Official Statistics and Research: an Evolving Relationship seen through the Classifications of Science

The relationship between scientific research and official statistics is currently a subject of great interest. There are several reasons for this new attention: First and foremost, the growing awareness that production and dissemination of official statistics requires a substantial input of scientific research for: 1) the construction and maintenance of conceptual frameworks, definitions and classifications; 2) structuring and restructuring production methods; 3) the analysis of results and the operation of quality control systems, etc.; Second, the cumulative pressures of increasing costs and decreases in public resources push for innovation in productive processes and products. The opportunities offered by new technologies and by the results of scientific research lead statisticians to look for new and closer connections with the world of research, and to promote research within the institutions of official statistics; The need for credibility, reputation and independence from political pressures becomes ever stronger, particularly when official statistics are used as explicit parameters for political or administrative action. Official statistics are called upon to consolidate the scientific foundation of its activity; to reinforce the professionalism of its practitioners; and to appeal to the authority of science and “truth”, seen as a distinct source of legitimisation, separate from the authority of the State and of democratic investiture; The orientation toward users, which more and more conditions the actions of those who provide public services, pushing official statisticians to establish a closer relationship with the centres of academic and non-academic research that have always been among the main users of statistical data; The relationship with centres of scientific research is also undergoing a consolidation because of the need for collaboration in data analysis and in the explanation and interpretation of phenomena. The conventional division of roles between official statistics and the world of research once saw the former as purely a producer, almost monopolising the supply of data, and the latter as simply a user of these data, in an almost exclusive monopsonistic position. This division no longer holds true: today official statistics is not the only producer of data, nor can official statistics be limited to providing raw data (just what does that mean?) without explaining their meaning and evaluating their quality for the interpretation of phenomena. Official statistics are therefore the first users of the data they produce. In the new relationship, collaboration and interaction with the world of research become an indispensable tool. Therefore, opportunities for partnership with the research sphere for the production, dissemination and analysis of data multiply. We must remember that the relationship between scientific research and official statistics is not at all new. On the contrary, from its origins official statistics has had an intermediary function at all new. These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.