Politics and Policies of Statistical Independence

Abstract

The issue of scientific and professional independence is a central concern of National Statistical Offices (NSOs). It pertains not only to scientific and methodological dimensions, such as definition of concepts, or design of surveys and questionnaires, but also to issues such as data release policies, definition of the overall statistical program, nomination or dismissal of the chief statistician, an NSO’s position within government architecture, as well as budgetary discretion. Independence is now an explicit feature of statistical acts and codes of practice in many countries and appears as a condition for securing the public’s trust. After describing the range of formal and informal means developed for protecting statistics independence in OECD countries, this chapter sketches three interpretations of the independence drive: the principal-agent, the independent expert and the technocrat-guardian models.