Comparative statistical analysis of official foreign trade results

Abstract

The problem of information reliability securing occurs when making international comparison, and it becomes a particularly pressing one when several sources are used. The official data on foreign trade results in a country nearly always contain discrepancies complicating the macroeconomic analysis. It is traditionally considered that import data are more reliable than the export data; however, there are no any more accurate methods for discrepancies assessment. The article presents a statistical method of comprehensive comparative analysis of foreign trade results based on two sources of information, aimed to assess existing digressions and to detect any patterns in change of their values. The method is based on mirror statistics for export and import commodity structures and is tested using the information on foreign trade in Russia and Norway. To exclude any infl uence of errors conditioned by specific correlation in characteristics of these countries’ customs accounting, only the official data from Norwegian sources, presented in different classification systems, were used. Refs 20. Figs 6. Table 1.