Abstract

The digitalization of our world has increased at unprecedented rate the collection, storage, analysis and use of data by almost everyone, everywhere. This offers great opportunities for statistical advancement. New data sources can improve the quantity and relative quality of statistics produced, which in turn help generate more timely actionable insights and strengthen decision-making. However, this does not come without concomitant risks related, for instance, to data privacy, data integrity and statistics quality, as well as data access and reliance on new data providers. These risks are exacerbated by the fact that the integration of new data sources in statistics production have not yet reached its maturity in terms of underlying methods, skills and governance mechanisms. How can this situation be best govern so that the benefits of this data (r)evolution largely overcome and address its underlying risks for statistics?

First, what we call “official statistics” need to adapt to this new paradigm to remain relevant. Statistical organizations’ role, capacity and expertise within the broader data sphere needs to be reinforced. For instance, this requires improved regulations and governance mechanisms to enable the access and use of non-traditional data sources by statistical organizations. In particular, the fact that many national statistics legislations have no or very limited provisions for the access of big data and/or Privately-held data for the generation of statistics needs to be addressed.

Second, better governance mechanisms are needed to exercise an adequate control over the use of these new data sources for statistics from the quality, data ethics and data privacy perspective. Indeed, the use of new data sources brings new challenges in these areas and failing to exercise appropriate control or manage those risks could severely damage statistical organizations’ reputation.

Finally, mechanisms are necessary to accelerate the advancement of new methods and tools as well as the development of adequate skills for the integration of alternative data sources for statistical purpose. In particular, investigating and consolidating the use of new data sources requires a much broader set of skills (e.g. data science, GIS and IT-computing) which are rather scarce in most statistical organizations and need to be managed strategically.

Several statistical organizations have however already taken steps towards these objectives and new practices are currently emerging. During this session, experts from a selected number of national, regional and international statistical organizations will discuss recent advancements made in the area of data and statistics governance with the aims to better harness the opportunities but also reduce the risks underlying the rapid growth of data being generated globally and the expansion of data ecosystem to a rising number of actors.