In recent years, the quantity of digital data created, stored and processed in the world has grown exponentially. The demand for statistical information has never been so apparent. For ‘official statistics’, the public informational infrastructure provided by statistical authorities, this new era offers not only essential opportunities but also manifold threats. In order for official statistics to function as a language for all kinds of societal interactions and decision-making, it is essential that the product ‘information’ is fit for purpose. This is an essential shift in perspective. Quality of official statistics needs to be seen with a much wider scope, going beyond the side of production, including the use side and analysing scientifically how these two sides are interacting in a dynamic relationship.

۱ Statistics and modernity

Essentially, statistics is the science of learning from data. Certainly, it is a modern technology that is part of the standards of today’s information age and society and is used in a wide array of fields. ‘Official statistics’ is one of these domains of practical applications of the ‘quantification as a social technology’ [۱] belonging to those with the longest history. Since the beginning of the nineteenth century [2], official statistics—as a child of the enlightenment—has grown and developed side by side with the different forms of the (modern) state. Desrosières uses the term ‘mutual co-construction’ for three interlinked phenomena, (A) a theory of the state (economy), (B) interventions of the state (policies) and (C) quantification of ‘variables’ specifically targeted by policy measures (statistics) [3]. Basically, official statistics can be defined by using three questions [4]: • Who? Normally, official statistics are produced and provided by statistical offices, i.e. public administrations • What? Statistical programme and priorities are prepared according to public sector standards (i.e. participation of civil society) with the final decisions partly taken in legislative procedures • How? Statistical methodologies are nowadays subject of international cooperation and manifested in statistical standards; high-level quality is assured through management systems and ethical codes. The relationship between official statistics as a technology and the society has been analysed by a relatively small community of historically interested scholars working in the intersection of statistics, sociology, political and historical sciences. Standard literature such as ‘The Politics of Large Numbers—A History of Statistical Reasoning’ [۵], ‘Trust in Numbers: The Pursuit of Objectivity in Science and Public Life’ [۱] or ‘The Mutual Construction of Statistics and Society’ [۶] provides for a profound understanding of this two-way dynamic interaction between official statistics on the one hand and societies, politics and economies on the other. This literature deals with the complexity of relationships, by highlighting the role of official statistics throughout historical episodes of the last two centuries in a couple of countries (France, England, Germany, USA, etc.); for other countries (e.g. Italy) corresponding literature exists [7,8]. ‘XComplexities’ is the term that is used by Longbing Cao, when he introduces his definition of data science: ‘data science = { statistics ∩ informatics ∩ computing ∩ communication ∩ sociology ∩ management ? data ∩ domain ∩ thinking }’[۹].