Consistency within and between official statistics is a key requirement of data users since it facilitates economic analysis, research and policymaking. This contribution briefly outlines the dimensions of consistency in the area of statistical accounting frameworks: consistency between accounting frameworks and source data, consistency within accounting frameworks (in the balance of payments, imperfect coherency results in errors and omissions), consistency between accounting frameworks (e.g. balance of payments, financial accounts and national accounts), and, ultimately, consistency between statistics of different countries (e.g. cross-border asymmetries). Based on this, it will discuss the arguments for and against strict consistency. The main part will deal with the practical steps needed to achieve consistency as well as the role of the CMFB and CMFB activities in this regard. These steps include: (1) consistent international manuals and classifications, (2) consistent interpretation of the manuals, (3) consistent implementation (including data sharing and estimations), and (4) consistent revisions. It turns out that much progress has been made over the last years but that statistical differences will not vanish altogether. However, user demand for consistency encourages official statisticians to learn from one another. This is a key factor driving further improvements in the quality of official statistics in many respects.
