Relations between statistical offices and governments

Abstract

In their relations with the rest of society (especially the world of politics) statisticians may encounter two types of problems: the statistical object may be handled without the caution and distance that are sometimes required; the statistical object may be exposed to radical criticism. This paper outlines that research and analysis can help to resolve these difficulties. It also addresses the more controversial issue of whether national statistical institutes themselves should engage in statistical and economic studies. While there are the advantages of an in-house research and analysis capability, some study areas must be excluded from the NIS's terms of reference in order to avoid an excessive closeness to the political sphere.