The OECD Smart Data Strategy was initiated in 2019 to modernise OECD’s data and statistics. Modernisation starts with improvements in the way data are disseminated and managed, moving away from obsolete tools, as well as from fragmented processes towards pooling of resources, mutualisation and automation of data tasks, as well as data harmonisation (“integrating the data cycle”). Modernisation is needed to enable scaling up of innovative projects using alternative data sources and new data techniques to increase the breadth, granularity and timeliness of the evidence supporting policy analysis (“embracing smart data”):
- Since more than a decade, the “data deluge” creates opportunities and challenges. New policy questions can be addressed as more granular, timelier data help complement existing statistics, enabling micro-level analysis, improving forecasts or combining sources for multidimensional insight and modelling. Demand for new evidence for policy requires tapping into more and more data sources, and leveraging modern data engineering and data science techniques, while ensuring that trusted quality of OECD output continues to be a core value.
- Since 2023, the “AI deluge” adds an additional layer of challenges and opportunities. Generative AI, specifically, is seen as an opportunity to increase productivity in data operations in different ways: assistance in the extraction of structured information from unstructured data; automation of (meta)data quality assurance tasks; meta-data driven automations of data workflows; improvement of code quality and coding productivity... Certainly, the data accessibility and data consumption paradigms are being overhauled, as external as well as internal users of our data are likely to consume them more and more through AI-based tools (the “conversation on data”). These opportunities come also with risks, that require OECD to adapt its policies and frameworks.
In order to keep OECD ahead of the curve and maintain its differentiating edge in evidence-based policy analysis, the OECD Smart Data strategy aims to create, consolidate or adapt the capabilities to create better data for better policies in the new context, in the areas of Data Governance, Data Skills, Data Quality, Data Sourcing, Data Platform, Data Dissemination & Engagement – all being transformed through the AI revolution. Priorities for the new cycle are the following ones:
- Re-engineer the “OECD data factories”, including by leveraging the cloud and experimenting, developing and deploying AI tools in the OECD data value chain, enabling automation and mutualisation of data tasks across OECD.
- Renew the data quality framework to maintain the OECD’s status as a recognised, trusted source of evidence for policy – and more broadly position (meta)data quality as the central value-add in an AI-impregnated world.
- Improve horizontal work on data @OECD by establishing more efficient ways of sourcing and collaborating on data in priority policy areas, and expanding the OECD data governance “toward a governance of contents”.
