Comment

5 Takeaways from the International Conference on Big Data for Official Statistics

Big data matters for official statistics. Nowhere is this more apparent than at the international gathering of the topic organized by the UN every year. This year’s conference had an air of urgency, taking a critical look at data collection during a global pandemic. The 6th International Conference on Big Data for Official Statistics was organised by Statistics Korea and the UN Global Working Group for Official Statistics, taking place from 31 August until 2 September. As we all know, information at such events is plentiful, which is why we put together our 5 key takeaways from the conference.

Positium colleagues tuned in to the conference from our meeting room.
1. There is no silver bullet


NSOs have adapted to the ongoing crisis by combining both traditional and newer methods. There is an understanding that in order to get a good picture of what’s going on, different datasets have to be combined – “There is no silver bullet”, as Frankie Kay (ONS) said. In order to understand the quickly changing world around us, both traditional and new big data sets, such as MPD, have to be combined. Just as Stefan Schweinfest from the UN Statistics Division said, in every crisis there is an opportunity. To be successful, developing new methods for data integration between new and old sources (like mobile phone data and surveys) is needed, Christophe Bontemps from UNSIAP noted.

2. Society needs fast, robust and reliable statistics

The COVID-19 has put the world to an unprecedented situation. NSOs throughout the world are tackling this challenge with newer and better methods. As Mart Mägi from Statistics Estonia said, during the crisis, governments need more frequent data to support the situation. He described how the use of mobile positioning data helped the Estonian government get daily mobility information in order to find out if the lockdown rules have been operational, and how the situation is in different parts of the country. Tom Smith (ONS) said that statistics and data has never been more important than now – we need trusted and robust data on the spread of the novel coronavirus. But even without the crisis, societies are developing, which raises the need for fast and reliable statistics, as May Offermans from Statistics Netherlands put it. Titi Kanti Lestari (BPS) stressed that society needs qualified and trusted statistics.


3. Security and trust are vital

When using different and new datasets, such as mobile positioning data, discussions about security and trust are vital for success. In Estonia, the use of MPD for mobility monitoring raised an important public discussion about privacy, trust and confidentiality, which is something that is always important to address. In the future, the challenge of data governance – how and by whom the data is obtained, kept, analysed and dissimilated is even more important, as Mart Mägi  (Statistics Estonia) put it. For interoperability, high and common security standards are crucial. Erwin Knippenberg from the World Bank stressed the importance to engage directly with the government and regulators about privacy topics, to make sure that there is understanding, ownership and political will for this kind of analysis. 

4. Collaboration is the key to success

For most of the world, big data is a novel area, where best practices are still evolving. Through the conference, the need for sharing knowledge, methodology and best practices was stressed numerous times by Claire Melamed (Global Partnership for Sustainable Development Data), Gemma Van Halderen (UNESCAP), May Offermans (Statistics Netherlands), David Johnson (ONS/Data Science Campus) and many others. Collaboration and discussions, both at the global level and with local stakeholders, are important since solving privacy, technical and other concerns takes a lot of energy and expertise. We shouldn’t have to reinvent the wheel in every country, as Erwin Knippenberg from the World Bank noted. We were happy to see that the UN handbooks on different topics, such as mobile positioning data, have been noticed and implemented successfully, which was mentioned by many parties like Jeong-ran Kim (Statistics Korea), Titi Kanti Lestari (BPS) and Bert Kroese (Statistics Netherlands). As Lestari highlighted, the MPD handbook offers information about quality assurance, sound methodology and privacy-preserving processing, which were often revealed as challenges of processing big or actually any kind of data. 

5. Big data is essential for achieving SDGs

Countless presenters showed how their use of big data in official statistics helps directly to measure the progress towards SDGs, whether it is to measure something that could not be measured before, or measuring it more effectively, more timely, in more detail. Dennis Mapa (The Philippine Statistics Authority) said that Philippines recognize big data and non traditional data sources for complementing the production of official statistics in order to strengthen their data ecosystem. Titi Kanti Lestari (BPS) explained that data is important for monitoring SDGs and policy making. For this cause, MPD offers more granular, frequent and timely data for SDGs, she said. Tracking such progress on global goals is essential and can be best achieved using a unified system. As stressed by Kecuk Suhariyanto (BPS), there is a need for a manual standard from the UN to maintain the comparability of data.


These 5 key messages resonated throughout the event. You can find the full programme along with presentations, available for download, at the official conference website. We are very happy to see that there is a high level global discussion on these vastly important topics.

What were your takeaways for this year’s conference? Send us a Tweet @positium.


PS! Check out the UN GWG Handbooks currently available: 



Add a comment

Email again: