OECD: Pushing the Boundaries of Data Sharing to Shape Better Public Policy with Power BI and its SDMX Power BI Connector

From improving economic performance to creating jobs, from promoting effective education systems to combating international tax evasion, the OECD provides policy analysis to leaders while facilitating knowledge sharing among countries and actors of change. To do this, it relies on a large community of statisticians and publishes up to 5 billion points data each year. Handling this kind of  volume of information is also a challenge.

While data sources are multiplying, the quality levels remain disparate, lengthening the time needed to create data products including dashboards and data visualisations. Added to this is the ever-increasing decentralization of data producers. The need for high-performance visualization tools capable of connecting, exploiting, and crossing factual data, and thus highlighting concrete solutions to today's major challenges, is becoming more and more pressing.

“With hundreds, thousands or even millions of lines of data stored in databases, how can we effectively tell a story," asks the open source community of official statistics, the Statistical Information System Collaboration Community (SIS-CC). “How do you access these huge volumes of data, how do you identify and understand patterns to support evidence-based decisions?” The OECD has responded to these questions with a four-letter answer: SDMX, or Statistical Data and Metadata eXchange, an ISO standard for open data, exchange, and driving standardised metadata modelling practices, making the data accessible to all from international statistical organizations, central banks or national statistical offices to government policy makers and journalists via a standard interface (API).

Power BI and SDMX: Crossing data sources to reveal new evidence

"At the OECD, we use the SDMX standard to restructure our data in a harmonized way" says David Barraclough, head of SDMX and data modelling at the OECD. All statistical domains can be represented in a semantically meaningful and harmonious way, explains the expert: employment data, trade data, education data, health data, population censuses, etc.

The SDMX Power BI connector developed by SIS-CC Open Source Community and now officially certified by Microsoft, makes it easy to export and share data. The result is data that is accessible to a much wider audience, end-users better equipped to create optimized and enriched data visualization and in-depth analysis of social and economic policies.

SDMX Power Bi Connector a proven value 

The SDMX Power BI connector has already enabled many concrete use cases. The Pacific Community, an international organization that contributes to the development of the 22 Pacific island states, adopted the connector to better understand the trade flows taking place in the region.  Using datasets from the Pacific Data Hub, a platform that centralizes all the region's data and meets the SDMX standard. The dashboard created via Power BI highlights inter-country trade relations, the amount of exports and the origin and destination of these flows.  

Tasked by G20, OECD experts collected data and developed a comprehensive analysis of sovereign debt issues, helping creditors enforce debt transparency principles.

The SDMX Power BI connector was also very useful during the Covid period: “to enable the tracking of recovery and vaccination data, but also to measure the impact of the pandemic on the labor market, notably on unemployment, working conditions and unpaid work” says Iulian Pogor, data architect and statistical tools engineer at the International Labor Organization (ILO).

One of the main advantages of the SDMX Power BI connector is the ability to link data from different sources without in depth technical training.. Intuitive use is a key ingredient for widespread adoption within the statistical community. When available using the same standard, with  data modelled according to SDMX good practice, it is possible to cross-reference them "and reveal new statistics", says David Barraclough.

Cutting by 90% the steps required to create dashboards 

Power BI integrated data is more readable and easily sharable. Users can graph data, build dashboards, and highlight relationships and correlations. "Power BI helps us analyze data, build powerful reports, quickly use them and cross-reference data from different sources”. The results can then be shared with end-users, "internal stakeholders, policy analysts, statisticians, government officials, students, journalists...", says David Barraclough.

For organizations, time savings are enormous: the Pacific Community, for example, estimated that using Power BI in combination with the SDMX standard reduced the time required to gather and process data to only 10% of the steps required with a custom script. "It's amazing how quickly a decision support dashboard can be built using Power BI combined with pulling data from our SDMX API" says Phil Bright, Manager Statistics Infrastructure and Dissemination,  Pacific Community.

The process is made even easier by the fact that SDMX has been certified through an iterative process with Microsoft and is now part of the platform's built-in connectors. "This is really great news, not just for the OECD but for the entire official statistics community" says David Barraclough. Within organizations where IT access is tightly controlled, the certification of the connector ensures direct, immediate, and secure access to the tool.