“ To retain trust in official statistics, the statistical agencies need to decide according to strictly professional considerations, including scientific principles and professional ethics, on the methods and procedures for the collection, processing, storage and presentation of statistical data. „
The work of official statistics is not just collecting data and crunching numbers. There is a whole chain of processes, from deciding what statistics are needed, hiring staff, designing surveys and identifying data sources, through gathering and processing data, to publishing and explaining the statistics.
Every link in this chain involves making decisions from among a range of options. In official statistics, these choices must be driven purely by professional motivations—never by political ones. While accountability to the taxpayer and value for money are also key considerations (as will be seen in Principle 3), this does not mean that cheapest is necessarily best. Scientific excellence, above all else, determines the methods official statisticians select to conduct their work. This includes statistical methods as well as those from the fields to which the statistics relate, such as economics, demography, and climate science, as well as ICT tools and modern communication techniques. Indeed, in fulfilling this principle, official statisticians strive not only to follow the best available scientific methods, but to work tirelessly to develop these methods.
As in any profession, applying professional ethics at every level from the individual staff members to the organization as a whole is essential to maintain the good reputation of the industry. Nowhere is this more true than for official statistics, where safeguarding this good image is crucial for maintaining public trust—which is essential both in order for respondents to continue providing the data needed for statistics, and in order that statistics are believed and used to their full potential.
United Nations Economic Commission for Europe
Palais des Nations, Geneva
