an invitation to join and influence a community. There is a lot to learn about how to make this work and the more agencies involved the greater the potential benefits. We are looking to engage more of the professional specialisms of National Statistical Offices in creating tools. Help clarify what you would need to use this Consider populating the investment intentions spreadsheets and help create the possibility of different forms of useful knowledge being exchanged. Invitation