Physica A: Statistical Mechanics and its Applications 525, pp. 761-770.
ISSN/ISBN: Not available at this time. DOI: 10.1016/j.physa.2019.04.042
Abstract: There is a widespread consensus that the national statistics on inflation were manipulated by the Argentinean government from 2006 to 2015. The best known tool to run a forensic analysis of this claim is to check for the validity of Benford’s law in the data series. We find that indeed, the inflation for that period fails to satisfy this statistical regularity. We further compare this behavior to that of Argentina’s inflation series for the same period but recorded independently of the government; to that of the national records of 1943–2006, as well as to historical series of other countries. We find again that Argentina in 2006–2015 is the only one in our sample that can be singled out as candidate for statistical manipulation. Alternative hypotheses for why the inflation series failed to satisfy Benford’s law can be formulated. One is that, it may be due to rounding price level figures to the significant digits. Or that it is due to changes in the base years which leads to splicing different series of general level of prices. We consider these alternative hypotheses and run simulations to assess them. We find that, independently of these possible changes in the underlying series of prices, the ensuing series of its variations, i.e. the series of inflation rates, always satisfies Benford’s law. Therefore we can claim that, indeed, inflation data was tampered with in Argentina for an entire decade.
Bibtex:
@article{,
title = "Tampering with inflation data: A Benford law-based analysis of national statistics in Argentina",
journal = "Physica A: Statistical Mechanics and its Applications",
volume = "525",
pages = "761--770",
year = "2019",
issn = "0378-4371",
doi = "https://doi.org/10.1016/j.physa.2019.04.042",
url = "http://www.sciencedirect.com/science/article/pii/S037843711930411X",
author = "Maximilano Miranda-Zanetti and Fernando Delbianco and Fernando Tohm{\'e}",
}
Reference Type: Journal Article
Subject Area(s): Economics, Statistics