National statistical organizations at a crossroads: Trust, independence, and democratic relevance in the AI Era

Paper by Anil Arora: “National Statistical Offices (NSOs) were established to serve essential state functions—taxation, mobilization, and administrative governance—but evolved into professional institutions that safeguard the public interest through impartial statistics. Today, despite sustained modernization, many NSOs face declining survey participation, fiscal pressure, political encroachment, and mounting competition from private-sector and AI-mediated data actors. This paper is intended to stimulate discussion within the international statistical community and those dependent on high quality data, information and insights to make decisions. It does not seek to exhaustively catalogue the many national statistical organizations, national statistical institutes (NSIs), and professional associations engaged in innovation and reform to increase trust, relevance and impact. Rather, it highlights illustrative examples of innovation, leadership, and risk to underscore a central proposition: modernization of methods and technologies is necessary but not sufficient. Sustained trust requires visible independence, ethical leadership, proactive engagement, and institutional courage. Drawing on international initiatives led by the United Nations Statistics Division, OECD, Eurostat, the International Association for Official Statistics (IAOS), the International Statistical Institute (ISI), and the American Statistical Association (ASA), among others, as well as modernization leadership within several advanced statistical systems, the paper proposes a comprehensive reform agenda for democratic resilience, notably in an emerging AI-mediated information ecosystem…(More)”.