Aplikace Benfordova zákona na vládní makroekonomická data: případ obchodní bilance
Alternativní metriky PlumXhttp://hdl.handle.net/11012/35187
MetadataZobrazit celý záznam
Purpose of the article: This article builds on research published in the article “Fact and Fiction in EU Governmental Economic Data”.The first section summarizes the basic theory of Benford’s Law. The possible applications of Benford’s Law in cases detecting manipulation of government economic data is dealt with in the practical section. Methodology/methods: An analysis of available literature and empirical studies is used in the theoretical section. We test hypotheses about the convergence of empirical data against Benford’s distribution in the practical section. The Mean Absolute Deviation and Z test are used as the test criterion. Scientific aim: The primary aim of the article is to summarize the current theory regarding Benford’s Law and the practical applications of this theory to governmental macroeconomic data. We compare the empirical results of the Czech Republic with the results of previous studies and discuss the possibilities and limitations of the utilization of the Benford law. Findings: Based on the critical values of the Mean Absolute Deviation and Z test for the first digit of the trade balance, the hypothesis that the occurence of the first digit converges to Benford’s distribution must be rejected. However, the second digit and the first and second combined we found to converge to Benford’s alias logarithmic distribution. Conclusions: The utilization in field of macroeconomic data is very limited because of limiting factors like the small size of the data sample, structural shifts in the economy, macroeconomic adjustment, general availability of data and finally because of the influence of utilized statistical tests. It can be used only as a complementary instrument for macroeconomic data control. The site audit still remains as the crucial instrument.
Zdrojový dokumentTrendy ekonomiky a managementu. 2014, VIII, č. 19, s. 45-51. ISSN 1802-8527.