11 Mar 2019 12:15 to 13:00
Session ID: 4
[Read more session reports and live updates from the OECD Going Digital Summit]
The Summit Issue Note initially framed the session around challenges that evidence-based decision-making faces in digital age due to how existing metrics and measurement tools are impacted by the fast pace of change of digital transformation.
Mr Paul Schreyer (Deputy Director for Statistics and Data Directorate at OECD) portrayed digital transformation as both an opportunity and risk for measurements, while his colleague Ms Alessandra Colecchia (Head of the OECD Economic Analysis and Statistics Division), explained that most data collected nowadays – such as those generated by machines – were not meant to serve as statistics. This introduces challenges related to the representativeness and quality of data for analytics, and to privacy, security, and costs (even if most data are free, one needs advanced skills and tools to curate and manage them). On the other hand, she noted, digital tools allow organisations to collect, verify, and disseminate data more easily, and there are already examples such as the UN Global Working Group on Big Data Project Inventory, that uses geolocation and smart meters for energy, as well as local examples in developing countries (such as tracking poverty reduction through observing satellite photos of roofs and how they get replaced with new materials). However, the adoption is still slow, she warned, as OECD data show that only some 10% of small and medium enterprises (SME) use predictive analytics, compared to about one third of big companies.
Responding to a question by the moderator, Mr Andrew Wyckoff (Director for Science, Technology and Innovation at OECD), about the role of cloud computing; Mr Anil Arora (Chief Statistician of Canada) stated that it can move SMEs to use data analytics, while Schreyer reminded that the volume of cloud services provided can help measuring growth in places where services are provided and offered.
Ms Rebecca Riley (Director of Economic Statistics Centre of Excellence of the UK), noted that the cloud also drives national statistics institutions (NSI) to adopt new data analytics approaches, yet emphasised the barrier for such adoption: skills that are largely missing, since data science techniques are commonly not learned within university curricula. Mr Arora added to challenges: public acceptance of the cloud, though critical, is still missing; there is a paradigm shift from collecting to accessing and managing data (including keeping privacy and confidentiality); and data analytics activities are now shifting from high skilled IT staff to statisticians and mathematicians.
Schreyer explained that NSIs are in transition from survey to data integration institutions, and should combine old and new approaches. She gave an example of the Swiss housing register which made an agreement with banks (that carry transactions and hold all the data) to enable statistic offices to crawl data in order to create house price indexes. Colecchia agreed that NSIs have to work closely with the private sector, and particularly with big cloud service providers which hold vast amounts of data, in order to overcome the challenge of analysing cross-border data flow. One can observe the traffic, but not the data within it, and particularly not the value of the data (because the value depends on how it is used, and by whom), Colecchia explained.
Asked by Wyckoff what is the role of the OECD, Mr Arora expressed belief that the OECD should advocate the role of the NSIs and help them to learn, adjust, and address the problem of multinational data. Riley emphasised OECD’s role in setting measurement standards and promoting them across countries, giving examples of the two recent studies. One on the ways countries calculate labour input needed to measure productivity, which highlighted differences in compilation methods across countries, and thus limitations with comparing the results. And another ongoing study on analysing microdata in countries, that are not accessible across countries. Schreyer supported the OECD’s role in harmonising methodologies and concepts so that country-to-country analysis are comparable, and saw it also as a platform for sharing tools and algorithms to be used internationally, for compiling information across national boundaries (with the help of NSIs who are working within national boundaries), and for leveraging conversation with big private players. Colecchia concluded with a reminder that the OECD has just developed a roadmap ‘Measuring the Digital Transformation: A Roadmap for the Future’ with nine actions, and invited all to implement parts of the roadmap.
By Vladimir Radunović