Abstract

Data from mobile network operators (MNOs) is emerging as a promising resource for improving the quality of official statistics, helping to refine public policies and inform public discussion. MNO data can be combined with two categories of data: on the one hand statistical data, gathered by NSIs primarily for statistical purposes (census, surveys…) and on the other hand non-statistical data, collected for alternate purposes and subsequently repurposed for statistical analyses: administrative data but also emerging sources like traffic loop data and satellite data.

Combining MNO data with other data sources has two major interests. The first one is to improve the general quality of produced statistics. Indeed, MNO data, primarily collected for network maintenance, may lack documentation on coverage deficiencies or network incidents. Adjusting counts through the incorporation of census or administrative data helps enhance the representativeness of these datasets. The second interest of this combination is to shed a new light on topics addressed by traditional official statistics. For instance, investigating inequalities in exposure to air pollution necessitates the fusion of MNO-derived daytime pollution data with air pollution data from specialized organizations and socio-demographic data collected by NSIs. In essence, every data source has inherent limitations, whether related to information coverage or representativeness, and combining data offers a strategic approach to mitigate these limitations.
This presentation is based on the work of Work Package 2 (WP2) within the ESSNet on “Trusted Smart Statistics – Methodological Development Based On New Data Sources”, funded by Eurostat. This research initiative focuses on developing methodologies for the integration of both MNO and non-MNO data. WP2 involves collaboration among NSIs from France, the Netherlands, Germany, Portugal, Romania, Sweden and Italy. Its objective is to establish a short list of the most promising non-MNO sources for integration with MNO data, providing a comprehensive review of their advantages, disadvantages, costs, gains, and sustainability through systematic identification.