Abstract

Official statistics are almost exclusively based on data collected through surveys and censuses on the one hand, and administrative records on the other. However, modern society and the economy are increasingly fuelled by data held by private organisations, such as transactions data, mobile networks data, sensor data, etc., that have so far hardly been used as a source for official statistics. This situation is undesirable and unsustainable for several reasons. Such privately held data are essential to understanding society and the economy. They can shed light on new phenomena such as the use of platforms and can provide details that surveys can never offer. Using such data for official statistics would significantly increase their timeliness and scope, while enhancing quality and also possibly reducing response burden and costs. In short, such data are indispensable for maintaining the high relevance of official statistics.

In this invited paper session, Mr. Mesenbourg will briefly describe the Committee on National Statistics efforts to build a vision for a 21st century national data infrastructure. The first report in a series of three described the need for a new data infrastructure, presented an initial vision, and described expected outcomes and key attributes. The report also discussed the implications of blending data from multiple sources as well as the organizational implications of cross sector data access. In his remarks, Mr. Mesenbourg will focus on why private sector data assets are an essential component of a new data infrastructure, the challenges and impediments of using private sector data for statistical purposes, and the necessary attributes of private sector data accessed and used in a new data infrastructure. The talk will conclude with suggestions regarding steps that will move the U.S. closer to realizing the promise of using private sector data to strengthen, improve, and transform the way the United States uses and benefits from better information.