Self-reported income data: are people telling the truth?

Abstract

Purpose

This paper aims to detect anomalous data in income reports of Argentina, including personal income – from a sample of households – and statements of public officials.

Design/methodology/approach

A widely known technique in forensic accounting – such as Benford’s Law – is used. The Chi-square test and the absolute mean deviation are used for verification. The databases consulted include the income declared by households in the Permanent Household Survey – for the 2003-2017 period – and the capital declarations of high-ranking public officials – for the period 1999-2017.

Findings

The results suggest that income reported in the Encuesta Permanente de Hogares do not follow a Benford´s distribution, and the degree of conformity with this decreases significantly between 2007 and 2015 – coincident with the intervention period of the Instituto Nacional de Estadísticas y Censos. Patrimonial statements of public officials present an acceptable level of compliance with Benford’s law, especially among those of the legislative branch (in more than 90% of cases) although to a lesser extent among officials of the executive branch.

Practical implications

The results suggest that income reports from the Permanent Household Survey, for the period 2007-2015, should be used with reservations because of their possible manipulation.

Originality/value

During the intervention of the official statistics institute in Argentina (2007-2015), the idea of lack of credibility of its reports has been disseminated. To date, however, there is no empirical evidence to support it related to income.

Citation

González, F.A.I. (2020), "Self-reported income data: are people telling the truth?", Journal of Financial Crime, Vol. 27 No. 4, pp. 1349-1359. https://doi.org/10.1108/JFC-08-2019-0113